Journal of Applied Engineering Science - 3 (2018), vol.16

IMPRESSUM J O U R N A L O F A P P L I E D E N G I N E E R I N G S C I E N C E (J A E S) The journal publishes original and review articles covering the concept of technical science, energy and environment, industrial engineering, quality management and other related sciences. JAES is Open-Access Journal that follows new trends and progress proven practice in listed fields, thus creating a unique forum for interdisciplinary or multidisciplinary dialogue. Electronic journal editing and transparent editorial and review policy are provided, with: • • • • •

Online paper submission and tracking of review process CrossRef: assignment of numerical identifiers (DOI) to assure greater visibility and accessibility CrossCheck: control for originality of submitted papers, to prevent plagiarism and duplicate publications Online-first publishing Automatic transfer of metadata to SCIndeks that support international protocols for data transfer

All published articles are indexed by international abstract base Elsevier Bibliographic Databases through service SCOPUS since 2006 and through service SCImago Journal Rank since 2011.

Serbian Ministry of Education, Science and Technological Development admitted the Journal of Applied Engineering Science in a list of reference journals as Leading national journal M51.

International Editorial Advisory Board Prof. dr Branko Vasiü, Faculty of Mechanical Engineering, Belgrade; Prof. dr Dragan Aleksendriü, Faculty of Mechanical Engineering, Belgrade; Prof. dr Dragan Milutinoviü, Faculty of Mechanical Engineering, Belgrade; Doc. dr Goran Vorotoviü, Faculty of Mechanical Engineering, Belgrade; Prof. dr Ivan Blagojeviü, Faculty of Mechanical Engineering, Belgrade; Prof. dr Jezdimir Kneževiü, MIRCE Akademy, United Kingdom; Prof. dr Jozef Aronov, Russian Research Institute for Certification, Russia; Prof. dr Miloš Kneževiü, Faculty of Civil Engineering, Podgorica, Montenegro; Prof. dr Miodrag Zec, Faculty of Philosophy, Belgrade; Prof. dr Mirjana Misita, Faculty of Mechanical Engineering, Belgrade; Prof. dr Mirko Vujoševiü, Faculty of Organizational Science, Belgrade; Prof. dr Nebojša Bojoviü, Faculty of Traffic and Transportation, Belgrade; Dr Nebojša Kovaþeviü, Geotechnical Consulting Group, United Kingdom; Prof. dr Nenad Ĉajiü, Faculty of Mining and Geology, Belgrade; Prof. dr Nenad Zrniü, Faculty of Mechanical Engineering, Belgrade; Prof. dr Radivoje Mitroviü, Faculty of Mechanical Engineering, Belgrade; Prof. dr Robert Bjekoviü, Hochschule Ravensburg-Weingarten, Germany; MSc Sonny Siniša Vidoviü, Energy Testing and Balance, USA Prof. dr Vesna Spasojeviü Brkiü, Faculty of Mechanical Engineering, Belgrade; Prof. dr Vladimir Popoviü, Faculty of Mechanical Engineering, Belgrade; Prof. dr Vukan Vuþiü, University of Pennsylvania, USA; Prof. dr Zdravko Milovanoviü, Faculty of Mechanical Engineering, Banja Luka; Prof. dr Željko Kamberoviü, Faculty of Technology and Metallurgy, Belgrade;

Publisher Institute for Research and Design in Commerce and Industry - IIPP; www.iipp.rs Copublisher Faculty of Mechanical Engineering – Belgrade University; www.mas.bg.ac.rs For copublisher: Prof. dr Radivoje Mitroviü Faculty of Transport and Traffic Engineering – Belgrade University; www.sf.bg.ac.rs For copublisher: Prof. dr Nebojša Bojoviü Editor in Chief Prof. dr Gradimir Danon, Faculty of Forestry, Belgrade; Executive editor Mr Nada Stanojeviü, Faculty of Mechanical Engineering, Belgrade; Technical editor

Darko Stanojeviü,

Faculty of Mechanical Engineering, Belgrade; Publishing Council Dragan Ĉorÿeviü, Belgrade Waterworks and Sewerage, Belgrade; Dušan Ĉuraševiü, Euro Sumar, Belgrade; Nenad Jankov, Power Plant Kostolac B, Kostolac; Miroslav Vukoviü, CBRE Serbia, Belgrade; Milutin Ignjatoviü, Institute for Transport and Traffic CIP, Belgrade; Editorial Office Miloš Vasiü, Miloš Dimitrijeviü, Ivana ûuk, Milica Mikiü, Ljubica Kneževiü; Printed by: Sigra star, Belgrade; Papers are indexed by SCOPUS UDC 33 ISSN 1451-4117 Online ISSN 1821-3197 www.engineeringscience.rs

Journal of Applied Engineering Science 16(2018)3

INSTRUCTIONS FOR AUTHORS Submission of the papers has to be done online, through journal e-service at http://aseestant.ceon.rs/index. php/jaes/login. For assistance during the process of submission and publication, please contact graphical editor Mr. Darko Stanojevic at dstanojevic@iipp.rs or +381 116300751. Every manuscript submitted to JAES will be considered only if the results contained in the paper were not already published, that are not currently in the process of publishing and not to be published in another journal. Each paper is sent to a review by two independent experts and the authors are obligated to adopt the observations and comments of the reviewers.

Articles presented at conferences may also be submitted, provided these articles do not appear in substantially the same form in published conference proceedings.

All articles are treated as confidential until they are published. Manuscripts must be in English free of typing errors. The maximum length of contributions is 10 pages.

THE FORMAT OF THE MANUSCRIPT The manuscript should be written in the following format: • A Title, which adequately describes the content of the manuscript.

• An Abstract should not exceed 250 words. The Abstract should state the principal objectives and the scope of the investigation, as well as the methodology employed. It should summarize the results and state the principal conclusions.

• Not more than 10 significant key words should follow the abstract to aid indexing.

• An Introduction, which should provide a review of recent literature and sufficient background information to allow the results of the article to be understood and evaluated.

• A Theory or experimental methods used.

• An Experimental section, which should provide details of the experimental set-up and the methods used for obtaining the results.

• A Results section, which should clearly and concisely present the data using figures and tables where appropriate.

• A Discussion section, which should describe the relationships and generalizations shown by the results and discuss the significance of the results making comparisons with previously published work. (It may be appropriate to combine the Results and Discussion sections into a single section to improve the clarity).

• Conclusions, which should present one or more conclusions that have been drawn from the results and subsequent discussion and do not duplicate the Abstract.

• References, which must be cited consecutively in the text using brackets [01] and collected together in a reference list at the end of the manuscript and in alphabetic order.

Units - standard SI symbols and abbreviations should be used.

Abbreviations should be spelt out in full on first appearance, e.g., variable time geometry (VTG). Meaning of symbols and units belonging to symbols should be explained in each case or quoted in a special table at the end of the manuscript before References.

Figures must be cited in a consecutive numerical order in the text and referred to in both the text and the caption as Fig. 1, Fig. 2, etc. Figures should be prepared without borders and on white grounding and should be sent separately in their original formats.

Pictures may be saved in resolution good enough for printing in any common format, e.g. BMP, GIF or JPG. Tables should carry separate titles and must be numbered in consecutive numerical order in the text and referred to in both the text and the caption as Table 1, Table 2, etc. The tables should each have a heading.

Tables should not duplicate data found elsewhere in the manuscript.

Acknowledgement of collaboration or preparation assistance may be included before References. Please note the source of funding for the research.

REFERENCES must be written in alphabetical order and in the following form: Journal: /Number/ (must match number in the text), Last name, Initial of the authors name, (Year of publication). Article title: secondary title. Title of the Journal (italic), volume number (number of the journal), page number Book: /Number/ (must match number in the text), Last name, Initial of the authors name, (Year of publication) Book title: secondary title, Place of publishing: Publisher.

Book chapter: /Number/ (must match number in the text), Last name, Initial of the authors name, (Year of publication) Chapter title: secondary title, Book title: secondary title, Place of publishing: Publisher, page numbers.

Internet source: /Number/ (must match number in the text), link to the page from which the text is taken, retrieved on (state the date) Journal of Applied Engineering Science 16(2018)3

CONTENTS

Johnny Robinson Zambrano Carranza, Stanislav Kovshov, Evgeniy Lyubin ASSESSMENT OF ANTHROPOGENIC FACTOR OF ACCIDENT RISK ON THE MAIN OIL PIPELINE PASCUALES - CUENCA IN ECUADOR

307 - 312

Aleksandr Petrovich Mokhirev, Marija Olegovna Pozdnyakova, Sergej Olegovich Medvedev ASSESSMENT OF AVAILABILITY OF WOOD RESOURCES USING GEOGRAPHIC INFORMATION AND ANALYTICAL SYSTEMS (THE KRASNOYARSK TERRITORY AS A CASE STUDY)

313 - 319

Martin Ficek SIMULATION AND MODELING IN CRISIS MANAGEMENT

320 - 327

Mikhail Smirnov, Yuri Andrianov, Victor Chernyakevich TECHNOLOGICAL MODERNIZATION OF FOREST ROADS CONSTRUCTION IN RUSSIA

328 - 332

Yuan-Shyi Peter Chiu, Hong-Dar Lin CREATION OF IMAGE MODELS FOR INSPECTING VISUAL FLAWS ON CAPACITIVE TOUCH SCREENS

333 - 342

Andrey Ostroukh, Nataliya Surkova, Oleg Varlamov, Valery Chernenky, Alexander Baldin AUTOMATED PROCESS CONTROL SYSTEM OF MOBILE CRUSHING AND SCREENING PLANT

343 - 348

Miloš Maljković, Ivan Blagojević, Vladimir Popović, Dragan Stamenković IMPACT OF THE DAMPER CHARACTERISTICS ON THE BEHAVIOR OF SUSPENSION SYSTEM AND THE WHOLE VEHICLE Katja Vogrinec, Miroslav Premrov EXPERIMENTAL AND ANALYTICAL STUDY OF THE INTER-STOREY HOLD-DOWN CONNECTIONS IN TIMBER-FRAME PANEL BUILDINGS

Institute for research and design in commerce & industry, Belgrade. All rights reserved.

349 - 357

358 - 367

Journal of Applied Engineering Science 16(2018)3

CONTENTS

368 - 373

374 - 382

383 - 390

391 - 397

Journal of Applied Engineering Science

Viktor Shebashev, Yuri Andrianov, Lyudmila Nizova, Anna Nikitina STUDENT SCIENCE AND TRAINING ENGINEERS

Nadezhda Zhdanova, Sergey Gavrytskov, Anna Ekaterynushkina, Julia Mishukovskaya, Julia Antonenko COMPREHENSIVE INTEGRATION AS AN EFFECTIVE WAY OF TRAINING FUTURE DESIGNERS AT TECHNICAL UNIVERSITIES (INTEGRATION AS A WAY OF TRAINING A DESIGNER)

Rudi Hariyanto, Sudjito Soeparman, Denny Widhiyanuriyawan, Mega Nur Sasongko CFD STUDY ON THE ABILITY OF A VENTILATED BLADE IN IMPROVING THE SAVONIUS ROTOR PERFORMANCE

Vladimir Lushpeev, Andrey Margarit OPTIMIZATION OF OIL FIELD DEVELOPMENT PROCESS BASED ON EXISTING FORECAST MODEL

398 - 403

Igor Petukhov, Lyudmila Steshina, Andrei Glazyrin APPLICATION OF VIRTUAL ENVIRONMENTS IN TRAINING OF ERGATIC SYSTEM OPERATORS

404 - 409

Alexander Ivanov, Alexei Gazin, Yulia Serkova A SIMPLE WAY TO ASSESS THE SPECTRAL LINES INFORMATIVITY OF A CHI-SQUARE MOLECULE IN ANALYZING SMALL SAMPLES OF BIOMETRIC DATA

410 - 415

Turkov Andrey, Abashina Natalia SAGS AND FREQUENCIES OF NATURAL OSCILLATIONS OF COMPOSITE TWO-LAYER ISOTROPIC PLATES IN CASE OF CHANGE OF THICKNESS OF ONE OF THE LAYERS

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

CONTENTS

Erwin Erwin, Adi Surjosatyo, Julianto Sulistyo Nugroho, Teuku Meurahindra Mahlia, Tresna Soemardi THE EFFECT OF HYBRID SAVONIUS AND DARRIEUS TURBINE ON THE CHANGE OF WAKE RECOVERY AND IMPROVEMENT OF WIND ENERGY HARVESTING

416 - 423

Robbi Rahim APPLIED POHLIG-HELLMAN ALGORITHM IN THREE-PASS PROTOCOL COMMUNICATION

424 - 429

Hossein Tahghighi, Mohammad Rashid Gholami NUMERICAL STUDY OF CONFINEMENT EFFECT OF FRP COATINGS ON BEHAVIOR OF RC FRAMES BY USING NONLINEAR ANALYSIS

430 - 440

Anatolij Semenovich Kalitvin, Vladimir Anatoljevich Kalitvin THE APPROXIMATE AND NUMERICAL SOLUTION OF ROMANOVSKIJ PARTIAL INTEGRAL EQUATIONS

441 - 446

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Journal of Applied Engineering Science

doi: 10.5937/jaes16-17019

Original Scientific Paper

Paper number: 16(2018)3, 533, 307-312

ASSESSMENT OF ANTHROPOGENIC FACTOR OF ACCIDENT RISK ON THE MAIN OIL PIPELINE PASCUALES – CUENCA IN ECUADOR Johnny Robinson Zambrano Carranza1, Stanislav Kovshov2*, Evgeniy Lyubin2 1 Escuela Politécnica Nacional, Ecuador 2 Saint Petersburg Mining University, Russia The pipeline infrastructure of Latin America countries, including the Republic of Ecuador, has a significant accident potential. Accidence rate is facilitated by the diversity of natural and climatic conditions in the region, military-political instability in its separate parts, and imperfection of the standards for the production of pipes and equipment.The technical condition of the pipeline systems operated in Latin America for 15 – 20 years is imperfect.

Depreciation or replacement of spent equipment and elements of the pipeline infrastructure is not being carried out at an adequate pace. Ecuador, for example, has a stable dynamics of increasing accidents at pipeline infrastructure facilities.

Therefore, an assessment of the risk of the condition of the main pipelines should be carried out. Anthropogenic risk is one of the most important accident factors (50 % of all risk on main oil pipelines in Ecuador). To evaluate it, the methodology of Federal Safety Manual is proposed. The anthropogenic risk is calculated for 3 main sections of Pascuales – Cuenca pipeline, which supports with oil the provinces of Cañar, Loja, Zamora Chinchipe and Morona Santiago. For the foothill area, the specific probability of accidents is the most, and for the coastal and flat areas less and equals to 0.15 and 0.14, respectively. The performed accident analysis will allow taking the necessary measuresto maintain the pipeline in working condition.

Key words: Pipeline, Risk assessment, Pascuales – Cuenca pipeline, Oils, Ecuador oil infrastructure, Anthropogenic risk factor.

INTRODUCTION Accidents on pipelines are accompanied by significant losses of transported raw materials and large-scale pollution of the nature. Fuel spillages cause soil degradation [Kovshov et al., 2015]. To restore the contaminated soils, additional measures are needed, e.g. biogenic [Kovshov, 2013; Kovshov&Kovshov, 2014]. Besides, highly aggressive mixtures are released during leakages into the environment causing significant damage [Pashkevich&Antciferova, 2013]. All this creates a very negative effect on the nature of Ecuador, significantly affecting both the coastal plains and areas adjacent to the Andes, as well as the Ecuadorian part of the Amazon rainforest [Bravo, 2007].

There are several ways to assess the anthropogenic risk on oil pipelines: phenomenological, deterministic, expert and probabilistic [Vtorushina et al., 2017]. It is promising to use neural networks allowing to reveal hidden rules in chaotic systems, which include technogenic risks that appear in the gas transmission systems of JSC Gazprom [Podavalov, 2008]. To assess the quantitative risk of oil pipeline failure, the combination of subjective (which is analytic hierarchy process) and objective (which is entropy method) weighting methods can be used [Guo et al., 2016]. It was proposed to combine entropy weight method with grey clustering method, which can divide risk indexes in several groups according to whitenization weight function [Hu et al., 2017]. It is noted the quantitative risk assessment technique is one of the most credible. Nevertheless, the adjustment of the method is

required, due to lack of the necessary data. Thereby, the pipeline comprehensive risk analysis algorithm was suggested [Gharabagh et al., 2009]. Such risk factors as third-party damage index, corrosion index, design index, and incorrect operations index are considered in detail [Muhlbauer, 2004]. It is reported, third-party activities affect significantly at accidence rate on oil and gas pipelines in China. Pipeline fragility – overall reliability method was used for quantitative risk assessment.

Proposed model helps to reduce costs and more accurately indicate the sections of the pipeline with high failure risk [Peng et al., 2016]. A fuzzy Bayesian belief network used for risk assessment in oil and gas pipelines showed that the most significant risk factors are construction defect, overload, mechanical damage, bad installation and quality of worker [Kabir et al., 2015]. Monte Carlo simulations for the probabilistic permanent fault displacement hazard were used to assess seismic risk on buried continuous pipelines [Cheng&Akkar, 2017]. Neural networks – fuzzy logic model was used for assessment the risk of biofouling on pipeline. Dependences of biofouling risk from roughness and temperature, as well as, from water content and velocity were received [Urquidi-Macdonald et al., 2014]. Modern systems for diagnostics of oil pipelines technical condition do not always satisfy the requirements of the oil industry. It seems promising, to consider the principle of the pipeline diagnostics system using a constant pulsating magnetic field, instead of imperfect wave intratubal methods [Proskuryakov&Dementev, 2016]. It is reported, third party activity is one of the factors with highest risk impact.

*Saint Petersburg Mining University, 21st Line, St Petersburg 199106, Russia, kovshovsv@mail.ru

307

Johnny Robinson Zambrano Carranza, et al. - Assessment of anthropogenic factor of accident risk on the main oil pipeline pascuales – Cuenca in Ecuador

Monte Carlo simulation and analytic network process model was used to test existing gas pipeline in Qatar and satisfactory results were received [El-Abbasy et al., 2015]. The factor of internal corrosion was assessed with Bayesian belief network approach. It was shown, internal corrosion depends on pipeline operating pressure and small diameter pipelines are the most vulnerable [Shabarchin&Tesfamariam, 2016].

A list of the main causes of accidents on main pipelines in Russia is shown in Figure 1.

anthropogenic impacts accounts for almost 50% of emergencies, the role of corrosion due to the lack of temperature and humidity jumps during the year is minimal. The factor of poor quality of pipes and conducting of construction works is about the same. Also in Ecuador, the role of the “human factor” in accidents is twice higher, which is explained by the lack of highly qualified engineering personnel serving engineering communications.

At the same time, the actual lack of own methods for assessing accident risk remains an important scientific and methodological problem of industrial safety in Ecuador.

Methods

Figure 1: Distribution of accidents causes on the main pipelines of Russia This assessment of the causes of accidents in the practice of industrial safety management systems in Russia formed the basis for the creation of various methods for assessing the risk of accidents on main pipelines.

For comparison, the data on spillage frequencies in Europe are given in Fig. 2.

The structure of the main causes of accidents in Ecuador differs quite significantly. The share of external

As a methodological basis for assessing the anthropogenic component of the accident risk on main pipelines in Ecuador, the use of Federal Safety Manual for assessing the risk of accidents at main oil and oil products pipelines is proposed.

Emergency situations on main oil pipelines can be characterized by a certain variety in the specific frequency of accidents both the all over pipeline and also in its separate sections . This division is based on differences in constructive and technological characteristics; design, construction and operation features; external natural conditions, economic and political environment. where kinf – coefficient of influence; ks – coefficient of strength, defined as a value, inverse to the ratio of the actual stock of strength of investigated section of the main pipeline, to the coefficient of strength of all pipeline; kl – coefficient considering the method of laying the pipeline. Coefficient considering the method of laying the pipeline. kl = 0.1 – on the sections made by microtun-

Figure 2: Cold pipelines spillage frequencies in Europe by cause (excluding theft) [Concawe report, 2017].

308

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Johnny Robinson Zambrano Carranza, et al. - Assessment of anthropogenic factor of accident risk on the main oil pipeline pascuales – Cuenca in Ecuador

nelling technology; kl = 0,4 – on the sections made by directional drilling; kl= 0,6 – on the sections executed by the technology “pipe-in-pipe” or with the use of the pipes, coated with concrete; and kl= 1 – on all other sections.

On each investigated section of the route, it is necessary to determine the value of the coefficient of influence kinf, showing how many times the specific frequency (probability) of accidents on the section differs from the average for this route. Coefficient of influence is determined using a standardized point evaluation system. For each value of the factor, with calculated or expert method, the number of points (on 10-point scale), expressing the degree of factor’s influence, is determined. Scorings of each investigated factor are analyzed through the formula:

where Bn is the point evaluation of the nth section of the main oil pipeline route, calculated as: I

J (i )

Bn = ∑∑ ρ i ⋅ qij ⋅ Bij i =1 j =1

Baver – average point evaluation of the main oil pipeline route, defined as:

where N is the total number of sections of the main pipeline considered.The value is determined on the basis of accident statistics in the organization operating the main oil pipeline. Also this parameter can be calculated by the formula:

where λaver – average accident rate at all main oil pipelines in the country for the last 5 years. kD – diametrical coefficient, which is determined from the accident statistics, depending on the nominal diameter DN of the pipeline. The values are taken in accordance with Table 1.

Table 1. Values of the coefficient kD for different nominal diameters of main oil pipelines DN, mm

1400

1200

1000

800

700

500

<500

kD

0.35

0.85

1.60

1.25

1.40

1.20

1.10

kreg– regional coefficient.

Results and Discussion The pipeline project Pascuales - Cuenca develops in an approximate length of 205 km in the regions of Guayas, Azuay and Cañar (Figure 3).

The main pipeline starts in the coastal zone near the city Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

of Guayaquil and ends with the Cuenca terminal, which stores the transported oil products, as well as liquefied petroleum gas. To maintain the required level of pressure in the pipeline, and to monitor its condition, three intermediate pumping stations are operated near the settlements of La Delicia, Ducur and Charcay.

Figure 3: Location of the Pascuales – Cuenca main pipeline The physico-geographical features of the zone of the main pipeline stretch and its relatively small area, make it possible to identify three sections, analyzed from the point of view of the accident risk: coastal, flat and foothill.

In accordance with the proposed methodology, the point evaluation of each of the allocated sections of the main oil pipeline route has the following sequence of actions.

Based on the accident statistics on the main oil pipelines, five groups of influence factors are identified, indicating the relative “contribution” ρi of each group (i from 1 to 5) in the total statistics of emergency failures using the weighting coefficient ρi (Table 2).

Table 2. Accident factors and their weighting coefficients Designation and name of a group of factors

Groupshare, ρi

Gr1

External anthropogenic impacts

0.60

Gr2

Corrosion factor

0.05

Gr3

Natural and climatic factor

0.05

Gr4

Constructive and technological factor

0.10

Gr5

Defects in pipe body and welds

0.20

Assessment of external anthropogenic impacts As can be seen from Table 2, of the proposed factors, the accident rate on main oil pipelines is determined on 60% by external anthropogenic influences.

This group includes the factors external to the investigated main pipeline, shown in Table 3, which affect the probability of damage from third party activity.

All over the length, the pipeline Pascuales–Cuenca has a certain specificity: in the coastal and flat areas, it

309

Johnny Robinson Zambrano Carranza, et al. - Assessment of anthropogenic factor of accident risk on the main oil pipeline pascuales – Cuenca in Ecuador

mainly passes under the ground, in the foothills – passes either over the surface or above the ground. Herewith, constantly occurring water obstacles are always overcome by overhead way.

Table 3: Factors of the group “External Anthropogenic Impacts” Designation and name of influence factor

Share in a group ρ1j

F11

Depth (in the ground) or height (above the ground) of pipeline laying

0.4

F12

Level of anthropogenic activity

0.2

F13

The danger of sabotage and tie-in with the aim of stealing oil and oil product

0.4

Then, averaged, the scoring value for the actual depth (height) of the laying is calculated by the following formulas: B11=1+25*(h-0.6)2= 1+25*(1-0.6)2=5 – for the foothill area (since h – the height of laying = 1m); B11=0.83*(1.8-h)= 0.83*(1.8-1.5)=0.25 – for flat and coastal areas (mean laying depth = 1.5m).

The level of anthropogenic activity is estimated based on the sum of the factors m, presented in Table 4.

All study areas are characterized by approximately the same population density, amounting to 55-60 people/ km2. Accordingly, the factor m = 3. The activity of construction works at the sites is different: for the coastal and foothill areas it can be estimated as high (=3), for the flat area - as moderate (= 2). All three districts are characterized by a small number of communications of other departmental affiliations (= 0.5) and the presence of road and rail sections (= 2).

Table 4: Level of anthropogenic activity m 1

2

3

4

310

Name of the component m

B12(m)

The average population density Hpop in a 3-km strip along the route

0 <Hpop< 50 people/ km2

0,06 *Hpop

Hpop> 50 people/km2

3

Frequency of construction and other infrastructure works carrying out

high

3

moderate

2

low

1

absence

0

large quantity

2

small quantity

0.5

Presence of other main pipelines and communications

Density of roads and railways

absence of extra-departmental communications

0

Roads presence

2

Roads absence

0

Thus, the coefficient B12= 3 + 3 + 0.5 + 2 = 8.5, for the coastal and foothill areas, and B12= 3 + 2 + 0.5 + 2 = 7.5, for the flat area.

B13 – point evaluation of the hazard factor of sabotage and tie-ins to the pipeline with the purpose of oil stealing for the conditions of Ecuador, can be determined using Table 5.

For all three sections, attempts to steal the oil were detected, but there were very few of them. Inspection of the pipeline route condition is carried out by air transport, there are no automated systems for detecting the tie-ins.

Thus, for all three sections, this criterion can be estimated as B13= 2.

Table 5: Danger of sabotage and tie-ins Name of the component In the investigated area, attempts to steal oil were not detected Attempts to steal oil were detected in the investigaed area. In order to prevent this, the main pipeline route is patrolled

B13(m) 0

Air inspection of the route is carried out from two to five times a week

2

Ground inspection of the route with a frequency of 2-3 times per week without air inspection of the route

5

Ground inspection of the route with a frequency of 1 time per week without air inspection of the route

8

The system of automated detection of tie-ins has been installed

0

In the investigated area, no measures of protection were taken in the presence of the fact of tie-ins or sabotages

10

Thus, it is possible to carry out a scoring of the factor “External anthropogenic impacts” for the three sections of the main oil pipeline: - for the coastal area: Bn=0.6*(0.4*0.25+0.2*8.5+0.4*2)=1.56; - for the flat area Bn=0.6*(0.4*0.25+0.2*7.5+0.4*2)=1.44; - for the foothill area Bn=0.6*(0.4*5+0.2*8.5+0.4*2)=2.7.

The average score for the whole route of the Pascuales – Cuenca main pipeline is Baver=(1.56+1.44+2.7)/3=1.9.

Hence, the coefficient of influence of external anthropogenic impacts: - for the coastal area kinf= 1.56 / 1.9 = 0.82; - for the flat area kinf = 1.44 / 1.9 = 0.76; - for the foothill area kinf = 2.7 / 1.9 = 1.42.

The average accident rate at the main oil pipelines in Ecuador over the past 5 years λaver= 0.13 accidents per Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Johnny Robinson Zambrano Carranza, et al. - Assessment of anthropogenic factor of accident risk on the main oil pipeline pascuales – Cuenca in Ecuador

year per 1000 km.

The nominal diameter DN of the Pascuales – Cuenca oil pipeline is 609 mm, therefore the diametric coefficient can be taken equal to 1.4.

The territory of Ecuador is rather small in area – 283,360km, so in calculations the regional coefficient can be neglected and assigned a value of 1.

Accordingly, the specific frequency (probability) of accidents over the entire pipeline =0.13*1.4*1=0.18.

Unfortunately, there is no reliable information on the parameters of the actual stock of strength of the main pipeline in open sources. In the absence of data, the strength factorks is taken to be 1.

The method of Pascuales-Cuenca oil pipeline laying cannot be unequivocally attributed to any of the variants proposed in the methodology, therefore, the coefficient considering the method of laying kl, can be averaged taken equal to 1.

Thus, the specific frequency (probability) of accidents at certain sections of the Pascuales-Cuenca pipeline: - for the coastal area λn=0.18*0.82=0,15; - for the flat area λn= 0.18*0.76=0.14; - for the foothill area λn=0.18*1.42=0.26.

In the practice of results representation of the specific accident rate evaluation in certain sections of main pipelines, it is customary to use bar graphs. Thus, the final form of accident analysis for the Pascuales – Cuenca pipeline sections can be presented in Figure 4.

Figure 4: Estimation of the accidents caused by external anthropogenic impact on separate sections of the Pascuales – Cuenca oil pipeline.

CONCLUSION Thus, according to the results of the study, the following conclusions can be drawn: 1. The share of the anthropogenic accident factor on the main pipelines in Ecuador accounts for more than 50% of the cases.

2. The physico-geographical features of the stretch area of the Pascuales – Cuenca main oil pipeline of Ecuador and its relatively short length allow to distinguish three Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

sections analyzed from the point of view of the accident risk: coastal, flat and foothill.

3. In accordance with Federal Safety Manual, the analysis and calculation of the main accident rates have been carried out, both along the entire Pascuales – Cuenca pipeline, and on its separate sections.

4. It has been established that the greatest accident potential caused by external anthropogenic impact is typical for the foothill area, which is primarily is connected with the overground pipeline laying factor in view of complexity of district areography.

ACKNOWLEDGMENTS The article was prepared with the grant support of the Academic Council of the Mining University on the basis of the results of the internship and scientific work at the National Polytechnic School of Ecuador.

REFERENCES Bravo, E. (2007). Los impactos de la explotación petrolera eneco sistemas tropicales y la biodiversidad. Acciónecológica, Vol. 24 (1), P. 35-42.

Cheng, Y., &Akkar, S. (2017). Probabilistic permanent fault displacement hazard via Monte Carlo simulation and its consideration for the probabilistic risk assessment of buried continuous steel pipelines. Earthquake Engineering & Structural Dynamics, 46(4), 605-620.

El-Abbasy, M. S., Senouci, A., Zayed, T., &Mosleh, F. (2015). A condition assessment model for oil and gas pipelines using integrated simulation and analytic network process. Structure and Infrastructure Engineering, 11(3), 263-281.

Gharabagh, M. J., Asilian, H., Mortasavi, S. B., Mogaddam, A. Z., Hajizadeh, E., &Khavanin, A. (2009). Comprehensive risk assessment and management of petrochemical feed and product transportation pipelines.

Journal of Loss Prevention in the Process Industries, 22(4), 533-539.

Guo Y., Meng X., Wang D., Meng T., Liu S., He R. (2016).

Comprehensive risk evaluation of long-distance oil and gas transportation pipelinesusing a fuzzy Petri net model. Journal of Natural Gas Science & Engineering, Vol. 33, P. 18 – 29, doi: 10.1016/j.jngse.2016.04.052.

Hu, Y., Liu, K., Xu, D., Zhai, Z., & Liu, H. (2017, August).

Risk Assessment of Long Distance Oil and Gas Pipeline Based on Grey Clustering. In Big Knowledge (ICBK), 2017 IEEE International Conference on (pp. 198-201).

IEEE. Kabir, G., Sadiq, R., &Tesfamariam, S. (2015). A fuzzy Bayesian belief network for safety assessment of oil and gas pipelines. Structure and Infrastructure Engineering, 12(8), 874-889.

Kovshov, S. (2013). Biological ground recultivation and increase of soil fertility. International Journal of Ecology & Development, 25(2), 105-113.

311

Johnny Robinson Zambrano Carranza, et al. - Assessment of anthropogenic factor of accident risk on the main oil pipeline pascuales – Cuenca in Ecuador

Kovshov, S. V., Garkushev, A. U., &Sazykin, A. M. (2015). Biogenic technology for recultivation of lands contaminated due to rocket propellant spillage. ActaAstronautica, 109, 203-207.

Kovshov, S. V., &Kovshov, V. P. (2014). Biogenic fixation of dusting surfaces. Life Science Journal, 11 (9), 401 – 404.

Muhlbauer, W. K. (2004). Pipeline risk management manual: ideas, techniques, and resources. Elsevier.

Pashkevich M.A., Antciferova T.A. (2013). Risk assessment of anthropogenic impact of the fueland energy complex. Journal of Mining Institute, Vol. 203, P. 225 – 228.

Peng, X. Y., Yao, D. C., Liang, G. C., Yu, J. S., & He, S. (2016). Overall reliability analysis on oil/gas pipeline under typical third-party actions based on fragility theory. Journal of Natural Gas Science and Engineering, 34, 993-1003.

Performance of European cross-country oil pipelines.

Statistical summary of reported spillages in 2015 and since 1971. Concawe, Brussels, 2017.

Podavalov,I.Yu. (2008). Analysis of methods of technogenic risk calculation in the operation of main gas pipelines. Journal of Mining Institute, Vol. 178, P. 82 – 85.

Proskuryakov, R.M., Dementev A.S (2016). The building a system of diagnosing the technical condition of the pipeline on the basis of continuous pulsed magnetic field. Journal of Mining Institute, Vol. 218, P. 215 – 219.

Shabarchin, O., &Tesfamariam, S. (2016). Internal corrosion hazard assessment of oil & gas pipelines using Bayesian belief network model. Journal of Loss Prevention in the Process Industries, 40, 479-495.

Urquidi-Macdonald, M., Tewari, A., & Ayala H, L. F. (2014). A neuro-fuzzy knowledge-based model for the risk assessment of microbiologically influenced corrosion in crude oil pipelines. Corrosion, 70(11), 1157-1166.

Vtorushina, A.N., Anishchenko, Y.V., Nikonova, E.D. (2017). Risk assessment of oil pipeline accidents in special climatic conditions. IOP Conf. Ser.: Earth Environ.

Sci. 66 012006.

Paper submitted: 31.03.2018.

Paper accepted: 30.07.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

312

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Journal of Applied Engineering Science

doi: 10.5937/jaes16-16908

Original Scientific Paper

Paper number: 16(2018)3, 534, 313-319

ASSESSMENT OF AVAILABILITY OF WOOD RESOURCES USING GEOGRAPHIC INFORMATION AND ANALYTICAL SYSTEMS (THE KRASNOYARSK TERRITORY AS A CASE STUDY) Aleksandr Petrovich Mokhirev*, Marija Olegovna Pozdnyakova, Sergej Olegovich Medvedev, Vladimir Olimbaevich Mammatov Lesosibirsk Branch of Reshetnev Siberian State University of Science and Technology, Lesosibirsk, Russia The article analyzes the availability of wood resources as a key indicator of the sustainability of logging operations.

This study leads to the conclusion that the process of assessing the availability of wood resources is associated with the collection and analysis of a large amount of data. At the same time, the more raw data is analyzed, the better the evaluation is. This article describes the experience of creating a software product based on geographic information and analytical systems designed for detailed assessment of the entire set of factors that determine the level of availability of forest resources. The results of the assessment of the availability of wood resources in the Krasnoyarsk Territory are presented. The results obtained will allow to increase the efficiency of the use of forest resources, to select the ideal technology and system of machines, to avoid environmental disturbance, to develop the optimal route for timber transportation which will eventually allow the forest user to make the largest profit from the timber products received.

Key words: Availability of wood resources, Profitability, Logging, Geographical information and analytical systems, Transport availability of wood resource INTRODUCTION Currently, the industrial development of forest resources in Russia faces significant difficulties. About half of Russia's forest resources are protected from logging on organizational and legal grounds, and the remainder is extremely heterogeneous and the economic conditions for its use are very different.

The division of forests according to the economic criterion is possible using the notion of "economic availability of wood resources".

This term is increasingly found in the research of modern scientists. A.G. Tretyakov [1] believes that economic availability cannot be considered separately from environmental availability R.V. Polshshvedkin and A.N. Mariev [2] propose to move away from costed assessment of wood and determine the economic availability of wood resources in the relative size (classes). At the same time among scientists there is no unequivocal definition of this notion.The availability of wood resources is a broad term that includes many indicators. However, in the final analysis, it characterizes the viability of the forest industry. Viable forest industry, in the first place, is characterized by high profitability, ensuring the competitiveness of timber products on the market.

Therefore, timely and accurate assessment of the availability of wood resources is a key moment in creating an efficient and stable timber industry complex.In order for the forestry industry to actively develop, it is necessary to use all available resources (not only raw materials, but also labor and economic resources).This requires to have full information about the territory used, such as:

forest resource base, human resources, the availability and development of infrastructure (including road network), relief features, climatic indicators, etc.

Thus, the possibility of obtaining an accurate assessment of the availability of forest resources is closely related to the large amount of data.This data should be systematized, updated and promptly processed.

THEORY AND EXPERIMENTS To date, the term "availability of wood resources" does not have a single definition. The authors distinguish several types of availability of wood.

1. The ecological availability of wood resources is expressed in limiting their use, as this can disrupt natural processes, undermine the ability of this resource to perform ecological functions, and reduce its ecological value. In some works [3, 4, 5] such forests are considered to be unavailable and forbidden to cutting by law. Mathematically, this can be expressed by formula 1.

(1) where Wea - environmentally available wood; Wfh - wood forbidden to cutting.

2. The technological availability of wood is an indicator that characterizes certain wood resources that can be used with a certain technology at a certain point of time. This type of availability is formed for each type of harvested wood products. For example, the technology for cutting sawlogs makes the production of wood chips unavailable.

* Lesosibirsk Branch of Reshetnev Siberian State University of Science and Technology, Office L-409, 410, 31, Krasnoyarsky Rabochy Av., Krasnoyarsk, Russia 660037, ale-mokhirev@yandex.ru

313

Aleksandr Petrovich Mokhirev, et al. - Assessment of availability of wood resources using geographic information and analytical system

3. The technical availability of wood is an indicator characterizing wood resources that can be used by a certain technique at a certain point of time. For certain natural and industrial conditions, a forestry machines with certain characteristics should be used.

Thus, the technical availability of wood will be positive under the following conditions (formula 2):

(2)

Where imax - maximum working gradient of the machine; igrad - maximum terrain gradient; δpressure- specific ground pressure; dp.t.- terrain load capacity; dmax- maximum diameter of trees processed by the working body; d - maximum diameter of trees; tmin , tmax - operating temperature range of the machine; Tmin , Tmax - ambient temperature range.

4. The transport availability of wood is an opportunity to transport wood to the delivery point by certain technical means at a certain time. It should be noted that wood resources can become transport available by changing the means of transporting timber or improving transport routes [6, 7]. This type of availability is closely related to technological and technical availability.

5. The economic availability of wood resources. This is a geographical location of the forest in relation to points of sale, which, with the current organizational and technical level of production, will provide the necessary level of profitability. Determination of economic availability of resources is achieved by measuring economic indicators such as the price of reproduction and its full cost price, including logging and transportation to the consumer. This type of availability is most often considered in various studies. At the same time, scientists offer different interpretations to the term "economic availability". To some extent, transport, technological and technical availability can be applied to economic availability, as today, with certain investments, it is possible to change the forestry equipment and technology of logging, create a transport infrastructure, etc. And this will change the transport, technological and technical availability of the resource. Mathematically, economic availability is expressed by formula 3: EA=I-Cl-Ct-Ci-R, Where EA - economic availability of wood resources;

314

(3)

I - income from the sale of logging products; Cl - logging costs; Ct - transportation costs of forest resources; Ci - costs for the creation and maintenance of the necessary transport infrastructure; R - rate of production profitability.

Based on the above definitions, it is possible to formulate a general definition for the term "availability of wood resources": Availability of wood resources is the opportunity of ecologically safe and economically efficient cutting of a certain type of wood products using certain forestry machinery and technology.

In order to obtain an objective assessment of the availability of forest resources of a country, region or individual forest plot, all significant factors should be taken into account. These are the factors, the change of which greatly influences on the level change of availability of wood [8]. These include: • Profit from logging production (from the use of forest resources). All objective costs that the producer bears in the process of logging operations will eventually be included in the cost of final products. At the same time, it does not matter what the final product is saw logs or timber.

• Transport availability of forest land. When transporting timber, several types of transport are used: water, rail and road. Each of them differs in cost and related factors of use.

• Cost of forestry work. Each tenant of a forest site is obliged to do quite costly reforestation activities in a certain amount at a certain time (depending on the regional policy and the type of forest plots).

Enterprises need access to structured information on the status and characteristics of forest areas (such as distance from industrial sites, possible ways of transporting wood to processing sites, tree species in the plot, the relief of the site, the transport routes on the site, the soil composition of the site, information on protective and nonoperational forests, etc.).

The most complete idea of the availability of forest on the area, its number, quality and location, can be obtained on the basis of the geographic approach. The most effective way to have all the necessary information and provide it in an available form to the enduser is the development of a specialized geographic information and analytical system (GIAS) for forest users [9].

The process of collecting and updating geographic information requires the use of material, technical and organizational resources.

However, using the readymade GIAS, the forest industry enterprise will be able to analyze the forest area of interest in a shorter time and with less cost, view the volume of the possible logging ("the allowable cut"), the places where timber is banned, the shortest and most economical transportation routes of wood to the places of its proJournal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Aleksandr Petrovich Mokhirev, et al. - Assessment of availability of wood resources using geographic information and analytical system

cessing, the seasonality of wood transporting from the logging area, to plan the development of logging areas, to determine the direction of cutting , define convenient location of loading and storage [10].

Thus, to date, there are some conditions for creating a geographic information resource necessary for preliminary assessment of the availability of forest resources, planning and modeling of industrial activities of forest industry enterprises.

The creating and updating of cartographic materials is an urgent issue and is one of the most important in ensuring effective management of forest resources. The specialized GIAS for forest users should contain the following elements: • information on wood growing stock located on the territory of the forestry area (divided into stocks of coniferous and deciduous species); • information on existing and projected transport routes suitable for wood transportation; • information on water bodies (DTM, maps); • information on forest use (tenants of forest sites); • information about the terrain (DTM, maps); • boundaries of the territory under analysis.

EXPERIMENTAL SECTION The climate of the Krasnoyarsk Territory is sharply continental: spring is long with sharp temperature fluctuations, summer is hot with sufficient moisture, autumn is short, winter is snowy and frosty.

The territory of the region is divided into two parts according to the type of relief. The western part is the eastern edge of the West Siberian Plain. It is a flathilly plain, where wide valleys of rivers alternate with watersheds.

Above floodplain terraces of rivers have depressions, filled with upland bogs. The eastern part of the region is located on the Central Siberian Plateau. The relief is characterized by the presence of hills and deep valleys with steep slopes.

In the south of the region there are island foreststeppes.

They are characterized by a flat and hilly terrain and fertile soils.

In the western part of the region dark coniferous forests predominate (spruce, fir, Siberian pine, Siberian larch).

Larchpine and pine forests grow in the eastern part.

There are 5 types of soils on the territory of the region.

They are divided in accordance with the load capacity.

Rocky ground have the highest load capacity. They predominate in the eastern part of the region, in the Central Siberian Plateau. Clay is characterized by the lowest load capacity. It is predominantly present in the river valleys of the West Siberian Plain. Sandy soils prevail in the Lower Angara region. Also, rocky soil and gravel can be found in the region in small quantities.

As a result of the research, valuable theoretical and experimental data were obtained. The practical aspect of Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

the study is presented below.

The development of a specialized GIAS for assessing the availability of wood resources, as well as planning and management of the logging process, involves the fulfillment of a number of tasks [8, 11]: Creating a database This is the most time consuming and informal part of the work. Its completeness and accuracy determine the quality of analysis and final products.

The first step in the creature of the GIAS project is the formation of database. Its basis is spatial data, along with their semantic environment. The GIAS provides a mechanism by which a digital model can be constructed from the spatial data of various sources [10].

When developing a digital database, the following steps are taken: • database design. This definition of the boundaries of the terrain, coordinate system, necessary layers of data, objects in each layer, attributes, methods of coding and organization of attributes; • input of the received data into the computer; • digitizing or converting data from other formats; • checking and correcting errors, creating topology; • input of attributive data and connection of attributes with spatial objects; • transformation of spatial data into real coordinates and joining of adjacent covers. For the development of the proposed GIAS, information is required on the exact boundaries of the forest areas, detailed information on transport routes, data on water bodies, data on tenants of forest areas, terrain data, information on wood growing stock, allowable cut volume and volumes of harvested wood, information on fire hazardous areas and water protection zones.

This information can be obtained from stateaffiliated organizations, research institutes, and industrial enterprises or on public Internet site. Information is processed on the ArcGIS platform.

Analysis of data As a result of the GIAS analysis of the territory, a series of thematic maps, graphs and tables are produced as the output data. This information is visualized and therefore is available for understanding. For this reason, special attention should be paid to the presentation of information. Maps can be presented as 2D or 3D and give information about some terrain and site characteristics or analysis of different indicators. Also, the maps can be static or animated. For GIAS analysis of the territory, geostatistical analysis methods based on interpolation, extrapolation and approximation of data and various cartographic image methods based on data classification are used. Based on various factors, it is possible to carry out a comprehensive assessment of the territory for its

315

Aleksandr Petrovich Mokhirev, et al. - Assessment of availability of wood resources using geographic information and analytical system

suitability for the assigned tasks, as well as to conduct its zoning, ranking and clustering. Evaluation of the dynamics of the development of a phenomenon or process, as well as its prediction, can be modeled on the basis of different time data.

Presentation of the analysis results The GIAS developed will allow the end user to specify the input parameters and obtain the required thematic maps by means of the Internet. In this case, the user does not need to have the skills to work in the ArcGISforDesktop software and understand the logic of the program, and highquality visualization of output data (maps) will allow using this product to a wide circle of users.

tle radar topographic mission) is used this is a radar topographic survey covering a large part of the Earth's territory. The SRTM data for the territory of interest can be downloaded from the official site. Similarly, other archives are identified and downloaded for the area studied. Further, from the downloaded rasters, a common mosaic of rasters is formed.

After processing the obtained SRTM data, a raster layer is obtained with visual data on the angles in degrees divided into three categories in accordance with the availability of different technical means (Fig. 2).

RESULTS The presented method is tested in the Krasnoyarsk Territory in the Russian Federation. Based on the data of the Forestry Regulations of the Krasnoyarsk Territory forestry areas and free data from the wms-server of Rosreestr, a polygonal cartographic layer containing information on the exact boundaries of the forest districts of the Krasnoyarsk Territory was developed. The attribute table of this layer contains information for each forestry area: "allowable cut", "actually harvested wood" and "undeveloped resources".

For realization of this project, roads were digitized (including logging roads, highways, branches and temporary roads, railways, platforms and dead ends). Digitization was carried out on the basis of satellite images and displayed on the computer desktop. This information allows you to determine the transport availability of forest areas (Fig. 1).

Figure 1: Image of transport availability of wood with overlap on a quarterly net Information on topography is necessary for planning the development of the logging site. For the development of this layer, free data from the SRTM mission (Shut-

316

Figure 2: Distribution of the forest area according to the slope of the terrain The map shows which slope prevails in a certain area.

By imposing information about the upcoming logging (the borders of the cutting area, the planned transport routes on it, technical and technological availability is assessed by a slope.

It is convenient to use maps with relief shading (shadow modeling) to assess the technological availability, planning of development of cutting areas, determining the direction of cutting, skidding and removal, and also to determine the most convenient location of loading and landing points (Fig. 3).

Figure 3: Terrain map with relief shading As a result of the processing of water bodies, a layer of polygonal Water Protection Zone objects is obtained.

Using the information of this layer, it is possible to assess the ecological availability of wood resources (it is posJournal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Aleksandr Petrovich Mokhirev, et al. - Assessment of availability of wood resources using geographic information and analytical system

sible to identify illegal logging and identify those areas where cutting of wood is forbidden).

Then, the areas leased by enterprises are digitized. The attribute table contains information about the tenant, the lease contract number, and the date of the lease termination. So, you can get information about unoccupied plots.

On the basis of free data on soil resources, a soil map of forest areas was created [10]. According to the information received, it is possible to determine the technical and transport availability of wood. On the basis of the received map it is possible to plan transport ways, seasonality of transport of wood, types of forest machines.

This information can be provided to the user in a digital form, in the form of highresolution maps (more than 40,000 to 20,000 pixels) or in the form of thematic atlases that are easy to view and do not require additional software.

The layers of the GIAS take into account the influence of all factors that determine the level of availability of wood resources. To determine the influence of factors, the method of peer review was used. In this case, several layers are intended to reflect one factor.

Using this information, a forest user can evaluate various types of availability of forest resources on a leased forest site in a timely manner, as well as solve the problems of selecting forest plots for future lease, delivery of raw materials and availability of information on forest resources.

The completeness and detail of the assessment will allow to increase the efficiency of wood resources use, to select the optimal technology and machine system, to avoid environmental disturbance, to develop an optimal route for wood transportation. This will allow the forest user to make the maximum profit from the received timber products.

The obtained results of the assessment of the availability of wood resources, using the developed program in the Krasnoyarsk Territory are presented in Table 1 and in Figure 4. Data on the area of forests and timber reserves are obtained from the State Forest Register dated 01.07.2014 (in respect of forests located within the boundaries of the Krasnoyarsk Territory).

As the initial data necessary for the functioning of the developed GIAS, the most typical conditions for enterprises of the Krasnoyarsk Territory are accepted: â&#x20AC;˘ manufactured products - saw logs; â&#x20AC;˘ logging technology: Scandinavian (harvester + forwarder) and Canadian (cutting and milling machine + skidder + delimbers); â&#x20AC;˘ transporting products: motor roads (KamAZ, MAN and Iveco trucks), waterways (if available). The territory within a radius of 10 kilometers from the existing logging roads is considered to be transport available (temporary forest roads can be laid).

Table 1: Results of assessing the availability of wood resources in the Krasnoyarsk Territory according to the developed geographic information and analytical system Area of forest fund lands cevered with forest vegetation (thousand hectares)

Wood growing stock (million m3)

105,054.1

11,510.9

Absolute value

61,958.2

7,709.7

%

59.0

67.0

Absolute value

78,348.9

9,154.9

%

74.6

79.5

Absolute value

61,156.9

7,061.6

%

58.2

61.3

Absolute value

40,253.0

5,367.0

%

38.3

46.6

Absolute value

35,943.0

4,755.5

%

34.2

41.3

Absolute value

26,021.2

3804.7

%

24.8

33.1

Index Overall value of the indicator in the region, including: Ecological availability

Technological availability

Technical availability

Transport availability

Economic availability Total available wood resources

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

317

Aleksandr Petrovich Mokhirev, et al. - Assessment of availability of wood resources using geographic information and analytical system

Figure 4: Diagram of availability of forest resources of the Krasnoyarsk Territory according to the developed geographic information and analytical system The results obtained indicate that only 24.8% of the forest area and 33.1% of the timber stock in the Krasnoyarsk Territory can be considered available for effective development. At the same time, the lowest value is observed in the category of "economic availability" - 34.2% and 41.3%, respectively. This is due to the fact that the definition of economic availability, as already noted, focuses on all other types of availability. Consequently, only ecologically, technologically, technically and transport available wood resources can be economically available.

Based on the obtained assessment of the availability of forest resources in the Krasnoyarsk Territory, it can be concluded that it is generally low. To increase the efficiency of forest use in the region, it is necessary to search causes and factors that reduce the availability of forest resources, as well as the introduction of new methods and technologies in industrial processes.

Methods and technologies for increasing availability of wood resources should be discussed by all interested parties or experts of a specific forest cutting enterprise.

Only with an integrated approach to this problem a really effective solution can be developed.

ACKNOWLEDGEMENT The reported study was funded by RFBR according to the research project № 18-310-00311

318

REFERENCES 1. Tret'jakov A.G. 2015. Forest Rent and Economic Accessibility of Forest Resources: methodological aspects: Journal of the Moscow State Forest University - Forest Bulletin, № 2, p. 153.

2. Polshvedkin R.V., Mariev A.N. 2006. Assessment of the Availability and Quality of Forest Resources in the Udorsk Region of the Komi Republic. Methodological approach: Use of Geoinformation Systems in Environmental Management and Environmental Protection in the Republic of Komi: Materials of the scientific-practical conference].

3. Velazquez-Marti B., Annevelink E. 2009. GIS Application to Define Biomass Collection Points as Sources for Linear Programming of Delivery Networks: Transactions of the ASABE, № 52 (4), pp. 10691078.

4. Sokolov V.A. 2013. Organization of sustainable use of Siberian forests: Forest management information,. no 2. pp. 52-59.

5. Sokolov A. P., Syunev V. S., 2014. A toolset of decision support systems for wood harvesting and forest bioenergy logistics in Russia: Renewable energy sources and clean technologies: Proceedings of 14th international multidisciplinary scientific geoconference SGEM 2014. - Albena: STEF92 Technology.

Vol.1. - pp. 3-10.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Aleksandr Petrovich Mokhirev, et al. - Assessment of availability of wood resources using geographic information and analytical system

6. Hashemkhani Z., S., Rezaeiniya, N., Zavadskas, K., E., Turskis, Z. 2011. Forest roads locating based on AHP and Coprasg methods: an empirical study based on Iran. E + M EKONOMIE A MANAGEMENT4: 6-21 7. Heinimann, H., R. 2017. Forest Road Network and Transportation Engineering-State and Perspectives.

Croatian Journal of Forest Engineering 38(2) 155173.

8. Mokhirev A.P., Pozdnjakova M.O., Aksenov N.V. 2017. Comparative Analysis of the Availability of Forest Resources of Logging Enterprises: Engineering Bulletin of the Don, Vol. 44, № 1 (44). URL: http:// www.ivdon.ru/ru/magazine/archive/n1y2017/3954 9. Mokhirev A. P., Egarmin P. A. 2011. Geographic Information System for Planning Optimal Development of Forest Fund: Sistems. Methods. Technologies, № 4 (12), pp. 172-176. URL: http://www.brstu.ru/static/ unit/journal_smt/docs/number12/172-176.pdf

10. Mokhirev A.P., Goryaeva E.V., Egarmin P.A. 2017. Creation of a Geoinformation Resource for Logging Production Planning: Bulle-tin of SGUGiT (Siberian State University of Geosystems and Technologies), Vol. 22, № 2, pp. 137-153.

URL: http://www.vestnik.ssga.ru/wp-con-tent/uploads/2017/07/%D0%A2%D0%BE%D0%BC-22%E2%84%96-2.pdf 11. Hosseini, S., A., Mazrae, M., R., Lotfalian, M., Parsakhoo, A. 2012. Designing an Optimal Forest Road Network by Consideration of Environmental Impacts in GIS. Journal of Environ-mental Engineering and Landscape Manage-ment, 20 (1): 58-66.

Paper submitted: 20.03.2018.

Paper accepted: 28.05.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

319

Original Scientific Paper

doi:10.5937/jaes16-16447

Journal of Applied Engineering Science

Paper number: 16(2018)3, 535, 320 - 327

SIMULATION AND MODELING IN CRISIS MANAGEMENT Martin Ficek* Tomas Bata University in Zlín, Czech Republic This article deals with the possibilities of modeling and simulation in terms of crisis management. The first part deals with simulations and modeling in the form of realized trainings of Integrated Rescue System (IRS) and crisis management (CM). It is focused on the most important aspects of modeling and simulation in CM conditions. In the following part, we are evaluating the selected SW modeling and simulation from the point of view of their possibilities and suitability for relation to CM. Based on expert consultation; tasks are defined that the SW should manage. In the last part, using the knowledge from the previous parts, the possible development in the area of SW modeling and simulation in CM is identified.

Key words: Simulation model, Models, Simulations INTRODUCTION Modern time influences a wide area of disciplines and crisis management is not the exception. One of the areas where this trend can be seen as the issue is training and preparation for extraordinary events (EE) and crisis situations (CS) - primarily the Fire Rescue Service of the Czech Republic (FRS) and the Integrated Rescue System (IRS). This is evidenced by the number of planned IRS trainings, when 65 were planned for the first half of 2017. Therefore, this article deals with simulations and modeling of leakage of dangerous substances.This actual problem is proven by the fact that 5 trainings are primarily focused on this issue. With respect to the amount of trainings (floods, fires, active shooter, and highly contagious diseases, etc.) is not an insignificant number, even though, the leakage of dangerous substances (DS) is included in other trainings as secondary consequences.

For example in these trainings: • February 7, 2017 –7 units of FRS and fire protection units (FPO) participated in the training of EE, namely DS ammunition from the new training hall, which was held in Třinec. [1] • June 29, 2017 – 6 units of FRS and FPO was training the EE with the leakage of dangerous substance (ammonia). This training was held in the premises of the MP Krásno in Ostrava-Martinov. [2] • July 31, 2017 – 7 units of FRS and FPO trained EE with DS (ammonia) which was leaking from multifunctional hall Polárka in Frýdek-Místek. [3] • March 3, 2017 – the MU with DS leakage of ammonia was trained in Bučovice. [4] Other examples of trainings will be given in the following chapter.

The issue of simulation and modeling of leakage of dangerous substances is dealt in the following publications.

Björnham Oscar et al.[5] discuss the risks of chemical leakage during transportation in a densely populated

area and its impacts. By analysis of fluid dynamics deals Sun et al. [6]. A direct comparison of two selected SWs is given in Ficek et al. [7]. From the point of view of simulation and modeling, Inanloo et al. [8] achieved very interesting results in modeling the combustion vapor explosion.Bernatik et al. [9] dealt with the much needed addition of models of computational fluid dynamics for common SW such as Terex or Aloha or physical modeling in aerodynamic tunnels which provides data that can be compared with the outputs of conventional SW. This issue is addressed also in scientific circles as can be seen in previous paragraph. Dangerous substances are used in products, processes, and procedures in many facilities and factories. Therefore, these substances must be stored and transported. Usually leaks occur at these stages. These leaks must be solved by forces and means of IRS.In order to better preparation for EEs with leakage of dangerous substances of the FRS and CM components, trainings, simulations and modeling are carried out. Therefore, the simulation and modeling possibilities, especially the possible developments in SW simulation and modeling for IRS, need to be addressed.

In this paper, selected IZS trainings focusing on leakage of dangerous substances will be analyzed by SWOT analysis. This paper will also review current SW tools for simulation and modeling of leakage of dangerous substances. Based on expert consultations with experts, a possible development will be proposed for SW simulation and modeling in crisis management aimed at leakage of dangerous substance.

Evaluation of realized trainings: February 7, 2017 - Training of the IRS components in the training hockey hall in Třinec, where 7 units of FRS and FPO practiced intervention in case of leakage of ammonia from the cooling device. DS leaked from a pipe in the cooling device in the engine room. This leak was detected by a sensor at a concentration of 300 ppm (this value exceeds the limit for ERPGs 2). In terms of simulating DS spreading, it is essential that the leakage was within

* Tomas Bata University in Zlín, nám. T. G. Masaryka 5555, 760 01 Zlín, Czech Republic, ficek@utb.cz

320

Martin Ficek - Simulation and modeling in crisis management

in the building. The exact amount of leakage and the size of the hole are unknown, but the precise substance is known. For the training, the approximate infiltration sector was predetermined. When we tried to simulate the propagation of the substance using two different SW, different results were achieved. The procedure of the units was highly professional; however, in terms of the simulated leakage of dangerous substance, the distribution of the substance was not fully taken into account. To improve this, we recommend to simulate the propagation of the substance by several SW and based on the result determine the propagation.

6 units of FRS and FPO participated in tactical training on the premises of the company MP Krásno in Ostrava-Martinov aimed at eliminating of the consequences of ammonia leakage from the cooling pipeline in the engine room took place on June 29, 2017. This involved two leaks of liquid ammonia in the building. Even in this case, FRS and FPO units are well prepared and trained and the intervention itself was professional. In terms of simulation, this situation is rather problematic. If a substance is spread in an object, it would be difficult to model two leaks that are intermingled for some time. Determining, for example, the resulting concentration would be problematic. During this training, the technical means, that prevented the propagation of the substance from the leakage site (engine room, compressor), should be used in the simulations to take into account the DS diffusion and interference. It would be useful to use SW with the possibilities of complex simulations that provide a relatively good result of the threat.

In the hockey hall Polárka in Frýdek-Místek took place an training focused on leakage of ammonia. It took place on January 31, 2017 and 7 units of FRS and FPO participated. In this tactical training, ammonia leaked from the cooling device (cracked valve) in the engine room.

Leakage was detected by a sensor at a concentration of 50 ppm. The intervention was highly professional and efficient. In terms of simulation, it is interesting that the deflectors have been set up outside, so they did not disguise the propagation of substance within the object.It may be advisable to use SW to simulate the spread of DS in the object and thus to work better with this data and to check the procedure to prevent DS from spreading within the building.

At the slaughterhouse in Bučovice, on 3 March 2017, training was conducted in which ammonia leaked after a technical defect on the cooling equipment. There were 5 units of FRS and FPO. The intervention itself was professional and effective. In terms of simulation, the influence of the propagation of DS in the object itself and the subsequent propagation outside the object was simulated. The simulation and modeling options showed how effective can be data obtained during training and how to adjust the upcoming training if we really expect the worst-case scenario.

The 4 trainings mentioned above, which took place in Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

the first half of 2017, were focused on the leakage of dangerous substances (ammonia). Although ammonia is most likely to leak in the Czech Republic, it would be appropriate to train the leakage of other substances in the future; however, the unit progress would be essentially the same or the same. In terms of modeling and simulation, it was found that the IRS does not use SW.The last example demonstrates that the usage of SW to model the propagation of DS in a building can contribute to the improvement of the training. This statement are based on simulations and analyzes (SWOT), but since it is only a training where there are no real leakage data, the author does not consider the analysis to be as significant to be stated in the article.

CONSULTATION WITH EXPERTS AND DEFINITION OF THE TASKS OF THE IRS The following information was obtained through personal meetings with IRS experts and Crisis Management staff.

Interestingly, the responses were largely coincided, that is why there are only few suggestions and answers.

Experts have defined very simple and generally on: • Transportation of resources.

• Assessing the situation at the site of intervention.

• Create an intervention base • Preparation of resources.

• Intervention.

• Finishing of intervention and return.

In the case of a leakage of a dangerous substance, the tasks of the units are to reduce the immediate risks and limit the extent of the accident in order to stabilize the situation: • exploration (if a dangerous substance accident is involved), • measures to save people and animals and closure of a intervention site, • calling assistance, including units specializing in an accident with a dangerous substance, • ensure a sufficient distance from the intervention site to the wind direction, • close the intervention site, identify the dangerous and outer zones, • exclude sources of initiation, • prepare resources for intervention, • if it is possible, prevent further leakage or spreading of a dangerous substance, • identification of a dangerous substance, provide information about its dangers, • if possible, take measures to capture or remove dangerous substances, • continuously evaluate the situation, • determine the possible amount of leaked dangerous substance,

321

Martin Ficek - Simulation and modeling in crisis management

• determine the size of the affected area, • identify the status and possibilities of their change, • identify the possibility of spreading hazardous substances, wind and weather, • determine the terrain and settlement density, • identify potential threats to surface or groundwater, • determine the rate of leakage of the dangerous substance and the rate of its spreading, • inform residents, relevant institutions and public authorities, • assess the need for population evacuation or other protection, • make a prognosis of the further development of the accident with a view to further escalation.

When dividing the intervention site into zones with a distinctive danger, it is at least the creation of: • dangerous zones, • outer zones and within them • the rear area, • the entrance space, • decontamination area.

As can be seen from the tasks designed for the intervention with the leakage of DS, it is obvious that SW simulations and modeling for DS leakage have some influence. In the previous task list are underlined those points where it is possible to use this software. Bold are then actions that are necessary to get the basic information needed for the proper function of each simulation and modeling software.

Further, consultations with experts showed that it is necessary to partially improve the user interface for the existing SW, but especially the accuracy of the calculation of the vulnerable area. Furthermore, the proposal for a wider database of substances and a more detailed description of these substances (mainly the possibility of chemical reaction of these substances with other substances). Generally, they would also glad if the SW were created for their needs. Expert requirements will be discussed in more detail in the following two chapters, namely: Evaluation of current SW and Identification of the possible development of SW. Evaluation of current SW The software must be assessed on the basis of their designation, i.e. in relation to the previous chapter, namely: identification of dangerous and outer zones, information on its danger of dangerous substances, continuous assessment of the situation, determination of the size of the affected area, assessment of the necessity of population evacuation or other protection,the prognosis of the further development of the accident with respect to the possibility of further gradation. In addition, it is necessary to take into account the determination of the program whether it is intended for intervention units or for crisis

322

and emergency planning.

These will be the parameters for SW evaluation For SW needs: Determination of the approximate amount of leaked dangerous substance, the state of DS, the possibility of spreading dangerous substance - wind direction and weather development, the terrain and population density, the rate of leaked substance and the rate of its spreading.

Because the individual software often operates on similar principles, there will be only given a summary description of only a few selected SW, two of them will then be selected, compared and evaluated.

Specifically, the software: • Aloha • Terex • Effect • Save II • Rozex alarm SAVE II – it is a numerical program focusing on the leakage of substances and their dispersion. It works on the principles of the Gaussian model and methodologies CPR 14 and CPR 18. The program allows modeling and simulating the following: continuous leakage followed by dispersion of toxic substance, leakage followed by dispersion of toxic substance, continuous leakage of the substance followed by explosion, substance leakage followed by explosion, and turbulent leakage.

EFFECTS – the program provides an estimate of possible physical effects of leakage of toxic and flammable substances. It uses Pasquill typing and wind speed; models operate with open smooth terrain, so it is necessary to specify the coefficient of surface roughness.

The Gaussian model is used to propagate neutral gas.

It is also possible to perform heavy gas modeling in the form of: instantaneous leakage, leakage of gas or dispersion by evaporation from the pool, dispersion resulting from turbulent/nozzle leakage. Gas dispersion in the atmosphere: Neutral gas dispersion model, heavy gas dispersion model, turbulent gas leakage model. Passive dispersion: Immediate leakage (up to 1 minute), semicontinuous leakage (1 to 10 minutes), continuous leakage (longer than 10 minutes).

ROZEX Alarm – the program provides leakage modeling of hazardous chemicals. The program database includes 10,000 DS with additional information. It works with up to 19 variants of scenarios. Again, the principle of maximum possible damage is applied. The program is designed to require minimum input data. For modeling atmospheric dispersion, the program offers the following basic options: • leakage of toxic substances - for neutral gas and heavy gas, • continuous leakage of toxic substances - for neutral gas and heavy gas, Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Martin Ficek - Simulation and modeling in crisis management

• continuous leakage of toxic substances through the opening - for neutral gas and heavy gas.

ALOHA – Areal Locations of Hazardous Atmospheres operates with statistical mathematical models (SLAB and DEGEADIS). The software shows the maximum concentration of NL and its range. It provides three basic options: cloud leakage, instantaneous leakage and “accidental” leakage.

Advantages: large database of substances, it is possible to connect SW with measuring devices.

Disadvantage: works only with nonreactive substances.

It works with the following input data: • Leakage data • Data on the state of the atmosphere • Data leakage source Output data: • The maximum leak rate - for liquids, it is the rate of evaporation, not the leakage rate.

• The average leak rate for at least 1 minute; for liquids, it is the rate of evaporation. For pressurized devices, a sudden initial leakage may result in overestimation.

• Total leakage of substance over a maximum of 1 hour.

• The maximum range of the hazardous zone in which the concentration reaches the specified value.

• Maximum concentration of leaked substance at any given location.

• The maximum dose at the selected site, which was absorbed by the body within 1 hour.

TEREX – works with chemicals and explosive systems.

The program is based on the conservative prognosis principle. The program needs relatively few input data: the amount of leakage, the wind speed, the air temperature, the type of environment, the cloudiness, and the

time of occurrence. For each substance, additional information can be found. This software works with the following models: TOXI - evaluates the range and shape of the cloud, PLUME - this is a long-lasting leakage of gas into the cloud, a long-lasting leakage of boiling liquid with rapid evaporation into the cloud, slow evaporation of the liquid from the puddle to the cloud. PUFF - gas leakage into the cloud, rapid flow of boiling liquid into the cloud.

TEROR - use of explosives. VCE, UVCE – leakage by explosion of a dangerous substance, such as: evaluation of the impact waves caused by the detonation of the mixture of substance with air. POOL FIRE, JET FIRE, FLASH FIRE and BLEVE - escape of combustible dangerous substance, these models evaluate the range of thermal radiation of fires.

• Advantage: Simple setup.

• Disadvantage: insufficient amount of chemicals in database.

EVALUATION USING SWOT ANALYSIS In order to obtain the data for the assessment, we compare actual situations and their threat zones (ERPG-3) with the results of simulations and models in SW Terex and Aloha. Although systems operate with different mathematical models, they can be compared in this way, because the systems use input data as variables, and mathematical equations themselves are predefined.

The student’s t-test is used for statistical evaluation.

Then, the statistics are applied to data from the table (real and ALOHA or TEREX), from which the hypothesis decision H0 is determined. H0 is the hypothesis that the mean values are equal.

According to the procedure, we find that the arithmetic mean of the difference between the measured reach of the substance and the simulated range in SW Aloha is -2.7989, the number of measurements is 10 and the variance s2 is 2.397465071.

Situation

Substance

Amount [kg]

Real [km]

ALOHA [km]

TEREX [km]

1

chlorine

188

0.431

7.2

0.462

2

ammonia

90

0.25

2.85

0.15

3

hydrogen chloride

55

0.56

0.49

0.57

4

methyl chloride

124

0.16

0.78

0.166

5

acetaldehyde

79

0.18

1.3

0.201

6

benzene

134

0.27

0.86

0.291

7

hydrogen fluoride

70

0.76

3.81

0.7

8

ammonia

50

0.17

2.8

0.194

9

formaldehyde

66

0.82

5.1

0.836

10

chlorine

89

1.3

7.7

1.319

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

323

Martin Ficek - Simulation and modeling in crisis management

Once these values have been matched to the equation

The result is t = 5.71625.

Tabulated critical value t1- α / 2(v), where v = n - 1 and α is chosen to 0.05, is 0.05, the conclusion therefore is that: t > t1 - α / 2(v). So there is a statistically significant difference.

We reject the zero hypotheses H0, i.e. the mean value of the measured values differs from the mean value of the simulated values using SW Aloha.

Applying the same procedure for SW Terex results in: Calculated value t = 0.021353 The tabulated critical value t1 - α / 2(v), where v = n - 1 and α is 0.05, is equal to 3.25.

Thus: t < t1 - α / 2(v).

So there is a statistically insignificant difference. We accept the zero hypotheses H0, i.e. the mean value of the real values is not significantly different from the mean value of the simulated values using SW Terex.

Table 1: SWOT analysis Aloha SWOT - Aloha Auxiliary

Noxious

Inner origin

Strengths: The system provides significant simulation capabilities.

Possibility to extend DS through an external database.

Possibility of connection with external measuring stations.

Weaknesses: The results can be distorted.

External origin

Opportunities: Work with 3D terrain models.

Terrain optimization.

Improvement of numerical models.

Threats: As a result of oversize results, results will not be taken into account.

324

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Martin Ficek - Simulation and modeling in crisis management

The quantification of this SWOT analysis is as follows: Strengths The system provides significant simulation capabilities Possibility to extend DS through an external database Possibility of connection with external measuring stations

Rating

Weight

Result

5

0.4

2

4

0.2

0.8

5

0.4

2

Total Weaknesses The results can be distorted

4.8 Rating

Weight

Result

-4.5

1

-4.5

Total Opportunities Work with 3D terrain models Terrain optimization Improvement of numerical models

-4.5 Rating

Weight

Result

4.5

0.3

1.35

4

0.2

0.8

5

0.5

2.5

Total Threats As a result of oversize results, results will not be taken into account Total Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

4.15 Rating

Weight

Result

-4

1

-4 -4

325

Martin Ficek - Simulation and modeling in crisis management

Internal Value = Strengths + Weaknesses = 4.8 - 4.5 = 0.3 External Value = Opportunities + Threats = 4.15 - 4 = 0.15 Result value = Internal value + External value = 0.3 + 0.15 = 0.45 SW Aloha has a relatively low result, due to overrated results, which are not suitable for use in CR. It is true that the principle of protection of the population is: “expect the worst, at least you will be prepared”, but SW Aloha is overrating the results so much that any intervention, according to the results, would be much more expensive and would lack in efficiency. However, SW Aloha is well-suited for crisis planning as it allows complex simulations to provide important information. Therefore, it should not be neglected.

The following section will focus on possible developments in the field of SW simulation and modeling of leakage of dangerous substances.

IDENTIFICATION OF POSSIBLE SW DEVELOPMENT Generally, it is necessary to deal with the possible developments in SW simulation and modeling. This is about refining models for calculating threat zones.This trend is valid for all SWs. Other trends are for certain groups of these SW. These trends include: involvement of 3D terrain models and air layers, simplifying the interface for enrolling and creating simulations, optimizing SW for field use, and for another group of optimizations, in order to create more complex simulations, and it would be appropriate for all SWs to be connected with other devices such as measuring and monitoring stations.

Expert consultations have led to the desire for SW to work primarily on devices (workstations, smartphones. etc), but would also communicate on a client-server basis, ensuring constant up-to-date information thusensuring functionality during server malfunction.Also, the ability to share results with other users could contribute to more effective interventions.

It is understandable that each SW will have specific properties and therefore specific possibilities for further development, but the goal is to identify possible developments in simulation and modeling software for leakage of dangerous substance.

CONCLUSION Leakage of dangerous substances is an extraordinary situation that is threatening almost every day. This is mainly due to the dependence of today's society on these substances. It is always necessary to transport and use DS in a variety of facilities and technological processes. Therefore, there is a need to plan for the situations of DS leakage. To plan for such situations, you need to know the threat zones. This is useful for SW, that models and simulates the leakage of DS and calculate the threat zone. From time to time there is a leak of DS, in this case the use of simulation and modeling software

326

to determine the threat zone.

Two SW were selected in the article, namely SW Aloha and SW Terex. Real situations were entered into these SWs and the results were compared with real measured data. The evaluation was carried out on two levels, in the form of a statistical part. In this section it was found that SW Aloha has statistically significant differences in the threat zones compared to the real situation. SW Terex has statistically insignificant differences in the same comparison.

The second comparison was done through SWOT analysis. It is focused on the entire SW. For each SW, the results differed, and generally the selected SW can be evaluated as relatively effective, but each has its primary purpose. SW Aloha is better to use in crisis planning, where there is plenty of time and allows more complex simulations. While SW Terex is more likely to be used during interventions, due to its relative accuracy and simplicity.

In the near future, possible developments should be in 3D modeling of terrain and air layers. Furthermore, creating an interface,that would simplify the creation of simulation during intervention, and at the same time enable comprehensive and complex simulation to be used for crisis planning.And finally, the possibility of connecting SWs and sensors or actors will be developed. The SW function is based on the hybrid principle client-server and SW on workstations.

It may seem that this area has already been explored, but the refinement of the models and the results of the simulations combined with the connection with sensors and actor models can make a significant contribution to a safer future.

ACKNOWLEDGEMENT This work was supported by the Internal Grant Agency of Tomas Bata University under the project No. IGA/ FAI/2018/014 REFERENCES 1. Hzs moravskoslezského kraje.První cvičení složek IZS v tréninkové hokejové hale v Třinci, fromhttp:// www.hzscr.cz/clanek/prvni-cviceni-slozek-izs-vtreninkove-hokejove-hale-v-trinci.aspx, accessed on2017-09-25.

2. Hzsmoravskoslezskeho kraje.Cvičení s únikem amoniaku v areálu v Ostravě-Martinově, fromhttp://www.hzscr.cz/clanek/cviceni-s-unikem-amoniaku-v-arealu-v-ostrave-martinove.aspx,accessed on2017-09-25.

3. Hzsmoravskoslezskeho kraje.V hokejové hale Polárka ve Frýdku-Místku unikal amoniak a zranil obsluhu, fromhttp://www.hzscr.cz/clanek/v-hokejove-hale-polarka-ve-frydku-mistku-unikal-amoniaka-zranil-obsluhu.aspx, accessed on2017-09-25.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Martin Ficek - Simulation and modeling in crisis management

4. Požáry.cz.Hasiči nacvičovali zásah při úniku čpavku v Bučovicích na Vyškovsku, fromhttps://www.pozary. cz/clanek/158139-hasici-nacvicovali-zasah-pri-uniku-cpavku-v-bucovicich-na-vyskovsku,accessed on2017-09-25.

5. Björnham O., Burman J., Parmhed O., Fureby C. 2014Crisis Management Modeling and SimulationLaboratory. In: Steyn D., Builtjes P., Timmermans R. (eds) Air Pollution Modeling and itsApplication XXII. NATO Science forPeace and SecuritySeries C: EnvironmentalSecurity. Springer, Dordrecht 6. SUN, H., NING, P., TANG, L. 2010. ApplicationofComputational Fluid Dynamics forManufacturedGasDispersion Modeling and LeakageConsequenceAnalysis. In: 7th International Symposium on

Safety Science and Technology (ISSST). Hangzhou, PEOPLES R CHINA. ISBN 978-7-03-029089-2.

7. FICEK, M., VIČAR, D., RAK, J., SVOBODA, P. 2016Usingthe SW modeling and simulatingtools in transport ofhazardouscargos. In: TRANSPORT MEANS 2016. Juodkrante, Lithuania: Kaunas university of technology, p. 862-865. ISSN 2351-7034.

8. INANLOO, B.et al., 2015. Explosionimpalsduring transport ofhazardouscargo: GIS-basedcharacterizationofoverpressureimpacts and delineationofflammablezonesforammonia. In: JournalofEnvironmental Management. s. 1-9. DOI: 10.1016/j.jenvman.2015.02.044. ISSN 03014797.

9. BERNATIK, A. et al., 2008. Modelling accidentalreleasesofdangerousGatesintotheLoirtropospherefrom mobile sources. In: ProcessSafety and EnvironmentalProtection. p. 198-207. DOI: 10.1016/j. psep.2007.12.002. ISSN 09575820.

Paper submitted: 05.02.2018.

Paper accepted: 16.04.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

327

Journal of Applied Engineering Science

Original Scientific Paper

doi:10.5937/jaes16-17071

Paper number: 16(2018)3, 536, 328 - 332

TECHNOLOGICAL MODERNIZATION OF FOREST ROADS CONSTRUCTION IN RUSSIA Mikhail Smirnov*, Yuri Andrianov, Victor Chernyakevich Volga State University of Technology (VSUT), Russia The paper considers the basic problems of construction of wood roads in Russia. Lack of sufficient quantity of roads functioning year round leads to annual isolation of almost 15 million inhabitants of forest settlements in autumn and in spring, absence of rhythm in timber-hauling process, increase in forest products cost. Two organizational methods of forest roads construction, namely, flow-line conveyor and non-flow-line conveyor ones are compared. According to the advanced flow-line conveyor method a complex stream is organized uniting specialized streams equipped with road machines and mechanisms for performing separate works. When the amount of work is small it is recommended to apply a non-flow-line conveyor method, either cyclic or inby.

The experience of constructing a road in a forest, which does not provide any headworks and is carried out by a dredge only, is considered. It allows to lower expenses for construction works.

Owing to the absence of stone road-building materials in the majority of forest regions of Russia a new design of a two-layer road covering from reinforced concrete slabs and the technology of its construction have been developed.

The comparison of the suggested technology of forest roads construction with existing domestic technologies and the Finnish one has shown its efficiency both in terms of money, and from the viewpoint of reduction of the number of necessary machines and equipment.

The technology suggested has passed industrial approbation on running logging enterprises.

Key words: Forest road, Technology, Construction, Reinforced concrete slab INTRODUCTION In the strategy of forestry development in the Russian Federation for the period till 2020 much attention is paid to forest roads construction. So, for the period specified it is planned to construct 10 000 km of forest roads and to allocate for this about one billion Euro from the federal budget. It is planned to build trunk-line roads using budget funds and to construct access routes to them using forces and facilities of large forests tenants. Creation of a special state cor-poration for construction of forest roads is considered and discussed in press.

The problem of forest roads construction is complex and many-sided [7-14]. Forest roads are used not only for timber hauling, frequently they provide development of the region as a whole and are important for solving social and other problems. For example, because of the lack of all-weather roads about 15 million of Russians living in forest villages remain literally cut off from the civilization in spring and autumn. Therefore local municipalities should take part in construction, maintenance and repair of trunk-line forest roads as well. It is known, that the better timber-hauling network is developed, the more the payment for forest fund rent conveyed to tenants by the state is. Accordingly, more funding can be allocated for forestry from the budget, including transport infrastructure. Thus, the forest roads system is the wealth bringing constant income and requiring development and maintenance.

Scientific research and global practice show, that forest resources become accessible and forest management successful if road system in a large forest amounts to no less than 10 m/ha of the forest area [1]. It should be noted, that in the majority of forest regions of the Russian Fed-eration the average density of road system makes 0.12 m/ha and is far from the above mentioned parameters [6]. Such state-of-the-art limits the availability of forest resources and possibility of carrying out operations on their reproduction, control and protection. The experience of forest fire fighting in 2010 convincingly testifies that because of impassability of Russia tremendous multi-billion damage was done to large forests, ecological and biological systems and the popu-lation.

DISCUSSION AND METHODS When constructing forest roads two methods of organization of work are used: flow-line conveyor and nonflow-line conveyor ones [1, 2]. The forest road is an engineering construction designed for movement of vehicles transporting cargoes and passengers. Its basic elements are earth roadbed, road dressing with trackway and road shoulders, engineering constructions (bridges, culverts, water diversion channels, etc.), road furnishing (road signs, marking, etc.). Using a flow-line conveyor method, which is considered to be basic, most progressive and theo-retically developed, a complex conveyor is organized, uniting specialized streams (squads, teams)

* Volga State University of Technology (VSUT), Russia, SmirnovaEV@volgatech.net

328

Mikhail Smirnov, et al. - Technological modernization of forest roads construction in Russia

equipped with road machines and mechanisms for performing different kinds of work. Only with small amounts of work, where complex conveyor is impossible to be organized, it is recommended to apply non-flow-line conveyor methods: cyclic or sectional.

To construct a road it is necessary to execute the interconnected complex of preparatory, basic and final works.

These works are carried out on easement area. The width of an easement area is adjusted by the Land code of the Russian Federation, 2.05.02-85 SNiP (Construction Standards and Regulations), and makes no less than 30 m for trunk-line roads. Sometimes the clearing width can amount to more than 40 m [1]. It is established, that occupying large territory (4-6 ha/km) forest roads reduce annual wood increment by 16-20 m3 per every kilometer and worsen ecological situation in general [2].

Therefore it is necessary to take measures to reduce the damage from land withdrawal, reducing the width of an easement area.

After cutting a lane and hauling the timber a traffic lane continues to be prepared: stumps are grubbed out or sheared off, a vegetative layer is removed, undergrowth and bushes as well as ground wood and stones are cleaned up, repeatedly laying-out the ground works.

Stumps and a vegetative layer are moved to the borders of a clearing. As a result of preparatory works execu-tion the traffic lane represents a closed at all sides site where water can gather and stand. According to the recommendations of the Norms [3] earth trenches or delves for drainage are arranged on its both sides 10-15 days before the beginning of erection of the earth roadbed. Thus, to perform preparatory works apart from harvesting machinery there will be required a set of vehicles consisting of a stumper, a bulldozer and a trench-cutting machine.

It is known, that on the average 1-2 culverts are erected within 1 km of a hauling road. For their construction reinforced concrete round rings and prefabricated portal blocks are recommended [3] to be put onto concrete foundation or gravel-macadam foundation pads. Construction of culverts using the existing technology consists of a set of operations: preparatory works connected with the delivery of materials and equipment, preparation of a site where the foundation will further be arranged, installation of blocks of pipes, waterproofing and other works. Such technology is material-intensive, labor-consuming and cost-demanding. Artificial constructions can make from 10 up to 25 per cent of road construction budget cost [2]. It should be noted, that polypropylene pipes has become widespread. Such pipes are made in Russia as well; they possess augmented mechanical stability, reliability within the conditions of temperature fluctuations, have rather small weight, and are produced in diameters up to 1137 mm. These pipes are widely applied on development sites, for example, in reclamation.

Road construction in the forest is complicated with the fact that it is conducted far from the building industry centers in underdeveloped and sparsely populated areas and is concerned with soft and swamp materials, deJournal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

ficiency or full absence of stone ones. According to the Forest Code in force, roads arrangement in large forests is a duty of tenants. However, market relations, competition in the commodity market, aspiration to increase profit make them minimize forest roads construction, maintenance and repair expenses, thus reducing transport component in the production cost.

Cost reduction can be achieved on the basis of foreign experience implementation. The experience of Finland where the greatest approximation to the model of sustainable forest man-agement is realized can be interesting and useful for Russian tenants. Construction of forest roads in Finland is carried out in a lane the width of which is only 12-14 m [15]. After tree felling in a lane and timber hauling no additional preparatory works are executed there and it is this point where significant reduction in cost of road construction is achieved. When timber has been hauled earth roadbed erection begins. The most essential difference of Finnish technologies from Russian ones is in the fact that the works are carried out not by a set of vehicles but by only one machine, the excavator. It excavates the ground and moves it from side ditches to earth roadbed. The ditches depth amounts to approximately 1.0 m, and the height of an earth fill reaches 0.5-0.6 m for waters to go downwards and the road upwards. The cross-section structure of such an earth roadbed is shown in Fig. 1.

Figure 1: Cross-section structure of an earth fill erected by an excavator from side ditches, m As the excavator is equipped with a bucket with a continuous cutting edge it also makes rough grading of the earth roadbed top, ditches, slopes of earth fills, as well as preliminary compacting of ground by its own weight of about 15-20 t depending on an excavator type. The same excavator, being a universal machine and having handling accessories (e.g., a set of slings), is also used for installing culverts of all-metal or other pipes. It is necessary to note, that application of an excavator, absence of restrictions on humidity and density of subsoils allow to essentially increase the duration of the road-building season which, using the existing technology, is feasible for only 78-108 working days. It will allow to use building machinery more effectively, to increase annual output and to reduce construction time.

Further, within several years, earth natural stabilization and compaction takes place and the traffic flow is banned there during summer periods. Then a 4.0 m wide and about 0.3 m thick single-layer gravel covering is ar-

329

Mikhail Smirnov, et al. - Technological modernization of forest roads construction in Russia

ranged on the dried roadbed, gravel being taken from local open cast mines without optimizing its structure.

Depending on subsurface hydrological conditions of the site, filtering, isolating or other layers can be used.

In the majority of forest regions of Russia natural stone road-building materials (gravel, macadam) are absent or extremely scarce. In our opinion, for such regions it is expedient to use modular reinforced concrete slabs as road covering. They can be produced with high quality at precast concrete factories all the year round and delivered to a place of laying down beforehand (e.g., in winter). More than half a century practice of modular design application has shown that such coverings, in comparison with others, allow quick increase in both the volume of forest road construction, and the volumes of forest exploitation.

However, the slabs utilized now and the constructions made from them demonstrate serious drawbacks limiting their application. The main drawbacks are as follows: • application of expensive (imported) coarse-grained aggregate, i.e. macadam or gravel, considerably increases slab cost; • imperfection of slab joints in a wheel guide causes the destruction of both the slabs, and the road and results in slabs breakdown, loss of road covering operational quality, additional expenses for repair and maintenance.

A new two-layer construction of modular road covering from slabs made of fine-grain (sandy) reinforced concrete [4-5] without the above mentioned drawbacks is developed and comprehensively investigated at the VSUT (Fig. 2).

In such a two-layer covering strengthened with welded joints the load from transport wheels is transferred by the top slab at least to two slabs of the bottom layer. As a result contact pressure in the zone of interaction of slabs with the sandy basis considerably decreases, their alignment along the slab length takes place. This excludes occurrence of dangerous plastic deformations in the sandy basis supporting the slabs, and in the underlying ground of the earth roadbed, considerably raising operational reliability of both the covering, and the road in general. Under adverse conditions of the locality (damp, wet places getting waterlogged, etc.) additional strengthening of the road design can be executed by applying a synthetic non-woven material, teplonit, laid down on the earth roadbed. The most used teplonit is 0.006 m thick and is produced as fabric up to 2.2 m wide in rolls up to 40 m long. It has an internal film layer behaving as isolating and connecting interlaminar layer. The film layer also distributes loadings over the surface of the bottom layer, prevents pushing, and increases mechanical durability of the material. One of the drawbacks of common non-woven materials made from synthetic fibers is the fact that they are pervious, which spoils their operational properties. Teplonit has a waterproof film layer which interferes with penetration of moisture. Besides, it consists of polypropylene fibers which are nonhydroscopic, environmentally friendly and light. Heat resistance of teplonit is high, about 1300С. It is convenient in operation and installation, easy to cut out, fastens to the surface, and can be lengthened in width. This material intended for water and heat isolation, is capable to protect the road construction against frosty soil expansion, which is important for operation of roads in conditions of the Russian Federation.

Teplonit is proved to be ecologically safe material, it is characterized by high chemical stability, is not subject to rotting, is much lighter and stronger than geotextile materials made from other chemical fibers and their mixes.

Teplonit prevents intermixing of sand underlayer with the ground of the earth roadbed, protects the latter against excessive moistening with abundant atmospheric precipitation and spring melting of snow cover.

RESULTS

Figure 2: Two-layer road covering from reinforced concrete slabs: 1 – upper layer slabs, 2 – lower layer slabs, 3 – slabs sand base, 4 – butt junctions, 5 – teplonit, 6 – road base, L – slab length

330

The comparative analysis of technological distinctions of the works carried out, technical equipment applied in existing and developed technologies of forest road construction is presented in summary Tab. 1.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Mikhail Smirnov, et al. - Technological modernization of forest roads construction in Russia

Table 1: Summary technological table of the work performed and machinery used under the technologies of wood road construction existing and being designed

1

Technological operation

Existing technology and machinery

Technology and machinery being designed

2

3

4

I Preparatory works 1

Tree felling and skidding, limbing, hauling

2

Stump, float stones, ground wood, undergrowth, etc. removal

3

Vegetative layer removal

Forest glade width up to 14 m (twoClearing width is to no less than 30 m fold reduction of harvesting machinery (harvesting machinery) volume of work) Bulldozer, stumper (the work is perBackhoe in the places of earth trenchformed on borrow pits on fill bases up to es, removed stumps are stacked into 0.5 m high) the road bed Bulldozer (he work is performed on borNot prefomed row pits on fill bases up to 0.5 m high) II Earth-moving works

4

Layer-by-layer development of ground in borrow pits and its moving to an earth-fill

Squad of 2-3 bulldozers

Earth-working is performed by an excavator in side ditches

5

Layer-by-layer earth fill soil spreading

Bulldozer

Not performed

6 7

Layer-by-layer landfill compacting by manifold roller passage with humidity Roller, laboratory and specific gravity control Earth roadbed top, sideslopes and Slop trimming machines, motorized road barrow pits grading grader

Not performed Excavator

III Water-carrying pipes construction 8

Site preparation

Bulldozer, diggers

Not performed

9

Pipe base preparation

Bulldozer, diggers

Excavator

10

Gravel-crushed stone subbase installation

Dump truck, bulldozer, diggers

Not performed

11

Pipe mounting

Track-type crane

Excavator with interchangeable equipment (slings) for all-metal or polypropylene pipe laying

12

Pipe waterproofing

Workers

Not performed

IV Road dressing construction 13

Placing water and heat isolating layer of teplonit

Not performed

Workers, work performance is possible in winter environment Dump trucks, motorized road grader (winter period is recommended to avoid tracking)

14

Delivery of sand and subbase installation (sub-base)

Dump trucks, motorized road grader, roller

15

Gravel material delivery

Dump trucks

16

Slabs delivery

17

Gravel surfacing arangement

Motorized road grader, light roller and super-compactor

Not performed

18

Road covering arrangement

Not performed

Hoist, welder, workers

Not performed Tractor-trailer trains, truck crane (works are executed in winter), workers

V Final works 19

Traffic lane remediation (removed vegetative layer and stubs cleaning off)

Excavator, dump trucks

Not performed

20

Finishing of a road (instalation of signs, crossingpoints provision, striping)

Moto transport vehicles, workers

Moto transport vehicles, workers

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

331

Mikhail Smirnov, et al. - Technological modernization of forest roads construction in Russia

CONCLUSION Analysing the data given in Tab. 1 it is possible to make certain conclusions. The tech-nology of forest road construction under consideration allows tenants to minimize capital investments on purchasing, leasing or renting road-building machines as well as current expenses on their maintenance; it allows to lower substantially the need in experts and workers, to reduce the payroll budget and other payments.

It is necessary to note, that the costeffective technology for forest road construction offered should pass wide approbation in working environment with the aim of bringing in necessary specifications and additions.

We have made and tested experimental batches of road slabs from fine-grain sandy concrete which have successfully passed factory (OAO KPD, Yoshkar-Ola) and industrial check on a working forest hauling highway of Maiski forest industry complex in the Kirov region of the Russian Federation.

In the forest regions of the Russian Federations not rich in natural stone materials it is expedient to use reinforced concrete slabs for road covering. The slabs are made with local sand and are applied in two-layer coverings developed by the authors.

REFERENCES 1. Ilyin, B.А. Theoretical bases of forest roads organization design: study guide for students / B.А. Ilyin.

Ex. ed. G.F. Grekhov – S-Pb.: LTA, 1992. -192 p.

2. Salminen, E.О. Forest transport: in 2 v. V. 1 Land transport: study guide / Ed. by E.О. Salminen. – M.: Akademia, 2009. – 368 p.

3. Technological rules and maps of timber hauling highways construction: in 2 v. V. 1. Technological rules.

– Leningrad: HYPROLESTRANS, 1975. – 209 p.

4. Chernyakevich, V.I. Designing and calculation of a multilayer road covering from rein-forced concrete slabs: monograph / V.I. Chernyakevich, N.N. Pushkarenko, L.M. Chernyakevich. – Yoshkar-Ola: MarSTU, 2007. – 177 p.

5. Smirnov, М.Yu. Quality perfection of fine-grain cement concrete for road slabs / М.Yu. Smirnov, V.I. Chernyakevich, N.N. Pushkarenko // Rational use of forest resources: Proc.of international research and practical conference, 20-22 April 1999. – Yoshkar-Ola: MarSTU, 1999. – Pp. 147-148.

6. Smirnov, M.Yu. Roads and transport in the forest / M.Yu. Smirnov, A.D. Gryazin // Forest Industry. – 1990. – No 7. – Pp.23-24.

7. Oregon Spatial Data Library. 2014. Transportation Data Framework Data. http://spatialdata.oregonexplorer.info/geoportal/catalog/main/home.page# (last accessed 25 May 2014).

332

8. Sander, H.A., D. Ghosh, D.V. Riper, and S.M. Manson. 2010. How Do You Measure Distance in Spatial Models? An Example Using Open-Space Valuation.

Environment and Planning. B: Planning and Design 37(5): 874-894.

9. Chung W, Stuckelberger J, Aruga K & Cundy TW. 2008. Forest road network design us-ing a trade-off analysis between skidding and road construction costs. Canadian Journal Forest Research 38: 439448.

10. Bont, L. G., H. R. Heinimann, and R. L. Church.

2012. Concurrent optimization of harvesting and road network layouts under steep terrain. Ann. Oper.

Res. 11. Chung, W., J. Stückelberger, K. Aruga, and T. W. Cundy. 2008. Forest road net-work design using a trade-off analysis between skidding and road construction costs. Can. J. For. Res. 38(3):439-448.

12. Bustamante and H. Musante. 2006. A combinatorial heuristic approach for solving real-size machinery location and road design problems in forestry planning. Oper. Res. 54(6):1017–1027.

13. Bont, L.G., Heinimann, H.R., Church, R.L., 2015.

Concurrent optimization of harvesting and road network layouts under steep terrain. Ann. Oper. Res. 232(1):41-64. doi:10.1007/s10479-012-1273-4 14. Cavalli, R., Grigolato, S., 2010. Influence of characteristics and extension of a forest road network on the supply cost of forest woodchips. J. For. Res. 15(3):202–209. doi:10.1007/s10310-009-0170-4 15. Kiiskinen, P., Savonen, X, Tomperi, T., 2014. Forest road construction. - Joensuu , Grano Oy, Mikkeli. – 51 p.

Paper submitted: 05.04.2018.

Paper accepted: 07.06.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Journal of Applied Engineering Science

doi:10.5937/jaes16-16888

Original Scientific Paper

Paper number: 16(2018)3, 537, 333 - 342

CREATION OF IMAGE MODELS FOR INSPECTING VISUAL FLAWS ON CAPACITIVE TOUCH SCREENS Yuan-Shyi Peter Chiu, Hong-Dar Lin* Department of Industrial Engineering and Management, Chaoyang University of Technology,Wufong District, Taiwan Touch screens (TSs) are commonly applied in many electronic appliances such as smartphones, tablets, etc. Currently, capacitive touch screens (CTSs) are the main touch technology of screen panels due to many excellent electronic properties. Problems exist in inspecting flaws inlaid in appearances of CTSs with structural patterns. Area flaws are a type of common visual defect that comprises dust, bubbles, ripple marks, and other flaws of bigger sizes. These flaws have the attributes of low contrast, brightness with slow changes, unusual and non-orientation forms, and sometimes both bright and dark flaws existing at the same time in a region. This paper suggests image models based on transformation filtering to inspect the area flaws on appearances of CTSs. We apply the Haar wavelet transform with flat zone filtering technique to eliminate the structural patterns of background by means of filtering an approximate sub-image of a breakdown wavelet domain image. Subsequently, the filtered image is reversely transformed to obtain a rebuilt image in spatial domain. Last, the rebuilt image with intensified flaws can be simply partitioned into three species (black flaws, gray flaws, and white background) by using a statistical interval estimation method.

Therefore, the intricate area flaws are precisely identified by the suggested scheme. We contrast our approach with three traditional methods with real samples under complex background and conduct quantitative comparisons. The effectiveness and accuracy of the developed image models are confirmed by expert assessments, as well as by comparative analysis with the known methods in the field of spatial localizations and production-related effects of flaw detection.

Key words: Computer-aided inspection system, Visual flaws, Capacitive touch screens, Image models, Transformation filtering, Filters, Imaging, Inspection INTRODUCTION According to the growth of smartphones, common touchtone phones are progressively being substituted to arouse a wave of touch screen devices. Touch screens (TSs) are adopted not only for mobile phones but also for computers, televisions, cameras, handheld game consoles and other 3C products. Thus, the increasing need for TSs encourages the expansion of the TS industry. Currently, touch technology of screen panels can be mainly divided into resistive (an earlier technology), capacitive, optical, electromagnetic, and ultrasonic types.

Since resistive touch screens (RTSs) are susceptible to scrape and fire, low light transmittance, and slow response shortcomings, the RTS products cannot satisfy the modern requirements for touch technology and thus have a market share much lower than that of the capacitive touch screens (CTSs). The CTSs, being an unshakable market leader in the TS industry, have the advantages of waterproofness, stain proofness, scratch proofness, fast response, anti-ultraviolet, etc.

The appearance quality of TSs influences their yield rates remarkably in manufacturing process. Lin et al. [1] reported a total flaw rate is 36.2% of 12 inspection tasks in the manufacturing process of a professional TS producer. Four production operations are related to the visual quality of sensing circuits, inclusive of TS clean, protect film attachment, sensing circuit pre-test and function film attachment. This flaw rate of the four operations is

14.1% that is roughly up to 38.95% of the total flaw rate in the production procedures. Accordingly, automated visual detection for the appearance flaws on the TSs is absolutely essential in the manufacturing process [2, 3].

CTSs are composed of transparent glass substrates, on the surface of which an oxide metal is regularly coated.

Figure 1 shows a CTS screen with duplicate modes of sensing circuits and an enlargement of part of a CTS surface. The CTSs have multi-layer structures and are categorized as structural textures. Appearance flaws influence the exteriors of CTSs as well as their durability, reliability and serviceability. It is a severe examination work when flaws are inlaid on appearances of CTSs having structural textures (sensing circuits). Small appearance flaws, regularly arising in the production process of TSs, arouse much more damages and losses when they emerge in electronic components than in manufacturing parts. Thence, to exist in current rival market of sci-tech products, TS producers cannot afford to neglect small appearance flaws [4, 5].

The appearance flaws are usually classified into two types: linear and areal. The areal type includes dust, water signs, foams and additional flaws of bigger dimensions. These flaws have the attributes of low contrast, brightness with gradual changes, irregular and non-directional shapes, and occasionally both bright and dark flaws existing at the same time in a region. This kind of flaws compared with the linear type is more complicate to identify its regularity.

* Department of Industrial Engineering and Management, Chaoyang University of Technology, Wufong District, Taichung 41349, Taiwan, hdlin@cyut.edu.tw

333

Yuan-Shyi Peter Chiu, Hong-Dar Lin - Creation of image models for inspecting visual flaws on capacitive touch screens

(a) (b) Figure 1: A CTS display screen (a) with duplicate modes of sensing circuits and (b) an enlargement of part of a CTS surface Thus, this paper proposes an automated inspection system to detect the areal flaws on TSs. Figure 2 shows a flaw-free image and two defective images with area flaws of CTS appearances having directional repetitive textures. The directional textures indicate grid forms with linear circuits in four ways: horizontal, vertical, and two diagonals. Those background sensing circuits make the flaw detection task harder when area flaws are inlaid in the appearances of directional textures. We thereby propose an image restoration approach based on WT to conquer these problems of automated areal flaw inspection of CTSs. The rest of the paper is organized as follows: First, we review the literature on optical methods of image processing for flaw detection. Second, we explain the proposed image models for inspecting flaws on CTSs. Third, we conduct the experiments and evaluate the performance of the proposed models with known methods. Four, we present the conclusion and the future work.

LITERATURE REVIEW Optical inspection of appearance flaws has turned into an important work for industries who exert to enhance output quality and process efficiency [6, 7]. Flaw inspection techniques compute a set of textural characteristics in a moving mask and seek for crucial partial changes among the feature vectors in spatial or frequency domains [8]. Lin and Chiu [9] developed a machine vision

system to find mass centers of chips, locate cutting lines and estimate process regulation plans for the automated and high-speed dicing of electronic passive components.

Adamo et al. [10] proposed a low-cost inspection system based on the Canny edge detection for online defects assessment in satin glass. Liu et al. [11] presented the method based on watershed transform methods to segment the possible defective regions and extract features of bottle wall by rules.

Automatic thresholding techniques have also been abroad utilized in the machine vision fields for automatic optical detection of flaws [12]. The Otsu method [13] is the representative approach of threshold techniques for common images regarding consistency and appearance measures. It means chosen threshold values maximizing the between-class variances of intensity histogram. It offers satisfying outcomes for thresholding an image having an intensity histogram of bimodal distribution. Ng [14] modified the Otsu method by choosing optimum threshold values for unimodal and bimodal distributions, and evaluated the effectiveness of this improved technique on flaw inspections. Navarro et al. [15] developed a sensor system with thresholding technology for inspecting blemishes in ship hull surfaces.

Three popular transforms, Fourier, Gabor and wavelet, are ordinary frequency transformations applied in texture analysis [16]. Nasira and Banumathi[17] utilized the Fourier transform and image processing to detect textile

Figure 2: Three CTS images having directional repetitive textures: (a) a flaw-free image; (b) two defective images with various area flaws

334

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Yuan-Shyi Peter Chiu, Hong-Dar Lin - Creation of image models for inspecting visual flaws on capacitive touch screens

flaws. Tsai and Hsiao [18] developed a wavelet transform method to detect flaws inlaid in uniformly textured surfaces. Lin [19] implemented a wavelet-based multivariate statistical method to detect water flaws with little gray level contrast on the surface of ceramic components. Li [20] used the wavelet transform to strengthen flaw areas and abstract discriminatory characteristics from reconstructed images and the support vector machine to categorize five major flaw classes on copper surfaces. Chang [21] emphasized on structure plan and execution of on-line textile flaw inspection system using Gabor filtering. Lin and Chiu [22] combined block cosine transform and grey relational analysis for blemish detection on domed surfaces in packets of light emitting diodes.

Directional textures with uniform modes are regularly observed on artificial objects, like manufactured parts, cloth textiles, and electronic components. Lu and Tsai [23] presented a machine vision system applying independent component analysis for automatic detection of tiny defects on LCD panel appearances including pure and intricate styles. Regarding flaws detection of TSs, Chen et al. [24] developed an automatic visual inspection device for analogical RTSs. This equipment used image processing techniques to detect flaws on surfaces of RTSs. The RTSs have the pattern of regular spacers in spatial domain and lead to the mode of repeated dots in Fourier domain. Lin and Tsai [25] suggested a Fourier transform based approach to detect linear flaws on CTS appearances. The linear flaws such as scratches and cracks are regular flaws with directional shapes.

This kind of flaws compared with the areal type is less complicated to identify its regularity. Hung and Hsieh [2] developed an adaptive model with the ability of learning online to separate defects by using the features of repetitive patterns of the sensing circuits. This approach has a limitation of shadow effect problem that influence on defect detection. Liang et al. [3] proposed a sparse representation-based approach to detect touch screen flaws in low-resolution images. They used a sparsity ratio of the sparse representation coefficients like a measure for distinguishing defective images. The positions of detected defects can only be approximately located. For a new type of compound CTS patterns without primitives nor periodicity, Jiang et al. [5] introduced a combined method of nonnegative matrix factorization with tolerance model for defect detection in such CTS patterns. This method can only be dependably used in such new type of CTS patterns. Therefore, the purpose of this research is to find efficient and effective techniques to automatically detect the appearance flaws on CTSs.

ed by means of filtering the approximate subimage of a breakdown image in wavelet domain. It is hard to exactly inspect area flaws inlaid in intricate directional textures.

Consequently, we propose an entire image reconstruction approach applying wavelet transform and flat zone filtering procedure for area flaw inspection on surfaces of CTSs. This technique does not proceed with the procedures of feature extraction and template matching.

MATERIALS AND METHODS

where 0 ≤ m ≤ (P-1), 0 ≤ n ≤ [Q/2]-1, and [ ] is the Gauss operator. The column transfer denotes as:

This study explores the area flaw detection techniques of the fashionable CTS components. The suggested method transforms a CTS testing image with four distinct ways of regular lines in background to wavelet domain. Then, the directional textures of background will be eliminatJournal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Image Modeling& Analysis This study develops a Wavelet Transform (WT) based flat zone filtering approach to detect area flaws of CTSs. The paper extends the earlier work [4] of the authors by a considerable amount of discussion and further experiments. When a CTS image with four distinct ways of line patterns of sensing circuits is transformed to wavelet domain, the directional textures of background can be eliminated by means of filtering the approximate subimage of a break-down wavelet domain image. The filtering scope is determined by a statistical interval. Within the range, the frequency elements will be replaced by the mean value of wavelet frequencies. Subsequently, the filtered image will be reversely transformed to obtain a rebuilt image in spatial domain. Last, the rebuilt image can be easily partitioned into three species (black flaws, gray flaws, and white background) by using a statistical interval estimation method and some characteristics of the identified flaws are abstracted.

Wavelet Transform WT offers an easy path to gain a multi-resolution depiction, where texture properties can be likely abstracted.

We apply the Haar WT to execute image transformation for frequency filtering since the advantages of Haar WT comprise partial image processing, easy computations, fast processing, memory saving, and numerous image properties [26-28]. The Haar WT is one of the easiest and fundamental WTs. A typical breakdown of a spatial domain image can be performed by early using 1D Haar WT to each row of pixel values and next carrying out another 1D WT to each column. This Haar WT can be calculated step-by-step by the average and half of differences of two adjacent pixels. Due to the transform notion of the 1D WT, the Haar WT could deal with a 2D image of (M x N) pixels in the later procedures. The row transfer expresses as: gR (m,n)= [(g(m,2n)+g(m,2n+1))/2], gR (m,n+[Q/2] )= [(g(m,2n)-g(m,2n+1))/2]

(1)

gC(m,n)=[(gR (2m,n)+gR (2m+1,n))/2], gC(m+[P/2],n)= [(gR (2m,n)-gR (2m+1,n))/2],

(2)

335

Yuan-Shyi Peter Chiu, Hong-Dar Lin - Creation of image models for inspecting visual flaws on capacitive touch screens

where 0 ≤ m ≤ [P/2]-1, and 0 ≤ n ≤ (Q-1).

In the above notations (Eqs. (1-2)), g(m,n) is a testing image,gR (m,n) the row transfer function of g(m,n), and gC (m,n) the column transfer function of gR (m,n). While gC (m,n) is as well the result of the wavelet breakdown of g(m,n), the final results of a Haar WT can be designated as: S(m,n)=gC (m,n), T1(m,n)=gC (m,n+[Q/2]), T2(m,n)=gC(m+[P/2],n), T3(m,n)=gC(m+[P/2],n+[Q/2]), (3) where 0 ≤ m ≤ [P/2]-1, and 0 ≤ n ≤ [Q/2]-1.

One layer of wavelet breakdown produces one approximate subimage and three detailed subimages including subtle compositions with horizontal, vertical and diagonal directions. An image is broken down by WT into one approximate subimage (S) and three detailed subimages (T1, T2 and T3). The four subimages with sizes of (P/2 x Q/2) pixels constitute the wavelet features. One-layer Haar wavelet breakdown is utilized to accurately find the pixels with the textural features. Multi-layer wavelet breakdown produces rougher expression of a testing image. A big number of breakdown layers will lead to the mixture effect for the flaws and may arouse localization deviations of the found flaws [29].

Transformation Filtering The wavelet approach transforms images into a depiction where presenting spatial and frequency properties. It is appropriate for depicting partial variations in a uniform textured image. For one layer of wavelet breakdown, we gain one approximate subimage and three detailed subimages including subtle compositions with horizontal, vertical and diagonal directions. Through carefully filtering the approximate subimage in distinct breakdown layers for inverse WT, the rebuilt image will eliminate periodical, repeated texture patterns and enhance exclusively regional flaws. A statistical decision interval method can then be applied to distinguish between flaw areas and uniform districts in a rebuilt image. This changes a hard flaw inspection problem in intricate textured images into an easy interval estimation problem in uniform images.

Due to the oscillations of frequency tendency and the properties of low and high frequency districts, we devise a flat zone filter pivoted at the origin of a 2D wavelet spectrum to filtrate primary low frequency elements of the spectrum image. A suitable band is early selected for the flat zone filter in the WT spectrum. The frequency elements inside the band of the flat zone filter (low frequencies) are then set to the mean value of frequency, and those beyond the filter (medium and high frequencies) are preserved. Last, we transform reversely the filtered WT image back to spatial domain. Selecting an adequate band based on the level of frequency fluctuations for the flat zone filtering task can remarkably intensify the flaw regions in spatial domain. This intensified effects

336

can be obviously viewed in the rebuilt images.

Contrast to ordinary band filtering, the suggested flat zone filtering utilize the mean value of overall frequency instead of zero to substitute the original frequency elements inside the chosen zone area in WT domain.

The flat zone filtering process is to decrease the variation between background and texture. The major intention is to eliminate the directional background textures having small changes in intensity, and preserve the area flaws having larger changes in intensity. We use the concept of statistical decision interval to calculate the flat zone area x̅ ±kσ and substitute all of the frequency elements inside the region by the frequency mean value x̅ . The procedure can be expressed as: WT(u,v)= x̅, if x̅ -kσ < WT(u,v) < x̅+kσ;

(4)

WT(u,v)= WT(u,v), otherwise; where WT(u, v) is a WT frequency image, x̅ and σ are the mean value and standard deviation of the frequency in WT(u, v), and k is a constant determined empirically. Figure 3 shows two 3D WT spectrum diagrams which are the before and after plots of the flat zone filtering are executed. Figure 3(a) is the 3D diagram of a WT spectrum and Figure 3(b) is the 3D diagram of a WT spectrum with flat zone filtering process. After the flat zone filtering operation, the frequency elements having larger frequency are intensified in WT domain.

(a) Figure 3: Two 3-D diagrams of WT spectrum which are before and after plots of the flat zone filtering applied: (a) the 3D diagram of a WT spectrum

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Yuan-Shyi Peter Chiu, Hong-Dar Lin - Creation of image models for inspecting visual flaws on capacitive touch screens

where L is a threshold for separating flaws from background, h is a control constant, μg' and σg' are a mean and a standard deviation of the rebuilt image with size M × M. The outcome three-level image G(x, y) for flaw segmentation is:

(6)

(b) Figure 3: Two 3-D diagrams of WT spectrum which are before and after plots of the flat zone filtering applied: (b) the 3D diagram of a WT spectrum after flat zone filtering applied Inverse Transformation and Segmentation After the appropriate zone is decided, the frequency filtering procedure can exactly identify the flaw-free low frequency districts and these frequency values are assigned to a mean value in wavelet domain. Then, we conduct reverse WT on the filtered frequency image to spatial domain for flaw segmentation later. In this study, we intend to eliminate most duplicate modes in the rebuilt image by choosing an appropriate zone in the approximate subimages for the mean value replacement.

Since structural textures may express high directivity, rebuilding the detailed subimages with directional emphasis distinct from those of the regular textures will delete most directional duplicate modes in an original image, and retain merely partial flaws in a rebuilt image. The directional duplicate modes will lead to a roughly uniform intensity, while the partial flaws will produce different intensities in the rebuilt image.

The rebuilt images have consistent intensities for pixels pertaining to homogeneous background areas, but they also give remarkably distinct intensities for pixels pertaining to heterogeneous flaw regions. The gray level variability in homogeneous areas is small, while the gray level variability in heterogeneous regions is very large compared with the whole rebuilt image. Thence, a statistical method is used to establish an interval for differentiating flaws from duplicate modes in the rebuilt image. The rebuilt image g' (x,y) will be roughly a uniform intensity image if a flaw-free testing image is processed.

The upper and lower interval limits (LL,LU) for gray level changes in the rebuilt image are denoted as: L=μg' ± h σg'

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

(5)

If an intensity value locates in the interval limits (LL,LU) then gray level is assigned to 255 (white) as a white backdrop. Otherwise, gray level is assigned to 0 (black) as a part of dark flaw if an intensity is less than LL and gray level is assigned to 127 (gray) as a part of gray flaw if an intensity is more than LU. Once a pixel with an intensity locates in the interval limits, this pixel is categorized as a homogeneous background component. Otherwise, it is categorized as a heterogeneous flaw component. If the flaw size to be inspected is usually very small compared with a whole testing image, μg' and σg' can be calculated straightly from the rebuilt image of the testing image to adapt lighting changes in the inspection surroundings.

These interval ranges are utilized to differentiate between uniform line modes and flaws in a rebuilt image.

The superior and inferior ranges of intensities in a rebuilt image are located at a distance hσg' from the mean μg'.

Moreover, choosing an appropriate control parameter leads to accurately distinguishing flaws from ordinary areas; however, an inappropriate control parameter outputs numerous incorrectly inspecting ordinary areas as flaws. A little constant value h provides a strict control and can lead to erroneous alerts. A great constant value h offers a loosen control and can produce lost alerts. A supervised appearance inspection problem is explored in this study. The developed learning systems are ordinary in computer vision and are proper for supervised environments in production sites. The number of WT breakdown layers and the size of flat zone filtering utilized for image rebuild are decided in advance from a texture model. The influence of number of multi-resolution levels and parameter settings of WT filtering process on inspection performance are assessed by trial and error in the experiments.

EXPERIMENTS AND DISCUSSION To assess the effect of the suggested method, assessments were carried out on actual CTSs supplied by a TS production plant in Taiwan. All trial samples were arbitrarily chosen from the production process of TSs. Figure 4 illustrates the structures of the exploratory circumstance where we capture a CTS sample to be taken as testing images in the laboratory. The CTS images (270) having thickness 0.78mm, where 132 have no flaws and 138 have diverse area flaws, were examined in our assessments. All images of the CTS appearance have the same size of 256 x 256 pixels and an intensity of 8 bits.

337

Yuan-Shyi Peter Chiu, Hong-Dar Lin - Creation of image models for inspecting visual flaws on capacitive touch screens

(a)

(b)

Figure 4: Structures of the exploratory circumstance where scanning a trial CTS sample: (a) equipment framework of experiments; (b) a trial CTS sample is put on testing platform

Figure 5: The user interface design of the implemented visual inspection system

338

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Yuan-Shyi Peter Chiu, Hong-Dar Lin - Creation of image models for inspecting visual flaws on capacitive touch screens

The suggested flaw detection arithmetic is implemented on a personal computer (CPU i5-3230M 2.6 GHz and 4GB RAM) by R2009b version of MATLAB language.

To numerically confirm the manifestation of the suggested method, we contrast the outcomes of our assessments contrary to those supplied by technical assessors (i.e. the ground truth). Three impartial measures: correct classification rate (CCR) and incorrect classification rates (α and β) were assessed for the effect of the area flaw detection methods. Statistical type I error (α) measures a probability of generating erroneous alerts, which judge regular areas as flaws. The area of regular regions judged as flaws is divided by the area of real regular regions to gain type I error. Statistical type II error (β) measures a probability of generating lost alerts, which invalid to alert actual flaws. The area of undetected flaws is divided by the area of real flaws to gain type II error.

The higher the effect assessment measures: (1-α), (1β), and CCR, the more exact the detection outcomes.

The correct classification rate CCR is denoted as: CCR = (C(1-α) + C(1-β)) / Ctotal * 100%

(7)

where C(1-α) is the pixel number of regular regions correctly detected as flaw-free areas, C(1-β) is the pixel number of actual flaw regions right detected as flaw areas, and Ctotal is the pixel number of a testing image.

Figure 5 displays a user interface design of the implemented system. The system interface shows the outcomes and differences executed by the suggested filtering procedure in distinct phases for inspecting area blemishes in CTS appearances. The image (1) and image (2) are a captured image and a corresponding gray scale image from a part of a trial sample. The image (3) is the transformed image of WT with the second breakdown layer and the image (4) depicts the filtered image after the flat zone filtering operation conducted in WT domain. The image (5) and image (6) are the rebuilt image of the filtered image and the resulting three-level images that show the detected black flaws in black and gray flaws in gray by the proposed inspection system, respectively.

The outcomes disclose that the area blemishes in CTS appearance are accurately separated in the three-level image, no matter of background with structural patterns.

Choices of important parameters WT breakdown of an image with textures in the appropriate level will effectively stress the partial flaws in a consistent appearance. To evaluate the influence of alternating number of breakdown layers on the rebuilt results, experiments present the rebuilt images from breakdown layers 1, 2, 3, and 4, separately. All these images are individually rebuilt from a filtered approximate subimage and three homologous detailed subimages with the Haar wavelet. Both images indicate that very small the number of breakdown layers (e.g. 1 and 2) cannot adequately segment flaws from the repeated texture patJournal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

terns and causes many erroneous alerts. Nevertheless, over large the number of breakdown layers produces the mixed effect of the flaws and results in lots of lost alerts. The number of breakdown layer 3 is more proper to emphasize flaws in the rebuilt image. Our experiments on a diversity of textured images have verified that WT breakdown layer 3 is normally adequate for this area flaw inspection application.

If a decisive thresholding changes, two paired measures, erroneous alert rate (α) and detection rate (1-β), describing the outcomes of a hypothesis testing will vary [30]. When different thresholding are applied, the sets of erroneous alert rates and detection rates are figured as points in a Receiver Operating Characteristic (ROC) curve. The upper-left corner on a ROC figure represents an ideal outcome having 100% detection rate and a 0% erroneous alert rate. The more the ROC curve moves toward the upper-left corner,the better the trial executes.

High-energy frequency elements related with repetitive line modes may arise approximately the primary zones in wavelet domain images. To remarkably filter out most consistent line modes and entirely retain the area flaws in spatial domain images,the frequency elements on the primary zones must be replaced by the mean value of the frequency elements from the wavelet domain images. The filtering wideness decides the districts of the zone neighborhoods will be filtered out for high-energy frequency elements. Experiments indicate the flaw detection effect of the suggested method with k value of 1.1 is better than those of the other k values. The flat zone filtering method with larger k value eliminates consistent line modes and partial flaws in the rebuilt image and lead to ignoring small flaws. It suggests the more exact regions of zone neighborhoods are filtered, the better the flaw inspection outcomes will obtain.

Contrast effects of distinct area flaw detection methods Three traditional methods commonly applied to defect detection are compared to contrast effects of flaw detection. To reveal the flaw detection outcomes, Figure 6 displays partial outcomes of inspecting area flaws by Otsu method [13], Iterative method [31], three-level method (two threshold values in spatial domain without frequency filtering), the suggested method, and ground truth, separately. The three spatial domain skills, the Otsu, Iterative and three-level methods, make many incorrect alerts on area flaw inspection. The frequency domain skill, the suggested method, inspects majority of area flaws and makes few incorrect alerts. Consequently, the frequency domain technique excels the spatial domain approaches in the area flaw inspection of the CTSs.

339

Yuan-Shyi Peter Chiu, Hong-Dar Lin - Creation of image models for inspecting visual flaws on capacitive touch screens

Figure 6: Resulting examples of area flaw inspection on CTS by: (a) Otsu method, (b) Iterative method, (c) three-level method, (d) proposed method, and (e) the ground truth To contrast the effects of different area flaw inspection methods, Table 1 sums up the detection outcomes of our tests. Three spatial domain skills and one frequency domain technique are assessed contrary to the outcomes by technical assessors. The mean flaw detection rates (1-β)of all testing images by these methods are, separately, 81.07% (Otsu method), 80.87% (Iterative method), 90.14% (three-level method), and 92.07% (suggested method). Nevertheless, the three spatial domain skills have remarkably larger erroneous alert rates (α), 26.70% (Otsu method), 20.41% (Iterative method), and 11.64% (three-level method). Otherwise, the suggested approach has quite smaller erroneous alert rate, 4.43%. In addition, the suggested approach has larger

CCR than do the other methods used to area flaw inspection of CTS samples. The average processing time for dealing with a 256 x 256 image is: 0.04 seconds by Otsu method, 0.06 seconds by Iterative method, 0.14 seconds by three-level method, and 0.16 seconds by the suggested method. Therefore, the suggested method conquers some troubles of inspecting area flaws on CTSs and outperforms in the ability of accurately differentiating area flaws from textured background.

Table 1: Contrast effects table of four flaw detection methods

340

Methods

α (%)

1-β (%)

CCR (%)

Time (sec.)

Otsu method

26.70

81.07

73.46

0.04

Iterative method

20.41

80.87

79.48

0.06

Three - level method

11.64

90.14

87.98

0.14

Proposed method

4.43

92.07

95.32

0.16

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Yuan-Shyi Peter Chiu, Hong-Dar Lin - Creation of image models for inspecting visual flaws on capacitive touch screens

Experiments on wavelet filtering effects To further assess the effect of WT filtering process in the suggested method, experiments with and without the flaw enhancement effect of filtering on CTSs are both executed. All testing images with and without the WT flaw enhancement procedure are separated by the three-level method to investigate how they differ in terms of area flaw inspection. Both of the three-level method and suggested WT-based filtering method achieve good performance in detecting area flaws of CTSs and the appearance flaws that hurt the product quality. Table 2 illustrates the production-related effects of appearance flaw detections by the present method, three-level method, and suggested method. The merits of the suggested method are fourfold: (1).Both of the three-level and suggested methods excel in the ability of accurately discriminating area flaws from ordinary areas. The suggested method has smaller erroneous alert rate and larger flaw detection rate than do the traditional methods. (2).

The two methods have larger CCRs than do the present method (human visual inspection) used to area flaw detection of CTS surfaces. (3).The retest rates for detecting appearance flaws of CTSs are as follows: 20.6% by present method, 7.7% by three-level method, and 2.6% by suggested method. The retest rate is the percentage of reinspection and retesting of CTSs that have undertaken rework or other corrections. The suggested method has separately almost three times and eight times smaller average retest rate than do the three-level method and present method. (4).The suggested method has the smallest material wastage rate 3.2% than do the present inspection method and three-level method because of the excellent flaw detection accuracy. The contrast outcomes of the experiments evidently illustrate the feasible and adequate property of the suggested WTbased filtering method in inspecting area flaws on CTSs.

methods of image processing for defect detection on touch screens and confirm the importance and urgency of the development. Then, we introduce a visual inspection system with image models based on transformation filtering for automatic detection of area flaws on CTS surfaces with structural textures. The duplicated line patterns of four directional textures in the testing images can be simply reduced by detecting the band region of an approximate subimage of a breakdown image in wavelet domain, assigning them to a mean value of frequency elements by the flat zone filter, and taking reverse WT to obtain rebuilt image. In the filtered rebuilt image of a CTSsurface, the periodic line regions will have a roughly uniform intensity distribution,while the flaw regions will be obviously preserved. A statistical interval is consequently estimated to establish the interval limits for differentiating among black flaws, gray flaws, and consistent line pattern background. Thus, the intricate area flaws can be precisely identified by the proposed system. The created methods of digital image processing allow to justify localization of flaws on CTS surfaces. Assessment outcomes indicate the suggested method reaches a larger 92.07% probability of accurately differentiating flaw areas from regular regions and a smaller 4.43% probability of mistakenly judging regular regions as flaws on repetitive textured appearances of CTSs. In the computation time of the suggested method, the time of taking forward and reverse WT requires almost 0.1 second averagely without searching any optimization algorithms for parameter determination in advance.

This computation time will be remarkably reduced when the parameter optimization has been executed and the WT is carried out in a hardware chip. Thence, the ways of further study will concentrate on investigating existing techniques to look for the most efficient and effective way for the proposed application.

ACKNOWLEDGEMENTS

CONCLUSION This study is devoted to the current scientific and technical research. We first reviewthe literature on optical

Authors deeply thank the Ministry of Science and Technology of Taiwan for sponsor of this study (under grant no. MOST103-2221-E-324 -036).

Table 2: The production-related effects of appearance flaw detections by the three methods Methods

Retest rate

Material wastage

Present method (Human visual inspection)

20.6%

16.7%

Three-level method

7.7%

4.8%

Suggested method

2.6%

3.2%

REFERENCES 1. Lin, Y.K., Huang, C.F., Chang, P.C. (2013). System reliability evaluation of a touch panel manufacturing system with defect rate and reworking. Reliability Engineering and System Safety, 118 (10), 51-60.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

2. Hung, M.H., Hsieh, C.H. (2015). A novel algorithm for defect inspection of touch panels. Image and Vision Computing, 41, 11-25.

3. Liang, L.Q., Li, D., Fu, X., Zhang, W.J. (2016). Touch screen defect inspection based on sparse represen-

341

Yuan-Shyi Peter Chiu, Hong-Dar Lin - Creation of image models for inspecting visual flaws on capacitive touch screens

tation in low resolution images.Multimedia Tools and Applications, 75(5), 2655-2666.

4. Lin, H.D., Li, J.M. (2014). Automated area defect inspection of touch panels using computer vision.

2014 Proceedings of the International Conference on Image Processing, Computer Vision, and Pattern Recognition (IPCV 2014), p. 16-22.

5. Jiang, C.C., Quan, Y.M., Lin, X.U. (2016). Defect detection of capacitive touch panel using a nonnegative matrix factorization and tolerance model. Applied Optics, 55(9), 2331-2338.

6. Murphy, J.N., Harris, K.D.,Buriak, J.M. (2015). Automated defect and correlation length analysis of block copolymer thin film nanopatterns.PLoS ONE, 10(7): e0133088.

7. Liu, J., Tang, Z., Zhang, J., Chen, Q., Xu, P., Liu, W. (2016). Visual perception-based statistical modeling of complex grain image for product quality monitoring and supervision on assembly production line.

PLoS ONE, 11(3): e0146484.

8. Huang, S.H., Pan, Y.C. (2015). Automated visual inspection in the semiconductor industry: A survey.

Computers in Industry, 66, 1-10.

9. Lin, H.D.,Chiu, Y.P. (2010). RBF network and EPC method applied to automated process regulations for passive components dicing, International Journal of Innovative Computing Information and Control, 6(11), 5077-5091.

10. Adamo, F., Attivissimo, F., Nisio, A.Di., Savino, M. (2009). A low-cost inspection system for online defects assessment in satin glass. Measurement, 42, 1304-1311.

11. Liu,H., Wang, Y., Duan, F. (2008).Glass bottle inspector based on machine vision. International Journal of Computer Systems Science and Engineering, 3(3), 162-167.

12. Sezgin, M., Sankur, B. (2004). Survey over image thresholding techniques and quantitative performance evaluation. Journal of Electronic Imaging, 13(1), 146-156.

13. Otsu, N. (1979). A threshold selection method from gray level histogram. IEEE Transactions on Systems, Man and Cybernetics, 9, 62-66.

14. Ng, H. F. (2006). Automatic thresholding for defect detection. Pattern Recognition Letters, 27, 16441649.

15. 15) Navarro, P., Iborra, A., Fernández, C., Sánchez, P., Suardíaz, J. (2010). A sensor system for detection of hull surface defects. Sensors, 10, 7067-7081.

16. Gonzalez, R.C., Woods, R.E. (2008). Digital Image Processing. 3rd Ed., Prentice Hall, New Jersey, USA. 17. Nasira, G.M., Banumathi, P. (2013). Fourier transform and image processing in automated fabric defect inspection system. International Journal of Computational Intelligence and Informatics, 3(1), 61-64.

18. Tsai, D.M., Hsiao, B. (2001). Automatic surface inspection using wavelet reconstruction. Pattern Rec-

342

ognition, 34, 1285-1305.

19. Lin, H.D. (2007). Automated visual inspection of ripple defects using wavelet characteristic based multivariate statistical approach. Image and Vision Computing, 25, 1785-1801.

20. Li, T.S. (2009). Applying wavelets transform and support vector machine for copper clad laminate defects classification. Computers & Industrial Engineering, 56, 1154-1168.

21. Chang, H.F. (2012). Design and implementation of real-time fabric defect detection system. Advances in information Sciences and Service Sciences, 4(21), 23-30.

22. Lin, H.D., Chiu, S.W. (2011). Flaw detection of domed surfaces in LED packages by machine vision system. Expert Systems with Applications, 38, 15208-15216.

23. Lu, C.J., Tsai, D.M. (2008). Independent component analysis-based defect detection in patterned liquid crystal display surfaces. Image and Vision Computing, 26, 955-970.

24. Chen, Y.C., Yu, J.H.,Xie, M.C.,Shiou, F.J. (2011).

Automated optical inspection system for analogical resistance type touch panel. International Journal of the Physical Sciences, 6(22), 5141-5152.

25. Lin, H.D., Tsai, H.H. (2012). Automated quality inspection of surface defects on touch panels. Journal of the Chinese Institute of Industrial Engineers, 29(5), 291-302.

26. Mallat, S.G. (1989). A theroyfor mulitiresolutionsignal decomposition the wavelet representation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 11(7), 674-693.

27. Shirazi, M.N., Noda, H., Takao, N. (2000). Texture classification based on Markov modeling in wavelet feature space. Image and Vision Computing, 18, 967-973.

28. Arivazhagan, S.,Ganesan, L. (2003). Texture segmentation using wavelet transform. Pattern recognition Letters, 24, 3197-3203.

29. Tsai, D.M., Chiang, C.H. (2003). Automated band selection for wavelet reconstruction in the application of defect detection. Image and Vision Computing, 21, 413-431.

30. Montgomery, D.C. (2009).Statistical Quality Control: A Modern Introduction,6th Edition, John Wiley & Sons, New York, NY, USA. 31. Nixon, M.S., Aguado, A.S. (2008). Feature Extraction and Image Processing.2nd Edition, Academic Press, Elsevier, Oxford, UK. Paper submitted: 19.03.2018.

Paper accepted: 04.07.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Journal of Applied Engineering Science

doi:10.5937/jaes16-15586

Original Scientific Paper

Paper number: 16(2018)3, 538, 343 - 348

AUTOMATED PROCESS CONTROL SYSTEM OF MOBILE CRUSHING AND SCREENING PLANT Andrey Ostroukh1*, Nataliya Surkova1, Oleg Varlamov1,2,3, Valery Chernenky3, Alexander Baldin3 1 Moscow Automobile and Road State Technical University (MADI), Moscow, Russian Federation 2 RI MIVAR, Moscow, Russian Federation 3 Bauman Moscow State Technical University (BMSTU), Moscow, Russian Federation The article research the problem of automating the mobile crushing and screening plant as a complex multi-level system. The analysis of the characteristics of the technological equipment and individual aggregates of the mobile crushing and screening plant is performed.A software for automated cone crusher control system, that allows to integrate a set of interconnected technological units in the control of one hardware, controller and human-machine interface on the touch panel in the operator station.Proposed human-machine interface based on a set of technical, informational, mathematical and software tools for controlling the mobile cone crusher, which ensures efficient operation of the technological control object as a whole.In accordance with the proposed concept, a complex automated system should provide an optimal level of automation for information collection and processing to form control signals and transfer them without loss and distortion to actuators in order to achieve efficient operation of the technological line for the crushing and screening plant.

Key words: Automation processes, Automated process control system (APCS), Crushing and screening plant, Jaw crusher, Cone crusher, Mnemonic scheme, Human-Machine Interface (HMI) INTRODUCTION Modern crushing and screening production of stone materials is a complicated technological complex with a set of operations including the delivery of raw material, its purification and pre-sorting, crushing during several stages including sorting the required commodity fractions into the stages of crushing, storing and transporting to the final consumer. Crushing and screening technological processes are among of the most important processes in construction materials formulation being, on the contrary, not highly effective.

In order to increase crushing and screening processing efficiency of stone materials one needs to develop new multi-level automatic process control systems of crushing and screening plants; in the last few years these plants become more and more transportable, e.g. movable from one site to another.

Modern automatic process control systems incorporate various elements brought together in order to reach some goal using branched interlocking bonds [1 - 6].

Taking into account the evolving modern concepts of "Internet of Things" and "Industry 4.0", the increasing number of industrial devices, machines and units are equipped with modern automation systems developed with the use of modern industrial controllers. Industrial devices, machines and aggregates have a permanent connection to the global network with used wireless communication channels. Control is performed with used Human-Machine Interface (HMI) implementing in software on interactive touch panels (touchscreens).

In the near future, machines (or rather their software) will

make many of the traditional deterministic decisions that a human would otherwise make, and artificial intelligence applications will facilitate even more abstract decisions through machine learning and deep-learning algorithms.

MIVAR computations as a scientific field began developing in the 80â&#x20AC;&#x2122;s years of the 20th century [17 - 19]. Developers faced a challenge of reducing the requirements for computational systems by using more sophisticated mathematical models. Three pillars of MIVAR technologies are innovative multi-dimensional databases, a unique MIVAR algorithm of information processing and a conceptual cognitive model of objects description and their relations.MIVAR-based technologies comprise the basis for evolutionary leap in robotic systems development: ranging from mechanical manipulators and simple mechanisms to reasoning systems and their social cooperation.

MOBILE CRUSHING AND SCREENING PLANT UNITS Crushing and screening plant processing kit is a branched continuous handling system consisting, for instance, of clatters, crushing station and several ribbon conveyors; the more complicated version includes dozens of equipment units included into one technological process and functioning automatically [7 - 15].

The choice of the crushing and screening plant technological scheme depends on the type of raw materials being processed, its physical characteristics, quality requirements and the purpose of the finished product. In addition, the crushing and screening plant design project depends on the required ratio of cubed grain, the

* Moscow Automobile and Road State Technical University (MADI), Moscow, Russian Federation, ostroukh@mail.ru

343

Andrey Ostroukh, et al. - Automated process control system of mobile crushing and screening plant

planned capacity, as well as the willingness of the crushing and screening plant owner to bear a certain share of the cost on restoration of the working capacity of its working bodies.

Research object is the Metso NW Rapid wheel-mounted transportable crushing and screening plant (Figure 1).

Figure 1: Metso NW Rapid wheel-mounted crushing and screening plant design project The relative performance of a crushing and screening plant can be calculated by the following formula:

where Q0i - relative productivity of all ni crushing aggregates of the i-th type; θ(ρ) - performance loss function.

The function θ(ρ) determines the loss of productivity of the crushing and screening plant due to the downtime of the crushing equipment with the corresponding production ρ = λ/μ, estimated by the ratio of the average number of unoccupied crushers to their total number.

The expression for the performance loss function can be written as:

where Mp - average number of unoccupied crushers; πk(ρ) - set of stationary probabilities of the system state; ni - number of crushers i-th type.

The task of optimizing the crushing and screening plant for maximum performance is written as:

Analysis of this expression shows that obtaining the maximum performance of the crushing and screening plant, the chosen configuration, is possible provided that the minimum θ(ρ).

Obviously, in order to obtain the maximum relative performance, the global minimum of the loss function, on the set of values of the production load ρ should tend to unity. When solving engineering problems, efficiency evaluations are usually formulated in direct or in reverse formulation.

The main units of transportable crushing and screening plant which provide high quality of crushing products as well as high performance, is: • Metso NW106 wheel-mounted transportable assembly (fig.2) equipped with Nordberg C106 roll-jaw crusher for primary crushing of stone products; • Metso NW220GPD wheel-mounted transportable assembly equipped with Nordberg GP220 cone crusher (fig.3) and four-deck vibratory screen with double-angle of DS series for the final crushing of stone materials and their sorting into specified fractions.

Mobile crushing and screening plant with jaw crusher Jaw crushers are used for primary and secondary crushing of solid and abrasive rocks (granite, diabase, sandstone). Crushing is performed by two 'jaws' - slabs, one of which perform oscillatory motions.

θ(ρ) = 1 + Mp Coefficient of idle crushing and screening plant:

Figure 2: Metso NW106 mobile crushing and screening plant with Nordberg C106 jaw crusher

344

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Andrey Ostroukh, et al. - Automated process control system of mobile crushing and screening plant

The width of the receiving hole determines the maximum possible size of the pieces loaded into the crusher.

Pieces of the initial material being loaded move downwards by gravity and are crushed in jaws. Large initial pieces are crumbled and processed, down to sand and dust. Pieces which are smaller than the size of the exit slot bz come out of the crusher, and the material remaining between the cheeks in the grinding chamber continues to be processed. Due to the complexity of the drive structure, a simple swing of the jaw is replaced by a wave motion that facilitates the movement of the material along the crushing chamber towards the exit slit.

The technical performance of jaw crushers is determined by the formula: Qt = 60 Vnμ

(1)

where V - is the volume of material falling out through the discharge gap per revolution of the eccentric shaft, m3; n = 200 ... 250 min-1 - eccentric shaft rotational frequency (swing frequency of the movable jaw); μ ≈ 0,3…0,65 – coefficient incorporating the presence of voids between the stones being crushed.

The maximum size of the crushing product is determined by the formula:

Mobile crushing and screening plant with cone crusher Cone crushers are the most effective for medium and fine crushing of strong and highly durable rocks. They carry out crushing continuously due to the rotation of the eccentric conical rotor - the "crushing cone" inside the outer stationary cone.

Cone crushers use continuous rotation instead of oscillations; unlike jaw crushers the change in clearance is made not simultaneously along the entire width of the gap, but alternately by its variation along the length. The working clearance in the cone crusher varies not simultaneously along the entire width as in the jaw crusher, but continuously along a circle, helping to improve the quality of crushing. The input and output gaps in the cone crusher are in the form of concentric rings. The maximum and minimum size of the output slot is set by the adjusting device.

(2)

where Kp - coefficient incorporating the configuration of the crushing jaw (triangular corrugations Kp=0,8; trapezoidal Kp=0,7); Kg - coefficient incorporating material properties (touchstone Kg=1,1; granite Kg=1,0); t - step of corrugations, mm (for approximate calculations t = bz); bz - slit width, mm; h - height of corrugations, mm (h ≈ t / 2); The weighted average size of the crushed product (mean size) is determined by the following formula: zсв = kdbz,

(3)

where bz - output slit width, mm; kd - weighted average coefficient, which is 0.65 for jaw crushers with a width of the receiving opening B = 600 mm and less; with the width of the receiving hole of B = 900 mm and more kd=0,8.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Figure 3: Metso NW220GPD mobile crushing and screening plantwith Nordberg GP220 cone crusher The size of the cone crushers product depends mainly on the size of the output gap and the strength of the rock.

The weighted average size of the crushing product zsv for soft, medium and strong rocks in fractions of the output gap bz of the cone crusher is determined by the formula: zсв = Kpbz,

(4)

where Kp=2,2÷2,8 (for soft rocks), Kp=2,6÷3,6 (for intermediate rocks), Kp=3,0÷5,0 (for solid rocks).

The technical productivity of cone crushers is determined in m3/s by material batch volume per one crushing cycle: Qt = 0,45 π n bz l D

(5)

where n - is the frequency of cone rotation, c-1, bz is the width of the output slit, m; l - is the length of the parallel zone of the output gap, m; D - is the diameter of the crushing cone base, m.

345

Andrey Ostroukh, et al. - Automated process control system of mobile crushing and screening plant

THE POWER SUPPLY AND AUTOMATION SYSTEM The power supply and automation system (Figure 4) is built on a modular principle, which provides the possibility of manufacturing reconstruction or its modernization.

The control commands of the crusher drives, the size of the output gap, the feeding conveyor, the screen, the conveyor of the finished products are formed on the operator's automated workstation where the program is executed in accordance with the specified algorithm for optimal control of the crushing process.

the chosen technological process, the conditions of the crushing and screening plant operation (planned production of crushing products or production under conditions of random demand).

Communication functions can be performed by wired communication with high reception and transmission interference immunity based on fiber optics, and in some cases also by wireless communication coupled with modern network industrial interfaces.

TOUCH SCREEN AND CONE CRUSHER CONTROL PANEL Metso IC automated control system for crushing and sorting process of stone materials (fig. 5) is developed on the basis of the SCADA system [16] and provides a simple and safe use of the crushing plant. The system allows you to control the feeder, crusher and conveyors with a single touch screen.

Metso IC automation solutions are designed to meet the customer's needs and the requirements of the crushing plants for continuous productivity, safety and easy monitoring of the crusher's parameters. Customers can be assured of the proper crusher functioning and minimal downtime under all the operating conditions thanks to the optimized start, stop and workflow algorithms that form the basis of the Metso IC automation system.

Figure 4: Nordberg GP220 Cone Crusher Power and Automation System The local control during automation is provided by information channels with the necessary number of sensors controlling the relevant parameters.

In case of the crusher they are the sensors of the drive speed (frequency), the size of the output gap, the force (moment) of crushing and energy being consumed.

The process of crushing is controlled by changing the size of the output gap and the drive speed regardless of the type of crusher (by adjusting the speed of the crusher drive) [9 - 12].

Simultaneous control of the crusher output gap size and the speed of the drive can be defined as two-dimensional which achieve the required value of the crusher capacity along with ensuring the required fractional composition of the output product.

Controlling of the output gap size alone is the main way to obtain the output of the required fractional composition and it can be defined as one-dimensional. It is the one-dimensional way of controlling the process of fragmentation that is mainly the subject of research in this dissertation work. The usage of this control method does not impose any restrictions on the crusher performance, considering that the required fractional composition is the main indicator that ensures the crushing and screening plant effectiveness.

The choice of the method for controlling the crushing process also depends on the crusher design features,

346

Figure 5: Touch screen and control panel of the Metso IC cone crusher Metso IC automation system allows you to optimize the crushing equipment performance. This is achieved through clearly defined safety control parameters such as oil temperature and pressure and drive power, which are indicators of the actual working crusher load. The Metso IC automation system monitors the crusher's condition and identifies faults and problems in advance.

Thus, it is possible to eliminate the problem without waiting for serious and costly consequences.

AUTOMATION SYSTEM COMPONENT INTEGRATION The upper level of crushing and screening plant automation is the automated system of crushing and screening production technological preparation for automation of all the production control processes.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Andrey Ostroukh, et al. - Automated process control system of mobile crushing and screening plant

The main task of the lower level automation is the organization of crushing process optimal control, the purpose of which is to provide the required fractional composition of the crushing product. The lower level automation task is the particular task of the general production automation task. At the same time, the lower level of crushing and screening plant automation can be considered as the automated control system for the technological process of crushing.

The Metso IC automation system can be easily connected to any general plant automated control system used for crushing and screening equipment (Figure 6).

Figure 6: Automation system component integration Such integration will provide the possibility of crusher centralized control as well as the whole enterprise.Also allow operator to safely monitor and regulate the crushing equipment operating parameters from one operator's console.

To increase the intelligence of the Automated Process Control System, the MIVAR technologies of creating logical artificial intelligence will be applied [16 - 19].

CONCLUSION In this way, the features of the functioning of the crushing and screening plant scheme elements have been determined due to the need to automate the crushing and sorting production.

Based on the design parameters of the crushers it is shown that the size of the crushing product, which is a certain weight average diameter of the crushed product grain is its main stochastic characteristic.

It is established that the task of the crushing process automatic regulation is to maintain the final product`s specified size and to maximize the energy supplied to the crushing aggregates by optimally loading the crushers.

The software for the automated process control system of crushing stone materials has been proposed, which makes it possible to integrate a set of interconnected technological units in the one hardware and technical complex management, the efficiency of their joint operation is largely determined by the conformance degree and the speed of the two-way information exchange.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

The developed system provided a significant operator`s comfort increase by providing him with a single user interface to manage various technological units from one operator station and an additional volume of information and services.

REFERENCES 1. Ostroukh, A.V. (2013). Systems of Artificial Intelligence in the Industry, the Robotics and the Transport Complex. Krasnoyarsk, RU: PublishingHouseScienceandInnovationCenter, Ltd. 2. Ostroukh, A.V. &Surkova N.E. (2015). Intelligence Information Systems and Technologies.Krasnoyarsk, RU: Publishing House Science and Innovation Center, Ltd. 3. Ostroukh, A.V. & Tian, Yu. (2014). Automated system for monitoring production-technological and organizational-economic activity of industrial enterprises. Instruments and Systems: Monitoring, Control, and Diagnostics, (3), 14-21.

4. Tian, Yu., Nguen, D.T., Chaudhary, R.R. &Ostroukh, A.V. (2014). Automated monitoring production technological and organizational - economic activities of an industrial enterprise. Automation and Control in Technical Systems, 3(1.2), 16-31. DOI: 10.12731/2306-1561-2014-1-16.

5. Ostroukh, A.V. & Tian, Yu. (2013). Integration of monitoring system components. Young Scientist, 6(10), 182-185.

6. Ostroukh, A.V. & Tian, Yu. (2014). Development of the information and analytical monitoring system of technological processes of the automobile industry enterprise. In the World of Scientific Discoveries, Series B,2(1), 92-102.

7. Ostroukh, A.V.,Gimadetdinov, M.K. & Popov, V.P. (2015). Selection Process Equipment for Automated Crushing Plant.Automation and Control in Technical Systems,4(2), 35-45. DOI: 10.12731/2306-15612015-2-4.

8. Gimadetdinov, M.K. &Ostroukh, A.V. (2014). List and Sequence of Solutions for Automated Crushing and Screening Production. Automation and Control in Technical Systems, 3(4), 55-61. DOI: 10.12731/2306-1561-2014-4-6.

9. Vorobeva, A.V., Ostroukh, A.V. Gimadetdinov, M.K., Wai Phyo Aung &Myo Lin Aung. (2015). Development of mathematical models and methods of optimal control of automated crushing and screening process.

IndustrialAutomaticControlSystemsandControllers, (1), 9-16.

10. Ostroukh, A.V.,Gimadetdinov, M.K., Borshch V.V.&Vorobeva, A.V. (2014). Development statistical modeling algorithm optimal control automated crushing and screening process. Industrial Automatic Control Systems and Controllers, (12), 3-10.

347

Andrey Ostroukh, et al. - Automated process control system of mobile crushing and screening plant

11. Gimadetdinov, M.K., Popov, V.P. &Ostroukh, A.V. (2015). Automation of Crushing and Screening Production with a Full Range of Technological Operations and Related Processes. Industrial Automatic Control Systems and Controllers, (9), 10-18.

12. Ostroukh, A.V., Vasiliev, Yu. E. &Kotlyarsky, E.V. (2015). Automated Control System Milling Unit for the Mineral Powders Plant. IndustrialAutomaticControlSystemsandControllers, (10), 3-10.

13. Ostroukh, A. V., Nedoseko, I. V., Surkova, N. E., &Bulatov, B. G. (2016). Automated Control System for the Milling Unit of Mineral Powders Plant. International Journal of Applied Engineering Research, 11(4), 2625-2628.

14. Ostroukh, A., Maksimychev, O., Nikolaev, A., Kolbasin, A., &Nedoseko, I. (2016). Development of automation of the drying and milling unit for the mineral powders plant. ARPN JournalofEngineeringandAppliedSciences, 11(9), 5717-5721.

15. Salniy, A.G., Kukharenko, V.N., Nikolaev, A.B.&Ostroukh, A.V. (2013). General Principles of SCADA Design. AutomationandControlinTechnicalSystems, 2(2), 8-12.

16. Varlamov, O.O. (2002). Evolutionary databases and knowledge for adaptive synthesis of intelligent systems. MIVAR Information Space. Moscow, RU: Radio and Communication, Ltd. 17. Varlamov, O.O. (2011). MIVAR: Transition from Productions to Bipartite Graphs MIVAR Nets and Practical Realization of Automated Constructor of Algorithms Handling More than Three Million Production Rules.https://arxiv.org/abs/1111.1321 18. Varlamov,O.O. (2015). Mivar Role and Place in Computer Sciences, Artificial Intelligence Systems and Informatics.Radio Industry,(3), 10-27.

19. Shadrin, S.S., Varlamov, O.O. & Ivanov, A.M. (2017).

Experimental Autonomous Road Vehicle with Logical Artificial Intelligence. Journal of Advanced Transportation. DOI:10.1155/2017/2492765.

Paper submitted: 05.11.2017.

Paper accepted: 14.07.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

348

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Original Scientific Paper

Journal of Applied Engineering Science

doi:10.5937/jaes16-17342

Paper number: 16(2018)3, 539, 349 - 357

IMPACT OF THE DAMPER CHARACTERISTICS ON THE BEHAVIOR OF SUSPENSION SYSTEM AND THE WHOLE VEHICLЕ Miloš Maljković*, Ivan Blagojević, Vladimir Popović, Dragan Stamenković University of Belgrade, Faculty of Mechanical Engineering, Belgrade, Serbia Damper asymmetry is a phenomenon for which there is no confirmed explanation in the literature, although it is present in almost all variants of telescopic dampers. This paper analyzes the influence of various characteristics of dampers, both symmetrical and asymmetrical, on the behavior of the suspension system and the whole vehicle.

Different tests have been carried out, and they simulate the different types of vehicle movement that are present in exploitation. The simulation was performed in CarSim 8 software package for the vehicle with chosen characteristics. Appropriate conclusions were drawn on the basis of the conducted research, i.e. the differences in the behavior of the suspension system and the vehicle depending on the selected characteristics of the damper are shown.

Key words: Damper, Asymmetrical damping, Vehicle behavior INTRODUCTION The damping ratio used in passenger vehicles is usually 40 to 60% higher in the case of bump than in the case of rebound damping, and very rarely can be completely symmetrical. Motorcycle dampers are characterized by even greater asymmetry. Asymmetry may vary depending on the operating conditions of the damper (compression or extension velocity). Sometimes, there is an expressive asymmetry in the dampers placed on the rear axle of passenger vehicles. This phenomenon is much less present with racing vehicles and motorcycles and one of the possible explanations is that greater bump damping, although it leads to reduced comfort, provides a better "sense of the road" to the driver.

Scientific research carried out so far has not yielded an acceptable explanation for the existence of asymmetry in dampers. The results of theoretical researches in terms of driving comfort request the dampers to be symmetrical, so this don’t provide a valid explanation.

It was found that the asymmetry of the damper is a direct consequence of its construction, i.e. that it is not caused by external influences.

One of the explanations of asymmetry states that the force of gravity which acts downwards on unsprung mass, requires greater damping in extension.

In a case where the wheel has no contact with the road, the gravity force is imposed as a significant influencing factor on the appearance of asymmetrical damping.

It is even stated that the cause of asymmetry is the greater presence of critical holes on the road than the bumps, which can be true, but it’s not proven. Even the opposite explanation is offered, which suggests that there are more large bumps so reducing the compression damping is necessary to absorb them, and therefore the reduction in extension damping is not necessary to overcome the large holes on the road.

The above claims can be justified only in the case of

high compression / extension velocities of the damper.

As an influential factor, in the case of previously stated claims, the ability of the tire to "fly" over short holes without the contact with the bottom of the hole is indicated, but a short bump must penetrate the tire profile fully. This argument is important only for small road irregularities.

The question arises - would we have completely symmetrical dampers if the roads would be without holes or bumps?

As a possible cause of asymmetry, it is further stated that passengers in the vehicle do not perceive the same way, in regard to comfort, the various abnormalities of the road.

It is also stated that for damper manufacturers it is more appropriate to have a certain asymmetry and that it is easier to achieve greater damping in the extension than in the compression. Also, there is a greater risk for the occurrence of cavitation of the working fluid in compression stroke [01] and some authors conducted experiments in which is shown that the thermal load of the working fluid has a great influence on damper characteristic [02].

As one of the possible causes of the damper asymmetry it’s stated that the damping needs to be higher in the case of extension, since when wheel come across bump, the damper, which is then in the compression stroke, is less involved in the energy absorption impact, this role is primarily taken by the elastic element (spring), all this in order to provide greater passenger comfort. The dissipation of accumulated energy, which takes place after the initial shock or compression of the damper, lasts longer due to the higher damping ratio [03]. The asymmetry of the damper, paired with the nonlinear kinematic response of the suspension system, can significantly change the dynamic response, in particular the return of the vehicle to an equilibrium position [04]. Some authors state that with an increase in asymmetry of the damper, there is also a greater variation of the wheel camber during the vehicle motion, with a non-linear relation with

* University of Belgrade, Faculty of Mechanical Engineering, Kraljice Marije 16, 11000 Belgrade 35, Serbia, mmaljkovic@mas.bg.ac.rs

349

Miloš Maljković, et al. - Impact of the damper characteristics on the behavior of suspension system and the whole vehicle

the travel of the suspension system [05]. Another conducted research on a real damper used in passenger car shows that real damper characteristic is asymmetric and nonlinear. It states that it is possible to find mean value of damping forces when using symmetric model of damper characteristic, but that it gives much lower values of damping force in extension and much higher values in compression than a real damper [06].

Therefore, the asymmetry of the damper should be considered by taking into account the following parameters: • road irregularities (bumps, holes, etc.); • different passenger sensitivity to vehicle movement upwards and downwards; • asymmetry of the damper as a design characteristic; • different tire behavior when crossing holes and bumps; • production costs [01].

This paper analyzes the behavior of a damper as an integral element of the suspension system, as well as the impact of its characteristics on the behavior of the suspension system itself and the whole vehicle. The analysis is focused on telescopic dampers.

RESEARCH ON THE TYPE OF DAMPING AND DAMPING RATIO OF THE DAMPER The goal of the study is to examine the influence of the selected characteristics of the damper, i.e. its damping ratio on the behavior of the suspension system and the whole vehicle while moving in different road conditions and in different characteristic maneuvers. This analysis was performed using different values of symmetrical and asymmetrical damping.

Another goal is to make appropriate conclusions on the behavior of the suspension system and the vehicle itself under the above mentioned conditions, i.e. the effect of damper characteristic change on handling, stability and comfort.

The effect of damping ratio change on the behavior of the suspension system and the vehicle was analyzed using the CarSim 8 software. For the needs of the simulation, a C segment vehicle is choosen with a sprung mass of 1274 kg and unsprung mass of 142 kg. The simulation of each of the selected tests is performed for four cases, each with a different characteristic of the damper.

The change in damper characteristic was made by changing the shape of the curve that shows damping force versus the velocity of the damper piston. The shape of the curve should be as smooth as possible to eliminate problems in exploitation and unnatural sensation to the driver.

The shape of the curve used in the first case is shown in Figure 1 and represents an initial asymmetrical damping.

Based on the diagram shown, it can be concluded that the bigger damping is present in the extraction (bottom left quadrant) than in the damper compression (upper

350

Figure 1: Initial asymmetrical damping right quadrant). This is in favor of the previously stated fact that the asymmetry is often present, and the characteristic used in this case is a basic setting for the subject vehicle, given by the software.

For the second case, the symmetrical characteristic of the damper is chosen, so that the extreme damping values of compression and extraction force correspond to the extreme values of damping force in compression stroke, as showed in Figure 1.

For the third case, the symmetrical characteristic of the damper was also chosen, but with the extreme values of the compression and extraction corresponding to the extreme values of damping force in extraction stroke, as showed in Figure 1.

For the fourth case, the reverse asymmetrical characteristic of the damper was chosen, in which the damping is bigger in compression. This case is only theoretical, because it is not usually present in exploitation, but it is necessary to analyze it.

In all four cases, a different curve slope is present for low and high piston velocity. Аt lower piston rates, damping force increases more rapidly compared to higher rates.

The first is to improve vehicle stability, and the second to better absorb the road irregularities.

Five different tests were performed with each of the damper settings described above.

Test 1 The first conducted test is a comfort test, and it is derived in two parts. In the first part of the test, the vehicle moves on a wavy road section with initially lower frequency, which increases later. The vehicle moves at a constant Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

MiloĹĄ MaljkoviÄ&#x2021;, et al. - Impact of the damper characteristics on the behavior of suspension system and the whole vehicle

speed of 40 km/h. The road profile is shown in Figure 2.

In the second part of the test, the vehicle moves along a cross-slope section, with a constant speed of 20 km/h.

The DLC (Double Lane Change) test was selected for the third test. The vehicle moves at a constant speed of 120 km/h. The nature of vehicle movement during this test, as well as the required dimensions of the test track are shown in Figure 5.

Test 4

Figure 2: Road profile of the selected road section

Figure 5: Double Lane Change test

Figure 3 shows the road profile for this part.

Test 2

In this test, the vehicle is crossing over a short transverse road bump is chosen as the fourth test. The vehicle moves at a constant speed (first at 40 and later at 80 km/h) and crosses the transverse bump that is 35 mm high and 400 mm long.

Test 5 For the fifth test, the vehicle moves at a constant speed of 50 km/h along an uneven 1200 m long road with small irregularities. The profile of the selected road section is given in Figure 6.

Figure 3: Profile of the cross-slope road section For the second test, the fishhook test is chosen and the nature of vehicle motion during the test is shown in Figure 4.

Test 3

Figure 6: Road profile THE ANALYSIS OF THE RESULTS Test 1 - first part Figure 7 shows the vertical acceleration of the vehicle's center of gravity for each of the four selected damper characteristics.

Figure 4: Fishhook test Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

351

Miloš Maljković, et al. - Impact of the damper characteristics on the behavior of suspension system and the whole vehicle

Figure 7: Vertical acceleration of the vehicle’s CG In the first part, where the frequency of oscillations is the lowest, most favorable vertical acceleration of the center of gravity of the sprung mass is present in the third case, when there is a higher value of the symmetrical damping.

The initial asymmetrical damping and lower symmetrical damping provide approximately the same change in vertical acceleration of the vehicle's center of gravity, with somewhat less extreme values achieved (initial asymmetrical damping).

Reverse asymmetrical damping ensures a change in vertical acceleration of a vehicle whose extreme values correspond to asymmetrical damping, but in the opposite direction. With reverse asymmetrical damping, there are certain unevennesses in the change of vertical acceleration, which are not present in other cases.

Observing the second part, where the frequency of the oscillations is increased, the most favorable vertical acceleration of the vehicle is present in the second case, hence the lower value of the symmetrical damping here gives the best results. The initial asymmetrical damping leads to somewhat higher vertical acceleration, while higher symmetrical damping value causes a significantly higher vertical acceleration.

Reverse asymmetrical damping provides maximum acceleration values for the vehicle's center of gravity between the values obtained by lower symmetrical and initial asymmetrical damping.

By analyzing the third part, where the frequency of oscillations is the highest, most favorable vertical acceleration of the center of gravity of the sprung mass is also present in the second case, therefore, the lower value of the symmetrical damping gives the most favorable results. The initial asymmetrical damping again results in

352

slightly higher vertical acceleration, while vertical acceleration achieved by the reversed asymmetrical damping is smaller. A higher symmetrical damping value causes a significantly increased vertical acceleration of the vehicle's center of gravity. At a higher value of symmetrical damping, a smaller oscillation amplitude is present in the third part relative to the second part, while in the first two cases it is the opposite.

In the case of reversed asymmetrical damping, the extreme values of vertical acceleration of the vehicle are almost identical in the second and third part of the chosen road section. А conclusion can be drawn that the lower symmetrical damping value gives the best results for such conditions of vehicle movement, which does not necessarily mean that the vehicle with this type of damper will provide the most favorable behavior in the various maneuvers and driving conditions.

It can be noticed that in the case of asymmetrical damping, the damper predominantly moves in the compression stroke, while movement in the extension part is present only on several occasions when the vehicle moves on a road section with a lower frequency of “waves”. A totally opposite situation of the above is present in the case of a damper with a reversed asymmetrical damping characteristic.

In other cases this movement is almost symmetrical throughout the test. The highest value of the piston movement is achieved by the compression of the damper with asymmetrical damping, and the lowest in the extension. It can be concluded that a higher damping value results in less movement of the damper piston.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Miloš Maljković, et al. - Impact of the damper characteristics on the behavior of suspension system and the whole vehicle

On the basis of conducted research, it can be concluded that with symmetrical damping, the displacement of suspension system is symmetrical upwards and downwards, while initial asymmetrical damping increases the upward movement. When using the reverse asymmetrical characteristic of the damper, the downward movement of the suspension system is increased. The total travel of the suspension system is smallest when using both asymmetrical characteristics of the damper, which can lead to a reduction in tire and suspension system elements wear.

Test 1 - second part Figure 8 shows vertical acceleration of the center of gravity of the vehicle. By looking at diagrams shown, it is noted that most unfavorable results for the comfort of the passengers (the highest values of the vertical acceleration of the vehicle) are achieved with a damper with an asymmetrical damping characteristic. The initial asymmetrical damping yields less favorable results than the reversed asymmetrical characteristic. The most favorable results are achieved by selecting a damper with a lower symmetrical damping value, while a higher symmetrical damping value provides slightly higher values of vertical acceleration of the vehicle's center of gravity.

On the basis of the obtained diagrams, it can be concluded that the damping characteristics do not have a significant role in reducing the roll of the vehicle, since the results achieved in all four cases are almost identical.

Figure 9 shows the pitch of the vehicle during the test.

It can be noticed that the change in vehicle pitch with two symmetrical damping characteristics is very similar, as opposed to asymmetrical damping. Obtained extreme

value of pitch is almost the same in first three cases, while maximum value achieved with the last case is a little bit higher. In the case of initial asymmetrical damping, the pitch of the vehicle changes the direction, whereas in the case of symmetrical damping this is not the case.

Reverse asymmetrical damping causes an additional increase in pitch, especially in the final stage of the test.

Test 2 On the basis of the obtained results, it can be seen that maximum values of the achieved yaw rate of the vehicle in all three cases range from 25 to 30 °/s.

At the beginning of the test there was a slightly higher yaw rate of the vehicle, when dampers with a lower symmetrical damping value are selected, while maximum yaw rate, after changing the direction of movement, is provided by dampers with a higher value of symmetrical and reverse asymmetrical damping. The difference in the nature of the change of the yaw rate is present, which shows that the choice of the damper characteristics and the maximum damping ratio is certainly important for a sudden change of the direction (performed in the first part of the test), while in the second part of the fishhook test, when the vehicle is moving in a constant radius, there are almost no differences in the behavior of the vehicle.

Figure 8: Vertical acceleration of the vehicle’s CG Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

353

Miloš Maljković, et al. - Impact of the damper characteristics on the behavior of suspension system and the whole vehicle

It can be seen that the maximum value of lateral force in contact between the tire and the road in the case of symmetrical and asymmetrical characteristics of the damper is same in the first part of the performed fishhook test, but there is a difference in the way this force changes.

In the second part of the fishhook test, when a second direction change takes place and vehicle continues to move without sudden changes in direction, the lowest value of lateral force is achieved when using a damper with a higher symmetrical damping value, while the highest value is achieved by the dampers with initial asymmetrical damping.

Dampers with a lower symmetrical damping value cause lateral forces whose intensity is between the afore mentioned damping characteristics.

The lateral force change over time has the most pronounced jumps or unevennesses in the fourth case, while the lower value of the symmetrical damping also leads to the appearance of certain unevennesses which are less pronounced than in the afore mentioned case.

Test 3 In all four cases, when an identical lateral deviation of the vehicle from the desired path was obtained and because of that it is not shown. This tells us that the damping ratio of the damper is not of great importance when the vehicle is found in such driving conditions, but primarily comes the stiffness of the springs as well as the geometric and mass parameters of the vehicle.

Теst 4 Figure 10 shows the vertical acceleration of the vehicle when the vehicle passes the transversal bump at a speed of 40 km/h.

It can be seen from the Figure 10 that extreme values achieved here in all four cases are almost identical, while the change in the vertical acceleration of the vehicle itself has certain variations depending on the selected damper characteristic. We can conclude that in all four cases the comfort is approximately at the same level, regardless of the chosen damper characteristic.

The extreme value of the longitudinal force is approximately the same in the case of initial asymmetrical damping and lower symmetrical damping whereas it is significantly lower at higher values of symmetrical damping. The lowest value of the longitudinal force is present in the reverse asymmetrical damping which can be considered the most favorable case. The change in the vertical force is approximately equal in all four cases, with identical extreme values achieved.

Figure 11 shows the vertical acceleration of the vehicle when the vehicle passes the transversal bump at a speed of 80 km/h.

Based on Figure 11, it is again possible to conclude that the extreme values achieved here in all four cases are almost identical, while the change in vertical acceleration of the vehicle itself also has certain variations depending on the selected damping characteristic. Again, we can conclude that in all four cases, the impact on the passengers, i.e. the achieved comfort, is approximately at the same level, regardless of the selected damping characteristic.

Figure 9: Vehicle pitch

354

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

MiloĹĄ MaljkoviÄ&#x2021;, et al. - Impact of the damper characteristics on the behavior of suspension system and the whole vehicle

By analyzing the achieved results of the research for different characteristics of the damper, it is possible to derive conclusions, so for the purpose of comparative analysis, the results achieved for each of the tests are evaluated and shown in the table.

The scores for each of the criteria considered during the performed tests are given in the range of 1 to 4 points (the best result achieved being 4, and the worst one being 1 point) for each of the selected dampers and are shown in Table 1. For tests performed at two vehicle speeds, the results for each of the considered criteria are given in the following format - the number of points achieved at the first speed value / number of points achieved at the second speed value, while the total score for each of the criteria is shown as the average number of points and is given in parentheses.

Corresponding weight factors were selected for each of the tests and the criteria considered. On the basis of the stated weight factors, the significance of each test, as well as the significance of each of the criteria within the selected tests is defined.

By assigning the appropriate points to each of the considered damper characteristics for each of the performed tests, almost equal results are obtained.

By analyzing the first part of test 1, it is concluded that when moving along the wavy road, the best results are obtained with the reverse asymmetrical damping which, as previously mentioned, is not present in the exploitation. This damping characteristic provides even better results in terms of vertical acceleration and pitch.

The lower symmetrical damping value gives slightly worse overall results for the subject test.

Although it provides the best results in terms of vertical acceleration, it is proving to be the worst possible solu-

tion for the vehicle pitch.

The initial asymmetrical damping ensures uniform behavior in terms of vertical acceleration and pitch of the vehicle, but in the overall score for the considered test it is worse than the lower symmetrical damping.

A higher symmetrical damping value gives the most unfavorable vertical acceleration while obtaining low pitch values. The achieved overall result for the test is at the level of initial asymmetrical damping.

In the second part of test 1, with the movement of the vehicle on the road with a cross-slope, the behavior of a vehicle with symmetrical damping characteristics is noticeably more favorable than a vehicle with asymmetrical dampers.

Both symmetrical damping values provide very good results in terms of reducing the vehicle roll. A lower symmetrical damping value provides a more favorable vertical acceleration of the vehicle's center of gravity, while higher symmetrical damping value provides a more favorable mitigation of the pitch.

By comparing asymmetrical characteristics, identical conclusions arise as well as in the case of symmetrical characteristics of the dampers, but it should be noted that the overall results achieved are less favorable.

In fishhook test, the vehicle achieved identical results for each of the selected characteristics, but there are various changes in lateral and vertical forces in the contact of the tire and the road. Also, the obtained maximum values of these forces are not the same. This leads to different tire deformations which certainly have a significant impact on its behavior. Bearing in mind that the tire, primarily itâ&#x20AC;&#x2122;s elastic characteristics, participates in the operation of the suspension system, the study of these effects requires additional attention in some future research.

Figure 10: Vertical acceleration of the vehicle Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

355

MiloĹĄ MaljkoviÄ&#x2021;, et al. - Impact of the damper characteristics on the behavior of suspension system and the whole vehicle

Test 3 (DLC test) shows that vehicle handling is at the same level with all selected damper characteristics. Due to the low impact of the damper on the steering behavior of the vehicle and the absence of differences in the parameters that depend on the behavior of the suspension system (vertical acceleration of the vehicle, pitch, etc.), this test is assigned a weight factor of 0.5.

When vehicle crosses a transversal bump (test 4), the higher symmetrical and reversed asymmetrical damping give completely opposite results with two different speeds of vehicle. Higher symmetrical damping value is shown to be a more unfavorable solution at lower speeds, while at higher speeds it is shown to be the most favorable. The completely opposite conclusion is made for reversed asymmetrical damping. The initial asymmetrical damping and lower symmetrical damping value show solid results at each of the selected speeds.

The overall results for this test show that each of the characteristics has been shown to be almost the same, but when selecting the optimum characteristic it should be kept in mind that the vehicle's favorable behavior is ensured in the broader range of speed and road characteristics (bumps, holes, etc).

Differences in the achieved amplitude of the vertical acceleration, as well as the achieved frequency when the vehicle is moving through a road with the small unevennesses, are barely noticeable.

characteristic for the vehicle under consideration will not lead to significant disturbance of the comfort or stability of the vehicle.

Reverse asymmetrical damping characteristic surpasses all the expectations, since such solution is not present in practice. In some segments it turned out even better than other configurations.

In the future, there is also a need to analyze cases with a higher mass of vehicle and to further investigate the allegations [03] that asymmetry can only be present on one of the axles, and that in this way the removal of vehicle vibrations is ensured and that it is almost at the same level as for vehicles using asymmetrical dampers on both axles.

In order to fully understand the phenomenon of asymmetry, it is necessary to examine in detail the construction of the damper itself, as well as the influence of the choice of the damping characteristics on the costs of development and production of dampers.

In practice, there are semi-active suspension systems that have the ability to change the damper characteristic in order to achieve a high level of comfort while maintaining the stability of the vehicle under different driving conditions. Тhe damper characteristic can be changed by adjusting the viscosity of magnetorheological or electrorheological fluid inside the damper, which provides a higher or lower value of damping ratio.

CONCLUSION

ACKNOWLEDGEMENT

Summing up all the results leads to the conclusion that there are no dramatic differences in the behavior of the vehicle depending on the choice of the damper characteristic. It can be said that the obtained results are fairly uniform and that the application of any type of damping

This paper presents some of the results obtained through the project supported by Serbian Ministry of Education, Science and Technological Development (TR 35045 "Scientific-Technological Support to Enhancing the Safety of Special Road and Rail Vehicles").

Figure 11: Vertical acceleration of the vehicle

356

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Miloš Maljković, et al. - Impact of the damper characteristics on the behavior of suspension system and the whole vehicle

Table 1: Evaluation of the results of the survey Weight factor of the test

Criterion

Initial Weight factor asymmetrical of the criterion damping

Lower symmetrical damping

Higher symmetrical damping

Reverse asymmetrical damping

Test 1 - first part

1

Vertical acceleration

0.5

2

4

1

3

Pitch

0.5

2

1

3

4

Total

1

2

2.5

2

3.5

Test 1 - second part

1

Vertical acceleration

0.33

1

4

3

2

Roll

0.33

4

4

4

4

Pitch

0.33

2

3

4

1

Total

1

2.33

3.67

3.67

2.33

Test 2 1

Stability

1

4

4

4

4

Total

1

4

4

4

4

Test 3 0.5

Lateral offset from designed path

1

4

4

4

4

Total

1

4

4

4

4

Test 4 1

Vertical acceleration

1

2 / 3 (2.5)

3 / 2 (2.5)

1 / 4 (2.5)

4 / 1 (2.5)

Total

1

2.5

2.5

2.5

2.5

Test 5 1

Vertical acceleration

1

4

4

4

4

Total

1

4

4

4

4

18.83

20.67

20.17

20.33

TOTAL POINTS REFERENCES

1. Dixon, J.C. (2007). Тhe shock absorber handbook (2nd ed). John Wiley and Sons, Chichester.

2. Demić, M., Diligenski, Đ., Milovanović, M. (2013). A contribution to research of the influence of degradation of vehicle vibration parameters on thermal load of shock absorbers. Journal of Applied Engineering Science, vol. 11, no. 1, 23-30.

3. Silveira, M., Pontes Jr., B.R., Balthazar, J.M. (2014).

Use of nonlinear asymmetrical shock absorber to improve comfort on passenger vehicles. Journal of Sound and Vibration, vol. 333, no. 7, 2114-2129, DOI:10.1016/j.jsv.2013.12.001.

4. Balike, K.P. (2010). Kineto-dynamic analyses of vehicle suspension for optimal synthesis (PhD thesis).

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Concordia University, Montreal.

5. Balike, K.P., Rakheja, S., Stiharu, I. (2010). Оptimization of asymmetric damper parameters of an automotive suspension for minimal camber angle variations. 12th International Conference on Advanced Vehicle and Tire Technologies, p. 57-64.

6. Ślaski, G. (2011). Damping parameters of suspension of passenger vehicle equipped with semi-active dampers with by-pass valve. Transport Problems, vol. 6, no. 2, 35-42.

Paper submitted: 04.05.2018.

Paper accepted: 25.07.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

357

Original Scientific Paper

doi:10.5937/jaes16-17164

Journal of Applied Engineering Science

Paper number: 16(2018)3, 540, 358 - 367

EXPERIMENTAL AND ANALYTICAL STUDY OF THE INTER-STOREY HOLD-DOWN CONNECTIONS IN TIMBER-FRAME PANEL BUILDINGS Katja Vogrinec1*, Miroslav Premrov2 1 Masiv DOO, Gornja Radgona, Slovenia 2 University of Maribor, Faculty of Civil Engineering, Transportation Engineering and Architecture, Maribor, Slovenia Timber-frame panel buildings have a very specific composition where the main challenge represent mechanical fasteners, which are unable to provide a fully rigid connection. The stiffness of the timber-framed walls is thus largely dependent on various factors that influences its stiffness, such as the bending and the shear flexibility of the composite wall element, the flexibility of the fasteners between the timber frame and the sheathing board along with the flexibility of the tensile and compressive support. Despite the fact that these contributions to the stiffness of the timber-framed walls are not negligible, they are not considered in Eurocode 5 standard for design of timber structures.

The current paper analyses the contribution of the tensile support and presents the experimental and analytical study of inter-storey hold-down connections in timber-framed panel construction system. Experimental tests are performed for two different hold-down connections appropriate for connecting timber-framed walls from the upper floor through the ceiling to the timber-framed-walls of the lower floor. Experimental results show that hold-down connections do not provide a rigid support conditions for the timber-framed walls and that their flexibility should be taken into account when calculating the overall horizontal stiffness of the timber-framed walls. Therefore, an analytical expression for determination of the stiffness of the hold-down is suggested for the hold-down connection with perforated strap. The formula can be used for analytical calculation of the stiffness of the timber-framed walls by taking into account the stiffness of the tensile support when a tested hold-down anchor is used.

Key words: Timber-frame panel structures, Timber-framed walls, Hold-down connections, Perforated strap, Angle brackets, Experimental study, Analytical expression, Stiffness INTRODUCTION Timber-frame panel construction system is one of the most widely used building system in timber construction in Europe. It originates from the American balloon-frame and platform-frame construction systems whose advantage is provedin factory prefabrication of the building elements assuring the so called ''ideal weather conditions'' where all building elements and details are carefully designed and precisely manufactured in the factory in accordance with work plans allowing contacts to align perfectly.

Due to its specific composition, timber-frame panel buildingsare very difficult to model. The main challenges are mechanical fasteners, which are unable to provide a fully rigid connection. The problem occurs when non-regular structures according to Eurocode 8 [1] appear, requiring more accurate analysis for lateral force distribution, as generation of the 3D model and use of a modal analysis.

An example of a simple mathematical model appropriate for 3D-modelling of timber-framed wallsusing a braced frame with one fictive diagonal is presented in Pintarič and Premrov [2] and an upgradeof the developed model in Vogrinec, Premrov and KozemŠilih [3]. The model employs a simple analytical calculation of the stiffness of the timber-framed wall embedded in the calculation of the fictive diagonal diameter. On the other hand, the presented model is simultaneously able to encompass

the influence of the stiffness of the sheathing material, the influence of the flexibility of the fasteners in the connecting plane between the timber frame and the sheathing board, the influence of flexibility of the tensile and compressive support and the influence of the door and window openings. Different analytical formulas for calculation of the stiffness of the timber-framed walls with the same analogy with summing deformations of the sheathing board, mechanical fasteners, timber frame and tensile and compressive support can also be found in Kessel [4], Casagrande et. al [5, 6], Sartori [7] and Hoekstra [8].Although the importance of including different contributions to the overall deformation of the timber-framed wall is addressed by different authors,the stiffness calculation of the timber-framed walls and the influence of the openings is not considered in the standard Eurocode 5 [9]. This article focuses on one of the contributions that influences the stiffness of the timber-framed wall and the building itself: the influence of the flexibility of the inter-storey hold-down connections.Even though the problem is discussed by different authors,only few experimental data and appropriate analytical formulas for the stiffness calculation of the hold-down anchor are available in the literature. Hence, the presented analysis deals with the first experimental test of inter-storey holddown connections which results in an analytical expression for one type of the connection.

* Masiv DOO, Spodnja Ščavnica 78a, 9250 Gornja Radgona, Slovenia, katja.vogrinec@outlook.com

358

Katja Vogrinec, et al. - Experimental and analytical study of the inter-storey hold-down connections in timber-frame panel buildings

TIMBER-FRAMED WALLS ANCHORING The required stability and load-bearing capacity of the timber-frame panel buildings in vertical and horizontal direction is attained with timber-framed walls, roof or floor diaphragms and the appropriate anchorage of the timber-framed wall elements. In fact, as written in Prion and Lam [10], horizontal roof and floor diaphragms are designed to distribute lateral loads to shear walls, which in turn carry this loads to the foundation. In timber-frame panel structures, the vertical bracing elements for carrying horizontal forces to the foundations are timber-framed walls. Their elements consist of a timber frame and single-sided or two-sided sheathing boards, attached to the timber frame with mechanical fasteners (usually staples). They can be produced as single-panel systems (Figure 1a) or as recently more commonly used macro-panel systems (Figure 1b). The timber frame is constituted by three timber studs and two timber plates (top and bottom) with the thermal insulation placed between timber elements as can be seen in Figure 1a.

Additionally, different types of boards can be used as sheathing material. In Europe, the most commonly used types of boards are fibre-plaster boards (FPB) and oriented strand boards (OSB). Besides the fire safety function (fibre-plaster boards), their primary purpose is to ensure the horizontal stability of the building if the appropriate fastening to the timber frame is made. Their load-bearing capacity differs in the type of a used sheathing board.

Premrov and Dobrila [11] present a numerical study,analysing the influence of OSB and fibre-plaster boards (FPB) on the racking resistance of timber-framed wall elements. The results of the developed semi-analytical model evidently demonstrate higher racking stiffness of the wall elements with FPB, while the racking resistance proves to be evidently higher in the case of using OSB boards. The authors indicate an important dilemma of using the best sheathing board with regard to the height and location of the building and recommend using OSB sheathing boards instead of fibre-plaster boards in extremely windy or seismic areas.

Figure 1: a) Composition of a single-panel wall element b) anchoring of the macro-panel timber-framed wall with shear and hold-down anchors to the foundations The horizontal load transfers from upper to lower floor and to foundations, and the prevention fromwall overturning is achieved with appropriate anchoring of the wall to the ground. These hold-down forces, needed to maintain equilibrium, can be provided either through vertical Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

loads from upper storeys or the roof, or they can come from a hold-down device attached to the vertical end stud (Prion and Lam [10]). In general, two different types of anchors are used for anchoring such a timber-framed wall. Shear anchors are used for shear forces appearing at the bottom of the wall and for sliding prevention, as well as for tension or hold-down forces and for the prevention of uplift of the wall hold-down anchors, as shown in Figure 1b.

Faherty and Williamson [12] claim that static equilibrium of the timber-framed wall requires the wall having a tension anchorage at the uplifting end. In practice, such an anchorage is needed at each end of the wall, since horizontal load can be imposed in either direction of the wall.

As can be seen in Figure 2,tensile and shear forces are transmitted through the anchors, caused by the horizontal forces on the wall elements.

Figure 2: Load transfer between roof/floor diaphragms and wall elements An anchor is usually a steel element (angle bracket, tape or plate) fastened with mechanical fasteners (typically nails) to the timber frame and with a screw to the concrete slab or on the ceiling or through the ceiling connection to the wall below. According to the purpose of use, we distinguish (i) anchors for anchoring the wall to the concrete slab and (ii) anchors for anchoring the wallsfromthe upper floor to the walls of the lower floor or to the ceiling.

At the same time, we distinguish (a) hold-down or tensile anchorsused to transmit tensile forces and (b) shear anchors used to transmit shear forces. In praxis, different types of wall anchoring can appear. Some examples of European types of anchoring are shown in Figure 3 for anchors used foranchoring the wall to the concrete slab and in Figure 4 for anchors used foranchoring the walls from the upper floor to the walls of the lower floor or to the ceiling. Some other types of anchoring used in the world, such as anchoring through the bottom plate of the wall, are not appropriate for European timber-frame panel structures if factory prefabrication of the elements is used. This type of anchoring requires thewall to stay opened until the anchoring is done at the building site.

359

Katja Vogrinec, et al. - Experimental and analytical study of the inter-storey hold-down connections in timber-frame panel buildings

Figure 3: Examples of timber-framed walls anchoring to the concrete slab with a) angle brackets for tensile loads, b) straps for tensile loads and plates for shear loads c) angle brackets for shear loads (©rothoblaas)

Figure 4: Examples of timber-framed walls anchoring to the ceiling and to the lower floor with a) angle brackets for shear forces, b) angle brackets for tensile forces, c) a perforated strap for tensile forces (©rothoblaas) As can be seen above, anchoring of the wall elementsplays an important role when designing timber-framed walls. Hereafter we will focus on the influence of the hold-down anchors on the behaviour of timber-framed walls. Besides their role of preventing uplift of the wall and ensuring the equilibrium,their flexibility has an important impact on the overall horizontal stiffness of the timber-framed wall. The reason lies in the fact that hold-down anchors do not provide totally rigid connection of the timber-framed wall with the ground and therefore some additional displacement of the wall appears on account of the flexibility of the hold-down anchors. Regarding typical wall configuration, Casagrande et al. [5] determine the average percentage of deformation due to each single contribution (sheathing-to-framing connection, rigid-body translation, rigid-body rocking, sheathing panel) for timber-framed and CLT walls. Obviously, timber-framed wall deformation is mostly made by the sheathing-to-framing connection and the hold-down connections.

Different types of connection systems for shear and hold-down anchoring of the timber-framed wallsto the foundations are experimentally investigated in order to derive their stiffness and strength parameters in Tomasi and Sartori [13]. Authors tested different shear and holddown angle brackets and suggested several improved shear anchors. Moreover, different hold-down angle brackets appropriate for the anchoring on concrete plate were investigatedfor hold-down anchors. Almost all of them showed similar stiffness values and only one type of anchor showed three times higher values of stiffness.

360

The authors also pointed to the fact that Eurocode 5 [9] does not provide a suitable formula for calculating the strength and stiffness of the hold-down anchor. Besides the deformation resulting from nailed connection, authors suggested to include two further contributions when determining the stiffness of the hold-down anchor: local deformation of the base steel plate and the elongation of the vertical steel flange. Authors also showed that in the case of OSB plates, where the anchors are not directly attached to the timber stud (as in case of fibre-plaster sheathing boards) but through the sheathing board, the interposition of the OSB does not affect the behaviour of the connections and the OSB plate can be ignored when calculating the bearing capacity of a hold-down anchor.

Explanations forthe anchoring of thetimber-framed walls, including the installation of the anchors, and experimental studies are also available in Kessel and Polatschek [14].They experimentally tested perforated steel plates with different dimensions and number of nails which can be used for anchoring the timber-framed wall to the ground, and proposed an analytical expression for the stiffness of the tested hold-down anchor. The stiffness of the anchor in vertical direction Kt is expressed as the sum of the stiffness of the nails in the form of: Kt = na * k

(1)

where na is the number of fasteners in the steel-to-timber connection and K is the slip modulus per shear plane per fastener. A simple analytical expression supported by a numerical example for hold-down anchor suitable for anchoring to the foundations is also proposed in zur Kammer [15], where the deformation of the tensile anchor is simply calculated as the sum of deformations resulting from: • deformation of the fasteners between the tensile anchor and the timber stud, • elongation of the anchor, • flexure of the bottom plate of the anchor, • elongation of the bolt anchor (where the connection between the bolt anchor and the concreteis not taken into account).

An analytical expression for hold-down anchors suitable for anchoring to the foundation is also discussed in Hoekstra [8], where the calculation method takes into account the stiffness of the steel and timber section, the stiffness of the fasteners, hole-clearance and the actual load level. The derived analytical expression is compared with experimental test available in the literature.

Consequently, the experimental studies and the analytical expressions are performed and derived for the hold-down anchors appropriate for the anchoring to the foundations. On the other hand, only few data can be found for the inter-storey hold-down connections where the timber-framed wall from upper floor is anchored through the ceiling to timber-framed wall in the lower floor (Figure 4b, c). According to zur Kammer [15] we can Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Katja Vogrinec, et al. - Experimental and analytical study of the inter-storey hold-down connections in timber-frame panel buildings

assume that the stiffness of the inter-storey connection with two angle brackets is half the size of the stiffness of the same angle bracket type used for anchoring to the concrete slab. Different hold-down anchoring details for connection of the timber-framed walls from the upper floor through the ceiling to the lower floor can be found for example in Kolb [16] and Holzrahmenbau [17]. An alternative approach for anchoring walls in multi-storey timber-framed buildings is discussed by Nelson, Patel and Arevalo [18],presenting a generic review of design issues of continuous tie-down systems for wood panel shear walls in multi-storey timber structures. This continuous tie-down system provides in comparison with traditional inter-storey hold-down connection an economical and structurally sound method of resisting overturning in multiple story wood framed buildings.

The following analysis describes an experimental and analytical studyof two ofthe most common types of inter-storey hold-down connections (hold-down connection with perforated strap or with two angle brackets) of the timber-framed walls. These two types are deliberately chosen because they allowthe direct connection of the timber-framed wall from the upper floor through the ceiling to the wall of the lower floor. Examples of this kind of anchoring can be seen in Figure 4b, c. In addition, analytical expression describing the stiffness of the investigated hold-down anchor is derived for the anchoring with perforated strap.

the interposition of the OSB sheathing boardsdoes not affect the behaviour of the connections and the OSB plate can be ignored when calculating the bearing capacity of a hold-down anchor, as mentioned earlier.

Figure 5: Experimental test set up for perforated strap hold-down connection (Type A)

EXPERIMENTAL AND ANALYTICAL STUDY Test configuration Our experimental study tests two types of most often used inter-storeyhold-down connectionsin praxis, namely: â&#x20AC;˘ Type A:perforated strap hold-down connection; hold-down connection where perforated strapis used and fastened to the timber studs in upper and lower floor with nails (Figure 4c) â&#x20AC;˘ Type B: hold-down connection with two angle brackets; hold-down connection where two angle brackets connected with threaded rod and fastened to the timber studs with nails are used (Figure 4b) For each type of hold-down connection, one type of an anchor is experimentally investigated.Details for both types of anchoring are designed according to the details used in praxis in order to simulate as accurately as possible the actual anchoring situation. These details have also been used in Kolb [16] and Holzrahmenbau [17].

The test set-up for both types of hold-down connection is shown in Figure 5 for perforated strap connection and in Figure 6 for two angle brackets connection. Hold-down anchors are fixed directly to the timber studs without the interposition of the sheathing board (as this is the case when fibre-plaster boards are used). When OSB sheathing boards are used,in practice hold-down anchors are fixed to the timber-framed wall with the interposition of the OSB plate. But according to Tomasi and Sartori [13], Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Figure 6: Experimental test set up for two angle brackets hold-down connection (Type B) In the experimental study, 14 test specimens are tested; 7 test specimens for type A connection and 7 test specimens for type B, respectively. For each type of connection first test specimen is used for the determination of the Fest with the use of preliminary monotonic static test. Other 6 specimens of each type of connection are tested using the standard EN 26891 [19] and its loading procedure.

Test specimens for both types of anchoring are composed of wood elements of timber quality C24 which arerepresenting the timber stud and bottom plate of the upper wall, ceiling and header joist and timber stud and top plate of the wall below the ceiling. Timber studs are interconnected with perforated strap BAN204050 (Simpson Strong-Tie) for Type A connection and two angle brackets WHT 340 (Rothoblaas) connected with thread-

361

Katja Vogrinec, et al. - Experimental and analytical study of the inter-storey hold-down connections in timber-frame panel buildings

ed rod M16 8.8 for Type B connection. Composition of test specimens is shown in Figure 5 for perforated strap connection and in Figure 6 for two angle brackets connection.

The displacements between timber studs of the upper and lower wall are measured separately on each side of the timber studs under the applied tensile load F. The difference between displacements is caused by the eccentric anchoring of the tested hold-down anchors to the timber studs and the partial release of the rotation at the point where test specimens are clamped into device with

steel jaws. The measured displacements on the side of attached hold-down anchor is named L2 displacements and the displacement on the opposite side of the timber stud L1 displacement. Both measuring sides are shown in Figure 5 and Figure 6.

Figure 7: Force-displacement (L1, L2) graphs for perforated straphold-down connection

Figure 8: Force-displacement (L1, L2) graphs for two angle brackets hold-down connection

362

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Katja Vogrinec, et al. - Experimental and analytical study of the inter-storey hold-down connections in timber-frame panel buildings

Analysis of the results Figures 7 and 8 show experimental test results for both types of anchoring for all test specimens. Force-displacements graphs are shown separately for displacements L1 and L2 resulting from the eccentricity of the anchoring. It can be seen that displacements on the side L2, where the anchor is attached, are smaller than on the sideL1, when the same load F is applied. This is a logical consequence of the eccentrically fixing of the hold-down anchor.

The Figure 9 shows a force-displacement graph with mean values of displacements L1 and L2 from all test specimens for both types of connections. Dashed lines of the graph in Figure 9 represent the average values of the displacements L1 and L2 for each type of the connection. The higher load-bearing capacity of the anchor is achieved by using a hold-down connection with two angle brackets, whereas a higher stiffness of an anchor is achieved by using a perforated strap hold-down connection.

Figure 9: Force-displacement graph: comparison between average values of displacements for both types of connections Tables 1 and 2 display an input data and measured values of v01, v02, Fmax according to the standard EN 26891 [19] for all test samples. For both types of testedhold-down anchors,a stiffness ks based on the test results is calculated according to EN 26891 [19] and is shown in Table 1 for perforated strap and inTable 2 for angle brackets for both measured displacements L1 and L2. Depending on the type of anchoringused,different failure mechanism occurred as can be seen from Tables 1 and 2 and in Figure 10 where different failure mechanisms of test specimens are shown. Two different types of failure occurred on the perforated strap; steel tape tearing and nails pulling are shown in Figure 10 a, b. In the case of two angle brackets, failure of the tested connection occurred as a result ofnailspulling and tearing (Figure 10c) and in two cases failure developed in the timber stud as the consequence of the defects in the stud. The cause of a failure was the presence of the gnarls in the wood. At this point a high concentration of the tensile stresses perpendicuJournal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

lar to the grain occurred, which led to a rapid and brittle failure of the test specimen.

Figure 10: Failure mechanisms for both types of connections: a) steel tape tearing, b) nails pulling, c) nails pulling and tearing, d) timber stud failure The importance of the contribution of the hold-down anchors on the stiffness of the timber-framed wall has already been highlighted in Casagrande et al. [5] where the average percentage of deformation due to each single contribution is shown for timber-framed walls. The value of the hold-down stiffness taken into the calculation was 5000 N/mm. The experimental tests of the hold-down anchors suitable for anchoring to the foundations has also been performed in Tomasi and Sartori [13] where almost all of tested types of hold-down anchors showed similar values of the stiffness (around 3000 N/mm).As can be seen from the experimental resultsin Tables 1 and 2, the stiffnessks of the tested inter-storey hold-down anchors is even smaller than the stiffness of the hold-down anchors used for anchoring to the foundationsmentioned above. As the flexibility of the inter-storey hold-down anchor is higher than the flexibility of the hold-down anchor used for anchoring to the foundations, its contribution to the horizontal stiffness of the timber-framed wall would be even higher than the contribution of the hold-down anchor used for anchoring to the foundations.The importance of including the contribution of the flexibility of the hold-down anchors to the overall horizontal stiffness of the timber-framed wall thus has even greater significance in the case of inter-storey hold-down connections.

As already emphasized in Tomasi and Sartori [13],an analytical expression which adequately describes the stiffness of the hold-down anchors is needed not only for hold-down anchors appropriate for anchoring to the foundations but also in the case of inter-storey holddown connections.

Derivation of the analytical expression for the stiffness of the perforated strap Analytical expression for the stiffness of a hold-down connection with perforated strapis derived by summing different contributions to the overall stiffness of the perforated strap in vertical direction. The total flexibility of the perforated strap and consequently its stiffness KtypeA is equal to the sum of: (i) contribution of the

363

Katja Vogrinec, et al. - Experimental and analytical study of the inter-storey hold-down connections in timber-frame panel buildings

flexibility of the nails fastened in the upper and lower stud of the timber-framed wall and (ii) contribution as a result of elongation of the perforated steel tape in the form of: (2)

1 / KTypeA = 1 / Knails, up + 1 / Knails, down + 1 / Ksteel

where Knails, up is the stiffness of the nails fastened to stud in the timber-framed wall of the upper floor, Knails, down is the stiffness of the nails fastened to stud in the timber-framed wall of the lower floor and Ksteel is the stiffness of the perforated steel tape.

It should be mentioned that Eurocode 5 [9] does not include any analytical expression for calculating the stiffness of the hold-down anchor, but it provides the formulas for calculating the joint slip with the help of the slip modulus Kser per shear plane per fastener. For nails without pre-drilling, the slip modulus Kser per shear

plane per fastener is calculated according to Eurocode 5 [9] as: (3)

Kser = (Ď m1.5 * d0.8) / 30

where d is the diameter of the nail in mm and Ď m mean density of the timber member in kg/m3. The stiffness of the n-nails Knails can now be calculated as the product of number of nails nnails and the slip modulus Kser per shear plane per fastener in the form of: (4)

Knails = nnails * Kser

Table 1: Perforated strap (type A) test results Sample

A2

A3

A4

A5

A6

A7

Failure mechanism

Nails pulling

Tape tearing

Tape tearing

Tape tearing

Tape tearing

Nails pulling

Type of connection

perforated strap

perforated strap

perforated strap

perforated strap

perforated strap

perforated strap

Type of the perforated strap

BAN204050

BAN204050

BAN204050

BAN204050

BAN204050

BAN204050

Manufacturer

Simpson Strong-Tie

Simpson Strong-Tie

Simpson Strong-Tie

Simpson Strong-Tie

Simpson Strong-Tie

Simpson Strong-Tie

Thickness of the tape [mm]

2.0

2.0

2.0

2.0

2.0

2.0

Width of the tape [mm]

40

40

40

40

40

40

Length of the tape [mm]

860

860

860

860

860

860

Nails

4.0 x 50 mm

4.0 x 50 mm

4.0 x 50 mm

4.0 x 50 mm

4.0 x 50 mm

4.0 x 50 mm

The number of nails per stud [pcs]

9

9

9

9

9

9

Diameter of the nail hole [mm]

5.0

5.0

5.0

5.0

5.0

5.0

Fmax [N]

24768

25333

25132

25229

252622

24005

v01 [mm]

1.22

1.81

1.94

2.47

1.84

1.87

v04 [mm]

6.21

7.15

8.59

5.83

7.88

8.56

vi, mod [mm]

6.66

7.12

8.86

4.49

8.07

9.05

ks [N/mm]

1488.09

1422.29

1134.66

2248.08

1252.69

1061.37

L1

ks [N/mm]

L2

v01 [mm]

0.45

0.46

0.39

1.28

0.49

0.52

v04 [mm]

3.08

3.15

3.77

3.02

3.19

3.57

vi, mod [mm]

3.51

3.59

4.50

2.32

3.60

4.06

ks [N/mm]

2819.56

2823.70

2232.10

4345.71

2805.18

2362.56

ks [N/mm]

364

1434.53

2898.30 Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Katja Vogrinec, et al. - Experimental and analytical study of the inter-storey hold-down connections in timber-frame panel buildings

Table 2: Two angle brackets (type B) test results Sample

B2

B3

B4

B5

B6

B7

Nails pulling and tearing angle brackets

Nails pulling and tearing angle brackets

Nails pulling and tearing angle brackets

Timber stud failure angle brackets

Nails pulling and tearing angle brackets

Timber stud failure angle brackets

Type of the angle bracket

WHT 340

WHT 340

WHT 340

WHT 340

WHT 340

WHT 340

Manufacturer

Rothoblaas

Rothoblaas

Rothoblaas

Rothoblaas

Rothoblaas

Rothoblaas

Thickness of the anchor [mm]

3.0

3.0

3.0

3.0

3.0

3.0

Width of the anchor [mm]

60

60

60

60

60

60

Length of the anchor [mm]

340

340

340

340

340

340

Threaded rod

M16 8.8

M16 8.8

M16 8.8

M16 8.8

M16 8.8

M16 8.8

Nails

4.0 x 60 mm

4.0 x 60 mm

4.0 x 60 mm

4.0 x 60 mm

4.0 x 60 mm

4.0 x 60 mm

The number of nails per anchor [pcs]

20

20

20

20

20

20

Diameter of the nail hole [mm]

5.0

5.0

5.0

5.0

5.0

5.0

Fmax [N]

55175

55420

62217

57407

63311

52920

v01 [mm]

4.20

4.34

4.19

3.83

3.06

3.46

v04 [mm]

19.77

18.95

19.45

17.35

16.49

15.03

vi, mod [mm]

20.76

19.48

20.34

18.03

17.92

15.44

ks [N/mm]

1063.36

1137.98

1223.55

1273.79

1413.36

1371.37

Failure mechanism Type of connection

L1

ks [N/mm]

L2

1247.24

v01 [mm]

2.50

2.65

2.64

2.32

2.09

2.40

v04 [mm]

10.46

10.68

10.52

9.91

9.39

9.14

vi, mod [mm]

10.61

10.71

10.51

10.13

9.74

8.99

ks [N/mm]

2079.22

2068.92

2368.41

2266.52

2600.96

2354.34

ks [N/mm]

The calculation of stiffness of the perforated steel tape takes into account the cross-section of the tape without the holes for the nailsAs,neto.In Hoekstra [8], detailed analysis of the force distribution between fasteners exposes linear hold-down force distribution over the fasteners which is valid for the elastic part of the force-displacement diagram of the hold-down. Accordingly, the whole length of the steel tape is not considered in the calculation.The reduced length of the steel tape taken into account Ls is now calculated as the actual length of the steel tape reduced by the value of half the length of the tape section, where nails are attached.

The stiffness of the steel tape Ksteel is now written in the form of:

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

2289.73

Ksteel = Es * As, neto / Ls

(5)

where Es is the modulus of elasticity of the steel tape, As, neto net cross-section of the steel tape and Ls reduced length of the steel tape.

According to the formulas (2-5),an analytical value of the stiffness of the perforated strap is calculated. 2 Ă&#x2014; 9 pieces of nails dimensions of 4,0 mm Ă&#x2014; 50 mm are used for the hold-down connection with perforated strap. The thickness of the steel tape is t = 2 mm, width of the tape is btape = 40mm and length L = 860mm. Diameter of the nails hole is 5 mm with 2 holes in a row. Nails are fastened to the timber stud in the length of 110mm on each side. For timber studs wood quality of C24 according to EN 338 [20] is used with mean density of Ď m = 420 kg/m3.

365

Katja Vogrinec, et al. - Experimental and analytical study of the inter-storey hold-down connections in timber-frame panel buildings

Stiffness of the nails Knails is now calculated according to equations (3,4) as: (6)

And stiffness of the steel tape according to equation (5) as:

(7) The analytically obtained stiffness of the perforated strap connection in the vertical direction is calculated according to equation (2) as:

Figure 11: Comparison between the experimental and the analytically determined stiffness of the perforated strapconnection CONCLUSIONS

(8) Comparison of the results Figure 11 shows a comparison between experimental tests results and analytical formula for the stiffness of the hold-down anchor where perforated strap is used.

Analytically obtained stiffness of the perforated strap is KtypeA = 3174,39 N/mm and the experimental value for the stiffness of the perforated strap on the side L2 where the strap is attached to the timber stud, is equal to ks = 2898,30 N/mm. As can be seen from the stiffness comparison in Figure 11,analytically obtained values are about 10% higher than experimentally obtained values what makes a good agreement between experimental and analytical test results. In analytical calculations both contributions were included in stiffness calculation; the flexibility of the nails fastened in the upper and lower stud of the timber-framed wall and the elongation of the perforated steel tape. If only deformation of the nails were taken into account when calculating the stiffness of the perforated strap, this would lead to overestimation of the stiffness of the hold-down anchor.Consequently, similar conclusions were drawn in experimental tests for hold-down anchors suitable for anchoring to the foundationsby Tomasi and Sartori [13].The presented analytical formula for the stiffness of the perforated strap can be used with analytical formula for the horizontal stiffness of the timber-framed walls presented in Vogrinec, Premrov and KozemĹ ilih [3]. However, in order to use conclusions on different hold-down anchors,these connections need further numerical and experimental studies.

366

Eurocode 5 [9] does not provide recommendationson calculation of the flexibility for the hold-down anchors and on inclusion of their contribution to the horizontal stiffness of the timber-framed walls and the building itself.

The first experimental test of the hold-down and shear anchors for the connections between timber-framed walls and foundations performed in Tomasi and Sartori [13] and derived analytical procedure in Casagrande et al. [5] already highlighted the fact that the contribution of the hold-down anchor should be taken into account when calculating the stiffness of the timber-framed walls. The experimental study of the inter-storey hold-down connectionspresented in this article shows similar conclusions.

The inter-storey hold-down connectionalso doesnot provide totally rigid support condition for the timber-framed walland its flexibility should be taken into account when calculating the stiffness of the timber-framed wall.In conclusion,the presented experimental results prove (Figure 9) that the load-carrying capacity and the stiffness of the resisting wall elements essentially depend on the type of the inter-storey connection and it is not negligible in the calculation of the stiffness of the timber-framed walls. In the presented experimental study,higher stiffness of the connection is obtained by using perforated strap, while higher load-carrying capacity is ensured with two angle brackets connection. Moreover, a good agreement between derived analytical formula and experimental results is achieved in the case of perforated strap.

The presented experimental studyis the starting point of the future studies of the inter-storey hold-down connections. However, additional experimental and numerical studies are needed for more general conclusions and appropriate derivation of analytical formulas. It would be of great importance to offerengineering practice the appropriate analytical formulas for different types of holddown connections. In addition, they should be included in European standards and should enable engineers to influence the stiffness and load-bearing capacity of the hold-down anchor and the building itself by simplevaryJournal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Katja Vogrinec, et al. - Experimental and analytical study of the inter-storey hold-down connections in timber-frame panel buildings

ing of the hold-down anchor type. Moreover, the influence of the inter-storey hold-down connections should be considered in the analysis of the entire timber-frame panel building under the influence of the horizontal forces. Therefore, the actual differences in the flexibility of the building and effect of the inter-storey hold-down anchoring on the distribution of the horizontal forces and horizontal displacements will be visible.

ACKNOWLEDGEMENTS Operation part financed by the European Union, European Social Fund.

REFERENCES 1. European Committee for Standardization. (2005).

Eurocode 8: Design of structures for earthquake resistance - Part 1: General rules, seismic actions and rules for buildings, EN 1998-1. Brussels.

2. Pintarič, K., & Premrov, M. (2013). Mathematical modelling of timber-framed walls using fictive diagonal elements. Applied Mathematical Modelling, 37, 8051-8059. doi:10.1016/j.apm.2013.02.050 3. Vogrinec, K., Premrov, M., & Kozem Šilih, E. (2016).

Simplified modelling of timber-framed walls under lateral loads. Engineering structures, 111, 275-284. doi:10.1016/j.engstruct.2015.12.029 4. Kessel, M.H. (2004). Vereinfachte Berechnung von scheibenartig beanspruchten Tafeln. In H.J. Blaß, K.J.H. Ehlbeck, & G. Steck (Eds.), Erläuterungen zu DIN. München: DGfH Innovations und Service GmbH. 1052: 2004-08.

5. Casagrande, D., Rossi, S., Sartori, T., & Tomasi, R. (2016). Proposal of an analytical procedure and a simplified numerical model for elastic response of single-storey timber shear-walls. Construction and Building Materials, 102, 1101-1112. doi:10.1016/j. conbuildmat.2014.12.114 6. Casagrande, D., Rossi, S., Sartori, T., & Tomasi, R. (2012). Analytical and numerical analysis of timber framed shear walls. In World Conference on Timber Engineering. Auckland, New Zealand.

7. Sartori, T. (2012). Structural behavior of timber framed buildings. Dottorato di Ricerca in Ingegneria dei Sistemi Strutturali Civili e Meccanici XXV ciclo.

Università degli Studi di Trento.

8. Hoekstra, T. (2012). Multi-storey timber-frame building. Delft: Delft University of Technology, Faculty of Civil Engineering and Geosciences. MSc Thesis.

9. European Committee for Standardization. (2004).

Eurocode 5: Design of timber structures - Part 1-1: General - Common rules and rules for buildings, EN 1995-1-1. Brussels.

10. Prion, H.G.L., & Lam, F. (2003). Shear Walls and Diaphragms. In S. Thelandersson & H.J. Larsen (Eds.), Timber engineering. (pp. 383-408). Chichester: John Wiley and Sons Ltd. 11. Premrov, M., & Dobrila, P. (2012). Numerical analysis of sheathing boards influence on racking resistance of timber-frame walls. Advances in Engineering Software, 45(1), 21-27. doi:10.1016/j. advengsoft.2011.09.012 12. Faherty, K.F., & Williamson, T.G. (1998). Wood Engineering and Construction Handbook, 3rd edition.

New York: McGraw-Hill Publishing Company.

13. Tomasi, R., & Sartori, T. (2013). Mechanical behaviour of connections between wood framed shear walls and foundations under monotonic and cyclic load. Construction and Building Materials, 44, 682690. doi:10.1016/j.conbuildmat.2013.02.055 14. Kessel, M.H., & Polatschek, T.M. (2009). Verankerung von HolztafelnTagungsband 21. Hildesheimer Informationstag Holzbau (21. HITH). HAWK Hildesheim.

15. zur Kammer, T. (2006). Zum räumlichen Tragverhalten mehrgeschossiger Gebäude in Holztafelbauart.

Braunschweig: Institut für Baukonstruktion und Holzbau der Technischen Universität Carolo-Wilhelmina zu Braunschweig.

16. Kolb, J. (2008). Systems in Timber Engineering. Basel: Birkhäuser.

17. Holzrahmenbau, Bewährtes Hausbau-System, 4.

Auflage. (2011). Köln: Bruderverlag.

18. [18] Nelson, R.F., Patel, S.T., & Arevalo, R. (2003).

Continuous tie-down systems for wood panel shear walls in multi story structures. Structure magazine, 1-16.

19. European Committee for Standardization. (1997).

Timber structures – Joints made with mechanical fasteners – General principles for the determination of strength and deformation characteristics, EN 26891: 1997. Brussels.

20. European Committee for Standardization. (2009).

Structural timber – Strength classes, EN 338: 2009.

Brussels.

Paper submitted: 13.04.2018.

Paper accepted: 10.07.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

367

Original Scientific Paper

Journal of Applied Engineering Science

doi:10.5937/jaes16-17072

Paper number: 16(2018)3, 541, 368 - 373

STUDENT SCIENCE AND TRAINING ENGINEERS Viktor Shebashev*, Yuri Andrianov, Lyudmila Nizova, Anna Nikitina Volga State University of Technology, Russian Federation The authors contemplate the forms of student science and their role in training engineers based on the documents of the International Labor Organization for Human Resources Management and Vocational Training: a case of technical universities in the Volga Federal District of the Russian Federation. The articles gives the dynamics of innovation-driven growth of Volga State University of Technology (Volgatech), students’ publication activity, students’ participation in the “UMNIK” programme and the number of incentives that the students were awarded at the Olympiads and competitions (with a breakdown by engineering professions). Based on monitoring, the authors define the tendency of Volgatech graduates’ employment (with a breakdown by integrated groups of majors of engineering graduates).

Key words: Student science, Student, Engineer, Production, HEI, Innovative methods of training, Employment of engineering graduates, Labor market INTRODUCTION Personnel training and retraining is a special issue in the work of the International Labor Organization. The organization broadly views the problem of human resource development, including such issues as governance development, career guidance and occupational rehabilitation. A number of conventions and recommendations substantiate the organization’s viewpoint. The recommendation No.57 “On Vocational Training” (1937) was one of the first normative acts, which gave definition of such training, described the basics for the content of programmes, andaddressed the issues of pre-vocational training and vocational training prior to and during employment.

The issues is best discussed in the Recommendation No.117 “On vocational training” (1962), the Convention No.142, supplemented by the Recommendation No.150 “On Human Resources Development” (1975), the Article 1 states: “Each Member shall adopt and develop comprehensive and co-ordinated policies and programmes of vocational guidance and vocational training, closely linked with employment, in particular through public employment services”.

The conventions and recommendations comprise agreed goals, principles, obligations of employers and workers in the field of policies and initiatives of vocational training. The basic principles of the training policy stipulate that: • all workers shall be given equal access to vocational training and employment should be ensured without discrimination on any grounds; • national policies and programmes are designed to encourage citizens to develop their ability to work, primarily in their own interests, as well as taking into account the needs of countries; • training systems should meet the needs of young people and adults for lifelong learning;

• national policies and training programmes should be designed and implemented by the government in close cooperation with entrepreneurs and workers organizations.

These principles are fundamental of the Constitution of the Russian Federation (Article 37), according to which every citizen has the right to be in command of his/her abilities to work.This right presumes special importance when there is a problem of training and further employing graduates by profession. Modern economic system deploys two types of employment technologies for engineering graduates: traditional and non-traditional (innovative). Traditional technologies are knownand stipulated by the law of the Russian Federation “On employment of the population in the Russian Federation” [1]. The main forms of employment promotion are: assistance to citizens in finding suitable work; information support on the situation in the labor market; arranging and holding job fairs; career guidance, job placement, and vocational training; psychological support, vocational training and professional development of unemployed citizens.

However, the above mentioned types of policy in the times of financial and economic crisis are losing their relevance, so, the methodology is poorly adapted to the modern economic system; traditional technologies do not encourage economic activity of engineering graduate, but rather provoke frustration in the chosen profession [2].

THEORY AND EXPERIMENT Nowadays the Russian government has launched the process to support the leading regional universities to facilitate their cooperation with local industries. More than 70 percent of universities have become active partners of the representatives of local industries, which stimulated the growth of regional economy.

The collaboration between academic institutions and various branches of industry includes: small innovative

* Volga State University of Technology, 424000, The Republic of Mari EI, Yoshkar-Ola, pl. Lenina 3, ShebashevVE@volgatech.net

368

Viktor Shebashev et al. - Student science and training engineers

enterprises, specialized departments, technology incubators, start-ups, as well as collaboration in the field of research and development.

Improved training of engineers is a priority took in the context of innovative economy [3].

Monitoring demonstrated that over the past six years, the highest proportion of graduates employed was in Russia and in the Volga Federal District (VFD) in 2010, while in the Republic of Mari El (RME) the highest proportion was in 2011. The lowest indicator for all Russian territories was in 2015 (76.8% - the Russian Federation, 74.4% the VFD, and 79.9% - the RME.) Generally, the share of graduate employment in the Russian Federation is 75%, including the region of residence - 64%.) [5].

The main reasons for the decline in employment are as follows: there is no balance between the labor market and the market of educational services – the universities educate graduates the local economy does not call for.

The latter results in the second problem – many specialists settle down in subject of the Russian Federation respectively.

Author's monitoring proved that the highest rate of employment in the region is typical of the Far Eastern Federal District (72%), the North-Western and the Volga Federal Districts (67%), and the lowest is - the Central (57%) and Siberian (58%) Federal Districts (Table 1).

Statistical surveydemonstrated that the highest level of employment in the Volga District is characteristic of Kazan National Research Technical University n.a. A. N. Tupolev (85%), Nizhny Novgorod State Technical University n.a. R.E. Alekseev (90%), M.T. Kalashnikov Izhevsk State Technical University (90%), Ulyanovsk State Technical University (90%). Meanwhile, the anal-

ysis showed that there are RussianHEIs with the minimum employment indicators, including the University of Russian Innovative Education (30%), Ulyanovsk Civil Aviation Institute n.a. the Chief Marshal of Aviation B.P. Bugaev (24%), and Tomsk State University of Control Systems and Radioelectronics (33%) [5].

The dynamics of the graduates’ employment mad it evident that, the most popular majors of training at Kazan National Research Technical University n.a. A.N. Tupolev are the following: Materials Engineering (94.7%), Techno sphere Safety and Environmental Engineering (94.2%), and the least popular are Methods and Technology of Ground Transport Vehicles (83.5%), Mechanic Engineering (85%). The most popular majors of training at Nizhny Novgorod State Technical University n.a. R.E. Alekseev are Aeronautic and Rocket-and-Space Equipment (97%), Mathematics and Mechanics (94%), while the least popular majors are Methods and Technology of Ground Transport Vehicles (85%), Materials Engineering (84%).

The most popular majors of training at Kalashnikov Izhevsk State Technical University are Computer Science and Computer Engineering (95%), while the least popular majors are Aeronautic and Rocket-and-Space Equipment (75%). The most popular majors at Ulyanovsk State Technical University are Photonics, Tool Engineering, Optical and Biotechnical Systems and Technologies (90%), where the least popular majors are Architectural Studies (56%).

Graduates’ training in the interests of specific enterprises based on cooperation and partnership should be prioritized [6].The Russian labor market can becharacterized by an excessive external mobility, which is substantiated by high labor turn-over rates (hiring and firing rates) and low rates of jobtenure.

Table 1: Graduates’ employment by federal districts

Share of employment by years Name of the federal district

2014

2015

2016

Share of the Employed in Share of the Employed in Share of the Employed in employed the region employed the region employed the region

Central Federal District

80%

62%

80%

58%

80%

57%

North-Western Federal District

80%

74%

80%

68%

80%

67%

Southern Federal District

75%

71%

75%

67%

75%

65%

North-Caucasian Federal District

55%

69%

55%

67%

60%

65%

Volga Federal Districts

80%

71%

80%

68%

80%

67%

Ural Federal District

80%

69%

85%

66%

80%

65%

Siberian Federal District

75%

58%

80%

60%

80%

58%

Far Eastern Federal District

75%

77%

80%

74%

75%

72%

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

369

Viktor Shebashev et al. - Student science and training engineers

Our hypothesis is that external mobility dominates (in comparison with the internal one) and is largely compensatory, being a reaction to the existing limitations for internal promotion [7].

This has negative impact not only on the labor market, but also on the development of the economy, especially on the industrial sector. In the case of the Republic of Tatarstan the level of graduates’ employment is 80%, 75% of which are graduates employed in the region, which contributed to a high rating on the socio-economic status among the subjects of the Russian Federation (5th in the rating). In the Perm region, the level of employment is 88%, 83% of graduates find work in the region; this results in the high 15th place in the Russian rating.

We are observing a contrary situation in the Republic of Mordovia, where out of 70%, 56% of graduates were employed in the region and, thus, the republic ranked 63rd among Russian districts. The Republic of Mari El ranks 69th in terms of socio-economic development, with 75% of graduates employed (64% of which are employed in the region [6].

Volga State Technological University sees enhancing the quality of engineering training as priority in the context of an innovative economy. First, this should be implemented and supported by innovative university infrastructure. Volgatech has Shared Knowledge Centre, eleven Research and Technology Centers based on unique laboratories, techno park, business-incubator, student design-engineering department, twenty-five small innovative enterprises, botanical garden-institute, and experimental forestry-enterprise. These structural units employ more than 100 students.Scientific and Educational and Scientific and Technological Centers are in the heart of the innovative university infrastructure.

Volgatech is a leader in the field of sustainable nature management based on nano-, bio-, energy-saving and infocommunication technologies.

Secondly, the university encourages students to use their innovative competencies. For that purpose teaching staff uses active forms of training, such as: business and role games, designing business plans, presentations, brainstorming, master classes, brain-rings, round tables (discussions, debates), case-study analysis, etc.

The engineering departments, as a rule, use topics related to the issues of specialist training for various manufacturing industries. There are business games “Job searching technologies and ways of employment”, “Labor market and market of educational services”, “Engineer is entrepreneur", etc. The Volgatech business incubator is in charge of training seminars and programmes such as: “How to Do Business” “Export, Import, Engineering, Finance, Logistics, Marketing”, “Fundamentals of Entrepreneurship”, “Fundamentals of Small Innovative Enterprises”, “Technology Transfer and Commercialization”.

The university takes part in the contest supported by the Ministry of Education and Science of the Russian Federation in the prioritized project “Modern Digital Education-

370

al Environment”, which aims at establishing a regional online learning center.

Thirdly, the university has committed to joining social partnership with real sector of the economy. University scientific and technological centers play a major role in increasing the volume of research and development work. The most important of which are: Center of Radar Systems and Complexes; Center of Sustainable Forest Management and Remote Sensing; Center for Automated Mechanical Engineering; Laboratory “New Technology”; Test laboratory of Furniture and Wood Products; Laboratory of Construction and Technical Expertise; Shared Knowledge Center “Ecology, Biotechnologies and Processes for Obtaining Environmentally Friendly Energy Carriers"; Center for the Development of Innovative Technologies in Construction; Student Design-Engineering Department; Center for Consulting in Management and Marketing; Center for Engineering and Industrial Design “Bioenergy”; Center for Expertise and Property Management.

Fourth, the university supports publication and project activities. The dynamics of the publication activity is given in Table 2, which shows that the students with engineering majors account for more than 70% of publications at the university.Students of the Institute of Civil Engineering and Architecture, as well as future specialists in the field of management and business are most active in publishing their research results [8].

The “UMNIK” programme has become very popular among students (Figure 1). Volgatech is among the top ten universities in Russia in terms of the number of projects financially supported by the Innovation Development Fund in the framework of the youth scientific and innovative competition “UMNIC” (in 2016 there were 237 of them).

Students' participation in grant activities is improving their competitiveness.The students of the Radio Engineering Faculty (over the last three years more than 30 grants), as well as the students of the Institute of Forestry and Nature Management (27 grants) are the most active in this activity. We are witnessing a steady increase in the number of medals, diplomas and certificates awarded to students at international, Russian and regional Olympiads and competitions. Figure 2 shows that there was a slight decrease in the number of incentives in 2016 compared to 2015, but there had been a significant increase compared to 2014.This is applicable to students of the Institute of Civil Engineering and Architecture, the Faculty of Radio Engineering, the Faculty of Information Technologies and Computer Engineering [4].

Students also participate in the International Festival of Innovations, Knowledge and Inventions “Tesla Fest” (Serbia); in the International Salon of Inventions “INOVA” (Croatia); in the International Forum of Innovation Development “Open Innovations”; and in the International Exhibition “Open Innovations Expo” and many others.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Viktor Shebashev et al. - Student science and training engineers

Table 2: Dynamics of students’ publications at Volga State University of Technology Engineering majors

2014

2015

2016

Institute of Mechanics and Machine Building (IMMB)

243

35

92

Institute of Forestry and Nature Management (IFNM)

50

64

154

Faculty of Radio Engineering (FRE)

88

82

132

Faculty of Information Technologies and Computer Engineering (FITCE)

34

29

52

Institute of Civil Engineering and Architecture (ICEA)

161

98

216

Faculty of Economics

223

97

325

Faculty of Management and Law

554

192

171

Total for engineering majors

1353

597

1142

Total at the University

1851

991

1605

Figure 1: Results of the “UMNIK”programme at Volga State University of Technology

Figure 2: The number of medals, diplomas and certificates awarded to students at the Olympiads and competitions Student science is a launching pad for scientific personnel training, thus, currently more than 200 PhD students study at Volgatech. Consequently, for 100 full-time students of the contingent, there are 3.55 PhD students. In 2016 the university implemented 23 PhD programmes Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

(for those who study 1-3 years according to the Federal State Educational Standard) and 25 PhD programmes (for those who study 4-5 years in compliance with the federal state requirements).

The study of the role of student science in the process ofengineering graduates training proved that the technology should be contin uing - from school to university. Thus, “My First Step into Science” and Russian student forum “Engineering Graduates are the Future of Russia's Innovative Economy” are annually held at the university.

Volgatech organizes the final rounds of the Russian Student Olympiad in Mathematics, Statistics, Robotics and Controller Programming, as well as the final rounds of the Internet Olympiads in 15 disciplines for university students in the Volga Federal District.

The accumulated experience in research is beingpractically used by Volgatech graduates who are in demand at the enterprises of the defense industry complex (hereinafter referred to as the DIC) in the Volga Federal District.

Currently, 400 students are enrolled in the programme “New Personnel of the Defense Industry Complex", 85% of which will be employed. More than two-thirds of engineers and technical support staff at enterprises of the Defense Industry Complex are graduates of Volgatech.

However, the problem of graduates’ employment is still seen as priority in Russia and the Republic of Mari El (Figure 3). In 2016 this indicator for Volgatech graduates fell down by 1.47 times (277 people in 2016 against 684 people in 2015).

The share of employed has decreased from 80% to 75% (table 3) according to the data of the Ministry of Education and Science of the Russian Federation.Specialists in Applied Geology, Mining, Oil and Gas and Geodesy; Machine Building; Techno sphere Safety and Environmental

371

Viktor Shebashev et al. - Student science and training engineers

Engineering; Electricity and Heat Power, as well as Agriculture, Forestry and Fishery are of particular concern.

According to the authors, the main reason for the decline in employment is the lack of balance between the labor market and the market of educational services – the universities educate graduates the local economy does not

versities are centers of space for innovations” which assures further development of the university as a leading regional engineering research and education center.

CONCLUSIONS

Figure 3: Dynamics of graduates’ employmentat the Volga State University of Technology call for.

To tackle the issue Volgatech has committed to joining social partnership with industrial enterprises of the republic and with real sector of the economy, thus ensuring republic’s innovative-driven growth.The university is a member in the prioritized project of the Russian Federation “Uni-

1. Further developing the open educational environment (infrastructure) to enhancestudents’ research activity.

2. Improving the quality of education based on competence models within professional standards by the use of interactive technologies.

3. Providing a symbiosis of theory and practice based on interaction with enterprises and organizations of leading sectors of the economy, including network programmes.

4. Supporting favorable conditions for the students’ participation in International, Russian and Regional Olympiads, Contests and Fora.

5. Remodeling the educational programmes (inclusion of the model of project training, updating the content of the educational programme with the inclusion of modules for training innovative entrepreneurship and commercialization of research outcomes to expand opportunities for the development of personal

Table3: Dynamics of engineering graduates’ employment atVolga State University of Technology Name of the integrated groups of majors

Share of emloyment 2014

2015

2016

Industrial Ecology and Biotechnology

x

85.7%

100%

Chemical Technologies

90.9%

85.7%

88.9%

Information Safety

68%

63%

82.4%

Electronics, Radio-Engineering and Communication System

76.8%

70%

80%

Methods and Technologies in Civil Engineering

76.5%

80.5%

80%

Methods and Technologies in Ground Transport Vehicles

59.6%

72.5%

75%

Materials Engineering

94.4%

86.7%

75%

Informatics and Computer Engineering

67.1%

75%

72.97%

Agriculture, Forestry and Fishery

73.7%

75.7%

71%

Heat-power Engineering

93.2%

79.1%

69%

Earth Science

61.5%

70.6%

68%

Techno Sphere Safety and Environmental Engineering

78.7%

69.2%

66.7%

Mechanic Engineering

82.4%

78.6%

64.6%

Applied Geology, Mining, Oil and Gas Geodesy

77.8%

100%

63.6%

Photonics, Tool Engineering, Optical and Biotechnical System and Technologies

x

58.3%

59.3%

372

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Viktor Shebashev et al. - Student science and training engineers

competencies (softskills)).

6. Developing e-education mechanisms, and increasing the number of applied bachelor's degree programmes, which lead to qualifications that are in demand in the region.

REFERENCES 1. Law of the Russian Federation No.1032-1, dated April 19, 1991 (amended on July 29, 2017) “On employment in the Russian Federation”.http://docs. cntd.ru/document/9005389 2. Nikitina A.S.,Nizova L.M. (2015) Socialnie indicatory rinka truda vipusknikov vuza [Social indicators of the employment market of university graduates].

Ingenernie kadri-buduschee innovatsionnoi ekonomik: materiali nauchno-prakticheskoi konferentcii.

[Engineering graduates are the future on the innovative-driven economy: proceedings of the research-practical conference], p.63-65.

3. Ovchinnikova N., Ovchinnikova O., Kolmykova T., Bredikhin V. (2018) New generation of regional universities in Russia. Journal of Applied Engineering Science, 16 (1), 132-141 4. Romanov E., Andrianov Y., Nizova L., Nikitina А., (2016) Engineering personnel training through the example of Volga State University of Technology, Mari El, Russia, Journal of Applied Engineering Sci-

ence, 14 (1), 36-43 5. The Ministry of Education and Science of the Russian Federation [Electronic resource]. Available at: http://vo.graduate.edu.ru/registry#/?year=2013&slice=1&page=1 (reference date: 26.11.2017); 6. Rating of socio-economic status of the subjects of the Russian Federation, results of 2016 [Electronic resource]. Available at: https://ria.ru/infografika/20170530/1495077131.html (reference date: 01.12.2017).

7. Gimpelson V.E. et al. (2016) The Pathways We Choose: Intra- and Interfirm Transitions.HSE Economic Journal. 2016,20(2), 201–242 8. Ivanov D.V., Andrianov Y.S., Nekhoroshkov P.A. (2017)Strategy of research development of the Russian Federation. Results of research activity at Volga State University of Technology in 2016.Volgatech, Yoshkar-Ola.

Paper submitted: 05.04.2018.

Paper accepted: 12.07.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

373

Original Scientific Paper

Journal of Applied Engineering Science

doi:10.5937/jaes16-18279

Paper number: 16(2018)3, 542, 374 - 382

COMPREHENSIVE INTEGRATION AS AN EFFECTIVE WAY OF TRAINING FUTURE DESIGNERS AT TECHNICAL UNIVERSITIES (INTEGRATION AS A WAY OF TRAINING A DESIGNER) Nadezhda Zhdanova*, Sergey Gavrytskov, Anna Ekaterynushkina, Julia Mishukovskaya, Julia Antonenko Nosov Magnitogorsk State Technical University, Russia The article deals with the possibilities of comprehensive integration as a way to upgrade the professional training of designers. The authors see a means of resolving some problems of design education in it. To realize this goal, a prognostic model was developed, linking all the components: the content, methodology of the teacher’s work, and student activities. This model was built on the basis of data obtained in the process of studying the activities of design practitioners, for which a survey was carried out and an interview was held with each of them. Particular attention was paid to the difficulties experienced by young designers in the first years of their work, as well as experienced professionals, watching their young colleagues’ activities.

The authors propose to use both “vertical” and “horizontal” integration in the learning process, since each has its own advantages. In this case, the formation of future designers’ necessary competences is completely ensured, which should be manifested in a holistic and flexible thinking capable of solving design problems of any complexity level. To check the availability of such thinking, a special task was developed in the field of environmental design. Students designed the product from construction waste, thereby solving the actual problem of recycling and reuse of old materials.

The purpose of this article is the authors’ desire to share their accumulated work experience, to give the pedagogical community the opportunity to discuss the results obtained for the further implementation of the integrative approach in the professional training of future designers.

Key words: Comprehensive integration, Professional training of the designer, Intersubject links, Designer’s personal qualities, Integration model, Activity, Designing INTRODUCTION Constant demand for designers has determined the intensive training of these specialists in various educational institutions, and the changing conditions for their work have caused adjustments of the competences formed.

It has led to a need for a structural analysis of the very professional activities of designers, learning the productive experience of other educational institutions, creating educational and methodological materials, and, most importantly, determining the directions for specifying the model of the future professional. Pedagogical integration has become one of the effective ways to solve all these problems. It is an important requirement of modern science and the development of civilization as a whole. To create a holistic picture of the world for future designers is possible only on the basis of an integrative approach to the whole educational process. However, pedagogical integration as a rather recent phenomenon has not yet been thoroughly studied and is being implemented not in all educational areas.

After studying the bibliographic sources and experience of some educational institutions, as well as the educational practice at our institute, the conviction appeared that in case an integrative approach to all parts of the educational process is implemented, the overall effect will be much higher than that from certain actions of these components.

The purpose of our experimental work was to introduce a prognostic educational model integrating separate parts of the learning process to form a complex thinking in the future designer, which would ensure the ability to carry out projects of different orientations at a sufficiently high level.

The implementation of this model was carried out for more than 10 years; at the end of each academic year, the teaching methodology, teaching tools and integrated parts were adjusted, and the members of the staff were partially substituted. At the same time, from year to year it narrowed, steady positions were identified, which ensured the effectiveness of the integrated process.

The experience of implementing an integrative approach in the educational process offered by a team of authors can be useful for all stakeholders.

LITERATURE REVIEW It cannot be said that the path chosen by us was entirely new. In the history of design education, there are positive examples of using integration and getting highly successful results.

This is primarily the work of the Bauhaus and the Ulm School of Design in Germany. The high school of building and artistic design – BAUHAUS, which existed from 1919 to 1933 in Germany, showed the first and most successful form of integration – theory and practice [01, 02].

* Nosov Magnitogorsk State Technical University, Russia, n4d.zhdanova@yandex.ru

374

Nadezhda Zhdanova, et al. - Comprehensive integration as an effective way of training future designers at technical universities (Integration as a way of training a designer)

The developed methodology of training future designers formed the basis for the organization of all higher educational institutions [03, 04]. Revived after the Second World War, design education at the Ulm School of Design made every effort to integrate science, technology and art, which was embodied in continuing and very successful experiments with form and materials [05, 06].

Good results were also obtained by British teachers who combined the project activities with the practical implementation of the developed subject [07]. In fact, it was also the integration of different types of activities only at the level of secondary vocational education.

In Russia, the idea of integrated design education came in the mid-90s of the 20th century and was characteristic of all levels of education [08, 09, 10]. One of the first to respond was the teachers who introduced the design into the general education schools [11], then the higher school [12, 13] was involved. Unfortunately, for a long time the experience gained in some educational institutions was not generalized, and most importantly it was not transferred from one stage of education to another [03].

In the mid-90s of the 20th century, the first state standards appeared in Russia, in which the most common approaches to the professional model of any specialist were put forward. It was them that constituted the basis for students’ education in higher educational institutions.

The fundamental difference was the transition to a competence model, which includes theoretical knowledge, practical skills, as well as certain personal qualities purposefully educated [14, 15]. Acquisition of the necessary competences can be achieved in different ways. One of them is comprehensive integration, that is, the unification in the educational process of various forms and kinds of activities, or the combination of previously separately existing parts of training into a single whole [13].

The integration, substantiated in pedagogy, made it possible to change attitudes towards it. First, the integration objectives were theoretically formulated and it became clear that they could be different. Then its main elements were defined. They are: • designed purpose (direction of integration); • the composition of the integration process (integration objects); • the form of pedagogical integration; • types of pedagogical integration; • system-forming factor of integration [16].

During the next twenty years the possibilities of integration for different levels and directions of education, including design, were tested [17]. Theorists proposed to consider two types of integration: vertical and horizontal.

For the design education, vertical integration seemed at first much more important. It provided for continuity between different levels of education, as well as interaction with various public institutions and enterprises; it became possible to link teaching design of schoolchildren and students with a single goal and a unified methodology.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

The old principle of succession and continuity has acquired a new dimension. If in the Soviet era, only some art schools could afford preparatory schools at faculties and institutes [18], then in the early 90s of the 20th century a rather large number of supplementary education institutions opened in the whole country, teaching one or another kind of design [11, 19], which indicated an increased interest in the profession and the importance of design skills. They were regarded as the main means of developing creative abilities; this point of view still dominates today [20].

The city of Magnitogorsk did not stay on the sidelines.

In the late 80-ies here a comprehensive school № 63 was launched, where the design was taught to all children from the first to the eleventh grade. Here the goal of integration was the unification of general and special education, checking how much the design activity contributes to students’ learning. The results of the activity were repeatedly discussed at scientific and practical conferences of different levels, and work experience was touched upon in a number of publications, including study guides for students and teachers [21].

However, it should be noted that vertical integration covered only a small percentage of schoolchildren, in fact only those whose learning was focused on the future professional work in the arts, because by the beginning of the century the interest in it began to fade, publications diminished progressively and they mainly described practical experience of certain educational institutions without theoretical generalizations.

When it comes to horizontal integration, the term interdisciplinary is often used, although, ultimately, integrated elements can be modules, whole units of courses, units of programs, projects, etc. The purpose and the system-forming factors are much more important in integration. In the process of implementing integration, the following rules should be adhered to: • definition of integrated objects; • highlighting factors that contribute to integration and interfere with it; • forecasting potential outcomes taking into account the needs and characteristics of participants in the educational process.

For higher education institutions, it is important to highlight those factors that contribute to or impede integration, and thus to achieve the final outcome. The most important factor is the demand for designers in the labor market. This factor can play both a positive and a negative role. Educational institutions constantly monitor it through interaction with the labor exchange and graduates. On the basis of the data, they specify on their own which specialist is needed today and will be in demand tomorrow that respond to requests of the country, region, certain city, or enterprise. This explains the significant differences in the professional training of graduates at different educational institutions [22].

375

Nadezhda Zhdanova, et al. - Comprehensive integration as an effective way of training future designers at technical universities (Integration as a way of training a designer)

The study of works by foreign authors has revealed that interdisciplinary education is an urgent issue at present.

Knowledge integration is involved in the learning processes in higher education [23, 24, 25]. The scholars argue that currently, “gaining an insight into the nature of interdisciplinary education may help when making design decisions for interdisciplinary education” [26]. They pay attention to certain parameters while analyzing interdisciplinary education. For instance, they speak about complicated, actual problems which are addressed to “in pedagogical formats such as problem-based learning formats, case studies, field-work, discovery or inquiry learning” [26]. Among other aspects related to interdiciplinarity in higher education which are touched upon in scientific articles are: interdisciplinarity in the engineering classroom [27], designing interdisciplinary postgraduate curricula (at the Research-intensive UK University) [28], interdisciplinary teaching and learning for diverse and sustainable engineering education [29], the important role of teachers in this process (‘the caring teacher’ in the context of the higher education environment) [30], etc.

Over the past years, many universities have incorporated sustainability into their education and research. The articles describe “the integration of competences for sustainable development in higher education” [31]; the way in which a multidisciplinary approach to teaching and learning sustainability has been included in learning programs and activities in engineering at the University of Surrey, UK [32]. In many papers the notions of sustainability and eco-design are interconnected and interrelated [33, 34, 35].

Integration of theory and practice in higher education is another aspect deserving attention [29, 36, 37]. A substantial provision for the integration of theory and practice in higher education is the synergy between the higher education establishment and industry. The annual research accomplished by the Confederation of Swedish Enterprise demonstrates that eight out of ten students prefer higher education to advance their qualifications on the employment market [38]. The so-called “work integrated learning (WIL)” is a strategic thrust for universities focusing to upgrade graduate employability in a labour market [39]. The “integrated design approach” is supported by students’ training in cross-disciplinary teams.

An example of this approach is the EU-project of IDESEDU “Master and Post-Graduate education and training in multi-disciplinary teams” [40].

Another aspect which was driven attention to is the research-based curricula. The turn to research-based education took place under the influence of the Bologna process and strong impact of such organizations as OECD, European Ministers of Education, Commission of the European Communities [41]. It presumes that students must actively participate and be socialized into the pursuance of the research. Research-based curricula aids at developing “21st century skills, such as active learning, critical reflection and problem-solving. Students should take a position as knowledge creators rather than

376

knowledge recipients” [41]. The United Nations Decade of Education for Sustainable Development conducted by the UNESCO (2005 – 2014) is worth mentioning too in enhancing the research-based education [42].

Taking into consideration the significance of environmental issues and the necessity and desire to cope with them relating to different fields, eco-design has become of paramount importance. This is a concept comprising “human sustainability priorities together with business interrelations. Its main objective in the improvement of product development methods is to reduce environmental loads” [43]. EcoDesign also comprises an aim “to use inspiration from a wider field of positive examples of smart products and methods, effective system solutions and attractive designs” [43].

While analyzing issues related to eco-design, the authors describe the development and use of various environmental improvements: the use of an alternative energy system [33], the environmental value of longer lasting products which induces the introduction of the ecological payback period [34], elaboration of life cycle design: from regular techniques to product type specific guidelines and checklists (for instance, in a case of developing guidelines for the eco-efficient design of NECTA vending machines) [35]; the management of mass products at the end-of-life [44], application of eco-design tools (eco-roadmapping, lifecycle assessments, etc.) [45] and many others [46, 47, 48, 49].

Various tools are applied when teaching and learning eco-design. Among them are design protocol analysis, which is a technique to comprehend designers’ cognitive procedures by studying series of observed results concerning their behavior. They apply audio, video, transcript data, sketches, sensor data and user logs with the usage of VizScribe, a visual analytics framework which helps analyze both the designer’s behavior and the design process [50]. An evidence-based approach grounded on virtualization and virtualization technology is a powerful teaching and learning tool in the system of higher education [51, 52].

The creation of web-based resources (for instance, www. design-behaviour.co.uk at the Department of Design and Technology at Loughborough University), is a supportive measure in centralizing and disseminating teaching material on the subject of Design for Behavioural Change [53].

Nowadays immersive Virtual Reality (VR) technologies are frequently applied as advanced training systems for future designers. These specialists are demanded to investigate different fields of engineering in order to master the relevant knowledge about the product, the problems of its operation, its impact on the natural environment, etc. Dealing the products that are taken away from usage is one of the urgent issues. One of the ways to fix these problems is taking into consideration these issues in the opening phase of the product design. Authors created the methodology to comprise recycling demands in the phase of the product design that was made with supJournal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Nadezhda Zhdanova, et al. - Comprehensive integration as an effective way of training future designers at technical universities (Integration as a way of training a designer)

port of the CAD 3D systems. The study of recycling of the chosen product designed in the CAD 3D system was carried out using Virtual Reality technologies [54]. Thus, future designers had possibility to upgrade their skills and knowledge in the sphere of eco-design “through the immersive trainings of virtual product design for recycling” [54].

Some scientific articles describe students’ various eco-design activities: greening of a campus community through 3R (reduce, reuse and recycle) waste management initiatives [55]; the first experiences of an undergraduate student team with eco-design activities [56], etc.

The conclusions of the above-mentioned publications admit the significance of advancing holistic tools for designers, recognizing that the synthesis of guidance, education and information, in addition to well-considered content, proper application and convenient access, are all important to their achievements [57].

MATERIALS AND METHODS Like every scientific research, our experimental work began with a theoretical analysis of information resources and with the search of the available practical educational experience.

The main method in our work was the modeling and the introduction into the educational process of the prognostic model of integration. At the beginning, an attempt was made to introduce existing educational models developed by other educational institutions.

In the late 80s of the 20th century, the teachers at Stroganov Moscow State University of Arts and Industry developed the first variant of the qualification characteristics of future designers – a document that is a concentrated verbal model of a specialist. Then such a model was understood as an image established documentarily, what kind of specialist he/she should be for a certain period of time, and besides, the means of its training were determined. “The model includes a certification of a specialist and the training format” [18]. In the certification the requirements of the society were recorded in relation to the sphere of labor activity; job duties, business qualities and intellectual properties. It was assumed that the system of training in a university should correspond to the socio-professional requirements recorded in the passport. When creating the first generation of state educational standards, this point of view was adhered to, therefore units were created based on the informative value of certain disciplines.

Over the past 15 years, state educational standards in Russia have been changed three times, which led to a change in the perception of the specialist’s model three times; today, as mentioned above, everyone has switched to a competency-based approach in the training of future designers.

Our experimental work was carried out in several stages:

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

• the first stage: the simulation of the educational process was carried out on the basis of the three units of competencies available in the state educational standard for the bachelor’s degree in major “Design”.

• the second stage: the identification of competencies which were missing, but demanded in the professional activity of designers; • the third stage: the simulation of the educational process was carried out with the inclusion of a new fourth unit of additional competencies.

After each stage, the model was tested and the obtained outcomes were discussed by the entire teaching staff.

The first stage began with the study of competencies and the creation of educational methodological material for its implementation. In all standards, there are three main units that combine competence in the direction of: • a unit of general cultural competence; • a unit of general professional competence; • a unit of professional competencies.

The unit of general cultural competencies is aimed at the development of generalized abilities that provide students’ general education, therefore, they are repeated in fields of education. This includes abilities to use the foundations of philosophical, historical, economic and legal knowledge in various fields of activity, the readiness of social and ethical responsibility for decisions to be made, knowledge of methods of self-organization and self-education.

The unit of general professional competencies is aimed at the formation of universal artistic abilities in the field of drawing, painting, sculpture, aesthetics and artistic creativity. Here, special attention is paid to the knowledge of folk traditions, modern trends in arts and issues of artistic synthesis, issues of color science, means of artistic composition, patterns and techniques for the reproduction of artistic design in certain materials. Particularly distinguished are the abilities to search, store, process and analyze information from various sources and databases that determines the student’s information and bibliographic culture.

The unit of professional competences contains the competences related to the types of activities chosen by every institution, but the analysis of more than three dozen curricula of different educational institutions that train designers demonstrates that almost every curriculum comprises the first three types: artistic, design and information technology. Everyone recognizes that artistic abilities have a decisive role in designing that figurative concepts contribute to the design of objects, goods, industrial designs, complexes and objects at a competitive level.

Visually, at first, our model was presented in the form of a triangle, in the center of which there was a student with the required competences. The above-mentioned units with main activities were based on each side. To confirm such a model, we decided to compare its competencies

377

Nadezhda Zhdanova, et al. - Comprehensive integration as an effective way of training future designers at technical universities (Integration as a way of training a designer)

with those that practicing designers use in their work and compare the model with the benchmark.

The second stage was devoted to the study of professional designers’ activities. Two sociological methods were used here: interview and questionnaires, as well as analysis of implemented projects. It was self-analysis and evaluation of their designers’ own activities.

The most effective form of collecting information is an interview with each designer and analysis of the success of his/her professional activity. Some of the required information was collected by the questionnaire method, which allowed a quick survey of personal data. At first, the respondent filled out his age, which educational establishment and when he graduated from, the number of years of professional experience, places of employment and the reasons for change of employment. The next section of questions dealt with the current state of affairs: the status of the enterprise (state or commercial), the specialization of the enterprise (environmental, architectural or graphic design, construction or advertising company, design bureau, private entrepreneur, etc.), job title and area of responsibility. In the next column, the respondent mentioned the major implemented projects, noting whether they retained the author’s concept and why they were modified. And, finally, the difficulties that he/she faced as a professional designer. Designers of different ages and length of work in the Urals region were surveyed. For more details, see N.S. Zhdanova’s publication [09].

Most of the respondents were interviewed according to the pre-elaborated and pre-developed questions with the timely recording of the obtained data. There were also improvisations that helped to clarify the nature of the difficulties of designers’ professional activity, especially in the first years of work. Along with the difficulties that are not relevant to this article, we have identified a fairly large range of problems of purely designer’s activity. The questionnaire survey showed that in our region only 52% of practicing designers have a design, architectural and artistic education, so the range of missing competencies was also quite large.

Particular attention was paid to design graduates at different universities. In the first five years of work, they complain about the lack of economic knowledge, allowing to launch small enterprises and correctly assess the economic costs for implementing a particular project; the lack of design and construction data, allowing to carry out re-planning of certain buildings and structures, and knowledge of specific art crafts and modern technologies.

As a result of the survey, there was another gap in the professional training of designers: a lack of willingness to participate and organize exhibition activities. In the field of design, exhibitions of design work and various proposals are often held. Sometimes they are accompanied by forums related to certain promising areas of design work.

It turned out that not every designer is ready to discuss and evaluate new project proposals. Even more often,

378

designers participate in the organization of exhibition displays for certain goods or services and sometimes personally participate in them. The ability to personally advertise products is by no means last skill in the designer’s work.

The interview with professional designers with a lengthy employment history included a number of issues related to the activities of their young colleagues. Despite some ethical difficulties in these surveys, it was possible to collect some useful information and add those missing abilities, the absence of which prevents from working properly by graduates in the first years after graduating from universities. This, above all, includes the ability to communicate with customers and an objective assessment of graduates’ own activities.

The third stage of experimental work began with the formation of the fourth unit of special, or in new terminology, additional competences. The unit system initially provided for integration, so immediately new components were integrated with the old ones.

Figure 1: The model of interaction of different units in designers’ professional training A model was created where all four units of the curriculum are interrelated (Figure 1). However, the mere introduction of additional disciplines is not enough; it is not sufficient for specialist’s professional training in case they are not included in the general system of education and upbringing. Right away, the content and methodology of some academic disciplines were adjusted, more time was devoted to professional and communicative activities, interpersonal communication, the ability to substantiate and defend their project proposals at each stage of their implementation.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Nadezhda Zhdanova, et al. - Comprehensive integration as an effective way of training future designers at technical universities (Integration as a way of training a designer)

The center of the model is the content of professional training, and the external background consists of four units of disciplines to study. Large arrows indicate the direct impact of certain activities on students’ competencies. Due to the integration, there are more close links between the units themselves and their components, as shown by semicircular arrows. By all means, no scheme reflects the entire completeness of the simulated phenomenon; it only makes it possible to show the most significant connections.

Based on the theoretical characteristics of integration, taking into account its elements, it is possible to create various integration objects. By changing the combinations of integration elements, it is possible to obtain fundamentally new types of integration. In our case, first of all, it contributed to the correct combination of team-based, independent and problem-oriented activity in the unit of general cultural competencies: artistic and creative, information-communicative and information-technological activities in the unit of general professional competencies; students’ project-oriented, information-technological and scientific and research activities in the unit of professional competencies. The fourth unit was discussed above, but here we want to note that the units themselves interact with each other.

In the process of implementing an integrative model, much depends on the teaching staff, which sets an example and teaches the professional language of design.

Intersubject integration requires of teachers to develop a single “pedagogical” strategy, the obligatory reciprocity of use of certain concepts and facts by them not only in the development of training courses, but also in communicating with students. These requirements should be considered not only essential, but also sufficient indicators for integrated teaching. The main thing that makes the course integrated is the long-term goal set in it and specific tasks aimed at its implementation, planned by all the teachers who carry out professional training. This goal is realized by them in students’ activities specially organized.

The basis for integration was the joint efforts of teachers to create students’ certain professional skills and proficiencies, overcoming isolated teaching of subjects, the generality of the selected topics, the similarity of the objects and phenomena studied, and the unity of the key ideas envisaged in the programs. In any case, the task of teachers who form the integrated unit of subjects is to find ways of processing the elements of content and structure of their academic disciplines to form students’ integral creative thinking.

RESULTS To check the achieved results, the students were asked to perform a complex task. They had to design furniture objects from recycled materials. As a rule, they were household or construction waste. The task led students to a new level of thinking – an ecological one. The idea Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

of recycled materials usage today unites a multitude of fashion trends in art and design, as well as hand-made techniques. Students, as a rule, are well aware of the issues of waste disposal, but have never carried out projects with their usage before.

The creative transformation of waste, as a rule, does not require additional production costs, but as a result, sometimes there are art objects, household products, accessories, clothes that surpass in their functional and aesthetic qualities the initial material. “Development of designer products from non-traditional, and, preferably, recycled materials has a lot of advantages, the most important of which is the value of the work being done” [58].

Five criteria are set for the designed products: wastefree performance, minimal technological processing of raw materials, functionality, ergonomics and aesthetic qualities. The criterion of aesthetics includes the concept of singularity, as a manifestation of creativity. The evaluation is made on a five-point scale.

The authors of the article did not consider it necessary to present students’ assessments here. In our opinion it is more important to note that they are actively initiating ecological design, but only some of them successfully complete it. It is here that integrative thinking is revealed, because the work requires an incredibly broad, flexible and creative approach, attracting all competencies and it is no longer important in which discipline, unit or program they were received.

Students’ works and their description are presented in another publication of the authors [58], which also emphasizes the role of integration, which profoundly changes the content of education, leads to changes in the methods of work and creates new teaching technologies.

It provides a completely new psychological climate for the teacher and student in the learning process and creates a favorable educational environment.

DISCUSSION The process of creating a predictive model of designers’ professional training was built in accordance with the model of practicing designers’ activity, for which they were surveyed and interviewed according to a pre-designed plan. The obtained data indicated some missing knowledge and skills, which are especially subversive in the first years of independent work.

However, it turned out to be very difficult for many designers to analyze their own activities, because, as psychologists point out, it takes place in a “minimized form”. In this regard, in our opinion, a large and objective information can be provided by videotaping the process of how the practicing designer performs a certain task. This can be included into our follow-up related to the experimental study of the design process. Beyond this article, there is some material on the educational significance of integration, without which it is impossible to train a designer.

379

Nadezhda Zhdanova, et al. - Comprehensive integration as an effective way of training future designers at technical universities (Integration as a way of training a designer)

CONCLUSIONS The long work of the teachers’ team led everyone to the conclusion that pedagogical integration is an effective means of educating a designer with a full range of competencies and developed professional qualities of the individual. This is achieved by introducing a predictive model of integration of all parts of the learning process.

Verification of the received competencies and qualities of thinking was carried out by means of the task of the increased complexity associated with the problems of using construction waste. The project is assessed on the following levels: modernization, innovation and inovation. Comparison of the proposed ideas over the years shows the growth of all indicators. Students find new original solutions, which indicates the development of creative and holistic thinking. Subsequent product manufacturing corrects some project proposals, making them more practical.

The result of integration was the effective formation of comprehensive integrated thinking in the future designers, which manifests itself in the ability to carry out projects of different kinds. The final result of this process is the high level of designer’s willingness to meet the demands of the profession in different conditions and at different levels.

ACKNOWLEDGEMENTS Work on the introduction of an integrated model for designers’ training could only take place with the support of the entire teaching staff at the department and the institute, to which we express our sincere gratitude.

We would also like to thank the Regional Office of the Designers Association of Russia, whose leadership helped implement the survey of practicing designers and organized meetings with them.

REFERENCES 1. Gropius, W. (1923). Idee und Aufbau des Staatlichen Bauhauses. Münch, Weimar.

2. Kentgens-Craig, М., Dessau, S.B. (1999). The Dessau Bauhaus Building. Birkhäuser, Basel – Berlin – Boston.

3. Kondratieva, K.A. (2005). Some conceptual foundations of design education: Stroganov Moscow State Academy of Design and Applied Arts. Moscow State Stroganov University of Design and Applied Arts, Мoscow.

4. Tkachenko, E.V., Kozhukhovskaya, S.M. (2007).

Design education in Russia. The world of education – education in the world. Interindustry Information Service, no. 1(25), 156-164.

5. 5. Druzhkova, N.I. (2012). The Ulm Higher School of Design (1953-1968). Pedagogy of Art, no.2, from http://www.art-education.ru/sites/default/files/journal_ pdf/drugkova_12_june.pdf, accessed on 2017-09-29.

380

6. Bauhaus-Chronik, B.M. (1952). Vom Bauhaus in Weimar zur Hochschule für Gestaltung in Ulm. Deutsche Universitätszeitung, no. 23-24.

7. Barlex, D., Pitt, J. (1999). Technological education in schools of Great Britain. School and Production, no. 5, 93-95.

8. Antonova, N.N. (2005). From the history of development of professional design education in Russia.

Scientific Information Bulletin of Doctoral Students, Post-Graduates, Students, vol. 2, no. 6, 200-205.

9. Zhdanova, N.S. (2007). Modeling the content of professional training for future designers. Bulletin of the Orenburg State University, vol. 11-2, no. 76, 52-57.

10. Tukhbatullina, L.M., Safina, L.A. (2013). Foreign experience in the application of project-based learning in polymer specialists’ training. Bulletin of the Kazan Polytechnic University, no. 3, 333-335.

11. Ladygin, E.V. (2000). Development of students’ creative abilities in the integrative active learning at the course “Fundamentals of Design”. Мoscow.

12. Banduristy, F.F. (2001). Optimization of teaching artistic design in the system of special training of the teacher of fine art in pedagogical educational establishments (universities). Prometheus, Moscow.

13. Yakovlev, I.P. (1987). Integration of the higher school with science and production. Leningrad State University Publishing House, Leningrad.

14. Tarasova, O.P., Yanysheva M.M. (2014). Formation of professional competencies of the future designer in educational and professional activity. Bulletin of the Orenburg State University, vol. 5, no. 166, 210215.

15. Fakhrutdinova, R.A., Akhmetova, G.P. (2014). Developing professional competence of students, future designers, in higher education training. Philology and Culture, vol. 2, no. 36, 298-301.

16. Bezrukova, V.S. (1994). Integration processes in pedagogical theory and practice. Russian State Vocational Pedagogical University, Ekaterinburg.

17. Aristova, U.V. (2007). Modeling of system of designers’ professional training in high school. Moscow Open Social University, Moscow.

18. Moscow School of Design. The experience of training specialists in MHPU. (1991). Stroganov Moscow State University of Arts and Industry, Мoscow.

19. Popugaeva, I.Ya, Ryazanov, V.N., Dolikhin, T.A. (2007). On pre-profile training on the basis of the institution of additional education of children. Education of Schoolchildren, no. 10, 40-44.

20. Babikova, V.V., Sokolov, M.V. (2017). Approaches to the development of creativity in design. Current Trends in Fine, Decorative and Applied Art and Design, no. 2, 170-175.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Nadezhda Zhdanova, et al. - Comprehensive integration as an effective way of training future designers at technical universities (Integration as a way of training a designer)

21. Zhdanova, N.S. (2010). Methods of teaching design in secondary school. Magnitogorsk State University, Magnitogorsk.

22. Sirotina, I.L, Schekochikhina, O.D. (2016). Digital arts as a factor in the optimization of professional development of a designer. Proceedings of the International Scientific and Practical Conference “Russian Creative Education in the Field of Digital Art in Accordance with EU Standards”, p. 233-242.

23. Flaig, M., Simonsmeier, B.A., Mayer, A.-K., Rosman, T., Gorges, J., Schneider M. (2018). Conceptual change and knowledge integration as learning processes in higher education: A latent transition analysis. Learning and Individual Differences, vol. 62, 4961, DOI: 10.1016/j.lindif.2017.12.008 24. Hotaling, L., Sheryll, R., Sheppard, K., Chassapis, C., McGrath, E. (2007). A paradigm for vertically integrated curriculum innovation – how curricula were developed for undergraduate, middle and high school students using underwater robotics. International Conference on Engineering Education – ICEE 2007, from http://ciese.org/papers/2007/icee2007_ LEGO_finalpaper.pdf, accessed on 2018-07-01.

25. Thomas, H. (2018). Powerful knowledge, technology and education in the future-focused good society.

Technology in Society, vol. 52, 54-59, DOI: 10.1016/j. techsoc.2017.09.005 26. Klaassen, R.G. (2018). Interdisciplinary education: a case study. European Journal of Engineering Education, DOI: 10.1080/03043797.2018.1442417 27. Richter, D.M., Paretti, M.C. (2009). Identifying barriers to and outcomes of interdisciplinarity in the engineering classroom. European Journal of Engineering Education, vol. 34, no. 1, 29-45, DOI: 10.1080/03043790802710185 28. Gantogtokh, O., Quinlan, K.M. (2017). Challenges of designing interdisciplinary postgraduate curricula: Case studies of interdisciplinary master’s programmes at a research-intensive UK university.

Teaching in Higher Education, vol. 22, no. 5, 569586, DOI: 10.1080/13562517.2016.1273211 29. Wachter, C. (2012). Interdisciplinary teaching and learning for diverse and sustainable engineering education. Béraud, A., Godfroy, A.S., Michel, J. (Eds.), GIEE 2011: Gender and Interdisciplinary Education for Engineers. Sense Publishers, Rotterdam, p. 4763, DOI: 10.1007/978-94-6091-982-4_5 30. Walker, C. , Gleaves, A . (2016). Constructing the caring higher education teacher: A theoretical framework. Teaching and Teacher Education, no. 54, 6576, DOI: 10.1016/j.tate.2015.11.013

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

31. Lambrechts, W., Mulà, I., Ceulemans, K., Molderez, I., Gaeremynck, V. (2013). The integration of competences for sustainable development in higher education: an analysis of bachelor programs in management. Journal of Cleaner Production, vol. 48, 65-73, DOI: 10.1016/j.jclepro.2011.12.034 32. Perdan, S., Azapagic, A., Clift, R. (2000). Teaching sustainable development to engineering students. International Journal of Sustainability in Higher Education, vol. 1, no. 3, 267-279, DOI: 10.1108/14676370010378176 33. Jansen, A., Stevels, A. (2006). Combining eco-design and user benefits from human-powered energy systems, a win–win situation. Journal of Cleaner Production, vol. 14, no. 15, 1299-1306, DOI: 10.1016/j. jclepro.2005.11.023 34. van Nes, N., Cramer, J. (2006). Product lifetime optimization: a challenging strategy towards more sustainable consumption patterns. Journal of Cleaner Production, vol. 14, no. 15–16, 1307-1318, DOI: 10.1016/j.jclepro.2005.04.006 35. Vezzoli, C., Sciama, D. (2006). Life cycle design: from general methods to product type specific guidelines and checklists: a method adopted to develop a set of guidelines/checklist handbook for the eco-efficient design of NECTA vending machines. Journal of Cleaner Production, vol. 14, no. 15, 1319-1325, DOI: 10.1016/j.jclepro.2005.11.011 36. Filho, W.L., Shiel, C., Arminda P. (2016). Implementing and operationalising integrative approaches to sustainability in higher education: the role of project-oriented learning. Journal of Cleaner Production, vol. 133, 126-135, DOI: 10.1016/j.jclepro.2016.05.079 37. Figueiró, P.S., Raufflet, E. (2015). Sustainability in higher education: a systematic review with focus on management education. Journal of Cleaner Production, vol. 106, 22-33, DOI: 10.1016/j.jclepro.2015.04.118 38. 38. Arnesson, K., Albinsson, G. (2012). Integration of theory and practice in higher education. International Journal of Educational Research, vol. 53, 370-380, DOI:10.1016/j.ijer.2012.05.002 39. Schuster, L., Glavas, Ch. (2017). Exploring the dimensions of electronic work integrated learning (eWIL). Educational Research Review, vol. 21, 5566, DOI: 10.1016/j.edurev.2017.04.001 40. Brunsgaard, C., Dvořáková, P., Wyckmans, A., Stutterecker, W., Laskari, M., Almeida, M., Kabele, K., Magyar, Z., Bartkiewicz, P., Op’t Veld, P. (2014). Integrated energy design – Education and training in cross-disciplinary teams implementing energy performance of buildings directive (EPBD). Building and Environment, vol. 72, 1-14, DOI: 10.1016/j.buildenv.2013.10.011

381

Nadezhda Zhdanova, et al. - Comprehensive integration as an effective way of training future designers at technical universities (Integration as a way of training a designer)

41. Afdal, H.W., Spernes, K. (2018). Designing and redesigning research-based teacher education.

Teaching and Teacher Education, no. 74, 215-228, DOI: 10.1016/j.tate.2018.05.011 42. Azzam Ismail, M., Keumala, N., Dabdoob, R.M. (2017). Review on integrating sustainability knowledge into architectural education: Practice in the UK and the USA. Journal of Cleaner Production, vol.

140, 1542-1552, DOI: 10.1016/j.jclepro.2016.09.219 43. Karlsson, R., Luttropp, C. (2006). EcoDesign: what’s happening? An overview of the subject area of EcoDesign and of the papers in this special issue.

Journal of Cleaner Production, vol. 14, no. 15-16, 1291-1298, DOI: 10.1016/j.jclepro.2005.11.010 44. Zwolinski, P., Lopez-Ontiveros, M.-A., Brissaud, D. (2006). Integrated design of remanufacturable products based on product profiles. Journal of Cleaner Production, vol. 14, no. 15, 1333-1345, DOI: 10.1016/j.jclepro.2005.11.028 45. Donnelly, K., Beckett-Furnell, Z., Traeger, S., Okrasinski, Th., Holman, S. (2006). Eco-design implemented through a product-based environmental management system. Journal of Cleaner Production, vol. 14, no. 15-16, 1357-1367, DOI: 10.1016/j. jclepro.2005.11.029 46. Boks, C. (2006). The soft side of eco-design. Journal of Cleaner Production, vol. 14, no. 15-16, 13461356, DOI: 10.1016/j.jclepro.2005.11.015 47. Johansson, G., Magnusson, Th. (2006). Organising for environmental considerations in complex product development projects: implications from introducing a “Green” sub-project. Journal of Cleaner Production, vol. 14, no. 15-16, 1368-1376, DOI: 10.1016/j. jclepro.2005.11.014 48. Karakaya, A.F., Demirkan, H. (2015). Collaborative digital environments to enhance the creativity of designers. Computers in Human Behavior, vol. 42, 176-186, DOI: 10.1016/j.chb.2014.03.029 49. Benson, E., Fine, P. (2010). Sustainable design education rethought: The case for Eco-Modernism. Design Principles and Practices, vol. 4, no. 6, 163-176.

50. Chandrasegaran, S., Badam, S.K., Kisselburgh, L., Peppler, K., Elmqvist, N., Ramani, K. (2017).

VizScribe: A visual analytics approach to understand designer behavior. International Journal of Human-Computer Studies, vol. 100, 66-80, DOI: 10.1016/j.ijhcs.2016.12.007

51. Klement, M. (2017). Models of integration of virtualization in education: Virtualization technology and possibilities of its use in education. Computers & Education, vol. 105, 31-43, DOI: 10.1016/j.compedu.2016.11.006 52. Uncles, M.D. (2018). Directions in higher education: A marketing perspective. Australasian Marketing Journal (AMJ), vol. 26, no. 2, 187-193, DOI: 10.1016/j.ausmj.2018.05.009 53. Lilley, D., Lofthouse, V. (2009). Sustainable design education – considering design for behavioural change. Engineering Education, vol. 4, no. 1, 29-41, DOI: 10.11120/ened.2009.04010029 54. Grajewski, D., Diakun, J., Wichniarek, R., Dostatni, E., Buń, P., Górski, F., Karwasz, A. (2015). Improving the skills and knowledge of future designers in the field of ecodesign using virtual reality technologies.

Procedia Computer Science, vol. 75, 348-358, DOI: 10.1016/j.procs.2015.12.257 55. Tangwanichagapong, S., Nitivattananon, V., Mohanty, B., Visvanathan, Ch. (2017). Greening of a campus through waste management initiatives: Experience from a higher education institution in Thailand. International Journal of Sustainability in Higher Education, vol. 18, no. 2, 203-217, DOI: 10.1108/ IJSHE-10-2015-0175 56. Ueda, E.Sh. (2018). Student team integrating aspects of sustainability in practical design education.

International Journal of Sustainability in Higher Education, DOI: 10.1108/IJSHE-08-2017-0136 57. Lofthouse, V. (2006). Ecodesign tools for designers: defining the requirements. Journal of Cleaner Production, vol. 14, no. 15-16, 1386-1395, DOI: 10.1016/j.jclepro.2005.11.013 58. Zhdanova, N.S., Zhdanov, A.A., Lymareva, J.V., Ilyasheva, E.V., Nemtseva, Yu.S., Zakharchenko, T. (2015). Design product projecting made of recycled materials. International Journal of Applied Engineering Research, vol. 10, no. 24, 45137-45141.

Paper submitted: 18.07.2018.

Paper accepted: 13.08.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

382

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Original Scientific Paper

Journal of Applied Engineering Science

doi:10.5937/jaes16-17189

Paper number: 16(2018)3, 543, 383 - 390

CFD STUDY ON THE ABILITY OF A VENTILATED BLADE IN IMPROVING THE SAVONIUS ROTOR PERFORMANCE Rudi Hariyanto1*, Sudjito Soeparman2, Denny Widhiyanuriyawan2, Mega Nur Sasongko2 1 Department of Mechanical Engineering, Merdeka University, Malang – Indonesia 2 Department of Mechanical Engineering, Brawijaya University, Malang – Indonesia This study aims to overview the ability of ventilated blades to improve the performance of the Savonius rotor based on CFD simulation. Rotor performance is analyzed by static torque, pressure profile, a airflow profile and vortex area.

The boundary conditions for all simulations use the assumption that the wind speed is 5 m/s and the environmental pressure is 1 atm or 101325 Pa. The CFD simulation results have strengthen the published experimental results.

Ventilation with the 5% gap width of blade diameter (the SV05 model) gives the best performance. Static torque of SV05 model is 23.8% higher than the conventional Savonius (SC) rotor. Based on pressure and airflow profile of CFD simulation results, ventilation on the blade can add the mass flow rate of air and make the lift force work early on rotating angle of 0o and 165°. Ventilation on the blade also able to improve the critical condition and vortex area as seen in the SC rotor. This is shown by a static torque from the CFD simulation results. The static torque value of the SV05 rotor are 69% and 73% higher than the SC rotor at the 165° and 0° rotation angle.

Key words: Ventilated blades, Savonius rotor, CFD Simulation, Vortex INTRODUCTION Rotor Savonius with ventilated blades or the ventilated Savonius (SV) rotor is a new innovation found by Rudi Hariyanto in 2016. Experimentally, the SV rotor was able to improve the efficiency of the SC rotor by 32% [1]. Ventilated blades do not change the basic shape of the rotor Savonius. The design of the Savonius windmill still has a very simple rotor construction so the manufacture and maintenance is easy. The blades can even be made from a split drum. The Savonius windmill is particularly suitable in areas with low wind speeds and is suitable for areas with fluctuating wind speeds [2].

Many researchers have attempted to improve the efficiency of Savonius windmill. There are researchers who focus on obtaining the optimum geometry of Savonius rotor. Variations in the number of blades, aspect ratio, end plates and overlap [3-5]. There is research that focuses on the modification of blade form, Kamoji [6] has been researching fishing rod or hook blades. Savonius rotor with hook-shaped blade is able to increase Cp from conventional rotor savonus by 3%. Saha and Rajkumar [7] examined the use of twisted blades on a Savonius rotor. The twisted blades can improve the efficiency 2.62% higher than the SC rotor that have semicircular blades.

In addition to experimental research, numerical simulation programs are also increasingly used in the Savonius rotor research. Numerical simulation programs are able to predict fluid flow profiles. So that numerical simulation programs can be used to predict the types of losses that occur in airflow during rotating of the blades. Thus a numerical simulation program can help researchers in designing research to reduce the loss of the fluid flow.

As an example of a numerical simulation program is

Computational Fluids Dynamic (CFD) simulation. CFDs program is widely used to strengthen the analysis of the experimental research. Nakajima [8]. Gupta et al. [9], Yaakob et al. [10] and Sukanta R. [11] utilize the CFD program to find out how the width gap between the blades or overlap on the efficiency of Savonius turbine. JL Menet & N. Bourabaa [12] and Thong, Z. & Dietmar R., [13] use a CFD program to simulate windmill characteristics based on the fluid flow profile and wind pressure distribution passing through conventional Savonius blades.

Alesandro et al. [14] and several other researchers present graphs from rotating angles vs Ct results from the CFD simulations. The shape of the graph is similar to the Kamoji’s graph [9] although the research variables studied differ. Savonius rotor has the smallest torque on the rotation angle on 150° – 165°. Rudi Hariyanto et al. [15] used CFD simulation to determine the effect of vortex on the Savonius’s rotor performance. The results showed that the vortex is inversely proportional to Savonius's rotor torque.. The lowest torque is generated when the vortex has the largest diameter on the rotation angle of 165°.

This paper aims to study the ability of ventilated blades especially the SV05 model in improving the performance of Savonius rotor based on CFD simulation. CFD simulation results are expected to be able to show if the making of fixed ventilation on Savonius blades will add the mass flow rate of air that worked on the concave surfaces of the blades. The ventilation on Savonius blades will make the lift force work early on the blade and improve a critical position on the SC rotor. For futher, this paper will present a comparison between changes in airflow profile, torque and pressure distribution on SC rotor and SV05 rotor. The change in the airflow profile that rotates

* Department of Mechanical Engineering, Merdeka University, Malang – Indonesia. Jl. Terusan Raya Dieng 62-64, Malang, Indonesia 65146, rudy.hariyanto@unmer.ac.id

383

Rudi Hariyanto, et al. - CFD Study on the Ability of a Ventilated Blade in Improving The Savonius Rotor Performance

the blade will explain why the ventilated blades of the SV05 model can improve the efficiency of the SC rotor by 32%.

TURBULENCE MODELING Basically, the flow field around a blade of rotor model is highly turbulent like a result of research from Rudi Hariyanto et al. [15]. Thus, the main features of turbulence must be considered while choosing the computational technique to solve of a turbulent flow over the rotor model. Therefore, the selection of the turbulence model plays an important role for obtaining the desired computational results. Fluent simulation program for Savonius rotor can use the k-ε turbulence models standard. K-ε turbulence model of the standard has been able to provide an accurate simulation results. The use of standard k turbulence model has been observed in the investigations of Pope et al. [16].

The transport equation for (k-ε) Standard Model are as follows: (1) And (2) Where, Gb is the generation of turbulence kinetic energy due to buoyancy; YM is the contribution of the fluctuating dilatation in compressible turbulence with the overall dissipation rate, calculated as described in Effects of Compressibility on in the k-ε Turbulence Models; C1ε, C2ε and C3ε is constant each worth 1.44, 1.92, and 0.09. σk and σε turbulent Prandtl numbers for k and ε each worth 1 and 1.3.

Figure 1: Cross-section 2D models of the ventilated Savonius rotor Experimental study from a fluid flow profile and pressure distribution on the blade of Savonius rotor would require sophisticated equipment and are too expensive if done privately. Using simulation software can help to do a manipulation becomes easier and cheaper. In this study, two simulation models will be compared. Two models are one model of a ventilated Savonius rotor and the conventional Savonius rotor model. The selected model of the ventilated Savonius rotor is the SV05 model which has the best performance of the experimental test results [1].

There are two methods for present the data analisys in this study. The first method is to manipulate fluid flow profile and pressure distribution with the help of CFD code Fluent 13.0. Fluent simulation programs have the option to provide some various models of turbulence k-ε containing formulations of transport for turbulence kinetic energy (k) and the rate of energy dissipation (e).

The complete two-dimensional computational model includes an inner of rectangle containing the rotor model.

The next step in the simulation effort is the generation of the computational mesh. Sizing of mesh use advanced size funtion on: curvature and proximate. Thus, two fluid mesh sections are clearly noticeable.

RESEARCH METHODOLOGY The modeling of ventilated Savonius rotor is carried out with the help of Mastercam modeling software. Figure 1 illustrates the sectional geometry of simulated model of the ventilated Savonius rotor. The geometry of a simulated model is made according to the size of themodel tested experimentally [rudi]. The simulatedmodel has the diameter of the blades or a semi-circle (d) = D/2, the blade thickness = 0.0015*D, the width of overlap (e) = D/10 mm, the long side of the ventilation further pushed towards the blade (s) = 0,02 to 0,2).d and the width of ventilation (a) = 0,075*D. The computational domain of the rectangle (L*L) is taken as 0.5 m × 0.5 m.

The axis of symmetry of the ventilation long side placed right at 15° from the outer end of the blade Savonius.

384

Figure 2: Grid generation around blade of the rotor model

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Rudi Hariyanto, et al. - CFD Study on the Ability of a Ventilated Blade in Improving The Savonius Rotor Performance

Figure 2 show if the grid node density utilized in the blade domain was higher than in the other domains. Mesh near the surface of the blade look smaller in size than that around the rotor. This shows the speed gradient that occurs on the surface of the blade has a very sharp angle.

Set - up boundary conditions used in CFD simulation is the inlet v = 5 m/s (constant), outlet: pressure outlet = 101325 Pa (constant). The boundary conditions that used for all simulations are same as the simulation study of Rudi Hariyanto et al. [15] and the experimental study of Rudi Hariyanto et al. [1]. Solution method includes the scheme: semi - (simple), the momentum: order upwind second. In the fluid flow analysis, the convergence criterion use to solve the Navier-Stokes equation by iteration.

Then the selected report of result is the static torque value.

The second method is to measure the vortex area using the Image-J program. In order to get credible and reliable data then set scale for the length of the line on each image of the simulation results. Figure 3 shows the blade diameter of 10 cm which is the reference standard of the line length used to obtain the vortex area.

rotor models capable to improve a critical condition in the conventional Savonius.

From simulated result, the static torque average of all ventilated rotor is 1,4% - 23,8% higher than conventional Savonius. As for the SV20 model looks to have the static torque lower than the conventional Savonius for a 15° – 150° rotation angle. But the static torque average of SV20 is 101.4% higher than conventional Savonius. Torque improvement by model SV20 at 165° and 0° (180°) angle rotation angles are capable of covering a torque drop at 15° – 150°. The simulation result is identical to Cp from the published experimental result.

Figure 4: Effect of ventilation gap of the Savonius blades on the static torque

Figure 3: Set the length scale of the line with the Image-J program RESULTS AND DISCUSSION Effect of The Ventilation on Blade to a Static Torque Figure 4 shows if the ventilation on the blade of the Savonius rotor able to improve a statis torque especially at 165° and 0o (180°) rotation angles. The 165° and 0° (180°) rotation angles are the critical condition of the SC rotor. At the 165o rotation angle, the ventilated blades able to improve the SC torque between 169% -197% and at 0° (180°) rotation angle between 157% -189%.

On both angles, the wider the gap of the ventilation increases the greater torque. This show if Ventilation of SV models are able to make the lifting force that has been working at 165° rotation angle. Ventilation is able to advance the lifting force of the wind so that the wind is able to rotate the rotor before and during the rotation angle 0° better. It proves if the ventilation on the blade of the Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Ventilation on the blade is also capable of shifting the peak torque of the conventional Savonius from rotation angle of 30° to 45°. At the rotation angle of 45°, the torque value of the conventional Savonius decreased by 7%. The highest torque at 45° rotation angle generated SV05 model with value 24% higher than torque from conventional Savonius. The SV05 model is also capable of generating static torque larger than the SC model at every rotation angle. The CFD simulation result proved to strengthen the experimental results. Thus, the ventilation on the SV05 blades can improve the efficiency of the Savonius rotor.

Pressure analysis on the concave surface of the Savonius blade The concave side of the blade is the most important part of Savonius rotor for producing a positive force to rotate the rotor. The positive force is not only produced on the concave surface of the advancing blade (AB) but also on the concave surface of the blade (RB). Figure 5 show if the ventilation on the blades of the SV05 model is capable of providing increased wind pressure on the concave surface of advancing blade by 8% - 37% than the SC model at rotation angle of 135°-45°. In a critical position at 165° and 0°, the SV05 is able to produce wind pressure 28% and 8% greater than the SC blade.

385

Rudi Hariyanto, et al. - CFD Study on the Ability of a Ventilated Blade in Improving The Savonius Rotor Performance

Figure 5: The pressure values on the concave surface of the blade of SC and SV At rotation angle of 60° – 120°, the pressure value on the concave surface of AB SV05 is lower 0% - 6% than the SC model. This is due to the difference between the air pressure in front of the concave surface of the blade and around the rotor. The pressure on the AB's concave surface is greater than around the rotor. So in part of the air will flow out through the ventilation gap. Thus the wind energy used to drive the concave surface becomes smaller. The greatest flow losses occur at rotation angle of 90°. However, the average pressure produced by the Advance Blade concave surfaces of the SV 05 model is 17% larger than the SC model when the rotor rotates 180 degrees.

The interesting data result is the increase of pressure on the concave surface of the RB-SV05. The increase in pressure is 15% - 281%. The increase in mean pressure on the concave surfaces of the RB-SV05 model is 108% when the rotor rotates 180 degrees.

Figure 6 show if concave surface of the SV05 returning blades (no. 1) has pressure value greater than the conventional Savonius. Ventilation on the SV05 blades can increase the wind pressure on both concave surface of the blades. This will certainly increase the positive force that works to rotate the rotor. If the wind velocity that passing through the blade surface is considered equal for of SC and SV05 models, then the mass flow rate at the SV05 blade surfaces are greater than the SC model.

This means that the ventilation on the SV05 blade effectively adds to the mass air flow rate that drives the blade at rotation angle of 135° – 45°.

In addition, the wind is also pressing on the convex surface of the blade and will produce a negative force.

Based on the color indicates if the pressure in front of the convex side (No. 2) of the SV05 is lower than SC blade.

Thus, the negative force on the SV05 blade (no 2) was smaller than the SC blade. This causes the SV05 rotors to rotate more easily than the SC rotor.

Streamline analysis on the concave surface of the Savonius blade Ventilations on the blade connect two regions that have a different of air pressures. At the rotation angle of 165°, the concave area of the blade has a lower pressure than the outside of the blade.

Figure 6: Pressure distribution on the rotation angle of 0° (A. SC model and B. SV05 model)

Figure 7: The pressure and streamlines profiles at the rotation angle of 165° (A. SC model and B. SV05 model)

386

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Rudi Hariyanto, et al. - CFD Study on the Ability of a Ventilated Blade in Improving The Savonius Rotor Performance

Based on the color, the value of the pressure in front of the concave area is negative because it is lower than 1 atm in both SC and SV05 models. Through the ventilation gap, the wind seemed to be sucked into the inside of the concave surface. The wind becomes concentrated in front of the ventilation gap and causes increased pressure in the ventilation gap (no. 1). Figure 7 shows if the red color on the blade (no. 1) of SV05 model looks wider than the SC model.

Figure 7 also shows if the ventilation directs the wind so that the wind follows the concave surface of the the advancing (AB) and returning blades (RB). The number of streamlines in front of the concave surfaces of the returning (no. 2) and advancing blades (no. 3) of the ventilated Savonius is seen more numerous and more dense than the conventional Savonius. The phenomenon is also shown by the results of CFD simulations for the rotation angle 135° – 45°. It shows if the mass flow rate of air in ventilated blade is greater than the conventional Savonius at the rotation angle of 165°.

Increased mass flow rate of air is able to increase the pressure on the concave surface of the returning blade.

Figure 7 shows if the returning blade of the SV05 model has a wider green color than the conventional Savonius model. This shows if ventilation on the blades of SV05 model are able to make the lifting force that has been working at 165° rotation angle. Ventilation is able to advance the lifting force of the wind so that the wind is able to rotate the rotor before and during the 0° rotation angle better. It also proves if the ventilation on the blade of the rotor models can improve a critical condition on the conventional Savonius.

Vortex Comparison In Front Of The Blade Concave Surfaces The concave surface of the blades is the most important part of Savonius rotor for producing a positive torque.

However, at a certain angle of rotation, in front of the concave surface of the blade occurs a vortex that an adverse for the rotor rotation. Figure 7 showed if the vortex is formed in front of the concave surface of the blades.

Vortex will absorb some of the kinetic energy of the wind.

This causes the pressure acting on the concave surface of the blade to be low.

Figure 8 show if vortex is formed at turn angle from 135°– 15°. At the rotating angle of 135°, the ventilation on the SV05 blades are able to improve air flow by eliminating vortex that occurred in front of the concave side of the returning blade. At the rotating angle of 165°, the ventilation on the SV05 blades are able to improve air flow by eliminating vortex that occurred in front of the concave side of the advacing blade (no. 3 on fig. 4).

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Figure 8: Vortex area in front of the concave surface on advacing and returning blades The larger the area of the vortex then the greater the pressure difference that occurs between the center of the vortex to the outside of the circular vortex. So the loss of kinetic energy due to it is absorbed vortex also getting bigger. Ventilation on the SV05 model effectively reduces the vortex area. Thus, the kinetic energy acting on the concave surface of SV05 is larger than SC model.

So SV05 is able to produce the positive torque larger than SC model.

CONCLUSIONS The CFD simulation results show if ventilation with a 5% gap width of the blade diameter (SV05 model) has the best static torque. SV05 model can improve the static torque of Conventional Savonius by 23.8%. Based on the vortex, streamline profile and distribution pressure of the simulated results, ventilation on the SV05 blade is able to increase the air mass flow rate and increase the pressure on the concave surface of the advance and returning blade by 17% and 313% larger than the SC model when the rotor rotates 180 degrees. So the vents on SV05 are able to make the rotating angle so that the SV05 rotates more easily than the SC model at both turn angles.

ACKNOWLEDGEMENTS The authors would like to thank the Studio of Design & Systems Engineering, Mechanical Engineering Department, Brawijaya University. Special thanks to Mr. Darto (lecturer of Mechanical Engineering Department, Merdeka University) who has taught the mastercam program.

387

Rudi Hariyanto, et al. - CFD Study on the Ability of a Ventilated Blade in Improving The Savonius Rotor Performance

Appendix SC Model SV05 Model

388

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Rudi Hariyanto, et al. - CFD Study on the Ability of a Ventilated Blade in Improving The Savonius Rotor Performance

SC Model SV05 Model

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

389

Rudi Hariyanto, et al. - CFD Study on the Ability of a Ventilated Blade in Improving The Savonius Rotor Performance

REFERENCES 1. Hariyanto, R., Soeparman, S., Denny, W., & Mega, N.S. (2016). Experimental Study On Improvement The Performance Of Savonius Windmill With Ventilated Blade. International Journal Of Renewable Energy Research (IJRER), Vol.6, No.4, 1403-1409. (www.ijrer.org/ijrer/index.php/ijrer/article/view/4641/ pdf). Retrieved from https://www.researchgate.net/ profile/Rudy_Hariyanto2/publication/317014862_ Experimental_Study_On_Improvement_The_Performance_Of_Savonius_Windmill_With_Ventilated_Blade/links/591f0ccfaca272d31bd3d02c/ Experimental-Study-On-Improvement-The-Performance-Of-Savonius-Windmill-With-Ventilated-Blade.pdf 2. Tian, W., Song, B., VanZwieten, J., & Pyakurel, P. (2015). Computational Fluid Dynamics Prediction of a Modified Savonius Wind Turbine with Novel Blade Shapes. Energies, 8(8), 7915-7929. 8(8): 79157929. doi:10.3390/en8087915 3. Ali, M.H. (2013). Experimental Comparison Study for Savonius Wind Turbine of Two and Three Blades at Low Wind Speed. Int. J. Modern Eng. Research, Vol.3 issue 5, 2978-2986. (https://pdfs. semanticscholar.org/3714/f618f873b6dace1f547dec907f9773d884f1.pdf). Retrieved from https:// www.researchgate.net/publication/286211255_Experimental_Comparison_Study_for_Savonius_ Wind_Turbine_of_Two_Three_blades_at_Low_ Wind_Speed 4. Saha, U.K., Thotla, S., & Maity, D. (2008). Optimum design configuration of Savonius rotor through wind tunnel experiments. Journal of Wind Engineering and Industrial Aerodynamics, 96(8-9), 1359-1375. 96(89): 1359-1375. doi:10.1016/j.jweia.2008.03.005 5. Hayashi, T., Li, Y., & Hara, Y. (2005). Wind Tunnel Tests on a Different Phase Three-Stage Savonius Rotor. JSME International Journal Series B, 48(1), 9-16. 48(1): 9-16. doi:10.1299/jsmeb.48.9 6. Kamoji, M.A., Kedare, S.B., & Prabhu, S.V. (2009).

Experimental investigations on single stage modified Savonius rotor. Applied Energy, 86(7-8), 10641073. 86(7-8): 1064-1073. doi:10.1016/j.apenergy.2008.09.019 7. Saha, U.K., & Rajkumar, M.J. (2006). On the performance analysis of Savonius rotor with twisted blades.

Renewable Energy, 31(11), 1776-1788. 31(11): 1776-1788. doi:10.1016/j.renene.2005.08.030 8. Nakajima, M., Iio, S., & Ikeda, T. (2008). Performance of Double-step Savonius Rotor for Environmentally Friendly Hydraulic Turbine. Journal of Fluid Science and Technology, 3(3), 410-419. 3(3): 410419. doi:10.1299/jfst.3.410

390

9. Gupta, R., Biswas, A., & Sharma, K.K. (2008). Comparative study of a three-bucket Savonius rotor with a combined three-bucket Savonius–three-bladed Darrieus rotor. Renewable Energy, 33(9), 1974-1981.

33(9): 1974-1981. doi:10.1016/j.renene.2007.12.008 10. Yaakob, O., Tawi, K.B., & Suprayogi, D.T. (2010).

Computer Simulation Studies on The Effect Overlap Ratio for Savonius Type Vertical Axis Marine Current Turbine. IJE Transaction A: Basic, Vol. 23, 79-88. (http://www.ije.ir/abstract/%7BVolume:23-Transactions:A-Number:1%7D/=1032). Retrieved from http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.468.833&rep=rep1&type=pdf 11. Roy, S., & Saha, U.K. (2013). Computational Study to Assess the Influence of Overlap Ratio on Static Torque Characteristics of a Vertical Axis Wind Turbine. Procedia Engineering, 51, 694-702. 51(): 694702. doi:10.1016/j.proeng.2013.01.099 12. Menet, J.L., & Bourabaa, N. (2008). Increase in the Savonius Efficiency via a parametric investigation. ENSIAME Journal,59313. (https://www.researchgate.net/publication/228957621_Increase_ in_the_Savonius_ro…. Retrieved from https://www. researchgate.net/publication/228957621_Increase_ in_the_Savonius_rotors_efficiency_via_a_parametric_investigation 13. Zhou, T., & Rempfer, D. (2013). Numerical study of detailed flow field and performance of Savonius wind turbines. Renewable Energy, 51, 373-381. 51(): 373381. doi:10.1016/j.renene.2012.09.046 14. D’Alessandro, V., Montelpare, S., Ricci, R., & Secchiaroli, A. (2010). Unsteady Aerodynamics of a Savonius wind rotor: a new computational approach for the simulation of energy performance. Energy, 35(8), 3349-3363. 35(8): 3349-3363. doi:10.1016/j. energy.2010.04.021 15. Hariyanto, R., Soeparman, S., Denny, W., & Mega, N.S. (2016). Analysis the Vortex Effect on the Performance of Savonius Windmill Based On Cfd (Computational Fluid Dynamics) Simulation and Video Recording. International Journal Of Renewable Energy Research (IJRER), Vol.6, No.3, 931-937 (www.ijrer. org/ijrer/index.php/ijrer/article/download/4063/pdf).

Retrieved from http://www.ijrer.org/ijrer/index.php/ ijrer/article/view/4063/pdf 16. Pope, K., Rodrigues, V., Doyle, R., Tsopelas, A., Gravelsins, R., Naterer, G.F., & Tsang, E. (2010). Effects of stator vanes on power coefficients of a zephyr vertical axis wind turbine. Renewable Energy, 35(5), 1043-1051. 35(5): 1043-1051. doi:10.1016/j. renene.2009.10.012 Paper submitted: 17.04.2018.

Paper accepted: 02.07.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Journal of Applied Engineering Science

doi:10.5937/jaes16-17218

Paper number: 16(2018)2, 544, 391 - 397

OPTIMIZATION OF OIL FIELD DEVELOPMENT PROCESS BASED ON EXISTING FORECAST MODEL Vladimir Lushpeev1*, Andrey Margarit2 1 Saint Petersburg State University, Saint Petersburg, Russian Federation 2 Gazpromneft NTC LLC, Saint Petersburg, Russian Federation Process of oil-and-gas field development optimization under the conditions of a mineral raw material base deterioration and increase in a share of hard-to-recover reserves is the integral part of commercial production stage, especially in the last stage of development. Decisions regarding the optimization of the development system with contour water flooding under the conditions of a high water-cut of well production need to be made using additional instruments for the decision making, such as 1-D, 2-D and 3-D models. Using of simulation does not exclude a participation of experts in such work and imposes great responsibility on them in making decisions. Searching for optimal decisions under the oil-and-gas field development optimization based on physic-mathematical models together with the participation of recovery and development experts is the basis for managerial decision making in oil-and-gas production companies. This article shows the principles of the oil-and-gas field development optimization based on the existing forecast model and describes an industrial example of such optimization instrument usage together with the participation of the experts.

Key words: Oils, Optimizer, Development processes, Oil field INTRODUCTION The results of simulating a development optimization project are only the implementation of certain conditions with some degree of approximation. Therefore, the model has an error, a risk and restrictions and does not take into account all effects, but only those that can be described with a selected degree of detail [9, 10].

In this connection the restrictions unaccounted in the simulation can be arose at the different stage of the project implementation, for example, an operating wells equipment capabilities, probable non-predicted effects under a complex configuration of a real reservoir or the development system.

To successfully solve the problems regarding the oil field management the organization of multidisciplinary teams involving the experts of different branches had been effectively realized in many oil-and-gas companies. However, such approach requires the expenditures of a significant quantity of labour resources and organizational capacities that can be ineffective.

As one of the methods to overcome such restrictions may be the expertise of the optimization project by production experts on site. It also seems appropriate to always adhere to a principle of preliminary testing of the proposed modifications at separate areas of the existing site within the framework of pilot production.

A similar approach to the implementation of the development optimization project was studied on the example of one of the oil fields in Western Siberia.

THEORY AND EXPERIMENTAL METHODS In order to set the optimization task the variables optimization method xϵRn, represented by the variety of test parameters of the well performance in certain time intervals, was chosen. The bottom-hole pressure (BHP) was used as the x variable for this task. A net present value (NPV) was chosen as an optimizable function of cost f(x) for relatively short time interval (one to three years):

(1) Where p0, clp, cwi – the oil cost (which already includes costs for separation and tax levies), lifting costs and costs for water injection, accordingly,q0j.k qlpj.k qwijk – oil and fluid production rates, water injection rate for each well j at every time interval k in the time period ∆tk, d – yearly discount rate.

A boundary conditions was also imposed, i.е. x was defined as x={xϵRn; xd≤x≤xu} , where xd and xu designate the lower and upper bounds accordingly. In addition to these boundary conditions the nonlinear restrictions were taken into account, which were mathematically described

by vector function k(x)ϵRnk, whose components have to be nonpositive for possibility of decision. These components in the task to solve depend on the value of recovery efficiency (RE) estimated for the same time interval as for the NPV (this RE should be greater than the average value for development). As a result the optimization task can be stated as follows: x*ϵ arg

f(x), provided that k(x)≤0xϵX

(2)

where X* designates any certain (local) optimal decision,

* St. Petersburg State University, 199034, St. Petersburg, Universitetskaya Embankment, 7-9, Russia, lushpeev035@gmail.com

391

Vladimir Lushpeev, et al. - Optimization of oil field development process based on existing forecast model

the one of the variety, and the exact value of nonlinear restriction is adjusted so that it is nonpositive [8].

OPTIMIZATION METHODS In view of the above stated conditions, it was decided to apply the optimization methods without using of derivatives (derivative-free). Most of performed experiments are based on Generalized Pattern Search technique, GPS [1, 3, 7]. As mentioned above, in the task of development control, as a rule, there are multiple optimums with the similar values of cost function that is used in this technique. Besides, the universal procedure Particle Swarm Optimization, PSO was tested on the one of the examples [2, 5, 6]. In both algorithms the analysis of the searching area was performed using a set of points (a pattern for GPS and a variety for PSO), which were defined for each iteration.

It is worth pointing out that time costs for calculations by methods without using the derivatives depend on quantity of optimizable variables. Therefore, the optimization of the oil fields at the late stage of development with relatively large quantity of wells is time-consuming.

The applied method assumes the searching of the optimal value qwik.(x), i.e. the injection capacity of well (that restricts the number of variables), which will correspond to maximum NPV value. Then the searching results shall be tested on heological-hydrodynamical model (HHDM) of the oil field.

Figure 1: recovery profiles for the examined oil field The maps of the development and the oil-filled thicknesses as of April 30, 2016 are shown on Figure 2.

THE OIL FIELD DESCRIPTION The optimization method was tested on the one of the oil fields of PJSC «Gazprom neft», located in Western Siberia. The deposit is the layer-uplifted deposit with elements of tectonic shielding. Reservoir is generally represented by sandy rock with beds of siltstone, hard clays and carbonate rocks, which divide the formation into 9-28 permeable partings. The reservoir’s plan dimensions are 12x4.3 km, the total thickness is 51 m. The average thickness of clean sand is 25 meters, the initial oil-filled thickness is from 1.1 m to 25.6 m. The average porosity for the field is 18.8%, the average permeability 23.6 mD, the average initial oil saturation - 61%.

Exploratory drilling confirmed the oil reserves in 1989, the development was begun in 2006. Oil recovery obtained on vertical exploration wells amounted 723 barrels per day, and on horizontal producing wells – 3,189 barrels per day. Maximum oil recovery was about 17,000 barrels per day from 2007 to 2012. Water flooding process happens effectively and without sharp breachings of displacement front as evidenced by a gradual increase in water-cut (see Figure 1).

Before the optimization of water flooding in May, 2016 the oil field was on the fourth development stage with the oil production rate of 6,440 barrels per day (see Figure 1). The stock has 19 production wells (including 16 horizontal wells) and 15 injection wells (one of them is horizontal well).

392

Figure 2: The maps of the development and oil-filled thicknesses, 30.04.2016 For timely monitoring and optimization the oil field’s HHDM was used. It contains 70x145x114 grid blocks (768,000 of them are active) with horizontal dimension of the grid block of 100x100 meters and the average thickness of 0.3 meter. The quality of the model history match is considered satisfactory for measures to optimize and monitor of the effect.

DEVELOPMENT OPTIMIZATION PROJECT ON HYDRODYNAMIC MODEL As a result of solving the water flooding optimization problem, three injection wells were identified for possible changes of the injection mode (see Figure 3). Main recommendation relates to the injection well INJ-2: optimal decision is achieved particularly due to the well shutting Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Vladimir Lushpeev, et al. - Optimization of oil field development process based on existing forecast model

in (i.е. reducing the injection capacity of the well by 100%). Additionally, it was recommended to increase the injection in INJ-1 and INJ-3 wells. The obtained decision can be considered as physically reasonable. High injection capacity of the INJ-2 well can impede the effective hydrodynamic interaction of the INJ-1 well with the adjacent production wells and, consequently, reduces the oil displacement efficiency. The shutting in of the INJ-2 well together with the increasing the injection of the INJ1 well can redistribute the filtration flows and involve the remaining reserves in the area of the specified wells. It was recommended to increase the injection in the INJ-3 due to the good efficiency of filtration between the INJ-3 well and the nearest production wells. The total injection capacity of the INJ-1, INJ-2 и INJ-3 wells should be reduced by 23%. It was assumed that involving of the remaining reserves along with the savings of the costs for the water injection and the lifting of liquid as a result exceeds losses caused by a decrease in formation pressure.

Figure 3: Proposed measures of optimization for water flooding control; the relative changes of injection capacity for each well INJ-1, INJ-2 и INJ-3 are shown.

DECISION MAKING BASED ON SIMULATION RESULTS Optimal decision was studied by the experts who are responsible for actual exploitation and the development control of the oil field. Only the injection restraining in the INJ-2 well by 50% was approved from the proposed optimization measures. The reasons of partial implementation of the proposed measures are the following: Infrastructure limitations There are significant technological limitations for redistribution of the injection using the ground infrastructure.

Every well pad has a separate system of water intake and water injection into the formation. The INJ-1 and INJ-2 wells are located at the different well pads; so the redistribution of the injection between them is technically impossible. Note that the injection capacity increase in INJ-1 well without the redistribution of the injection from Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

the other wells is also impossible due to the ultimate capacity of the water intake system in the well pads.

Detailed and timely consideration of such factors becomes possible due to the existence of an integrated model of the oil field. However, the creation, updating and usage of that model for decision making in practice result in significant increase in computational complexity and labour costs. Comparison of an efficiency of the decision making based on the models of various levels of detailing can become the focus for further researches.

High level of uncertainty It is obvious that predictable positive effect with high degree of possibility shall to be confirmed in case of its value does not exceed the level of incoming uncertainties.

For this certain case there is a high degree of risk (which was not estimated quantitatively) that existing activities will have a negative impact on the development performances: • when increasing the injection in the INJ-3 well there is high possibility of the breaching of the displacement front to Т3 point (bottom hole) of the PRO-5 well; • a full shut-off of the INJ-2 well which provides about 30% of the injection at that area of the oil field can lead to more significant decrease of the formation pressure than the predicted one based on HHDM due to the fact that such prolonged shut-offs of the INJ-2 well have not been carried out and, consequently, the model can not be validated for this type of measures.

Therefore, a limited version of the suggested optimization recommendations was applied at the oil field - pilot-industrial works to limit the injection capacity of INJ-2 well by 50%. This allowed making rather small changes in the modes, measuring the dynamics and, depending on the result obtained, expanding or refusing the measures taken. This analysis can be one of the key moments for the estimation of the effect of the full INJ-2 well shut-off. This well has a key role in forming of the water flooding system in the studied oil field region, and therefore, the serious reasons are required for its full shut-off.

RESULTS OF THE OIL FIELD EXPERIMENT IN THE WATER FLOODING CONTROL As a result, it was decided to install a flow nipple in the INJ-2 well for a double reduction in the injection capacity.

Figure 4 shows the actual monthly operational reports (MOR) a year before the start of the experiment and for 8 months after. The area with the production wells PRO1, ... -5 and the injection wells INJ-1, -2, -3 is defined as the «area of the experiment» (the region with the highest predictable potential of the optimization). A significant decrease in the injection has been observed from May, 2016 (the start of the experiment), and the recovery stabilization and the water cut reduction are also observed.

In order to reasonably estimate the effect of the water

393

Vladimir Lushpeev, et al - Optimization of oil field development process based on existing forecast model

flooding optimization the parameters of the experiment area and the rest part of the oil field were compared. It is necessary because at the entire oil field, including in the experimental region, the operator continuously performs regular activities for the development monitoring. It is assumed that the effect from these activities is uniformly distributed over all wells of the oil field including the wells of the experiment area. Let’s start by considering the effect of decreasing in an injection capacity.

It is shown in Figure 5 that the injection capacity in the experiment area is compared with another one for other wells. Due to the different quantity of the wells in these two groups (3 versus 12), the different scales are used.

In order to simplify the effect estimation the total injection rates during the year preceding the experiment (from May, 2015 to April, 2016) were approximated on the mean value to determine a “basis” injection capacity.

As it is shown in the Figure 5, the value of this parameter is 7,090 barrels per day.

Figure 4: The results of the oil field experiment: a cumulative oil and fluid recovery and water cut of the PRO-1, ...

-5 and a cumulative injection for INJ-1, INJ-2 and INJ-3 wells. A year before the start of the experiment and for 8 months after.

Figure 5: Comparison of total injection from the INJ-1, INJ-2 and INJ-3 wells (the experiment area), with the remaining 12 injection wells

394

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Vladimir Lushpeev, et al. - Optimization of oil field development process based on existing forecast model

A variance of the injection liquid volume equal to 390 barrels per day (it is nor shown in the Figure), can be an approximate estimation of the error of the “basis” value.

According to a mean reduction of the injection capacity (1,660 barrels per day), the effect obtained is in the range of 311,000 to 502,000 barrels within 8 months (245 days) of the experiment. However, this estimation can not be considered as the final result: as stated above, it is required to take into account the operator’s continual optimizations over all wells of the oil field. To do this it is required to repeat the similar operation for wells which are not in scope of the experiment region. The decreasing in an injection capacity by “operator” is varied from 40,000 to 304,000 barrels: this effect shall be normalized by equal quantity of wells (3 wells) and be deducted from the previous one. Consequently, a “net” effect is in the range from 235,000 to 492,000 barrels and provides the appropriate decreasing of the operational costs.

Figure 6 shows the decrease in dynamics of fluid recovery for both areas of the oil field (a corresponding water cut for each area is also shown in the Figure). If we repeat the above described estimation process we will find that average decreasing of the fluid volume recovered in the experiment area and in the other part of the oil field is 19,000 barrels and 241,000 barrels, accordingly. Fluid rate in the remaining area is vastly larger than

the corresponding rate in the experiment area (even if we take into account a normalization coefficient that in this case is taken to be equal to 2.6). In this regard it was estimated that the optimization considered can provide relatively small increase in the fluid production volume and we evaluate this increase to be about 73,700 barrels. Assuming that the water cut in the experiment area decreases somewhat more noticeable than in the remaining part of the oil field, this result can point at higher relative oil recovery for the project optimization with respect to a general approach.

Oil recovery performances for both areas of the oil field are shown in Figure 7. In order to estimate the effect based on the oil recovery, the oil recovery is approximated by linear function which is often a rational approximation for a short terms (as well it is an approximation of exponential decline rate which in the scale of 1 year in this case does not distinguished from the linear). By means of such estimation it was found that average increase in a volume of oil recovery produced relative to the initial level for the experiment area and the rest of the oil field is 25,300 barrels and 24,700 barrels, respectively. If taking into account that the total oil recovery in the second area is about 1.9 times greater than in the first one, then we will obtain the effect approximately equals to 12,000 (it is better to round well in the end) barrels of oil.

Figure 6: Fluid recovery and water cut over “the experiment area” (PRO-1…-5) and over the remaining 14 development wells of the oil field Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

395

Vladimir Lushpeev, et al. - Optimization of oil field development process based on existing forecast model

Figure 7: Dynamics of oil recovery for the wells of “the experiment area” (PRO-1,…-5) and for the remaining 14 development wells of the oil field. Linear decline rate which is a good approximation for short period of time is shown for both groups of the wells.

Overall results of the experiment show that the application of the approach presented in this article is economically reasonable. The increase in oil recovery and the decrease in costs for water injection prevail over the slightly higher lifting costs.

CONCLUSIONS This article presents an example of optimization of the oil field development, estimated on the basis of the existing hydrodynamic model, and the stages of its implementation and results of the experiment are shown.

Under the existing conditions of the project implementation, the importance of the expert assessment by the specialists in the oil field, which are responsible for management of the field development, is presented. During the implementation of the optimization project they performed the risk analysis that led to adjustment of scheduled. It is probably that in case of more complicated integrated model and a qualitative uncertainty analysis, it would be possible to take this into account beforehand, but creating such a model requires much more organizational and time resources [4].

396

The results of the water flooding optimization experiment were recognized as positive, however, the analysis of the results showed that the model does not have a high predictive power, and it was decided to refuse the full shutting in of the INJ-2 well, leaving the current mode.

Besides, this experiment is a good example of initiating a cyclic working process based on the optimization of the development in a mature oil field: significant changes in the operating modes of the wells, which may be contrary to the usual approach to planning, lead to useful responses in the dynamics of the field factors. These results can be used in the updating and validation of HHDM and the transition to the next optimization iteration.

REFERENCES 1. Audet, C. and Dennis Jr., J.E. (2002) Analysis of generalized pattern searches. SIAM Journal on Optimization, 13(3), 889–903.

2. Eberhart, R.C. and Kennedy, J. (1995) A new optimizer using particle swarm theory. Proceedings of the Sixth International Symposium on Micromachine and Human Science, 39–43.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Vladimir Lushpeev, et al. - Optimization of oil field development process based on existing forecast model

3. Echeverría Ciaurri, D., Isebor, O.J. and Durlofsky, L.J. (2011) Application of derivative-free methodologies to generally constrained oil production optimization problems. International Journal of Mathematical Modelling and Numerical Optimisation, 2(2), 134–161.

4. Integrated Modelling Approach as Estimation Tool for Well Regimes and Gathering Network Impact on Oil Rim Development (Russian), 182007-RU SPE Conference Paper – 2016.

5. Kennedy, J. and Eberhart, R.C. (1995) Particle swarm optimization. Proceedings of IEEE International Joint Conference on Neural Networks, 1942– 1948.

6. Onwunalu, J. and Durlofsky, L.J. (2010) Application of a particle swarm optimization algorithm for determining optimum well location and type. Computational Geosciences, 14, 183–198.

7. Torczon, V. (1997) On the convergence of pattern search algorithms. SIAM Journal on Optimization, 7(1), 1–25.

8. Van Essen, G., Van den Hof, P.M.J. and Jansen, J.D. (2011) Hierarchical long-term and short-term production optimization.109 SPE Journal, 16(1), 191–199.

9. Lushpeev V.A., Tsiku Y.K., Sorokin P.M. (2014) Welltest during synchronous-separate operation. Life Science Journal 2014; 11(12s):351-353] (ISSN:10978135). http://www.lifesciencesite.com. 73 10. J.R. Fanchi, (2000). Integrated approach. ANCO «Institute of computer researches».

Paper submitted: 20.04.2018.

Paper accepted: 20.06.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

397

Journal of Applied Engineering Science

Original Scientific Paper

doi:10.5937/jaes16-17382

Paper number: 16(2018)2, 545, 398 - 403

APPLICATION OF VIRTUAL ENVIRONMENTS IN TRAINING OF ERGATIC SYSTEM OPERATORS Igor Petukhov*, Lyudmila Steshina, Andrei Glazyrin Volga State University of Technology, Russian Federation The paper discusses the approach to the training of ergatic system operators, which implies the application of virtual reality elements. It presents the review of the existing models of the virtual environment application, offers a system model of the operator's response, systemizes the main structural and functional elements of personal experience in virtual environments.

Key words: Ergatic control systems, Human-operator, Virtual reality, Virtual environment INTRODUCTION Most of the contemporary technological objects can be classified as ergatic systems (ES). The main peculiarity of such systems is their sociopsychological aspects that make them different from typical man-machine systems.

The modern technology causes the development of significantly altered ergatic functions. The implementation of virtual and alternate realities ensures the higher effectiveness of the ES operator training and professional activities. At the same time, their operational gnostic and practical functions alter, and their working practice transforms, too.

In this context, many issues of the arrangement of the ES operators' working practice in virtual environments are not yet understood, the optimization of the training process in virtual environments requires further discussions, the confident data on the efficiency of the operators trained with the application of a virtual content are missing.

The goal of this paper is to study the possibilities of virtual environments within the training of ES operating personnel.

THEORETICAL ANALYSIS The virtual reality (VR) essentially differs from other software systems, first of all, by its strong psychophysiological impact on a person. It deals with a so-called presence effect or immersion into the environment [1] that is empirically proved, in particular, by physiological measurements: the body temperature, muscular tension, cardiovasculare cardiogram, etc. [2], and is determined by multimodal environmental characteristics [3, 20].

It should be noted that the subjective sensation of immersion depends on many factors like concentration, imagination, proneness to immersion and detailed elaboration of images [4]. In addition, different components of immersion can make different impacts on the relevant types of tasks [5] and transfer the knowledge from the VR into the real environment [6].

Unfortunately, most of the current models consider the

user as a secondary element of the system or do not consider them at all. The focus of the semantic profile moves on the detailed elaboration of the imitation process [7] or on the software or hardware [8].

It its turn, VR is able to reproduce complex communicative connections [9], to imitate an internal psychological state [10], to activate cognitive functions [11], reproduce professional practices [12].

However, beside a positive effect this technology has a negative effect either; for example, cognitive dissonance. This phenomenon appears in the person mind due to discrepancies between their previous experience and the present situation or due to logic inconsistency of the virtual reality and a physical object, according to L. Festinger. In this situation a key role belongs to the depth of immersion into the virtual reality and the level of human-machine interaction. At the same time, the implementation of operators' commands in the virtual reality has a certain difficulty [13]. In this connection, the study of interface interaction between an operator and a technical system in the virtual reality with the aim of creating a better effect of the physical presence and better control using adaptive interfaces is rather relevant at the moment.

Currently, ensuring situational awareness (SA) is considered as the main problem of the ergatic control system (ECS). Operators are able to make effective decisions providing a proper operating process only if they have a required situational awareness. [14, 15].

Therefore, there is a number of questions connected with the definition of the subjective experience of a person's presence in the environment, revealing key elements and categories of such presence, and developing a human-oriented model for effective training simulation, etc.

THE REVIEW OF THE CURRENT MODELS OF VIRTUAL ENVIRONMENT APPLICATION In comparison with traditional software interfaces VR comprises a great variety of objects, interactive techniques, behavior scenarios, communicative practices, etc. All this causes problems in the virtual environment

* Volga State University of Technology, Russian Federation, PetuhovIV@volgatech.net

398

Igor Petukhov, et al. - Application of virtual environments in training of ergatic system operators

Figure 1: The I3 and I4 models [19] modeling process that are related to the realism of the environment (design and implementation of the 3D content) [16], the co-existence of virtual and physical objects that can make an information exchange between each other [17], the interaction of the environment and a person (human characteristics, psychophysiological features, multimodal interaction) [18], etc.

In general, the VR model can be described through conceptual elements with semantic connections. In this context, the most representative model is I3 [19] that can be completed up to the I4 model shown in Figure 1.

It should be emphasized that generic categorization of conceptual elements cannot give a complete idea of the functioning and integration of the environment with the user. Also, there is neither division between a virtual environment, a technical system and a user nor defined two-sided communicative connections.

Another approach to the VR modeling defines the framework of the virtual environment that is characterized as external in relation to the user. This approach establishes

Figure 2: The simple diagram of the human-machine interaction, where (a) is a VR system, (b) is a virtual environment, (c) is a user, and (d) is interactive communication [16]

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

a two-sided communicative connection between the user and VR. Therefore, the key difference is the introduction of a mediator that integrates the environment with the person. [16].

The VRID (Virtual Reality Interface Design) methodology presents a more formalized system of functional connections. This approach outlines the main functional levels of a virtual environment using technical and software decomposition. VRID deploys design templates, namely MVC (Model-View-Controller) [18].

According to this approach, the VR system can be described through three interconnected levels; (i) software; (ii) data input control; (iii) an interface (objects and data).

On the other side, the VRID methodology renders the user as a source of inputs that come to the interaction module. There is no clear division between a virtual environment, technical equipment, and other elements.

Semantic connections do not give a complete idea of the functional loading of a system. Therefore, we need a more focused approach to modeling, which could take into account a set of specific objectives being attained in the environment and present the detalization of the process.

One of the first attempts to outline and describe human and technical elements that establish independent spheres of interaction was made within the Conceptual VR Model (CVRM) [21].

This model represents the user as a complex system of input and output data processing. The emphasis is made on the functional loading of perceptive and muscular elements that provide the interaction with VR within the psychological (cognitive) circuit. The work efficiency in the environment depends on the individual ergonomics, work experience and sensibility to the virtual environment. CVRM moves the main focus of structuring the functional connections on the user and restricts the virtual environment by communication connections in the form of effectors and sensory responses. At the same

399

Igor Petukhov et al. - Application of virtual environments in training of ergatic system operators

time, this model does not take into consideration the objective sphere of the virtual environment application, where different scenarios and manipulations can be applied.

The abovementioned can help make preliminary conclusions on the criteria of modeling the virtual environment.

The existing models do not represent the interaction process between a person and a virtual environment in full. They lack important factors of gaining subjective experience of the presence in the environment, which can make a great impact on the environment efficiency.

We should take into account the simulation sphere of the environment application, which should comply to the intended use of the simulator and reproduce the specificity of training and simulated work in the virtual reality.

THE MODEL OF A VIRTUAL ENVIRONMENT FOR THE TRAINING OF ERGATIC SYSTEM OPERATORS The analytical study of system models and the pathways of the operator's actions makes it possible to reveal the most significant stages and specificity of operator's functions that influence the efficiency of the operating activity through: • accepting and perceiving information (the initial stage of receiving information from different sensory sources) [22]; • evaluating and processing information (finding out the most important information and, then, its analysis and evaluation) [23]; • making a decision (planning a certain activity) [24]; • implementing the decision made (physical movement) [24].

One of the key elements of different models of operator's responses is reasoning or conclusion that make the ground for making a final decision [25, 26]. Making a final decision is based on the analysis of individual facts of workings, which are transformed into an integrated pattern of the workings [27].

Therefore, it is essential to include a mechanism of accumulation of priority information to the degree when we get the possibility to compare this information with some familiar situation or model of an operator's behavior. It is necessary to give a detailed description of the cognitive and motor levels and to clarify the control circuit in the model of operator's responses.

According to the proposed model shown in Figure 3, an operator receives information through sensory canals and by means of a sensory analysis forms a situational model prototype (SMP). Then, an access to the Conceptual Models (CM) database (DB), which has been formed by an operator earlier, and a content-addressable search are performed using key attributes of the situational model. It should be noted that DB comprises gained experience (mental templates) of planning and responding to different situations of the operating activity.

This process fills ‘gaps’ of the situational model and it is converted into an integral situational model (ISM) which implies a description of the current situation, a forecast for its development, generation of response variants, an analysis and selection of a response variant in accordance with the current criteria. A newly-formed integral situational model (ISM) is added to the Conceptual Models database for the further usage. Then, a scheme of motor response (CMR) is created using a set of motor programs included into the database, and a motor program (MP) is generated and implemented. Further, this cycle is repeated.

Thus, the proposed system model accumulates the information flow of the background knowledge at the cognitive level. If there is 'enough' information for making certain decisions, such information is instantiated in ISM. In this case we can create a simulation environment that would take into account ISM and the process of extracting background knowledge and compare them with the real situation. Therefore, the reproduction of subjective perception of the professional environment is seemed to be of a higher priority than the copying of a working scenario without consideration of an operator's internal state.

Figure 3: Thesystemmodelofan operators' response

400

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Igor Petukhov et al. - Application of virtual environments in training of ergatic system operators

Subjective experience of the presence and actions in the virtual environment depends on numerous conditions.

K. Stanney et al. [3] state that the efficiency of the user in the virtual environments is, first of all, affected by the human factor. The main emphasis is made on psychic movements, human perception, multimodal interaction, parameters of objectives connected with mental processes.

D. Boyd [25] describes the connection between the efficiency of the virtual environment and the individual subjective experience and, in addition, the proven experience of the VR application for research into effective human visual perception in the real life [6].

N. Pares and R. Pares [16] define the human factor using the concept of virtual subjectiveness that determines the effectiveness of different objective attainment and the design of the virtual environment.

The essential element of the perception of a virtual environment is its representation as the latter establishes the balance between the physical equipment and the media content at the level of the reproduction of human metaphors and the recognition of cognitive patterns most often used in the everyday practice. The probability of the recognition of cognitive patterns depends on the confidence level of the representation of the virtual environment presented by its attribute groups shown in Table 1.

The breach or incorrect transfer of the patterns may affect the environment efficiency in part of realism and generation of the psychophysiological state of a trainee.

The first group of attributes (spatial attributes) allow the operator to orient in the space and receive information on the controlled objects, their location, speed and the distance.

Beside the spatial parameters there are principles of physics and physical phenomena that determine the dynamic state of certain objects and the environment in general. If the spatial attributes mostly help orient in the environment, than the physical ones help perform certain

actions with objects and the environment. This results in the evaluation of the situation and making certain decisions. In addition, the physical elements have the potential of a subjective psychophysiological impact and are often used for the simulation of hazardous environments.

The object simulating attributes of the representation specificate the certain object simulating activity and draw the user's attention to it with the aim of effective training; they build the main logic of training, namely: succession of actions, accuracy of manipulations, detailed visualization of the objective sphere, a possibility of a flexible control over the situation.

The emotional and psychological attributes help influence the emotional state of the user and simulate their behavior in emergency situations, in the rush and monotonous work modes.

The social grouping attributes ensure the interaction of users in the virtual environment, possibilities of their communication, group work, and in fact are determinative for the ergatic system.

The variation of priorities of representation attributes gives the opportunity to simulate a wide range of ergatic systems in the virtual environments with emphases on different objectives depending on the targeted function.

CONCLUSION Application of virtual environments for the training of ES operators has its distinct advantages. First of all, they are associated with the reduced cost of training, the diversity of cognitive patterns of response and motor programs for emergency work modes. On the other side, application of virtual environments has a certain risk connected with inadequate responses.

According to the proposed system model of the operator's response, the confidence level of the conceptual model (the correspondence of the ES state to the reality) formed by the operator in the virtual environment is provided through the confidence level of the formation of representation attributes.

Table 1.

Representation attribute

Description

Spatial

the distance and scale, sizes of objects, speed and movement of objects (of the virtual content)

Physical

reproduction of principles of physics and physical phenomena similar to natural, simulation of time intervals

Object simulating

simulation of the certain practical activity connected with the training process and simulation activity details

Emotional and psychological

detalization of environment parameters connected with the influence on emotional distress or phobic reactions

Social grouping

formation of the environment of the collaborative group activity of operators (users) in the virtual environment, communicative aspects

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

401

Igor Petukhov et al.- Application of virtual environments in training of ergatic system operators

In this context, we can control these attributes in dependence of the target function of training, for example, for simulation of group interaction the social grouping and object simulating attributes are of the highest priority.

The development of the model of representation that takes into account all attributes of representation is a promising direction for research and, at the same time, involves the greatest difficulty and scientific interest. This model determines a set of objective and psychological references that should synchronize the situational model prototype (SMP) and integral situational model (ISM).

The next level objective is the development of the models of interaction between the user and the technical system in the virtual environment, the operator's gaining subjective experience in the virtual environments and development of the conceptual model describing a current state of the ergatic system.

The research outcomes are obtained by the support of RF Ministry of Education and Science grant No. 25.1095.2017/4.6.

REFERENCES 1. Dinh H. Q. et al., (1999). Evaluating the importance of multi-sensory input on memory and the sense of presence in virtual environments. U: VirtualReality 1999, IEEE. 222-228.

2. Mestre D.R., & Fuchs P., (2006). Immersion etprésence. U Fuchs P, Moreau G, Berthoz A, &Vercher JL (Ur.), Le traité de la réalitévirtuelle. Paris: EcoledesMinesdeParis 309-38.

3. Stanney K.M., Mourant R. R., , & Kennedy R. S., (1998). Human factors issues in virtual environments: A review of the literature. Presence: TeleoperatorsandVirtualEnvironments, 7(4), 327-351.

4. Laha B., & Bowman D., (2012). Identifying the benefits of immersion in virtual reality for volume data visualization. U: Immersive visualization revisited workshop of the IEEE VR conference. 1-2.

5. Bailey J, Bailenson JN, Won AS, ,& et al., (2012).

Presence and memory: immersive virtual reality effects on cued recall. U: The International Society for Presence Research Annual Conference, Philadelphia, Pennsylvania. 24-26.

6. Davis, A. (2015). Virtual Reality Simulation: An Innovative Teaching Tool for Dietetics Experiential Education. TheOpenNutritionJournal, 9 (Suppl 1-M8), 65-75.

7. Aggarwal R. et al., (2007). Proving the effectiveness of virtual reality simulation for training in laparoscopic surgery. Annalsofsurgery, 246(5), 771-779.

8. Psotka J., (1995). Immersive training systems: Virtual reality and education and training. Instructionalscience, 23(8), 405-431.

9. Prasolova-Førland, E., Fominykh, M., Darisiro, R., Mørch, A.I., & Hansen, D. (2014). Preparing for In-

402

ternational Operations and Developing Scenarios for Inter-cultural Communication in a Cyberworld: A Norwegian Army Example. U: LectureNotesinComputerScience. 118-138.

10. Cha M. et al., (2012). A virtual reality based fire training simulator integrated with fire dynamics data.

FireSafetyJournal, 50, 12-24.

11. Hanson K., & Shelton B. E., (2008). Design and Development of Virtual Reality: Analysis of Challenges Faced by Educators. EducationalTechnology&Society, 11(1), 118-131.

12. Wang P. et al., (2005). Designing a virtual reality simulator for neurosurgery. U: The seventh international conference on virtual reality, Laval, France. 19-22.

13. Gribova V.V., Kleschev A.S., ,Krylov D.A., , MoskalenkoPh. M., , Timchenko V.A., , Fedorischev L.A., &Shalfeyeva E.A., (2016). The Base Technology for Intelligent Services Development with the Use of IACPaaS Cloud Platform. Part 3. An Interface Development and an Example of Applied Services Creation..SoftwareEngineering, 7(3), 99-107.

14. Endsley M. R., (1995). Toward a theory of situation awareness in dynamic systems. Human Factors: The Journal of the Human Factors and Ergonomics Society, 37(1), 32-64.

15. Endsley M. R. & Jones W., (2013). Situation awareness. U The Oxford Handbook of Cognitive Engineering. (str. 88-108).

16. Parés N.,&Parés R., (2006). Towards a model for a virtual reality experience: the virtual subjectiveness.

Presence: TeleoperatorsandVirtualEnvironments, 15(5), 524-538. doi:10.1162/pres.15.5.524 17. Cakmak H, &Kuhnapfel U., (2000). Animation and simulation techniques for vrtraining systems in endoscopic surgery. U: EurographicsWorkshoponAnimationandSimulation. 173-185.

18. Tanriverdi V., & Jacob R. J. K., (2001). VRID: a design model and methodology for developing virtual reality interfaces. U: The ACM symposium on Virtual reality software and technology. 175-182.

19. Al-Ahmari A. M. et al., (2016). Development of a virtual manufacturing assembly simulation system.

AdvancesinMechanicalEngineering, 8(3), 1-13. doi:10.1177/1687814016639824 20. Burdea G., Richard P., &Coiffet P., (1996). Multimodal virtual reality: Input-output devices, system integration, and human factors. InternationalJournalofHuman-ComputerInteraction, 8(1), 5-24.

21. Latta J. N., Oberg, D. J. (1994). Conceptual virtual reality model. IEEE Computer Graphics and Applications, 14, 23-29.

22. Petukhov I., (2011). Temporal Aspects of Human-machine Interaction in the Perception of Visual Information. U: ACHI 2011, The Fourth International Conference on Advances in Computer-Human Interactions, Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Igor Petukhov et. al - Application of virtual environments in training of ergatic system operators

Gosier, Guadeloupe, France. 43-47.

23. Petukhov, I.V.,Rozhentsov, V.V., &Aliev, M.T. (2007).

On the accuracy of evaluations of temporal characteristics of visual perception. Bulletin of Experimental Biology and Medicine, 144(2), 267-268.

24. Petukhov I., Steshina L., Kurasov P., &Tanryverdiev I., (2016). Decision Support System for Assessment of Vocational Aptitude of Man-machine Systems Operators. U: IEEE 8th International Conference on Intelligent Systems (IS'16). Sofia, Bulgaria.778-784.

25. Nilsson D., &Kinateder M., (2015). Virtual Rality Experiments-The Future or a Dead End. U: 6th InternationalSymposiumHumanBehaviourinFire, Cambridge. 13-22.

26. Whitworth B. (2010).Simulating space and time // arXiv preprint arXiv:1011.5499.

27. Boyd D. (2000).Depth cues in virtual reality and real world //Department of Computer Science. Providence, Rhode Island, Brown University. Vol. 26.

Paper submitted: 08.05.2018.

Paper accepted: 17.07.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

403

Original Scientific Paper

doi:10.5937/jaes16-18435

Journal of Applied Engineering Science

Paper number: 16(2018)2, 546, 404 - 409

A SIMPLE WAY TO ASSESS THE SPECTRAL LINES INFORMATIVITY OF A CHI-SQUARE MOLECULE IN ANALYZING SMALL SAMPLES OF BIOMETRIC DATA Alexander Ivanov1*, Alexei Gazin2, Yulia Serikova3 1 Penza Scientific Research Electro-Technical Institute, Russian Federation 2 Lipetsk State Pedagogical P. Semenov-Tyan-Shansky University, Russian Federation 3 Penza State University, Russian Federation The prerequisites for reducing the test sample chi-square Pearson test size from 400 to 32 or fewer examples while maintaining its power are considered. The urgency of the problem results from the fact that when learning and testing the biometric identification means to identify the personality, it is not possible to use large volumes of learning and test samples. The conditions under which the chi-square test on small samples from the continuous distribution of values becomes a discrete distribution of values are formalized. Normal and uniform laws of values distribution use histograms with uniform intervals, which accurately relate the central intervals of the histogram to the mathematical expectation calculated on the test sample. 16 experiments shown that the chi-square-synchronized test built on histograms with four equal intervals has a discrete probability spectrum consisting of only 20 significant spectral lines. A simple method for estimating the informativity of each of the important spectral components is proposed. Traditional statistical assessments can be strengthened by the following deeper level of the spectral components analysis of small samples of biometric data. The second deeper level of statistical processing should be substantially more powerful. Under the same conditions, the computational informativity increases from 2.22 bits to 24.95 bits due to the transition from simple continual calculations to discrete calculations of high computational complexity.

Key words: Chi-square Pearson test, Small samples of biometric data, Quantum effects of continuums representation by small samples, Distributivity, Histogram, Spectral lines INTRODUCTION - THE PROBLEM OF PERFORMING STATISTICAL ASSESSMENTS ON SMALL SAMPLES OF BIOMETRIC DATA In practice, data is always low in amount. This problem is especially acute when solving the problem of biometric identification of an individual [01, 02, 03]. A person is a very complex object of high dimensionality, which significantly complicates statistical multidimensional estimates. The same problems arise in the transition from a person's personal biometrics to collective biometrics of groups of people. For example, when testing new drugs, it is required to involve a sufficiently large number of biometric data of patients with the necessary pharmacistsâ&#x20AC;&#x2122; disease. Then they should be treated, and then, having typed the necessary amount of statistics, confirm the harmlessness of the new pharmacological drug. This approach to testing new drugs usually takes several years.

Obviously, other things being equal, a two or three times reduction in the amount of the test sample can reduce the time of testing the biometric data two to three times and, accordingly, shorten the time for the withdrawal of pharmaceutical preparations to the market. That is, the correct reduction in the test samples volume is one of the effective ways to reduce the cost of drugs and reduce the time of their withdrawal to the market.

Statistical estimates often use the chi-square Pearson test because it knows the analytical description of the asymptotic distribution of values:

(1)

where m is a positive integer of degrees of freedom, calculated through the number of columns of the histogram - k from the formula m = k-3, is the Euler gamma function.

Standardized recommendations for the application of the chi-square test [04] require that for each column of the histogram an average of about 5 or more samples be counted, and the sampling itself would be from 200 to 400 experiments. In biometrics, it is not always possible to obtain such a large sample. In connection with this, the actual task is to reduce the volume of the test sample to 16 examples while maintaining the power of the Pearson chi-square test. Naturally, the asymptotic relation (1) ceases to work with such small samples and other mathematical constructions must be created in its place.

QUANTUM EFFECTS ARISING FROM THE REPLACEMENT OF STATISTICAL CONTINUA BY SMALL SAMPLES OF BIOMETRIC DATA Let us suppose that there is a samples set of the normal distribution of values, containing 16 experiments.

Next let us build a histogram of data, consisting of 4 columns. The columns width of the histogram will be chosen acording to the following rule:

* Penza Scientific Research Electro-Technical Institute, Russian Federation, alexander.ivanov.04@bk.ru

404

Alexander Ivanov, et al. - A simple way to assess the spectral lines informativity of a chi-square molecule in analyzing small samples of biometric data

square values from this data: (2) Obviously, the histogram synthesizing operation is nothing else than an operation of quantizing a continuous normal continuum by a quantizer with 4 output states.

Naturally, such a quantizer generates quantization noise of a continuous normal distribution. This situation is shown in Figure 1 below.

Figure 1: The effect of quantizing the continuous density of the values distribution by representing it with a histogram (upper part) and quantization noise (the lower part of the Figure) Obviously, quantization noise causes the ratio (1) to stop working. It is also obvious that it is possible to improve the situation, for example, by smoothing the quantization noise [05, 06, 07] with a simple digital filter. In this case, the selection of the parameters of the digital smoothing filter will never allow for approaching the relation (1), because the smoothing filtering will lose the independence of the given chi-square functionals [08].

NUMERICAL MODELLING OF CHI-SQUARE DISTRIBUTION LAWS FOR SMALL SAMPLES There were no powerful computers at the Pearson's times. Today the situation is different; it is possible to use the software generator of many samples from 16 experiments. Further it is possible to calculate a set of chi-

(3) where bi is the number of experiments that hit the i-th interval of the histogram, Pi is the expected theoretical probability of falling into the i-th interval of the histogram under the normal distribution law.

Further, it is possible to construct the densities of chisquare values for the normal law of pseudo-random numbers and the uniform law of the original data. As a result, two curves are obtained for a million implementations, shown in Figure 2.

It can be seen from Figure 2 that the final density distributions of the values are non-monotonic. There are significant periodic bursts superimposed on some smooth chisquare distributions with different number of freedoms. It is obvious that the number of freedoms for normal initial data will be much less than the analogous number of freedoms for uniform initial data.

There is an illusion that the imposed oscillations are of a random nature and can be eliminated by increasing the volume of the number, accounted for the implementation of samples of 16 experiments. One can repeat the experiments and one will get the same result. It is possible to increase the number of realizations in tens or even hundreds of times, however, the vibrational components do not decrease their amplitude. All this testifies to the non-random (deterministic) appearance of the vibrational components of the distribution density values of Figure 2.

Vibrational superimpositions on smooth distribution functions of values are a consequence of the periodic noise quantization. The quantization noise envelope (the lower part of Figure 1) is periodic; it is the periodic nature of the quantization noise that leads to the appearance of a periodic component of the distribution density of Figure 2. Due to the partial dissynchronization of the periodic components of the quantization noise, there is a slight attenuation of the amplitude when data is accumulated, but complete elimination of the vibrational components for small samples of 16 examples never occurs.

Figure 2: Chi-square distribution functions for normal source data (continuous line) and for raw data with a uniform law of distribution of values (dashed line) Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

405

Alexander Ivanov, et al. - A simple way to assess the spectral lines informativity of a chi-square molecule in analyzing small samples of biometric data

At the weakening point of the periodic components of the quantization noise [05, 06, 07], averaging can be done differently and emphasize the periodic components present in the quantization noise [09, 10, 11, 12, 13].

Quantization noise synchronization is significantly improved if the maximum and minimum of the second and third histogram intervals are tied to the mathematical expectation of biometric data sampling: where E(x) is the mathematical expectation of each sample, over which the chi-square functional is calculated.

Then the width of the four intervals of the histogram should be calculated by the following formula: (4)

where is the standard deviation of each sample involved in the calculation of chi-square values distribution.

(5) The fulfillment of the synchronization conditions (4) and (5) results in the fact that the continuous spectrum of the continuum of the chi-square test states becomes a discrete spectrum. An example of the spectral lines location of chi-square functionals for normal data and uniform initial data is shown in Figure 3. Physics [14, 15] and chemistry [16] study the spectra of hydrogen, oxygen, lithium, sodium and other substances molecules. The spectrum of the output states of the chi-square functional (3) can be considered as some mathematical molecule with four allowed orbitals and 16 electrons on them.

Figure 3: The position of the spectral lines of chi-square functionals (fine lines) for normal data and (thick pale lines) for initial data with uniform distribution Figure 3 shows that the spectral lines amplitudes of normal and uniform data are different. In fact, one is dealing with two different spectral portraits of uniform and normal data. Different technologies can be used to recognize portraits. In particular, artificial neural networks trained according to GOST R 52633.5 [17] can be used.

The standardized learning algorithm [17] is convenient because it has linear computational complexity and it is possible to estimate the end result of the operation of a standard neural network. In particular, one can estimate the informativity of each spectral component as a module of the logarithm of the amplitudes ratio of identical spectral lines:

It follows from relation (6) that spectral lines with the same amplitude (for example, spectral line No. 15 in Table 1) have zero informativity. On the contrary, the spectral component with the number 13 will have the maximum informativity. It follows from relation (6) that spectral lines with the same amplitude (for example, spectral line No. 15 in Table 1) have zero informativity.

(6) On the contrary, the spectral component with the number 13 will have the maximum informativity.

Table 1: The Spectral Lines Amplitudes of Chi-Square Molecule for Normal and Uniform Input Data S/N

1

2

3

4

5

6

7

8

9

10

x

0.0

0.28

0.98

1.16

1.2

1.56

1.78

1.9

1.98

2.2

Norm.

0.128

0.168

0.130

0.043

0.071

0.103

0.061

0.037

0.011

0.036

Equal to 0.058

0.068

0.074

0.014

0.144

0.053

0.102

0.060

0.003

0.077

S/N

11

12

13

14

15

16

17

18

19

20

x

2.26

2.42

2.57

2.72

2.94

3.06

3.12

4.18

4.44

4.5

Norm.

0.015

0.007

0.002

0.042

0.025

0.056

0.003

0.003

0.005

0.056

Equal to 0.005

0.035

0.018

0.018

0.025

0.114

0.005

0.004

0.007

0.114

2

2

406

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Alexander Ivanov, et al. - A simple way to assess the spectral lines informativity of a chi-square molecule in analyzing small samples of biometric data

In the limit, the final informativity of the finite neural network solutions can be estimated as the sum of all particular informativities along the 20 spectral lines of Table 1.

the situation is different, one can quite spend several hours of desktop computer work to reduce the requirements for the size of a reliable test sample.

OBTAINING DATA FOR CALCULATING THE SPECTRAL LINES OF THE SAMPLE CHI-SQUARE FUNCTIONAL PORTRAIT FROM 16 EXPERIMENTS

EVALUATION OF THE SAMPLE CHI-SQUARE FUNCTIONAL INFORMATIVITY FROM 32 EXPERIMENTS IN THE CONTINUUM AND DISCRETE VERSIONS

In order to obtain a portrait of the output states spectrum of a chi-square molecule with 16 freedoms, millions of (7) realizations with 16 experiments in each are required. In order to obtain many implementations, the authors take one large sample of 32 experiments. If one randomly selects a small subsample of 16 experiments from a large sample of 32 experiments, then one gets millions of different options: Further, let us calculate the chi-square functional (3) on each of the millions of subsamples. As a result, a vector of millions of the chi-square molecule states is obtained.

Statistical processing of these states will enable data similar to the rows in Table 1.

(8) It is extremely important that with a deeper statistical processing of the data in the place of one or two easily computable chi-square functionals (3), the one is compelled to calculate hundreds of millions of such functionals. There is a significant increase in the computational complexity of algorithms for deeper statistical analysis.

That is, the algorithms for analyzing the spectral images of small samples may prove to be more powerful than simpler classical algorithms only because they initially require much more computing resources. In Pearson's time, there was no significant computing power. Today

The sample of 32 experiments is small and, therefore, cannot be used for reliable analysis according to standardized requirements [04]. In this connection, it is necessary to involve a numerical experiment built on a million samples from 32 experiments obtained from software generators of pseudo-random numbers with normal and uniform distribution laws. The probability distributions curves of chi-square functionals are shown in Figure 4 for 32 experiments and 4 columns of the histogram.

The informativity of the chi-square test for a sample of 32 experiments is determined by calculating equiprobable errors of the first and second kinds PЕЕ=P1=P2=0.215.

Figure 4 shows the equiprobable errors point marked with a dotted line. The sample chi-square functionals in 32 experiments is: If the estimate (7) and the estimate (9) are compared, a huge gain in the two probabilities ratio is obtained. The probability of an error in classical continuum calculations takes the value of PЕЕ=2-2.2=0.215, transition to the spectral lines analysis reduces the probability of er(9) rors up to PЕЕ=2-24.9=0.000000003. There is a 222□4 000 000-fold gain on reduction of errors of the first and the second kind. This is precisely the stimulus to force researchers to move from a relatively simple classical continuum statistical processing to a more complex image-processing of spectral portraits of small samples.

Moreover, it can be shown that the gain can be even greater if one goes from using histograms with 6 columns instead of histograms with 4 columns. In this case, the number of significant spectral states approximately doubles, which is shown in Figure 5.

Figure 4: The densities of distribution values of the chi-square functionals for the normal (solid line) and uniform (dashed line) laws Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

407

Alexander Ivanov, et al. - A simple way to assess the spectral lines informativity of a chi-square molecule in analyzing small samples of biometric data

Figure 5: The discrete spectrum of chi-square test conditions, which has almost doubled the number of spectral lines with an increase in columns in the histogram As a result, estimate of errors of the first and second kind probability in the transition from histograms with 4 columns to histograms with 6 columns should decrease significantly with PЕЕ=2-24.9 to PЕЕ=2-49.8 (the exponent value should at least double).

CONCLUSION It should be noted that the gain estimate from the continuous probability densities analysis to the discrete spectral lines analysis is an upper-bound estimate. Real gain will be always lower. This is due to the fact that the spectral lines of the chi-square molecule are correlated (linked together [18]). Correlation links between the data arise because a set of samples in the size of 16 experiments are obtained from one sample in 32 experiments. However, even this circumstance should not stop the researchers of such statistics. Even if the theoretically possible gain of 4,000,000 times is reduced to a technically realizable gain of 40 times, the programming costs and additional power consumed by the computers will be justified when processing small amounts of expensive biometric data.

Gain from a deeper statistical processing of the second level should always be present. This gain can be estimated more reliably only after the implementation of the corresponding software product. As a matter of fact, in the offing, there is a technical opportunity to exchange the complexity of the software product and the time for solving problems on the computer on the reliability of the statistical decisions it makes. Even today, when calculating the standard deviation and the correlation coefficients, it is possible to reduce the requirements for the test sample volume from 1.2 to 3 times. In this article, the authors tried to demonstrate that the potential for further reducing the requirements for the test sample size is much larger.

408

REFERENCES 1. Volchikhin, V.I., Ivanov, A.I., Funtikov, V.A. (2005).

Fast learning algorithms for neural network mechanisms of biometric-cryptographic information protection. Publishing House of the Penza State University, Penza.

2. Malygin, A.Yu., Volchikhin, V.I., Ivanov, A.I., Funtikov, V.A. (2006). Fast testing algorithms for neural network mechanisms of biometric-cryptographic information protection. Publishing House of the Penza State University, Penza.

3. Yazov, Yu.K. (2012). Neural network protection of personal biometric data. Radio Engineering, Moscow.

4. Federal Agency on Technical Regulating and Metrology. (2001). R 50.1.037-2002. Recommendations on standardization. Applied statistics. Rules for verifying the agreement between the experimental and the theoretical distributions. Part I. x2 type criteria.

5. Serikova, N.I., Ivanov, A.I., Kachalin, S.V. (2014).

Biometric stats: smoothing histograms based on small training sample. Scientific Journal of Science and Technology, vol. 3, no. 55, 146-150.

6. Serikova, N.I. (2015). Assessment of the likelihood of the normal distribution hypothesis by the Gini criterion for smoothed histograms constructed on small test samples. Questions of Radio Electronics, vol. 1, 85-94.

7. Ivanov, A.I., Akhmetov, B.B., Serikova, Yu.I. (2016).

Strengthening the power of the chi-square test with tenfold increase in freedoms of statistical computations on small test samples. Reliability and Quality of Complex Systems, vol. 4, no. 16, 121-127. DOI: 10.21685/2307-4205-2016-4-17

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Alexander Ivanov, et al. - A simple way to assess the spectral lines informativity of a chi-square molecule in analyzing small samples of biometric data

8. Akhmetov, B.B., Ivanov, A.I. (2016). Multidimensional statistics of essentially dependent biometric data generated by neural network emulators of quadratic forms. LEM, Almaty.

9. Akhmetov, B.B., Ivanov, A.I., Serikova, N.I., Funtikova, Yu.V. (2015). The discrete nature of the chisquare test distribution for small test samples. Bulletin of the National Academy of Sciences of the Republic of Kazakhstan, vol. 1, no. 353, 17-25.

10. Kulagin, V.P., Ivanov, A.I., Gazin, A.I., Akhmetov, B.B. (2016). Cyclic continuum-quantum computing: Strengthening the Chi-Square test power on small samples. Analytics, vol. 30, no. 5, 22-29.

11. Volchikhin, V.I., Ivanov, A.I., Serikov, A.V., Serikova, Yu.I. (2017). Quantum superposition of the state discrete spectrum of mathematical correlation molecule for small samples of biometric data. Mordovia University Bulletin, vol. 27, no. 2, 224-238. DOI: 10.15507/0236-2910.027.201702.224-238 12. Volchikhin, V.I., Ivanov, A.I. (2017). Neural Network Molecule: a Solution of the Inverse Biometry Problem through Software Support of Quantum Superposition on Outputs of the Network of Artificial Neurons.

Mordovia University Bulletin, vol. 27, no. 4, 518-529.

DOI: 10.15507/0236-2910.027.201704.518-529 13. Volchikhin, V.I., Ivanov, A.I., Gazin, A.I., Bannih, A.G. (2017). Conditions of obtaining the discrete kurtosis

spectrum of statistical distributions of biometric data for small samples. Journal of Computational and Engineering Mathematics, vol. 4, no. 4, 53-63. DOI: 10.14529/jcem170405 14. Nilson, M., Chang, I. (2006). Quantum calculations and quantum information. Mir Publishers, Moscow.

15. Filippov, V.V., Mitsuk, S.V. (2017). Modelling magnetoresistance effect in limited anisotropic semiconductors. Chinese Physics Letters, vol. 34, no. 7, 077201. DOI: 10.1088/0256-307X/34/7/077201 16. Stepanov, N.F. (2001). Quantum mechanics and quantum chemistry. Mir Publishers, Moscow.

17. State Standard R 52633.5-2011. (2011). Data Protection. Information Protection Technique. Automatic Learning Neural Network Converters Biometry-Code Access. Standartinform, Moscow.

18. Ivanov, A.I. (2016). Multidimensional neural network processing of biometric data with software reproduction of quantum superposition effects. Penza Scientific and Research Electronic Technical Institute, Penza.

Paper submitted: 27.07.2018.

Paper accepted: 14.08.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

409

Original Scientific Paper

doi:10.5937/jaes16-15297

Journal of Applied Engineering Science

Paper number: 16(2018)2, 547, 410 - 415

SAGS AND FREQUENCIES OF NATURAL OSCILLATIONS OF COMPOSITE TWO-LAYER ISOTROPIC PLATES IN CASE OF CHANGE OF THICKNESS OF ONE OF THE LAYERS Turkov Andrey*, Abashina Natalia 1 Orel State University, Russian Federation The article presents the results of a research of a composite isotropic two-layer plate of a square outline. We conduct the research for a number of plates with the variable thickness of layers and different conditions of fixing: pivotally supported and rigidly jammed on a circuit. By results of numerical researches we define the maximum sag of W0 from action of a static load and the frequency of natural cross oscillations ω of composite plates. By the results we define the proportionality coefficient K in a formula of the prof. V. I. Korobko. By the results of researches we construct the diagrams of change of the maximum sag and frequency of natural oscillations in case of change of thickness of layers of a composite plate.

Key words: A composite plate, A maximum sag, Frequency of natural oscillations, Proportionality coefficient K

INTRODUCTION The composite plate represents the construction consisting of separate layers which are connected by means of compliant mechanical bindings. Such systems are often used in case of strengthening of structures by upbuilding of a layer that is dictated by increase in loading in case of reconstruction, change of operating conditions of the building, etc.

Design of modern buildings and constructions, and also their reconstruction and strengthening is inseparably connected with the analysis of their toughness, rigidness and stability under the influence of static and dynamic loads. The models of somestructures are presented in the form of composite plates with different boundary conditions and their researches are most fully reflected in [1, 2]. Such scientists as V. I. Korobko, A. V. Turkov and K. V. Marfin [3 … 11], E. V. Karpova [12], N. S. Abashina [13] were engaged in researches of correlation of the maximum sags and frequencies of natural oscillations of composite plates of different forms. In case of the solution of these problems bynumerical methods we often face the question of accuracy of the received results. For plate finite elements it depends, in particular, on density of a mesh spacing of the initial plate. Besides, the quantity of final elements of construction is often dictated by a problem of the best approximation of an outline of a plate.

There is a fundamental dependence of professor V. I. Korobko [3] connecting the maximum sag of W0 in case

of action of a uniformly distributed load q and frequency of natural cross oscillations of a solid isotropic plate ω: * Orel State University, Russian Federation, elena-ogu@mail.ru

As the expression standing in the right part doesn't depend on the stiffness propertiesof a plate, it is possible to make the assumption of applicability of this correlation to composite plates.

Use of analytical dependence about correlation of the maximum sag and oscillation frequencies simplifies the decision of designer tasks. Nowadays there have been made researches of two-layer composite plates of different outline with the identical thickness of layers [5 …

12] with use of the fundamental regularity offered by V. I. Korobko. At the same time the number of questions of deforming of composite plates in case of a static and dynamic load, in particular, the tasks of correlation of the maximum sag and frequencies of natural oscillations of two-layer plates at different thickness of layers and different boundary conditions, represents both theoretical, and practical interest.In this connection researches in this field of construction mechanics areof current interest.

RESEARCHES OF THE TWO-LAYER PLATE A model structure is a two-layer square plate with the side of 1 m. At the first stage calculation of plates is performed (at first – pivotally supported, and then – rigidly jammed on a circuit), the low layer of each of which has constant thickness of 10 mm, and the upper one changes and makes 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 30, 50 and 100 mm. At the second stage plates with the similar thickness of layers and conditions of fixing are considered, at the same time the thickness of a high layer (10 mm) is invariable.

Both layers during calculation were broken into 400 finite elements of 50х50 mm in size, in each of them shift connection is established. Cross bindings and bindings of shift were put into construction nodes. Rigidness of

410

Turkov Andrey, et al. - Sags and frequencies of natural oscillations of composite two-layer isotropic plates in case of change of thickness of one of the layers

cross bindings is constant and makes EAc = 644 kN that corresponds to a steel dowel with a diameter of 2 mm. All characteristics of layers are accepted for a plate of chipboard: average density ρ=720 kg/m3, elastic modulus in case of a bend of E=2600MPas. For dynamic calculation the weights in nodes were gathered according to volume weight of material of layers and the cargo area of a node taking into account plate layer thickness. In case of static calculation the uniformly distributed load of q=1 of kN/

sq.m was applied to a high layer. The researches were conducted by the finite-element method. The distance between layers was accepted equal to distance between gravitational centersof layers. The diagram of setting of cross bindings and shift bindings is given below – “Figure 1”, plates models – “Figure 2”.

Numerical researches were conducted by means of the

Figure 2 – Conditions of fixing of a plate: hinge fixing (a) and rigid fixing (b)

Figure 1 – Plate finiteelement

Table 1: Results of calculation of hinge-supported plates with different thickness of layers (the first stage of calculation) Thickness Thickness of of a high a low layer layer hlow, mm hhigh, mm

hhigh/hlow

Maximum sag W0, mm

Frequency of the main tone of natural oscillations ω, 1/с

Frequency of the main tone of Coefficient K in W0, mm natural oscillations FEM ω, 1/с Divergence of K Ktheor 247.7138 from Ktheor,

Theoretical Coefficient value K in FEM Ktheor

Divergence of K from Ktheor, %

Theoretical value

1.582

-1.0

1.56554

1.582

-1.04

256.3577

1.56492

1.582

-1.1

1.344859

265.0514

1.56013

1.582

-1.6

1.4

1.207723

273.7952

1.56458

1.582

-1.1

10

1.5

1.088417

282.5877

1.56450

1.582

-1.1

16

10

1.6

0.9841364

292.4268

1.57542

1.582

-0.4

17

10

1.7

0.8925969

300.3102

1.56492

1.582

-1.1

18

10

1.8

0.811921

309.2350

1.56524

1.582

-1.1

19

10

1.9

0.7405551

318.1986

1.56561

1.582

-1.0

20

10

2

0.6772047

328.1292

1.57494

1.582

-0.4

30

10

3

0.310998

418.7003

1.57021

1.582

-0.7

50

10

5

0.099168

606.4788

1.57575

1.582

-0.4

100

10

10

0.0165543

1098.145

1.58109

1.582

-0.1

10

10

11

10

12

10

%

1.503294

13

10

1.3

14

10

15

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

411

Turkov Andrey, et al. - Sags and frequencies of natural oscillations of composite two-layer isotropic plates in case of change of thickness of one of the layers

program SCAD. During calculation the maximum sags of W0 from action of a permanent load of q=1kN/m2 uniformly distributed on plate surface, and the frequency of the free cross oscillations ω in the no-load condition were defined.

RESULTS OF RESEARCHES Results of calculation of plates are given in Tables 1 – 4.

According to the tables the diagrams of change of sags and oscillation frequencies in the plates under study, “Figure 3”, “Figure4 ” and coefficient of proportionality K of Professor V. I. Korobko fundamental dependence (1) “Figure 5” were constructed.

CONCLUSION As a result of a numerical research it is defined that with the increase in thickness of one of the layers of a composite plate, the frequency of its free oscillations increas-

es and the sag decreases that proves the increase of rigidness of a plate. The method of fixing of a plate gives different results in the conducted experiment: more rigid are the plates which are rigidly jammed on a circuit – that is well noticeable in the diagrams –“Figure 3”, “Figure 4” (the values of frequencies are bigger, and the number of sags are less, than at pivotally supported ones). The relative positioning of plate layers, different in thickness, slightly influences the results; however, it is possible to mark that the smaller sag characterizes plates which layer of a bigger thickness is located on top, and this difference is more noticeable when calculating plates which have a changeable thickness of a layer much more than constant: 50 mm, 100 mm (the difference reaches 1.09% in case of a hinge fixing and 3,04% in case of rigid jamming). The proportionality coefficients K defined in the calculation differ from analytical values Ktheor. within 1.6% in case of a hinge fixing and within 5% in case of a rigid jamming.

Table 2: Results of calculation of hinge-supportedplates with different thicknessof layers (the second stage of calculation) Thickness Thickness of of a high a low layer layer hlow, mm hhigh, mm

412

hhigh/hlow

Maximum sag W0, mm

Frequency of the main tone of natural oscillations ω, 1/с

Theoretical Coefficient value K in FEM Ktheor

Divergence of K from Ktheor, %

10

10

1

1.902618

239.1168

1.56651

1.582

-1.0

10

11

0.909

1.687378

247.7138

1.56554

1.582

-1.0

10

12

0.833

1.503297

256.3577

1.56492

1.582

-1.1

10

13

0.769

1.344863

265.0515

1.56458

1.582

-1.6

10

14

0.714

1.20773

273.7952

1.56446

1.582

-1.1

10

15

0.667

1.088425

282.5877

1.56451

1.582

-1.1

10

16

0.625

0.9841465

291.4268

1.56468

1.582

-1.1

10

17

0.588

0.8926088

300.3102

1.56494

1.582

-1.1

10

18

0.556

0.8119348

309.235

1.56527

1.582

-1.1

10

19

0.526

0.7405708

318.1986

1.56565

1.582

-1.0

10

20

0.5

0.6772224

327.1984

1.56606

1.582

-1.0

10

30

0.333

0.3110359

418.7003

1.57040

1.582

-0.7

10

50

0.2

0.0992484

606.4788

1.57702

1.582

-0.3

10

100

0.1

0.0167369

1098.145

1.59853

1.582

1.0

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Turkov Andrey, et al. - Sags and frequencies of natural oscillations of composite two-layer isotropic plates in case of change of thickness of one of the layers

Table 3: Results of calculation of rigidly jammed plates with different thickness of layers (the first stage of calculation) Thickness of a high layer hhigh, mm

Thickness of h / high a low layer hlow hlow, mm

Maximum sag W0, mm

Frequency of Theoretical the main tone of Coefficient value natural oscillations K in FEM Ktheor ω, 1/с

Divergence of K from Ktheor, %

10

10

1

1.259987

296.3286

1.59322

1.610

-1.0

11

10

1.1

1.104124

308.9617

1.5936

1.610

-1.0

12

10

1.2

0.9694215

322.2326

1.59443

1.610

-1.0

13

10

1.3

0.8530879

336.0734

1.5956

1.610

-0.9

14

10

1.4

0.7526142

350.4212

1.59697

1.610

-0.8

15

10

1.5

0.6657761

365.2191

1.59848

1.610

-0.7

16

10

1.6

0.5906274

380.4161

1.60006

1.610

-0.6

17

10

1.7

0.5254822

395.9671

1.60167

1.610

-0.5

18

10

1.8

0.4688918

411.8326

1.60326

1.610

-0.4

19

10

1.9

0.4196182

427.9777

1.60482

1.610

-0.3

20

10

2

0.3766072

445.6363

1.61549

1.610

0.3

30

10

3

0.1473327

617.5049

1.61798

1.610

0.5

50

10

5

0.0388027

985.7546

1.62886

1.610

1.2

100

10

10

0.0054736

1942.727

1.63615

1.610

1.6

Table 4: Results of calculation of rigidly jammed plates with different thickness of layers (the second stage of calculation) Thickness of a high layer hhigh, mm

Thickness of h / high a low layer hlow hlow, mm

Maximum sag W0, mm

Frequency of Theoretical the main tone of Coefficient value natural oscillations K in FEM Ktheor ω, 1/с

Divergence of K from Ktheor, %

10

10

1

1.259987

296.3286

1.59322

1.610

-1.0

10

11

0.909

1.104128

308.9617

1.5936

1.610

-1.0

10

12

0.833

0.969429

322.2326

1.59445

1.610

-0.9

10

13

0.769

0.8531

336.0733

1.59562

1.610

-0.8

10

14

0.714

0.75263

350.4212

1.597

1.610

-0.7

10

15

0.667

0.665795

365.2191

1.59853

1.610

-0.6

10

16

0.625

0.59065

380.4161

1.60012

1.610

-0.5

10

17

0.588

0.525507

395.9671

1.60175

1.610

-0.4

10

18

0.556

0.468921

411.8326

1.60336

1.610

-0.3

10

19

0.526

0.41965

427.9777

1.60494

1.610

-0.2

10

20

0.5

0.376641

444.3721

1.60648

1.610

-0.2

10

30

0.333

0.147393

617.5049

1.61864

1.610

0.5

10

50

0.2

0.038906

985.7546

1.63321

1.610

1.4

10

100

0.1

0.005671

1942.727

1.69524

1.610

5.0

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

413

Turkov Andrey, et al. - Sags and frequencies of natural oscillations of composite two-layer isotropic plates in case of change of thickness of one of the layers

Figure 3: The diagram of change of the maximum sags and frequencies of natural oscillations depending on change of thickness of a high layer of a two-layer composite plate at a constant thickness of a low layer (10 mm) in the conditions of a hinge fixing and rigid jamming

Figure 4: The diagram of change of the maximum sags and frequencies of natural oscillations depending on change of thickness of a low layer of a two-layer composite plate at a constant thickness of a high layer (10 mm) in the conditions of a hinge fixing and rigid jamming

Figure 5: Diagrams of change of coefficient K depending on change of thickness of one of the layers of a two-layer plate in comparison with its theoretical value REFERENCES 1. Rzhanitsyn A. (1976) Raschet sostavnyh plastinok s absolyutno zhestkim i poperechnymi svyazyami [Calculation of composite plates with absolutely tight cross couplings]. Researches on the theory of con-

414

structions, issue XXII, M.: Stroyizdat, pp. 120-133. (in Russian) 2. Rzhanitsyn A. (1986) Sostavnyesterzhniiplastinki [Composite rods and plates]. Moscow: Stroyizdat, 316 p. (in Russian) Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Turkov Andrey, et al. - Sags and frequencies of natural oscillations of composite two-layer isotropic plates in case of change of thickness of one of the layers

3. Korobko V.I. (1989) Ob odnoj “zamechatel'noj” zakonomernosti v teorii uprugih plastinok [About one “remarkable” regularity in the theory of elastic plates].

The news of higher education institutions. Construction and architecture, no. 11. pp. 32-36. (in Russian) 4. Korobko V., Korobko A. (2010) Stroitel'naya mekhanika plastinok: Tekhnicheskaya teoriya [Construction mechanics of plates: Technical theory]. Moscow: publishing house “Spektr”, 410 p. (in Russian) 5. Marfin K. (2015) Vzaimosvyaz' maksimal'nyh progibov isobstvennyh chastot poperechnyh kolebanij sostavnyh plastin na podatlivyh svyazyah [Correlation of the maximum sags and natural frequencies of cross oscillations of composite plates on pliable communications]. Candidate thesis for in Engineering Sciences, Orel, 150 p. (in Russian) 6. Turkov A. (2008) Vzaimosvyaz' zadach dinamiki I statiki sploshnyh I sostavnyh derevyannyh konstrukcij [Correlation of tasks of dynamics and statics of continuous and composite wooden constructions].

Doctoral thesis in Engineering Sciences, Orel, 386 p. (in Russian) 7. Turkov A., Marfin K. (2011) Progiby I chastity sobstvenny hkolebanij sostavnyh kvadratnyh izotropnyh plastin s razlichnymi granichnymi usloviyami pri izmenenii zhestkosti svyazej sdviga [Deflections and frequencies of natural oscillations of composite square isotropic plates with different boundary conditions in case of change of rigidness of shift bindings]. Construction and reconstruction, Orel: FSBEE HPE “State University – UNPC”, no. 4, pp. 38-42. (in Russian) 8. Turkov A., Marfin K. (2011) Tochnost' rezul'tatov chislennyh issledovanij kvadratnyh sostavnyh izotropnyh plastin na podatlivyh svyazyah pri razlichnom kolichestve konechnyh ehlementov [The accuracy of results of numerical researches of square composite isotropic plates on pliable bindings in case of different quantity of terminal elements]. Construction and reconstruction, Orel: FSBEE HPE “State University – UNPC”, no.6, pp. 43-49. (in Russian) 9. Turkov A., Marfin K. (2013) Issledovanie progibov i chastot sobstvennyh kolebanij kruglyh transtropnyh plastin [The study of deflections and frequencies of

natural oscillations of round transversely isotropic plates]. News of higher educational institutions. Forest journal, no. 4, pp. 66-71. (in Russian) 10. Turkov A., Marfin K. (2014) Metodika provedeniya ehksperimenta dlya staticheskogo i dinamicheskogo ispytaniya izotropnyh sostavnyh plastin na podatlivyh svyazyah [The procedure of carrying out an experiment for static and dynamic test of the isotropic composite plates on pliable bindings]. Construction and reconstruction, no. 4, pp. 31-36. (in Russian) 11. Turkov A., Marfin K. (2015) Eksperimental'nye issledovaniya sostavnyh kruglyh plastin na dinamicheskie I staticheskie nagruzki [The pilot studies of composite circular plates for dynamic and static loads].

Construction and reconstruction, no. 3, pp. 60-66. (in Russian)Marfin K. (2015) Vzaimosvyaz' maksimal'nyh progibov isobstvennyh chastot poperechnyh kolebanij sostavnyh plastin na podatlivyh svyazyah [Correlation of the maximum sags and natural frequencies of cross oscillations of composite plates on pliable communications]. The thesis for a degree of Candidate of Technical Sciences, Orel, 150 p. (in Russian) 12. Turkov A., Karpova E. (2015) Issledovanie koehfficienta zhyostkosti dlya treugol'noj sostavnoj izotropnoj plastiny v zavisimosti ot eyo osnovnoj chastity kolebanij pri raznoj zhyostkosti svyazej sdviga [The study of a seam stiffness coefficient for a triangular composite isotropic plate depending on its basic frequency of oscillations in case of different rigidness of shift bindings]. Construction mechanics and calculation of constructions, no. 2, pp. 66-69. (in Russian) 13. Turkov A., Abashina N., Karpova E. (2016) Progiby i chastity sobstvennyh kolebanij sostavnyh rombicheskih izotropnyh plastin, sharnirnoopyortyh po konturu pri izmenenii zhyostkosti svyazej sdviga [Deflections and frequencies of natural oscillations of composite rhombic isotropic plates pivotally supported on a circuit in case of change of rigidness of shift bindings]. Construction and 14. Semenov A., Gabitov A. (2005) Proektno-vychislitel'nyjkompleks SCAD v uchebnomprocesse. Chast' I. Staticheskijraschet [The SCAD project computer system in educational process. Part I. Static calculation]. Moscow: ASV publishing house, 152 p. (in Russian)

Paper submitted: 10.10.2017.

Paper accepted: 03.07.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

415

Original Scientific Paper

Journal of Applied Engineering Science

doi:10.5937/jaes16-16582

Paper number: 16(2018)3, 548, 416 - 423

THE EFFECT OF HYBRID SAVONIUS AND DARRIEUS TURBINE ON THE CHANGE OF WAKE RECOVERY AND IMPROVEMENT OF WIND ENERGY HARVESTING Erwin Erwin1* Adi Surjosatyo2 Julianto Sulistyo Nugroho2 Teuku Meurahindra Mahlia3 Tresna Soemardi2 1 University Sultan Ageng Tirtayasa, Banten, Indonesia 2 University Indonesia, Depok, Indonesia 3 University Tenaga Nasional, Selangor, Malaysia The energy crisis encourages the development of renewable energy; one of the potential renewable energy is wind.

In the field of wind turbine there is a two-way development of the utilization of wind energy, first by making a large wind turbine, the second by making a wind farm energy with a relatively small wind turbine.This hybrid VAWT wind turbine (Sultan Wind Turbine) is designed to work optimally on a farm array, on a wind turbine farm array will always cause a wake effect that will reduce overall wind turbine and farm array performance, an investigation with a CFD simulation is required to predict how far the wake effect will be before farm array build.The use of simulation software has been widely used to predict the effects of this wake, and experiments in the laboratory have also been done to predict the effects of a wake as well.This study'spurpose is to predict the distance area of the recovery wake behind the wind turbine, this distance which will be the reference distance between wind turbine units and determining the density of the turbine in a farm. Simulation using Computational Fluid Dynamics (CFD), with a method of Multi Frame Reference (MRF). Analysis using descriptive and inferential method in statistics such as mean, Kolmogorov-Smirnov Z and KruskalWalis test.From the analysis of simulation results and data processing descriptively and analytic statistic, it can be concluded from the data given, the distance of x/D=4, wind speed has recovery to the value near the input speed and no significant change to x/D= 9. Then it can be concluded that the distance between two wind turbines that can be used is a distance of 3.6 meters.These data suggest that the hybrid farm array VAWT savonius and darrieus have a higher power density compared to HAWT. From this power density calculation the hybrid VAWT has a greater electrical potential up to 300 percent compared to the HAWT farm array.

Key words: CFD, Wake effect, Power density, Farm array, Statistics INTRODUCTION Dabiriet. al. argues that the utilization of the small VAWT farm array can address the growing challenges of energy demand [1, 2]. Wake effect for HAWT causing the absorbed wind energy to drop by 31% at a distance of 3x blade diameter, and down to 35% at a distance of 5x the diameter of the blade [3], for VAWT at a distance of 2.5 x diameter, wind speed is reduced to only 41% of the original speed [4]. Many similarities between wake in VAWT and wake in HAWT [5], but the difference in recovery wake distance between the two types of turbine is not stated directly.Another research using 2D wake model for optimization of wind turbine layout [6].

The placement pattern and distance of a turbine in a wind farm is very important to improve the performance of the farm array. Research using methods such as genetic algorithm optimization [7], wind farm layout using biogeography-based optimization [8], grid-like layout [9] can significantly improve the performance of the farm array.

Investigation to co and counter-rotating 2 arrays and 3 arrays has been done to get the best efficiency [10] and propose for 2 array use backward counter-rotating and for 3 array use forward counter-rotating, Dabiri use

forward counter rotating of 2 arrays [1], in this research using forward counter-rotating 2 arrays to get small wake spreading rate.

Several parameters have been studied that affect the efficiency and power generated by wind turbines such as tower height, rotor optimization and use of lightweight materials can be done to improve overall farm efficiency[11]. Another way to improve farm efficiency is by reducing recovery distance by increasing the tip speed ratio, also a concern to Cp while increasing tip speed ratio [12].

Sultan wind turbine, a hybrid savonius-darrieus [13] use as a prototype of hybrid VAWT in this project research.

This hybrid is proposed to use in hybrid with biomass to improve the performance [34], another hybrid also introduced using solar and biomass for rice drying [35].

Wake arising from the single turbine will affect the overall wind farm performance [14] then the wake of each turbine should be reduced. Farm array will cause wake effect[3, 4, 14], wake effect greatly affect wind turbine performance and wind turbine performance will affect farm performance, therefore it is necessary to investigate the recovery effect of the wake as a reference in designing shape and size of the distance between turbine in the

* University Sultan Ageng Tirtayasa, Jalan Raya Jakarta Km 4, Panancangan, Cipocok Jaya, Kota Serang, Banten 42124, Indonesia, erwin@untirta.ac.id

416

Erwin Erwin et al. - The effect of hybrid savonius and darrieus turbine on the change of wake recovery and improvement of wind energy harvesting

farm.In this study, wake recovery zone distance will become a reference in determining the distance and the shape of two wind turbines in a farm.

There has been no research that discusses the effects of wake generated by hybrid wind turbines VAWT especially hybrid savonius-darrieus and its influence on power density and comparison with HAWT.Ghosh A, did a CFD simulation research about combination of savonius and darrieus [22], but has different structure with our hybrid savoniusdarrieusVAWT. This study aims to predict the wake recovery distance of the hybrid VAWT turbine savonius-darrieus, this position will be a reference in designing the shape and size of this VAWT hybrid farm array.

Governing equation The continuity and conservation of momentumequations: (1) (2) Multiple Reference Frame (MRF) formulation The MRF approach is used in models with internal frames that rotate on a stationary reference frame.

Absolute speed in the internal frame reference uses ί notation for inertia and r for rotation: (3)

METHODOLOGY This research will use CFD to simulate fluid movement around the turbine and wake effect that arise. Data of simulation result will be analyzed by the statistic method to compare significance differences of each distance group of data.

Computational Fluid Dynamic A numerical study investigating hybrid wind turbine savonius-darrieus has been performed using RANS (SST k-ω) turbulence modelling [15-17]. The turbine's rotation is handled by the numerically robust Multi-Reference Frame technique, simulation with the CFD Fluent Software.

The MRF assumes that the volume used has a constant rotational speed with the surface of revolution with nonwall boundaries type. MRF performs a rotational simulation and observes the result at the equivalent position of the rotor. Also assumes a weak interaction between the volume of MRF and the volume around the stationary.

Some earlier researcher using MRF approaches to model the inner rotating reference frames in a stationary computational mesh and outer reference frame for the full wind turbine rotor simulations, MRF is also used in wake simulation on HAWT even in its result, sliding mesh method is better than MRF, but MRF is efficient in computing time[18]. From the simulation results can be seen a visualization of vortices or wake by using iso-surface of q-criterion or lambda2.

Mathematical Model Statement of the problem The physical problem being examined is the fluid flow through double VAWT hybrid wind turbines savonius and darrieus. This study examined the expansion of wake and initial predictions with the Multi Reference Frame CFD approach.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Acceleration equation expressed as: (4) By using equation (2) : (5) Rotating frame ofreference in incompressible RANS equations as absolute velocityat the inertial frame of referenceis: (6) (7) At the rotating frame of reference is: (8)

(9) StatisticalTesting Statistic method has been utilized in many fields in science, the combination of simulation and statistics is a powerful tool in drawing conclusions on cases that are difficult to decide with other methods.

Statistical methods are widely used in designing experiments of a simulation, such as full factorial research [19-21],a statistic is a powerful tool to analyze and make some conclusion from scattering data, a statistic used to estimate CFD result for drag prediction [30]. Combination of CFD and statistic technique has been used to guessspreading of airborne sea salt [31].

The Kruskal–Wallis test used to test statistical differences between lumen areas from CFD result in the medical application of stent coronary artery [32].

417

Erwin Erwin et al. - The effect of hybrid savonius and darrieus turbine on the change of wake recovery and improvement of wind energy harvesting

Descriptive And Deductive statistics that discuss how scattered postulate a set of data in a form that is easy to the analysis quick to provide information, presented in the form of tables, graphs, concentration values and spread values.

Mean An average is a very frequently used cantering measure.

The advantage of calculating the average is that the number can be used as a representation or representative of the observed data. The average is sensitive to the existence of extreme values or outliers.

(10) Inferential Analysis Kolmogorov-Smirnov Z Test,is a nonparametric test to see the normality of the data, if data normal, then continued with One Way ANOVA Test on the parametric test, and if not normal, it will be tested with Kruskal Wallis Test.

The Kruskal Wallis test is a rank-based nonparametric test whose purpose is to determine whether there is a statistically significant difference between two or more independent variable groups in the numerical dependent variable (interval/ratio) and ordinal scale.

(11)

The next test is a comparison between each group to see significant differences between groups. Here followed by Test Two independent data groups with Mann Whitney U Test, is a non-parametric test used to know the median difference of two free groups, if the dependent variable data scale is ordinal or interval / ratio but not normally distributed.

(12) Power density Power density is a measure of electrical energy that can be generated per unit area of the farm array of wind turbine, the commercial HAWT flux limits theory is 2-3 Watts/m2, while commercial VAWT can be up to 8 Watts/ m2.[1] (13) Power coefficient of VAWT is higher â&#x2030;&#x2C6;6% than Betz Limits which is usually analyzed as the coefficient limit on HAWT [33].

RESULTS AND DISCUSSION For statistical analysis required preparation of data to be analyzed, the data in the form of wind speed values from the simulation results at each point that has been determined. This data will also be displayed in the form of speed contour graphs in each field.

Figure 1: Simulation layout and field position of wind speed data retrieval

418

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Erwin Erwin et al. - The effect of hybrid savonius and darrieus turbine on the change of wake recovery and improvement of wind energy harvesting

Figure 1 shown the experimental setup and wind data retrieval in a simulation. There are 9 fields, where each field has 5 lines of lines that line up vertically, and each line there are 50 points of data collection of wind speed.

Incoming velocity is varying from 3 m/s to 9 m/s, with an outlet is ambient pressure.

The wind velocity data used in the simulated inlet is taken from a range of wind data measured directly in the field.

The wind data sample was taken using an anemometer and data acquisition withan Arduino for 24 hours on August 3, 2017, the wind speed ranging from 0 m/s to 11 m/s.

Simulation data With regard to the measured wind velocity contours, for simulated wind speeds set from 3 m/s to 9 m/s, the temperature and pressure are set to default. The working fluid is air. With variations of rotational speed of the rotor from 80 rpm to 175rpm. Domain sized is 12.000mm x 5696mm x 5350 mm.

Another setup, speed measurement distance ranging from 1 times to 9 times the diameter of the rotor darrieus.

In Figures. 2, Speed contour in x/Dd= 4 in two elevation positions, one in the middle of darrieus and another in the middle of savonius, with a wind speed of 6 m/s. The two graphs above show that savonius has a higher wake effect than darrieus. So the idea of varying the position of the savonius on the shaft arises to reduce the wake effect that resonates on the farm array if the position of savonius at the same level.

Analysis Taking into account the 3D contours of wake from 3 m/s up to 9 m/s above, it supports 2D data result, that the wake effect on savonius is greater than that of darrieus.

The wake effect also gradually disappears moving to the top-sided with increasing distance from the shaft, only it is difficult to determine at what distance position there has been recovery wake.

[23] analyzes and sets wake region based on a guest on an analysis dimensional graph without using any tools to analyze and determine wake region position. In this analysis, statistical analysis tools are used to determine the wake region position based on 3D graphics and data from simulated CFD results.

From the result of the test of data normality (Kolmogorov-Smirnov Z Test) with a result of Pval (0.000) <0,05 indicates that data is not normal, so do follow-up testing with Kruskal Wallis Test. Tests with 9 groups yielded Pval (.971) > 0.05 indicated as individuals, no statistically significant group of data.

The next test with 2 groups of data with Mann Whitney U, the result can be viewed in table 1.

From the result of the test of data normality (Kolmogorov-Smirnov Z Test) as a result of Pval (0.000) <0,05 indicates that data is not normal, so do follow-up testing with Kruskal Wallis Test. Tests with 9 groups yielded Pval Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

(.971)> 0.05 indicated as individuals, no statistically significant group of data.

The next test, with 2 groups of data analyzed with Mann Whitney U, shows that for the data presented, there is a significant difference between the data group distance x/ Dd=1-3 with the data group x/Dd= 4-9.

From two analyses, descriptive and Mann Whitney U, showing that at a x/Dd= 4, the wind velocity that descends behind the shaft recover and rises to an initial velocity such as before the shaft, or wakes affected zone only happen up to a x/Dd=3. The division of these two zones can be seen in Figure 3.

Wake that occurs behind the turbine, there is one important region that moves to the top, this is because of the position of the savonius that is at the top of the rotor. Whereas if the only darrieus then there are two parts of the region that one moves up and the second moves down [24].

For rated of hybrid VAWT at 500 watts, and turbine spacing x/D = 4, show that Power Density is about 8.29 W/m2.

Discussion So far from the results of simulations and experimental results on wake, it will be quite difficult to determine the boundary area between wake effect and the recovery wake area. This research aims to determine the distance of wake recovery in hybrid VAWT. Dabiri states that the wake recovery VAWT is much shorter than HAWT the and power density of VAWT is higher than HAWT[1, 25]. This statement, however, is slightly different from that of other researchers suggesting that the characteristics of VAWT and HAWT have many similarities, but do not specifically state the equations on wake recovery distance [5]. Several other studies have suggested that VAWT wake recovery occurs at a distance ofx/Dd = 3-5[4, 23]. Slightly different from VAWT's research in this paragraph, this study investigates the wake recovery of VAWT hybrids, and for determining the wake recovery distance to compare the analysis of simulation results and statistical analysis. The results of this study are important because the distance obtained will determine the number of wind turbines in a farm array and the power density performance of the VAWT hybrid. Hybrid savoniusdarrieus VAWT hass been analised by using CFD [22] with different structure with tis research and the result is wake recovery at savonius is less than darrieus, slightly different with this research result that shown wake recovery at savonius is bigger than darrieus. This happen because different position of savonius, in Ghosh A research savonius placed at bottom and darrieus at top, in this research savonius is placed at top side of the structure, and form vortices result shown that savonius wake is moving to upside of the frame.

419

Erwin Erwin et al. - The effect of hybrid savonius and darrieus turbine on the change of wake recovery and improvement of wind energy harvesting

Figure 2: Contour wake speed in center of darrieus and savonius at x/D = 4, for wind speed 3-9 m/s

420

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Erwin Erwin et al. - The effect of hybrid savonius and darrieus turbine on the change of wake recovery and improvement of wind energy harvesting

Table 1: Statistical analysis Data group

Mann-Whitney U

Limit Pval

Conclusion

(1-4 and 5-9)

(.230)

>0.05

there was no statistically significant difference

(1-6 and 7-9)

.447

>0.05

there was no statistically significant difference

(1 and 2-9)

.000

< 0.05

there is a statistically significant difference

(1-2 and 3-9)

.000

< 0.05

there is a statistically significant difference

(1-3 and 4-9)

.000

< 0.05

there is a statistically significant difference

Figure 3: Descriptive test, average mean velocity at every x/Dd CONCLUSIONS The statistical method descriptive and inferential method, has been implemented to analyze the simulation results, and it has been shown that based on the above analysis, that at the x/Dd= 4 of the turbine or at the distance 3.6 meters from hybrid VAWT at the front, significantly for the given data indicates that each wind turbine will get an average wind speed equally at this distance, and it is expected that each wind turbine can produce the same electric rotation and energy on every wind turbine.

The analysis also using post processing to examvorticesfor wind speed 9m/s with the Q Criterion[26]method shows the same trend that recovery has begun to occur about x/Dd = 4. This result is compatible withthe single VAWT simulation result which provides wake recovery Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

occurring in the range of x/Dd = 5 [23], and based on experiment, wake recovery from straight-bladed Vertical Axis Wind turbine occurs in the range x/Dd = 3.3 and x/ Dd = 4.5 [4] which much shorter than HAWT wake recovery that happens in x/Dd= 15 [27-29].This indicates that with the area of the wind sweep cross-section similar to the HAWT, hybrid VAWT has a higher power density than HAWT. Based on the calculation of Power density its show that for hybrid VAWT produce power density 8.29 W/m2over 300 percent than HAWT.

421

Erwin Erwin et al. - The effect of hybrid savonius and darrieus turbine on the change of wake recovery and improvement of wind energy harvesting

REFERENCES 1. Dabiri JO. Potential order-of-magnitude enhancement of wind farm power density via counter-rotating vertical-axis wind turbine arrays. Journal of Renewable and Sustainable Energy. 2011;3:043104.

2. Robert WW, Sebastian L, John OD. Fish schooling as a basis for vertical axis wind turbine farm design.

Bioinspiration & Biomimetics. 2010;5:035005.

3. Bartl J, Pierella F, Sætrana L. Wake Measurements Behind an Array of Two Model Wind Turbines. Energy Procedia. 2012;24:305-12.

4. Li Qa, Maeda T, Kamada Y, Ogasawara T, Nakai A, Kasuya T. Investigation of power performance and wake on a straight-bladed vertical axis wind turbine with field experiments. Energy. 2017.

5. Rolin VFC, Porté-Agel F. Experimental investigation of vertical-axis wind-turbine wakes in boundary layer flow. Renewable Energy. 2018;118:1-13.

6. Gao X, Yang H, Lu L. Optimization of wind turbine layout position in a wind farm using a newly-developed two-dimensional wake model. Applied Energy.

2016;174:192-200.

7. Emami A, Noghreh P. New approach on optimization in placement of wind turbines within wind farm by genetic algorithms. Renewable Energy. 2010;35:155964.

8. Bansal JC, Farswan P. Wind farm layout using biogeography-based optimization. Renewable Energy.

2017;107:386-402.

9. Serrano González J, Trigo García ÁL, Burgos Payán M, Riquelme Santos J, González Rodríguez ÁG. Optimal wind-turbine micro-siting of offshore wind farms: A grid-like layout approach. Applied Energy.

2017;200:28-38.

10. Lam HF, Peng HY. Measurements of the wake characteristics of co- and counter-rotating twin H-rotor vertical axis wind turbines. Energy. 2017;131:13-26.

11. Bukala J, Damaziak K, Kroszczynski K, Krzeszowiec M, Malachowski J. Investigation of parameters influencing the efficiency of small wind turbines. Journal of Wind Engineering and Industrial Aerodynamics.

2015;146:29-38.

12. Hezaveh SH, Bou-Zeid E, Lohry MW, Martinelli L. Simulation and wake analysis of a single vertical axis wind turbine. Wind Energy. 2017;20:713-30.

13. Erwin Erwin SW, Erny Listijorini, Rina Lusiani, Tresna P Soemardi. Development of the Third Darrieus Blade of Sultan Wind Turbine for Low Wind Speed.

Applied Mechanics and Materials. 2015;758:7.

14. González-Longatt F, Wall P, Terzija V. Wake effect in wind farm performance: Steady-state and dynamic behavior. Renewable Energy. 2012;39:329-38.

15. Zuo W, Wang X, Kang S. Numerical simulations on the wake effect of H-type vertical axis wind turbines.

Energy. 2016;106:691-700.

422

16. Mohamed MH, Ali AM, Hafiz AA. CFD analysis for H-rotor Darrieus turbine as a low speed wind energy converter. Engineering Science and Technology, an International Journal. 2015;18:1-13.

17. Nobile R, Vahdati M, Barlow JF, Mewburn-Crook A. Unsteady flow simulation of a vertical axis augmented wind turbine: A two-dimensional study. Journal of Wind Engineering and Industrial Aerodynamics.

2014;125:168-79.

18. Liu J, Lin H, Purimitla SR. Wakefield studies of tidal current turbines with different numerical methods.

Ocean Engineering. 2016;117:383-97.

19. Rukthong W, Weerapakkaroon W, Wongsiriwan U, Piumsomboon P, Chalermsinsuwan B. Integration of computational fluid dynamics simulation and statistical factorial experimental design of thick-wall crude oil pipeline with heat loss. Advances in Engineering Software. 2015;86:49-54.

20. Prošek A, Končar B, Leskovar M. Uncertainty analysis of CFD benchmark case using optimal statistical estimator. Nuclear Engineering and Design.

2017;321:132-43.

21. Hemsch MJ. Statistical Analysis of Computational Fluid Dynamics Solutions from the Drag Prediction Workshop. Journal of Aircraft. 2004;41:95-103.

22. Ghosh A, Biswas A, Sharma KK, Gupta R. Computational analysis of flow physics of a combined three bladed Darrieus Savonius wind rotor. Journal of the Energy Institute. 2015;88:425-37.

23. Chowdhury H, Mustary I, Loganathan B, Alam F. Adjacent Wake Effect of a Vertical Axis Wind Turbine.

Procedia Engineering. 2015;105:692-7.

24. Posa A, Parker CM, Leftwich MC, Balaras E. Wake structure of a single vertical axis wind turbine. International Journal of Heat and Fluid Flow. 2016;61:7584.

25. Matthias Kinzel QM, John O. Dabiri. Energy exchange in an array of vertical-axis wind turbines.

Journal of Turbulence. 2012;13:13.

26. Sedaghatizadeh N, Arjomandi M, Kelso R, Cazzolato B, Ghayesh MH. Modelling of wind turbine wake using large eddy simulation. Renewable Energy.

2018;115:1166-76.

27. Miao W, Li C, Pavesi G, Yang J, Xie X. Investigation of wake characteristics of a yawed HAWT and its impacts on the inline downstream wind turbine using unsteady CFD. Journal of Wind Engineering and Industrial Aerodynamics. 2017;168:60-71.

28. Naderi S, Torabi F. Numerical investigation of wake behind a HAWT using modified actuator disc method. Energy Conversion and Management.

2017;148:1346-57.

29. Hu D-m, Du Z-h. Near Wake of a Model Horizontal-Axis Wind Turbine. Journal of Hydrodynamics, Ser B. 2009;21:285-91.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Erwin Erwin et al. - The effect of hybrid savonius and darrieus turbine on the change of wake recovery and improvement of wind energy harvesting

30. Hemsch, M. J. (2004). "Statistical Analysis of Computational Fluid Dynamics Solutions from the Drag Prediction Workshop." Journal of Aircraft 41(1): 95103.

31. Suto, H., Y. Hattori, et al. (2017). "Computational fluid dynamics simulation and statistical procedure for estimating wide-area distributions of airborne sea salt considering local ground conditions." Structure and Infrastructure Engineering 13(10): 1359-1371.

32. Migliori, S., C. Chiastra, et al. (2017). "A framework for computational fluid dynamic analyses of patient-specific stented coronary arteries from optical coherence tomography images." Medical Engineering & Physics 47: 105-116.

33. ThĂśnniĂ&#x;en, F., M. Marnett, et al. (2016). "A numerical analysis to evaluate Betz's Law for vertical axis wind turbines." Journal of Physics: Conference Series 753(2): 022056.

34. Erwin, E., P. S. Tresna, et al. (2018). "Design optimization of hybrid biomass and wind turbine for minapolitan cluster in Domas, Serang, Banten, Indonesia." IOP Conference Series: Earth and Environmental Science 105(1): 012010.

35. Satria, D., Haryadi, et al. (2016). "Design of drying chamber and biomass furnace for sun-biomass hybrid rice-drying machine." AIP Conference Proceedings 1717(1): 050015.

Paper submitted: 17.02.2018.

Paper accepted: 16.08.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

423

Original Scientific Paper

doi:10.5937/jaes16-16557

Journal of Applied Engineering Science

Paper number: 16(2018)3, 549, 424 - 429

APPLIED POHLIG-HELLMAN ALGORITHM IN THREE-PASS PROTOCOL COMMUNICATION Robbi Rahim* University Malaysia, School of Computer and Communication Engineering, Perlis, Malaysia Three-pass Protocol is a method or technique that can be used by 2 (two) sender and recipient of the message to communicate with each other using XOR function, the problem that occurs is when the communication process there is parties who can know the messages sent from the sender to the recipient. To solve the problem we need an algorithm in this case Pohlig-Hellman algorithm, the use of Pohlig-Hellman algorithm on Three-pass Protocol ensures the security of messages sent by the sender to the receiver because each sender and receiver uses keys p, e, and d which is a number random and d is the inverse modulo of p and e of the sender and receiver, the results of this research suggest that it is impossible for the attacker or cryptanalyst to know the correct message quickly despite having adequate computer resources.

Key words: Algorithms, Recivers, Three-Pass Protocol, Pohlig-Hellman Algorithm, Sieve of Eratosthenes, Little Theorem Fermat INTRODUCTION The key protocol is a widely used exchange model where keys are distributed directly to recipients [1], the key distributions performed are usually performed on algorithms that have the symmetrical types [1], [2]. Cryptographic algorithm consists 2 (two) types that are symmetric and asymmetric [3]–[6], an symmetric algorithm use same key and on the asymmetric keys are used differently for each encryption and decryption[3]. The key in asymmetry type are not distributed, it’s different with symmetry type but in symmetrical computing calculations are much faster than asymmetric[1], [2], [5], [6].

Asymmetric type cryptography has a distinct disadvantage to the symmetrical type where main problem lies in key distribution, in asymmetric the problem is the key that must be long to improve security and require complex and long[7]–[10].Implementation of cryptographyprotocols without key exchanges is still an area of less concern, the strength of the keyless cryptography protocol is based on padding and the exchange of keys generated especially in using prime number[11], [12], one of the algorithms using cryptography without key exchange is the three-pass protocol[2], [13].

Three-pass protocol is a framework that allows senders to send encrypted messages to recipients without having to distribute keys to message recipients[14], called a three-pass protocol because senders and recipients do not need to exchange keys and communication is perform in three directions where both parties each use the key[14]. The three-pass protocol scheme enables various types of cryptographic algorithms to be implemented, Pohlig-Hellman[13] is one of the cryptographic algorithms that can be used in the three-pass protocol scheme. Pohlig-Hellman are not type asymmetric and asymmetric algorithm because both encryption and de-

cryption keys must be kept secret[13]. Pohlig-Hellman is different from Diffie-Hellman, Diffie-Hellman is a key exchange protocol that allows computers to generate similar secret keys for both systems and using public key as process in encryption and decryption[1], [15], [16]. The use of a three-pass protocol scheme with Pohlig-hellman is to cover the weaknesses of an algorithms that still using symmetric and asymmetric keys in the text message security process or by using session key.

METHODOLOGY Security in a cryptographic protocol is very important [12], this is because many attacks on the protocol due to the selection of wrong algorithms in cryptographic protocol. Modular arithmetic [17], greatest common divisor [13], [17], euclidean algorithm, prime number [11] and inverse modulo[18] are used on the application of Pohlig-hellman algorithm and three-pass protocol.

Modular Arithmetic Modular arithmetic is used in the process of encryption and decryption of the Pohlig-Hellman algorithm [13]. Encryption can be done to calculate the value of the message raised with the value of encryption key obtained then by doing modulo at predetermined prime values, the formula as follows: (2.1) Greatest Common Divisor Greatest Common Divisor is used in the Pohlig-Hellman algorithm at the time of the determination of additional keys [13], [17]. The conditional additional key must be a member of the odd number in which the GCD between the odd number and the totient value obtained must be

* Volga State University of Technology, 424000, the Republic of Mari El, Yoshkar-Ola, pl. Lenina 3, SmirnovMY@volgatech.net

424

Robbi Rahim - Applied pohlig/hellman algorithm in three-pass protocol communication

1. In the notation can be written (2.2) Greatest Common Divisor or GCD of the number of a and b is the largest integer d such that d | a and d | b. In this case we state that GCD (a, b) = d. Suppose that in determining GCD (5.2) = 1. It is found that the value of a is 5 and the value of b is 2.

Euclidean Algorithm This algorithm is used in the Pohlig-Hellman algorithm for the determination of the value of additional key numbers[19]. Suppose that there are two non-negative integers m and n where m ≥ n, the Eulidean algorithm can find the largest common divisor of m and n.

Modulo Invers If a and m are relatively prime and m> 1, then we can find the inversion of a modulo m. The inversion of a (mod m) [13], [20], [21], also called inversion multiplication, is an integer a-1. a* (a - 1) ≡1 (modm) (2.3) The relative prime definition it is known that GCD (a, m) = 1, and according to the equation there are integers p

and q, such that: p*a + q*m = 1 implying that: p * a + q * m ≡1 (modm) Since qm ≡ 0 (modm) then the value of p * a ≡ 1 (modm), this variance means that p is the inverse of a (modm), the above process is used to find the inverse value which will be used Pohlig-Hellman encryption and decryption process.

Pohlig-Hellman The concept of encryption on the Pohlig-Hellman Algorithm is similar to the RSA algorithm. Basically this algorithm is one asymmetric algorithm because it uses different keys for encryption and decryption [22]. In the Pohlig-Hellman algorithm does not use the public key concept because the key can be used at the time of encryption and decryption so it must be kept confidential[13], like in the RSA algorithm to be able to perform encryption and decryption were calculated by using formula as below: C = Pe mod n P = Cd mod n Provide that the value of e * d ≡ 1 Based on function that used in this implementation of Pohlig-Hellman ini Three-Pass Protocol, Figure 1 display how function key and encryption process.

Figure 1: Pohlig-Hellman Process in Three-Pass Protocol

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

425

Robbi Rahim - Applied pohlig/hellman algorithm in three-pass protocol communication

Three-Pass Protocol Three Pass Protocol is a similar process of sending and receiving messages without key distributions and exchange keys so that the owner and recipient of the message does not worry that the message will change or

read by a third party [17]. The Three Pass Protocol guarantees the absence of a key exchange between the party performing the encryption and decryption, each party having a private encryption key and a private decryption key [2], [13], [14], [18].

Figure 2: Three-Pass Protocol Scheme Process RESULTS AND DISCUSSION An experiment Pohlig-hellman algorithm with Three-pass protocol on communication is done gradually, first determining the key p, e and d on the sender and receiver of the message.

Table 1: Key of p, e and d Sender and Receiver Sender

Receiver

p = 761351

p= 761351

e= 116733

e= 410551

d=229097

d=336601

Getting e value can be done by using Euler's totient function, here is the process: p = 761351 totient p = 761350 Random e (1 < e < totient p -1) e = 719498, GCD(761350,719498) = 2 e = 598658, GCD(761350,598658) = 2

426

e = 482416, GCD(761350,482416) = 2 e = 11832, GCD(761350,11832) = 2 e = 51575, GCD(761350,51575) = 25 e = 116733, GCD(761350,116733) = 1 Based on the above function can be value e = 116733 due to the result e (1 <e <totient p -1) = 1, and this process applied to sender and receiver e value. Getting a d value is not much different from the e value function and also using the Extended Euclidean algorithm to get a relatively prime d value.

The next test is the communication process done with the Three-pass protocol with Pohlig-Hellman algorithm after the known value of p, e and d, the message to be secure is â&#x20AC;&#x153;Robbiâ&#x20AC;?.

1. Sender encrypt message First step sender encrypt message and get the result as Table 2.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Robbi Rahim - Applied pohlig/hellman algorithm in three-pass protocol communication

Table 2: Sender encrypt message Char

Decimal

Encrypt

Result

R

82

82 ^ 116733 mod 761351 = 38715

38715 to byte array = [59][151][0][0]

o

111

111 ^ 116733 mod 761351 = 161101

161101 to byte array = [77][117][2][0]

b

98

98 ^ 116733 mod 761351 = 458699

458699 to byte array = [203][255][6][0]

b

98

98 ^ 116733 mod 761351 = 458699

458699 to byte array = [203][255][6][0]

i

105

105 ^ 116733 mod 761351 = 731164

731164 to byte array = [28][40][11][0]

All byte array convert to base64 and sent to receiver [59][151][0][0][77][117][2][0][203][255][6][0][203][255][6][0][28][40][11][0] = O5cAAE11AgDL/ wYAy/8GABwoCwA=

2. Receiver encrypt message Message = O5cAAE11AgDL/wYAy/8GABwoCwA= The message will be decode to decimal and get the re-

sult as: [59][151][0][0][77][117][2][0][203][255][6][0][203][255][6] [0][28][40][11][0]

Table 3. Receiver decrypt message Decimal

Encrypt

Result

[59][151][0][0] to int = 38715

38715 ^ 410551 mod 761351 = 395200

395200 to byte array = [192][7][6][0]

[77][117][2][0] to int = 161101

161101 ^ 410551 mod 761351 = 417066

417066 to byte array = [42][93][6][0]

[203][255][6][0] to int = 458699

458699 ^ 410551 mod 761351 = 301152

301152 to byte array = [96][152][4][0]

[203][255][6][0] to int = 458699

458699 ^ 410551 mod 761351 = 301152

301152 to byte array = [96][152][4][0]

[28][40][11][0] to int = 731164

731164 ^ 410551 mod 761351 = 268029

268029 to byte array = [253][22][4][0]

All byte array convert to base64 and sent to sender again [192][7][6][0][42][93][6][0][96][152][4][0][96][152][4][0][253][22][4][0] = wAcGACpdBgBgmAQAYJgEAP0WBAA= 3. Sender decrypt message Message = wAcGACpdBgBgmAQAYJgEAP0WBAA= The message will be decode to decimal and get the result as: [192][7][6][0][42][93][6][0][96][152][4][0][96][152][4][0] [253][22][4][0] 4. Receiver decrypt message Message = yqUDAAfjBgBHrgEAR64BAJ9xAQA= The message will be decode to decimal and get the result as: [202][165][3][0][7][227][6][0][71][174][1][0][71][174][1][0] [159][113][1][0]

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Based on the above test results, the encryption and decryption results of the three-pass protocol are done twice for both the sender and the recipient of the message, the power of using the three-pass protocol is in the Pohlig-hellman algorithm which involves the inverse modulo process and the extended Euclidean algorithm and also using a key generator algorithm of up to 1,000,000 bits, in theory this would make it hard to read the messages sent and takes n billion years to decipher.

427

Robbi Rahim - Applied pohlig/hellman algorithm in three-pass protocol communication

Table 4: Sender decrypt message Decimal

Decrypt

Result

[192][7][6][0] to int = 395200

395200 ^ 229097 mod 239050 to byte array = [202] 761351 = 239050 [165][3][0]

[42][93][6][0] to int = 417066

417066 ^ 229097 mod 451335 to byte array = [7] 761351 = 451335 [227][6][0]

[96][152][4][0] to int = 301152

301152 ^ 229097 mod 110151 to byte array = [71] 761351 = 110151 [174][1][0]

[96][152][4][0] to int = 301152

301152 ^ 229097 mod 110151 to byte array = [71] 761351 = 110151 [174][1][0]

[253][22][4][0] to int = 268029

268029 ^ 229097 mod 761351 = 94623

94623 to byte array = [159] [113][1][0]

All byte array convert to base64 and sent to receiver again yqUDAAfjBgBHrgEAR64BAJ9xAQA= Table 5: Receiver decrypt message Decimal

Decrypt

Result

[202][165][3][0] to int = 239050

239050 ^ 336601 mod 761351 = 82

82 = R

[7][227][6][0] to int = 451335

451335 ^ 336601 mod 761351 = 111

111 = o

[71][174][1][0] to int = 110151

110151 ^ 336601 mod 761351 = 98

98 = b

[71][174][1][0] to int = 110151

110151 ^ 336601 mod 761351 = 98

98 = b

[159][113][1][0] to int = 94623

94623 ^ 336601 mod 761351 = 105

105 = i

CONCLUSION The implementation of Pohlig-hellman algorithm to the three-pass protocol can be done well, the use of algorithms such as euler totient, GCD, Rabin Miller, Extended Euclidean in the encryption process and decryption Pohlig-hellman algorithm can improve security especially from the key side used. The development of the Pohlig-helman algorithm is particularly possible at the time encryption and decryption combined with other algorithms or also the Three-pass protocol can be upgraded by combining with other protocols such as secret sharing or blind signatures.

REFERENCES 1. Abdalla, M., & Pointcheval, D. (2005). Simple Password-Based Encrypted Key Exchange Protocols.

Lectures Notes in Computer Science, 3376, 191– 208. https://doi.org/10.1007/978-3-540-30574-3_14 2. Al-Khalid, R. I., Al-Dallah, R. A., Al-Anani, A. M., Barham, R. M., & Hajir, S. I. (2017). A Secure Visual Cryptography Scheme Using Private Key with Invariant Share Sizes. Journal of Software Engineering and Applications, 10(01), 1–10. https://doi. org/10.4236/jsea.2017.101001

428

3. Alam, M. I., & Khan, M. R. (2013). Performance and Efficiency Analysis of Different Block Cipher Algorithms of Symmetric Key Cryptography. International Journal of Advanced Research in Computer Science and Software Engineering, 3(10), 2277–128.

4. Blumenthal, M. (2007). Encryption: Strengths and Weaknesses of Public-key Cryptography. CSRS 2007, 1–7. https://doi.org/PA 19085 CSC 3990 – Computing Research Topics 5. Bruce, S. (1996).

Applied cryptography.

John Wiley & Sons. https://doi.org/10.1017/ CBO9781107415324.004 6. Guo, M., Bhattacharya, P., Yang, M., Qian, K., & Yang, L. (2013). Learning mobile security with android security labware. In Proceeding of the 44th ACM technical symposium on Computer science education - SIGCSE ’13 (p. 675). https://doi. org/10.1145/2445196.2445394 7. Hoffstein, J., Pipher, J., & Silverman, J. H. (2014).

Diffie–Hellman key exchange. An Introduction to Mathematical Cryptography, 65–67. Retrieved from http://www.math.brown.edu/~jhs/MathCrypto/SampleSections.pdf Journal of Applied Engineering Science 16(2018)2

Robbi Rahim - Applied pohlig/hellman algorithm in three-pass protocol communication

8. Li, N. (2010). Research on diffie-hellman key exchange protocol. In ICCET 2010 - 2010 International Conference on Computer Engineering and Technology, Proceedings (Vol. 4). https://doi.org/10.1109/ ICCET.2010.5485276 9. Mesran, M., Abdullah, D., Hartama, D., Roslina, R., Asri, A., Rahim, R., & Ahmar, A. S. (2018). Combination Base64 and Hashing Variable Length for Securing Data. Journal of Physics: Conference Series, 1028, 012056. https://doi.org/10.1088/17426596/1028/1/012056 10. Nurdiyanto, H., & Rahim, R. (2017). Enhanced pixel value differencing steganography with government standard algorithm. In 2017 3rd International Conference on Science in Information Technology (ICSITech) (pp. 366–371). IEEE. https://doi.org/10.1109/ ICSITech.2017.8257140 11. Nurdiyanto, H., Rahim, R., & Wulan, N. (2017).

Symmetric Stream Cipher using Triple Transposition Key Method and Base64 Algorithm for Security Improvement. Journal of Physics: Conference Series, 930(1), 012005. https://doi.org/10.1088/17426596/930/1/012005 12. Putera, A., Siahaan, U., & Rahim, R. (2016). Dynamic Key Matrix of Hill Cipher Using Genetic Algorithm. International Journal of Security and Its Applications, 10(8), 173–180. https://doi.org/10.14257/ ijsia.2016.10.8.15 13. Rahim, R. (2017). Man-in-the-middle-attack prevention using interlock protocol method. ARPN Journal of Engineering and Applied Sciences, 12(22), 6483– 6487.

14. Rahim, R., Dahria, M., Syahril, M., & Anwar, B. (2017). Combination of the Blowfish and LempelZiv-Welch algorithms for text compression. World Transactions on Engineering and Technology Education, 15(3), 292–297.

15. Rahim, R., & Ikhwan, A. (2016a). Cryptography Technique with Modular Multiplication Block Cipher

and Playfair Cipher. International Journal of Scientific Research in Science and Technology (IJSRST), 2(6), 71–78.

16. Rahim, R., & Ikhwan, A. (2016b). Study of Three Pass Protocol on Data Security. International Journal of Science and Research, 5(11), 102–104. https:// doi.org/10.21275/ART20162670 17. Rahim, R., Winata, H., Zulkarnain, I., & Jaya, H. (2017). Prime Number: an Experiment Rabin-Miller and Fast Exponentiation. Journal of Physics: Conference Series, 930(1), 012032. https://doi. org/10.1088/1742-6596/930/1/012032 18. Schneier, B. (1996). Applied Cryptography: Protocols, Algorithms, and Source Code in C, Second Edition. Network. https://doi.org/10.1016/S0740624X(96)90083-0 19. Sklavos, N., Papadomanolakis, K., Kitsos, P., & Koufopavlou, O. (2002). Euclidean algorithm VLSI implementations. In Proceedings of the IEEE International Conference on Electronics, Circuits, and Systems (Vol. 2, pp. 557–560). https://doi.org/10.1109/ ICECS.2002.1046226 20. Sriadhi, S., Rahim, R., & Ahmar, A. S. (2018).

RC4 Algorithm Visualization for Cryptography Education. Journal of Physics: Conference Series, 1028(1), 012057. https://doi.org/10.1088/17426596/1028/1/012057 21. Uchoa, A. G. D., Pellenz, M. E., Santin, A. O., & Maziero, C. A. (2007). A Three-Pass Protocol for Cryptography Based on Padding for Wireless Networks.

In 2007 4th IEEE Consumer Communications and Networking Conference (pp. 287–291). IEEE. https:// doi.org/10.1109/CCNC.2007.63 22. Yang, L., Wu, L.-A., & Liu, S. (2002). A quantum three-pass cryptography protocol. In Proceedings of SPIE - The International Society for Optical Engineering (Vol. 4917, pp. 106–111). https://doi. org/10.1117/12.483035

Paper submitted: 14.02.2018.

Paper accepted: 17.08.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

429

Original Scientific Paper

doi:10.5937/jaes16-16866

Journal of Applied Engineering Science

Paper number: 16(2018)3, 550, 430 - 440

NUMERICAL STUDY OF CONFINEMENT EFFECT OF FRP COATINGS ON BEHAVIOR OF RC FRAMES BY USING NONLINEAR ANALYSIS Hossein Tahghighi1*, Mohammad Rashid Gholami1 University of Kashan, Kashan, Islamic Republic of Iran Confinement of both existing and newly constructed reinforced concrete (RC) frames by fibre reinforced polymer (FRP) has been commonly used in recent decades. This is because of its ability to enhance the behavior of RC frames under lateral loading.The purpose of this research is to perform a numerical study on effect of FRP confinement on strength and ductility of RC frames using OpenSEES software.After evaluating the accuracy of numerical model results by comparison with experimental results, the behavior of FRP-resistant frames is discussed.The 8-story studied building is reinforced in various ways by GFRP and CFRP composite sheets.Results of nonlinear analysis show that un-wrapped frame exhibits low resistance and less energy absorption.However, the loading capacity and the performance level have improved significantly in the frames strengthened with FRP materials.Also, increased capacity and reduced drift present more amount of increment in cases which CFRP materials are used.

Key words: Reinforced concrete, FRP materials, Confinement, Flexural retrofitting, Nonlinear analysis INTRODUCTION The widespread destructions of concrete structures in past earthquakeshave shown that identification and strengthening of vulnerable structures gain more importance in order to mitigate human and economic casualties.A comprehensive overview of traditional seismic rehabilitation techniques was presented by FEMA-547 [1].

Conventional techniques which utilize braces, jacketing or infills, as well as more recent approaches including base isolation and supplemental damping devices, have been considered to strengthen the building and as such improve its behavior. In the past decade, an increased interest in the use of advanced nonmetallic materials has been reported.

Among the variousrehabilitation methodsin RC buildings, the use of Fiber Reinforced Polymers(FRP) has been widely considered in recent years.FRP is a compound material with very high tensile strength which consists of two parts.The first part consists of fibers that are the load bearing part toburden of bearing loads.The adhesive or resin forms the second part, where gains the role of maintaining the fibers together and transferring power to all parts.Typical types of FRP used in structural reinforcement are carbon (CFRP) and glass (GFRP).Other advantages of FRP materials include low specific gravity, high corrosion resistance and ease of use. Where disadvantages are sudden failure and its vulnerability to fire and heat.

FRP materials are used to increase confinement, which is a desired and favorable condition in concrete.Therefore, the tension-strain behavior of concrete confined with FRP improves and increases significantly [2-6].FRP confinement causes the performance of concrete columns * University of Kashan, Isfahan province, Kashan, Islamic Republic of Iran

to increase significantly due to improved specification of confinedconcrete.Saadatmanesh et al. [7] performed a test on 5 circular RC columnsfrom base of a bridge.The samples were designed to comply with bridge seismic regulations ofbefore 1971and the area with high earthquake risk.Test results showed thatstrength and ductility increased significantly incolumns which were wrappedby GFRP composite strips in the plastic hinge area.Chaallal et al. [8] studied the effects of concrete strength characteristics, ratio of length to column section widthand number of CFRP layers on confinement by examining 90 short columns with square and rectangular sections under axial load.They showed that confinement in both cross sections increased compressive strength and ductility, although this is less than of circular columns.They also showed that increase in number of layers improved the performance of column and this performance in columns with low-strength concrete was betterthan columns with normal-strength or high-strength concrete.

Balsamo et al. [9] conducted a study on a four-story RC frame with columns and beams covered with FRP.Their research showed that the frame wrapped with FRP could withstand against an earthquake 1.5 times the design earthquake intensity.In experimental studies by De Luca et al. [10], the effect of FRP sheets on axial force and column deformation was investigated.In these studies, several RC columns with square and rectangular sections were constructed in real scale and were rehabilitated with FRP sheets of glass and basalt and were applied under pure axial load.It should be noted that these specimens were designed according to the regulations of the 70's.The results of mentioned studies showed that although FRP sheets increase the axial strength of concrete, but they are more effective in increasing the concrete strain capacity.

430

Mohammad Rashid Gholami et al. - Numerical study of confinement effect of FRP coatings on behavior of RC frames by using nonlinear analysis

The study by Wang et al. [11] provides test data and relevant analysis on seismic behavior of deficient circular RC columns wrapped with CFRP under high axial load. Their experimental results indicated that the use of CFRP resulted in remarkable improvement in ductility, energy dissipation, and strength of the deficient lateral confinement RC columns. In another study, Ronagh and Eslami [12] used FRP to confine columns at the critical zones of an 8-storey poorly-confined frame. Their analytical results showed that the seismic performance and ductility increased substantially.Following the aforementioned researches, the purpose of this paper is to study the behavior of RC structures strengthened with FRP. FRP sheets have been used to increase the flexural strength of buildings whichinclude the possibility of plastic hinge.One of the objectives of present study is to compare the effects of type and condition of FRP coating on force capacity and displacement of RC frames.Therefore, GFRP and CFRP materials are used in different conditions includingfully FRP-covered frames, covered only in beams, covered only in columns and covered only in plastic hinge areas.It should be mentioned that studied models are evaluated by nonlinear static analysis (Pushover) with the help of OpenSEES software [13].

STRESS-STRAIN MODELS OF FRP-CONFINED CONCRETE The first stress-strain models of concrete encapsulated

by FRP were obtained from developing the models provided for steel-confined concrete by Mander and Priestley [14]. However, the researchers later concluded that development of confined models made for steel are not

suitable for FRP-confined columns, due to the different mechanical properties of steel and FRP. For this reason, distinctbilinear models are provided for stress-strain curve of concrete confined with FRP [2, 3].Besides, other different stress-strain models have been proposed by other researchers, where in most of them behavior of FRP-confined concrete is assumed to be in the same upstream branch as notconfinedconcrete [6-4].This is considered since of insignificance confinement provided

by FRP due tolow side-strain ofconfined concrete atinitial loading stages.

Lam and Teng [15] proposed a new design-oriented stress–strain simple model for concrete confined by FRP wraps with fibers only or predominantly in the hoop direction based on a careful interpretation of existing test data and observations. In the meantime, this model captures all the main characteristics of the stress–strain behavior of concrete confined by different types of FRP. In addition, for unconfined concrete, this model reduces directly to idealized stress–strain curves in existing design codes. The basic assumptions of this simple model are: (i) the stress–strain curve consists of a parabolic first portion and a straight line second portion, as given in Figure 1; (ii) the slope of the parabola at εc=0 (initial slope) is the same as the elastic modulus of unconfined concrete Ec; (iii) the nonlinear part of the first portion is affected to some degree by the presence of an FRP jacket; (iv) the parabolic first portion meets the linear second portion smoothly (i.e. there is no change in slope between the two portions where they meet); (v) the linear second portion ends at a point where both the compressive strength and the ultimate axial strain of confined concrete are reached.

Figure 1: Stress–strain relationship for FRP confined concrete [after 15].

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

431

Mohammad Rashid Gholami et al. - Numerical study of confinement effect of FRP coatings on behavior of RC frames by using nonlinear analysis

Based on the assumptions listed above, the proposed stress–strain model for FRP-confined concrete is given by the following expressions: (1) where εc is axial strain, fo is the intercept of the stress axis by the linear second portion, εcu is the ultimate strain. The parabolic first portion meets the linear second portion with a smooth transition at εt, which is given by: (2) where E2 is the slope of the linear second portion, given by: (3) where fcc is compressive strength of confined concrete.

STUDIED REINFORCED CONCRETE BUILDINGS In current paper, an 8-story RC structure is studied with a moment frame system that is strengthened invarious ways with CFRP and GFRP materials.

Original RC Frame In this section, the building of primary RC Frame is introduced, such that one of its internal frames is used for analysis.The studied 8-story moment frame has three spans of 5 meters in length, where distance between frames and height of floors are 5 and 3 meters, respectively.Concrete design of building is carried out by [12] and based on average ductility criteria of ACI 02-318 [16].Figure 2 represents the plan and view of building.

Further, Figure 3 and Table 1 depict the dimensions and details of reinforcement in beams and columns of 8-story RC moment frame, where designdead load and live loadare 10 and 30 kN/m, respectively.Concrete compressive strength ( ) and rebar yield strength (fy) are considered as 25MPa and 420MPa.Moreover,shear reinforcement steel bars are chosen of 10 mm diameter.Assuming that the building is located in type III soil and region with high seismic hazard, seismic loads arecalculated in accordance with the 2800 regulation [17].Due to the regularity of building, torsional effects of seismic loads are ignored.

Table 1: Beams, Columns and reinforcement details of the studied frame Element

Story No.

B (mm)

H (mm)

D (mm)

d' (mm)

Ast

As

A's

Transverse Steel spacing

Column

1,2 3,4 5,6,7,8

600 600 500

600 600 500

540 540 540

60 60 60

16 Ø25 16 Ø18 16 Ø16

-

-

150 150 125

Beam

1,2,3,4 5,6 7,8

500 500 500

500 500 500

540 540 540

60 60 60

-

6 Ø25 6 Ø22 6 Ø18

4 Ø25 4 Ø22 3 Ø18

100 100 100

Table 2: Mechanical properties of composite sheets [10, 18] Composite type

Tensile strength, fr(MPa)

Ultimate tensile strain, Efr

Tensile modulus, Ef (MPa)

Thickness, tf (mm)

CFRP GFRP

3900 3241

0.0155 0.045

240000 72397

0.165 0.589

FRP-strengthened RC frame In following study, both GFRP and CFRP composites are used to compare the efficiency of FRP materials in strengthening of RC structures.In accordance with design philosophy of poor beam-strong column in seismic design codes, bending capacity of columns at junction of beam to column should be greater than total bending strength of beams.To meet this requirement, Tables 2 and 3 represent the characteristics and number of required GFRP and CFRP composite sheets in frame

432

members, respectively.According to Table 3, the thickness of composite sheets in beams and columns is considered to be same, but forfifth and sixth floors’ column which requiremore sheets.Moreover, to create equal modeling conditions, the number of FRP layers is calculated in such a way that the strength of GFRP and CFRP materials be equal.In other words, the product of tensile modulus (Ef), ultimate tensile strain (Efr) and thickness (tf) of a composite should be equal to other composite’s tensile strength (fr).

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Mohammad Rashid Gholami et al. - Numerical study of confinement effect of FRP coatings on behavior of RC frames by using nonlinear analysis

Figure 2: Configuration of RC building model: (a) plan and (b) elevation

Figure 3: Typical sections: a) column and b) beam [after 12] It is worth pointing out, by assuming a proper method for implementation of FRP coatings, the separation of com-

posite sheets in modeling has been neglected [19].

Table 3: Number of composite sheets in the GFRP and CFRP retrofitted frame [12] Element

Columns

Story No

7-8 5-6 3-4 1-2

7-8 6 5

No. of GFRP layers No. of CFRP layers

4 4 2 2 6 9 6 6

- - 2 2 - 6 6 6

NUMERICAL ANALYSIS Concrete, steel and FRP are introduced using uniaxial materials by parameters such as compressive strength, strain, modulus of elasticity and yield stress.According to Figure 4, for introduction of concrete, the uniaxialMaterial Concrete01 command is used based on the Kent-Park model [20] and the steel02 uniaxialMaterial command is employed for longitudinal reinforcement, based on the Giuffre-Menegotto-Pinto model [13].In addition, the Lum and Tang model [15] has been applied for the concrete enfolded with FRP. It is necessary to explain that the effect of number and spacing of transversal reinforcements or crumbs is taken into account by using the amount of confinement in concrete whichseparates the concrete into confined and unconfined parts.After the introduction of required materials, fiber section is hired to describe Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Beams 1-4

the beams and columns.As displayed in Fig. 5, sections wrappedwith FRP materials consist of 3 parts, including FRP coating, confinedconcrete with cross-reinforcement (concrete core) and concrete confined with FRP (concrete coating) and finally, a nonlinear beam-column element with widespread plasticity distribution is utilized for definition of beam and column elements [13].

The analyzes of present work were carried out using non-linear static analysis method (Pushover) in OpenSEES [13].In Pushoveranalysis, the gravity load is assumed to be constant and the lateral load is applied to the structure gradually increasing as long as the displacement of control point is achieved to the target value.

Besides, the total dead load plus 20% of live load is considered in Pushover analysis in accordance with Iran's seismic regulation[17], where reverse triangular distribution in height was considered as lateral load model.

433

Mohammad Rashid Gholami et al. - Numerical study of confinement effect of FRP coatings on behavior of RC frames by using nonlinear analysis

Figure 4: The stress-strain relationship for: a) Concrete02 material and b) Steel02 material [13]

Figure 5: An example fiber section [13] Further, the P-Δ effect in OpenSEES software is considered by using the geomTransf PDelta command in nonlinear analyzes.

For a comprehensive nonlinear analysis, consideration of the shear failure in members is necessary. However, past studies have reported that even for under-designed RC buildings, the shear demand is significantly lower than the shear capacity in both beams and columns and that no shear failure would occur [21, 22]. Therefore, in this study, only inelastic flexural behavior of structural elements was considered through concentrated rigid plastic hinges at the ends which are susceptible to inelastic behavior.

Obviously, nonlinear time history analyses may provide most capabilities to explore FRP strengthening effects for RC buildings. Furthermore, earthquake ground motions close to a ruptured fault can be significantly different from far-fault records. Near-fault motions are noticeably influenced by the forward directivity when the fault rupture propagates toward a site [23, 24]. Mortezaei et al. [25] presented the results of typical existing RC buildings subjected to far-fault and near-fault ground motions and the potential improvements achievable after FRP retrofitting of the buildings. Thus, the detailed evaluation of the effect of FRP rehabilitation on the seismic performance of existing RC buildings against comprehensive earthquake ground motion database can be a matter of further research.

434

VERIFICATION EXPERIMENTS In current section, multipleexperimental samples ofRC column confined with CFRP materials are introduced firstly and then the analysis results of numerical model are evaluated and compared withexperimental results.

Introducing experimental samples The studied columns consist of 8 circular RC columns on a scale of one-third which their reinforcement details were designedin accordance withregulations’ criteriaof 80's. [11]Two RC columns without FRP strengthening were used as control samples to evaluate the effect of CFRP-retrofitting on six other rehabilitated columns in plastic hinge area, whereall columns have a diameter of 180 mm and length of 1260 mm.The concrete coating thickness is 8 mm and six longitudinal bars of 12 mm diameterand cross-sectional reinforcement of 4 mm diameter are utilized incolumn. In addition, spacing of transverse reinforcement inmiddlespan and at both ends ofsamples are 120 mm and 60 mm, respectively.It is worth noting that CFRP coatings with different lengths are wrapped on smooth and polished surface of column.

Column details,specimens’ specificationsandcharacteristics ofmaterials used forcolumns are representedin Figure 6 and Tables 4 and 5, correspondingly.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Mohammad Rashid Gholami et al. - Numerical study of confinement effect of FRP coatings on behavior of RC frames by using nonlinear analysis

Figure 6: RC column specimens details, mm [after 11] Table 4: The details of test specimens [11] Specimens

fc' (MPa)

P/fc'Ag

wrapping length (mm)

Retrofit detail

C1 C2 C 1n 1 C 1n 2 C 1n 3 C 2n 1 C 2n 2 C 2n 3

54.8 71.2 54.8 54.8 54.8 71.2 71.2 71.2

0.45 0.45 0.45 0.55 0.65 0.45 0.55 0.65

200 320 320 200 320 320

Control Control 4-layer 4-layer 4-layer 4-layer 4-layer 4-layer

Table 5: Material properties of the test specimens [11] Materials

Yield strain

Tensile modulus (GPa)

Tensile strength (MPa)

Longitudinal steel Hoop CFRP

0.00179 0.002 0.018

198 201 230

354 402 3430

Validation of numerical models Figures 7 and 8 displayPushover curves obtained from experimental results [11] and simulation ofpresent work for all un-wrapped and wrapped columnsfor concrete compressive strength of 54.8 MPa (type C1) and 71.2 MPa (type C2), respectively.As it can be seen, unwrapped circular concrete columns bear a slight displacement indicating they are brittle, where in the case of RC circular columnsrehabilitated with FRP materialsthe Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

displacement has increased, which impliesmoreductility.

Besides, by increasing axial load, maximum lateral load of strengthened columns increases. Another worth-mentioning point is, with increase of coating length in circular FRP-strengthened RC columns the strength increases as well. In addition, higher compressive strength ofconcrete makes the FRP-rehabilitated columnsmore durable against base shear.

435

Mohammad Rashid Gholami et al. - Numerical study of confinement effect of FRP coatings on behavior of RC frames by using nonlinear analysis

Figure 7. Comparison of capacity curves from finite element modeling technique with experiments of wrapped and unwrapped RC columns (fc'= 54.8 MPa)

Figure 8. Comparison of capacity curves from finite element modeling technique with experiments of wrapped and unwrapped RC columns (fc' = 71.2 MPa) RESULTS AND DISCUSSION Pushover analysis results of RC frames confined by various types of CFRP and GFRP materials described in Section 3 are compared and discussed in following sections.

Comparison of capacity in un-wrappedframes with wrappedones Curves of Pushover analysis inframes rehabilitated with GFRP and CFRP compared to the conventional frame are demonstrated in Fig. 9.As represented in Figure 7, there is a significant increase of load capacity and bearing strength in FRP-wrapped frames, compared to the original frame, where indicates the high efficiency of these material. Therewith, CFRP-reinforced frame is more resistant and shows higher bearing capacity (load bearing) compared to the GFRP- rehabilitated frame,

436

which is due to higher tensile strength of carbon fiber than glass fiber.

Comparison of CFRP effect on beams and columns In current section, results of CFRP-wrapped RC frameâ&#x20AC;&#x2122;s retrofitting are examined in three different methods, which include totalstrengthened of all beams and columns in frame, strengthened only in beams and strengthened only in columns, while it is assumed that cracking does not occur in beam-column joint area.As Figure 10 shows, enhancement of strength and stiffness is higher in only-columns strengthened frame thanonly-beams strengthened frame. Even though, use of CFRP simultaneously in beams and columns increases the strength and stiffness of frame much more.This suggests that simultaneous use of FRP materials in beams and columns is more effective than only one of them hired in retrofitting.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Mohammad Rashid Gholami et al. - Numerical study of confinement effect of FRP coatings on behavior of RC frames by using nonlinear analysis

Figure 9: Comparison of pushover curves of the original and rehabilitated 8-story frames (fully wrapped FRP)

Figure 10: Comparison of pushover curves of the original and three different types of the rehabilitated 8-story frames Comparison of CFRP coating effect on plastic hinges The other cases studied in this researcharecomparison of capacity curve instrengthened frame’s plastic hinge areaswith capacity of a frame which its total members have been rehabilitated with CFRP materials. Hence, the modeling of these two methods in OpenSEES software is almost identical, with a single difference in connection-strengthenedframe model in whicheach frame element (beamand column) is divided into three parts, including one part in middle ofelement without FRP and two FRP-covered parts at beginning and end of element (joints), wherethe plastic hinges are made.The minimum length of CFRP coating at plastic hinge’s formation area is usually considered equal to beam depth [12]. Howbeit,length of applied composite sheets in beams and columns is defined 1.5 times the thickness of beam (75 cm) in this study, in order to eliminate the possibility of plastic hinge’s displacement. Thus, the schematic design ofstrengthening inframe’s joint areas is depicted in Figure 11.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Further, Figure 12 displayspushover curves of the frame rehabilitated with CFRP in joints and throughout the frame.It can be observed that although the frame overall retrofitted with CFRP in beams and columnshas a higher resistance than framerehabilitated only in joints, but this increment is not significant, implying that in order to save time and money, it is possible to ignore the slight increase in overall strengthening and only beam-column joints strengthening using FRP materials.

Comparison of driftin wrapped and un-wrapped frames Effect of FRP use on relative drift of RC frame’s floors is studied such thatan 8-story un-wrapped frame has been compared with two cases wrapped by glass and carbon fiber. According to Figure 13,floordriftof FRP-wrapped frames has been reduced significantly. While, this reduction is higher in frame with carbon fiber than the glass fiber frame. As shown in Fig. 13, the drift in un-wrapped frame has exceeded the proposed 2.5% allowable limit in seismic regulations [17] and reached up 3.5%.

437

Mohammad Rashid Gholami et al. - Numerical study of confinement effect of FRP coatings on behavior of RC frames by using nonlinear analysis

Figure 11: Scheme of FRP strengthening in an interior joint [after 12]

Figure 12: Comparison of pushover curves of the rehabilitated 8-story frames Besides, in FRP-wrapped frames, drift is reduced and within the permitted range, indicating the effective effi-

ciency of FRP materials in improving the performance of rehabilitated frames.

Figure 13: Comparison of inter-story drifts of the original and rehabilitated 8-story frames (FRP at joints)

438

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Mohammad Rashid Gholami et al. - Numerical study of confinement effect of FRP coatings on behavior of RC frames by using nonlinear analysis

CONCLUSIONS

REFERENCES

In currentresearch, results of nonlinear analysis of RC frames are presented in OpenSEES finite element software, regarding confinement effect of concrete which is due toaction oftransversal bar onbeam and column and covering effect ofFRP in four modes of coverage in column total height, beam total length, simultaneous beam and column total lengthand finally coating only in plastic zones of beam and column.Verification of numerical model in column specimens showed that stress-strain modelis able to determine the capacity curves of wrapped concrete, under axial and lateral loadings. The results of analysis are presented with aim of comparing the effect of confinement methods on frame behavior, which several outstanding results are referred in following: 1. FRP coating in plastic joints area improves ductility and energy absorption ofcolumns. Thus, bygreater length ofFRP coating,strength and ductility ofcolumns rises up to some extent.

2. Strength incrementin beams and columns that are overall wrapped with FRP is not remarkable in comparison to the one that is reinforced in only plastic hinge. Therefore, it is possible to ignore the overall retrofitting of frame members due to cost saving.

3. CFRP-reinforced frames increase the strength of structure more than the frames strengthened with GFRP materials.

4. Simultaneous strengthening of RC framesby FRP materials in beams and columns perform better than only-beams or only-columns strengthening.

5. Inter-story drifts decrease in frames rehabilitated with glass and carbon fiber, which leads to a structure with a more ductile behavior.In addition, drift reduction is greater when using carbon fiber.

1. [FEMA-547. Techniques for the Seismic Rehabilitation of Existing Buildings. Federal Emergency Management Agency (FEMA), USA, 2006.

2. Xiao, Y., Wu, H. Compressive behavior of concrete confined by carbon fiber composite jackets. Journal of materials in civil engineering, 12 (2): 139-146, 2000.

3. Karbhari, V.M., Gao, Y. Composite jacketed concrete under uniaxial compression-verification of simple design equations. Journal of materials in civil engineering, 9 (4): 185-193, 1997.

4. Harajli, M.H., Hantouche, E. and Soudki, K. Stressstrain model for fiber-reinforced polymer jacketed concrete columns. ACI structural journal, 103 (5): 672, 2006.

5. Wu, G., Lu, Z., Wu, Z. Stress–strain relationship for FRP-confined concrete cylinders. In Proceedings of the 6th international symposium on FRP reinforcement for concrete structures (FRPRCS), Singapore 552-60, 2003.

6. Wei, Y.Y., Wu, Y.F. Unified stress–strain model of concrete for FRP-confined columns. Construction and Building Materials, 26(1): 381-392, 2012.

7. Saadatmanesh, H., Ehsani, M. R., Jin, L. M. Seismic Strengthening of Circular Bridge Pier Models with Fiber Composites. ACI Structural Journal, 93(6): 936947, 1996.

8. Chaallal, O., Shahawy, M., Hassan, M. Performance of axially loaded short rectangular columns strengthened with carbon fiber-reinforced polymer wrapping.

Journal of Composites for Construction, 7(3): 200208, 2003.

9. Balsamo, A., Colombo, A., Manfredi, G., Negro, P., Prota, A. Seismic behavior of a full-scale RC frame repaired using CFRP laminates. Engineering Structures, 27(5): 769–780, 2005.

10. De Luca, A., Nardone, F., Matta, F., Nanni, A., Lignola, G.P., Prota, A. Structural evaluation of full-scale FRP-confined reinforced concrete columns. Journal of Composites for Construction, 15(1): 112–123, 2011.

11. Wang, Z. Y., Wang, D.Y., Sheikh, S.A., Liu, J.T. Seismic performance of FRP-confined circular RC columns. International Conference on FRP Composites in Civil Engineering, Beijing, China, 2010.

12. Ronagh, H.R., Eslami, A. Flexural retrofitting of RC buildings using GFRP/CFRP–A comparative study.

Composites: Part B, 46: 188-196, 2013.

13. OpenSees. Open system for earthquake engineering simulation. Pacific Earthquake Engineering Research Center, University of California, Berkeley, USA, 2016.

14. Mander, J. B., Priestley, M. J. N. Theoretical stressstrain model for confined concrete. Journal of Structural Engineering, 114(8): 1804-1826, 1988.

ACKNOWLEDGMENTS The authors would like to thank M. Alborzi for valuable comments on the numerical analyses and great help during this research.Discussions with M. Forughi are also appreciated.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

439

Mohammad Rashid Gholami et al. - Numerical study of confinement effect of FRP coatings on behavior of RC frames by using nonlinear analysis

15. Lam, L., Teng, J.G. Design-oriented stress–strain model for FRP-confined concrete. Journal of Construction and Building Materials, 17 (6): 471-489, 2003.

16. ACI Committee. Building code requirements for structural concrete (ACI 318-02) and commentary.

American Concrete Institute, International Organization for Standardization, 2002.

17. BHRC. Iranian code of practice for seismic resistant design of buildings (Standard No. 2800). Building and Housing Research Center, Tehran, Iran, 2014.

18. Mahini, S.S, Ronagh, H.R. Web-bonded FRPs for relocation of plastic hinges away from the column face in exterior RC joints. Composite Structures, 93(10): 2460–2472, 2011.

19. Mostofinejad, D., Mahmoudabadi, E. Grooving as alternative method of surface preparation to postpone debonding of FRP laminates in concrete beams.

Journal of Composites for Construction, 14(6): 804– 811, 2010.

20. Kent, D.C., Park, R. Flexural members with confined concrete. Journal of the Structural Division, 97 (7): 1969-1990, 1971.

21. Di Ludovico M, Prota A, Manfredi G, Cosenza E. Seismic strengthening of an under-designed RC structure with FRP. Earthquake Engineering Structural Dynamics, 37:141–62, 2008.

22. Inel M, Ozmen HB. Effects of plastic hinge properties in nonlinear analysis of reinforced concrete buildings. Engineering Structures, 28: 1494–502, 2006.

23. Tahghighi H. Earthquake fault induced surface rupture–A hybrid strong ground motion simulation technique and discussion for structural design. Earthquake Engineering and Structural Dynamics 40: 1591-1608, 2011.

24. TahghighiH. Simulation of strong ground motion using the stochastic method: Application and validation for near-fault region. Journal of Earthquake Engineering 16: 1230-1247, 2012.

25. Mortezaei, A., Ronagh, H.R., Kheyroddin A. Seismic evaluation of FRP strengthened RC buildings subjected to near-fault ground motions having fling step.

Composite Structures 92: 1200–1211, 2010.

Paper submitted: 16.03.2018.

Paper accepted: 14.09.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

440

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Original Scientific Paper

Journal of Applied Engineering Science

doi:10.5937/jaes16-18433

Paper number: 16(2018)3, 551, 441 - 446

THE APPROXIMATE AND NUMERICAL SOLUTION OF ROMANOVSKIJ LINEAR PARTIAL INTEGRAL EQUATIONS Anatolij Semenovich Kalitvin1*, Vladimir Anatoljevich Kalitvin1 1 Lipetsk State Pedagogical P. Semenov-Tyan-Shansky University, Russian Federation The study of Markov chains with two-way coupling leads to the solution of linear partially integral equations of the second kind in the space of functions continuous on the square. A characteristic feature of the equations is the permutation of variables for the unknown function under the integral sign and integration over part of the variables.

Equations of such types are not Fredholm integral equations and for their study a well-developed theory of Fredholm integral equations of the second kind canâ&#x20AC;&#x2122;t be directly applied. The equations considered in the article we call partially integral equations of Romanovskij, who first obtained them in the study of Markov chains with two-way coupling and studied these equations in the case of continuous kernels. An explicit solution of partially integral Romanovskij equations can be found in rare cases, and therefore the problem of studying approximate and numerical methods for solving such equations is vital. When using approximate and numerical methods, it should be taken into account that the linear partially integral operator in the Romanovskij equation is not completely continuous, and the direct application of methods associated with the complete continuity of operators for its solution requires justification. The justification of approximate and numerical methods for solving linear partially integral equations of Romanowskij is given in the annotated paper. The paper contains theorems on the solvability of equations, results on various approximate and numerical methods for their solution, the theorem on the solution of linear partially integral equations by Romanovskij, using the method of mechanical quadratures, together with an estimate of the rate of convergence of a numerical solution to an exact solution of this equation.

Key words: Linear integral equations, Linear equations of Romanovskij, Partial integrals, Approximate and numerical methods INTRODUCTION Markov chains play an extremely important role in the study of various problems of technology, genetics, physics and other problems.

The problems of doubly connected and multiply connected Markov chains are reduced to integral equations first studied in the case of continuous given functions by V.I. Romanovskij [01]. Solutions of such equations can be found explicitly in rare cases, so it is important to develop approximate and numerical methods for their solution.

Approximate and numerical methods for solving the inhomogeneous integral Romanovskij equation: (1) are being considered in this article.

The singularity of equation (1) is connected with the rearrangement of variables for the unknown function under the integral sign and integrating it in one of two variables.

Because of this, the operator M in equation (1) is not integral (since the criterion of A.V. Bukhvalov [02]) on the integral representation of a bounded linear operator is not fulfilled) and is not completely continuous.

Fundamentals of the theory of equation (1) with a continuous kernel are constructed by V.I. Romanovskii [01], using methods analogous to the method of Fredholm determinants, in [03] studied more general classes of linear

integral equations of Romanovskii type with partial integrals and kernels of more general types. We note that the Fredholm property of equation (1), whose kernel is a continuous function by (t,s) with values in the space of summable functions, is established in [04] for the case of a space of continuous functions.

The conditions for the equivalence of equation (1) to the Fredholm integral equation of the second kind and the conditions for the invertibility of these equivalent equations in the space of functions continuous on the square are contained in Theorem 1; in Theorem 2 conditions for the unique solvability in the space of continuous functions of equation (1) and the conditions for which equation (1) either has no continuous solutions, or has more than one continuous solution are given.

Next, we study the approximate solution of the invertible equation (1). This equation is replaced by a linear partially integral Romanovskij equation with a degenerate continuous kernel close to the kernel m. The solution of the approximate equation is constructed explicitly, and the estimate of the error of the approximate solution of equation (1) is given.

For other methods of approximate solution, equation (1) is reduced to a system of Fredholm integral equations of the second kind with an additional condition for solutions or is transformed into a two-dimensional integral Fredholm equation of the second kind with a degenerate kernel.

*Lipetsk State Pedagogical P. Semenov-Tyan-Shansky University, Lenina str. 42, Lipetsk 398020, Russia; akalitvin@inbox.ru

441

Anatolij Semenovich Kalitvin et al. - The approximate and numerical solution of romanovskij linear partial integral equations

Numerical schemes for solving equation (1) with a continuous kernel are constructed. The integral equation (1) is replaced by a system of linear algebraic equations, whose solution tends to the solution of equation (1) for an unrestricted refinement of a segment [a, b] .

Another method of numerical solution of equation (1) is associated with the replacement of this equation by a system of linear integral equations of the second kind with an additional condition and the replacement of this system by a system of linear algebraic equations with an additional condition on the solution. Theorem 3 shows the conditions under which the solution of equation (1) can be found by the method of mechanical quadratures, and the rate of convergence of the numerical solution to the exact solution is given.

RESULTS AND DISCUSSION CONDITIONS ON THE SOLVABILITY OF EQUATION (1) Let D= [a, b]x[a, b], C=C(D), be the space of functions continuous on D, L1, be the space of summable functions on [a, b], C(L1) be the space of continuous vector-valued functions (t, s) → z(t, s, ., .)

L1 (D) and let m C(L1).

By [03], the operator M acts and is bounded in C. As noted above, the operator M is not a completely continuous integral operator in C. However, the operator M2 is a completely continuous integral operator in the space C [05].

Indeed, applying the Fubini theorem, we establish the equality:

∈

∈

Applying the operator (I+M)-1 to both sides of equation (3) to the left and taking into account that ± 1 ∈/ σ ( M ), we get that x(t,s) is the unique solution of equation (1). Thus, in the case under consideration, equations (1) and (2) are equivalent and have a unique solution x(t,s).

If, however, 1 is an eigenvalue of the operator M2 and ± 1 ∈/ σ ( M ), then by (3) the operator I+M is invertible in C. Applying the operator (I+M)-1 to both sides of equation (3), we obtain the equivalent equation (1).

Thus, it is valid Theorem 1. Let m ϵ C(L1) and f be an arbitrary function in C. Then the following assertions hold: a. if 1∈ / σ(M2) then in C the equation (1) and the Fredholm integral equation of the second kind (2) are equivalent and invertible; b. if 1ϵσ(M2) and 1∈ / σ(M2) then in C the equations (1) and (2) are equivalent.

We note that the equation x=Ax+f with linear bounded operator A in C is considered (here and below) invertible in C if the operator I-A is invertible on C. Suppose that the condition of Theorem 1 is satisfied.

Then in C equations (1) and (2) are equivalent.

Let k (t , s, σ , σ 1 ) = m(t , s, σ ) m(σ , t , σ 1 ), where 1 mϵC(L1). Since the kernel k (t , s, σ , σ 1 ) ∈ C ( L ( D) where 1 mϵC(L ) is the space of continuous vector-valued functio 1 ns (t , s ) → z (t , s,⋅,⋅) ∈ L ( D ) and C(L1(D)) is realized as a tensor product of the spaces C and L1(D) with a cross-norm, which coincides with the norm in C(L1(D)) [06], and the set of continuous functions from the space C(DxD) is everywhere dense in C(L1(D)), then: (4)

1

It can be verified directly that m(t , s,⋅)m(⋅, t ,⋅) ∈ C ( L ( D )), where C(L1 (D)) denotes the space of continuous vec1 tor-valued functions (t , s ) → z (t , s,⋅,⋅)E∈ L ( D). . Then M2 is a completely continuous integral operator in C. Consequently, Fredholm’s theorems hold for equation (1) with kernel from C(L1).

Suppose that 1 is not an eigenvalue of the operator M2.

By the theorem on the spectrum map ± 1is not a point of the spectrum of the operator M : ±1∈/ σ ( M ). Therefore, (1) has a unique solution x(t, s). Therefore, x(t, s) is the unique solution of the integral equation:

where:

~

and kj and k j are continuous functions on D and the functions kj are linearly independent, and the functions ~ k j are orthonormal in the usual sense.

Substituting (4) into (2) we obtain, (5) where

(2) Conversely, let equation (4) have a unique solution x(t, s). Equation (2) can be written in the form: (3) where I is the identity operator on C. Since equation (3) with a completely continuous operator M2 has a unique solution, then 1 ∈ / σ ( M ) . Then the operator I+M is invertible in C .

442

(6) By virtue of [07] (7) Journal of Applied Engineering Science 16(2018)2

Anatolij Semenovich Kalitvin et al. - The approximate and numerical solution of romanovskij linear partial integral equations

Therefore, the operator I-K0 has a bounded inverse operator in C and

small perturbations [08], there is a ϵ>0, such that equation:

(7)

(11)

where

is invertible for Solutions of equations (1) and (11) can be written in the form:

and

By virtue of Then equation (5) can be written in the form (8)

we get

Taking (7) into account, we obtain an integral equation with a degenerate kernel: (9)

(12) where c is a constant. If now (13)

where

then from (12), (13) and the formulas for the norm of the Romanowskij operator [03], we have

Assuming

Thus, the operators and differ little in norm, if the kernels of equations (3) and (13) are sufficiently close.

Taking (12) into account, we obtain the following estimate: (14)

we obtain a system of linear algebraic equations (10) Thus, it is valid.

Theorem 2. Let m ϵ C(L1) and f ϵ C. Then the following assertions hold: a. if the principal determinant of system (10) is not equal to zero, then equation (1) has a unique solution continuous on D; b. if the principal determinant of system (10) is zero, then in C equation (1) either has no solutions, or has a finite number of linearly independent solutions.

AN APPROXIMATE SOLUTION OF EQUATION (1) An approximate solution in C of equation (1) with a kernel from C(L1) and a continuous function f(t,s) is a rather effective replacement of the kernel by a degenerate one.

We assume that equation (1) with kernel m ϵ C(L1) and function f ϵ C is invertible in C. By virtue of the stability of the invertibility of equations with respect to sufficiently Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

It shows that the number is sufficiently small if (13) holds. The application of the estimate (14) is related to the estimate of the constant c. In the general case, effective estimates of the constant c are unknown. However, any known upper bounds for the numbers and lead to an upper estimate for the constant c.

Let us cite one such estimate. Similarly to [03], the equalities

where If now

are some functions.

where the known functions

443

Anatolij Semenovich Kalitvin et al. - The approximate and numerical solution of romanovskij linear partial integral equations

then by virtue of the estimate of the norm of an operator of Romanovskij type with partial integrals in C [03], from (14) implies the estimate

We show that when the kernel of equation (1) is replaced by a degenerate kernel, the Romanovskij equation is obtained, whose solution is constructed explicitly.

Suppose that in (11)

solution can be found as the union of the solutions of systems obtained from (22) for each fixed j=1,...,n. Since equation (11) has a unique solution, each of these systems has a unique solution. Consequently, system (22) has a unique solution. Substituting this solution of system (22) into (21), we obtain The only solution of equation (11) is now obtained by substitution of the found in (18).

Another method of approximate solution of equation (1) is connected with the transition to an equivalent problem for the system of linear integral equations of Fredholm of the second kind with a parameter and the subsequent approximate solution of this problem.

Indeed, let

(15) where lj, mj, nj are continuous functions on [a, b].

Substituting (15) into (11), we obtain (16)

Then

We set and equation (1) can be written in the form of the system (17) (23)

Then (18)

Fredholm integral equations with parameter t in which the unknown function satisfies the additional condition

Substituting (18) into (17), we obtain the system (19) where

Assuming (20) in view of (19) we obtain (21) Substituting (21) into (20), we obtain the system

(24) If 1 ∈ / σ ( M ), then problem (23) / (24) is equivalent to equation (1), has a unique continuous on D solution (y(t, s), z(t, s)).

Thus, under the condition 1 ∈ / σ ( M ), the approximate solution of equation (1) reduces to an approximate solution of the system (23) and verification of equalities (24), understood as approximate equalities.

Another method of approximate solution of equation (1) for ± 1 ∈ / σ ( M ) is associated with the transition to the Fredholm integral equation of the second kind (2) and the replacement in (2) of the kernels by formula (4), in which the kernel k0 (t , s, σ , σ 1 ) is chosen equal to zero, and the sum is chosen so that:

(22) where

Thus, the integral equation of Romanowskij (11) with the degenerate kernel (15) reduces to system (22), whose

444

where ε > 0 is an arbitrarily small number. As a result, we obtain the Fredholm integral equation of the second kind with a continuous degenerate kernel: (25) Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

Anatolij Semenovich Kalitvin et al. - The approximate and numerical solution of romanovskij linear partial integral equations

where the function g(t, s) is determined by the formula (6). For sufficiently small ε > 0 , equation (25) is invertible and is solved in the standard way.

NUMERICAL SOLUTION OF EQUATION (1) We consider equation (1) with continuous given functions f(t, s) and m(t , s, σ ), where t , s, σ ∈ [a, b].

The following approximation scheme is justified by V.I. Romanovskij in [01] and can be used for the numerical solution of equation (1).

The segment [a, b] is divided into parts of length δ by points: We set and let ∆ denote the determinant of the system of linear equations (26)

The verification of equality (24) reduces to estimating the smallness of the number The approximate values of the solution of equation (1) are calculated by the formula: by sufficiently small δ We note that the direct application of quadrature formulas to equation (1) with continuous given functions f(t,s) and m(t,s,σ) causes difficulties due to the fact that the operator M in equation (1) is not completely continuous, and the well-known arguments of the mechanical quadrature method for Fredholm integral equations use the complete continuity of integral operators, which determine such equations.

2 However, if 1 ∈ / σ ( M ), then the method of mechanical quadratures is applied not to equation (1), but to the equivalent reversible equation (2) with a completely continuous integral operator M2. This uses the cubature formula:

If now n → ∞, then just as in Fredholm theory, system (26) approximates equation (1), and its solution tends to the solution of equation (1) [01].

Thus, an approximate numerical solution of equation (1) can be found as a solution of system (26). We note that this solution (26) is obtained under the condition ∆ =/ 0..For sufficiently large n this condition means that 1 ∈/ σ ( M )

where

Another method for the numerical solution of equation (1) is based on the numerical solution of problem (25)/ (26) with the use of quadrature formulas. For example, using the formula of left rectangles, the segment [a,b] splits into n equal parts by points

be realized.

(28)

It is assumed that the quadrature process (28) converges: for any function fϵC(D) the condition

Equation (2) can be written in the form (29)

where where and the system (25) is replaced by the system

(27)

and g(t,s) is a function (8). Setting t=tp, s=sq replacing the integral by the formula:

in (29) and

where where The system (27) is solved for each fixed i=0,1,...,n-1, its solution reduces to solving n systems of linear algebraic equations [09]. Since for each fixed tϵ[a,b] the system (23) is a system of linear integral equations with completely continuous integral operators, then for n → ∞ the solution ( yij( n ) , zij( n ) ) of the system (27) tends to ( yij , zij ) where yij = y (ti , s j ), zij = z (ti , s j ).

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117

and rpqPQ is the remainder, we get the system, after discarding the remainders in the equations of which we will have a system of equations (31) where xij = x (ti, sj).

445

Anatolij Semenovich Kalitvin et al. - The approximate and numerical solution of romanovskij linear partial integral equations

By [10], we have Theorem 3. Let the following conditions hold: 1. for each P and Q the coefficients ijPQ of formula (30) are positive and there exists a number G such that γ ijPQ ≤ G 2. the process (28) converges; 3. X0ϵC is a solution of equation (29).

γ

Then for sufficiently large P and Q the system (32) has the solution

and the rate of convergence is estimated by inequalities

c1RPQ ≤

max

1≤ p ≤ P ,1 ≤ q ≤Q

| x pq − x0 (t p , sq ) |≤ c2 RPQ ,

where c1 and c2 are positive constants,

The analytic approximation of x pq (t , s ) to the solution ~ x (t , s ) of equation (29) is naturally defined by:

We note that Theorem 3 was established in [11].

CONCLUSION To solve the partially Romanovskij integral equation (1), it is possible to use other methods of numerical solution of integral equations. However, when applying such methods directly to the equation (1), one should take into account the absence of complete continuity for the operator M. If the applied method is connected with the complete continuity of the integral operator, then by applying this method to the Romanovskij integral equation (1) this method next substantiate for equation (1) directly, or apply it to equation (2) with a completely continuous integral operator, or apply it to the numerical solution of problem (23) / (24) for the system of Fredholm integral equations second kind with parameter t and the completely continuous integral operators.

3. Kalitvin, A.S. (2007). Integral equations of Romanovskii type with partial integrals. Lipetsk State Pedagogical University, Lipetsk.

4. Appell, J., Eletskikh, I.F., Kalitvin, A.S. (2004). A note on the Fredholm property of partial equations of Romanovskij type. Journal of Integral Equations and Applications, vol. 16, no. 1, 1-8, DOI: 10.1216/ jiea/1181075256 5. Zabrejko, P.P., Koshelev, A.I., Krasnosel’skii, M.A., Mikhlin, S.G., Rakovshchik, L.S., Stecenko, V.J. (1968). Integral equations. Nauka, Moscow.

6. Levin, V.L. (1969). Tensor products and functors in categories of Banach spaces, defined by KB-lineals.

Transactions of the Moscow Mathematical Society, vol. 20, 43-82.

7. Kalitvin, A.S., Frolova E.V. (2004). Linear equations with partial integrals. C-theory. Lipetsk State Pedagogical University, Lipetsk.

8. Krejn, S.G. (1971). Linear equations in the Banach space. Nauka, Moscow.

9. Kalitvin, A.S., Kalitvin V.A. (2009). One method for the numerical solution of the integral equation of the Romanovskij double-connected Markov chains. Survey of Applied and Industrial Mathematics, vol. 16, no. 1, 115-116.

10. Vainikko, G.M. (1967). Galerkin’s outrageous method and the general theory of approximate methods for nonlinear equations. Journal of Computational Mathematics and Mathematical Physics, vol. 7, no.

4, 723-751.

11. Kalitvin, V.A. (2018). Some methods for the numerical solution of the Romanovskij partial integral equation. Proceedings of the International Conference "S.G. Krejn’s Voronezh Winter Mathematical School - 2018", p. 229-233.

ACKNOWLEDGMENTS The work is published with assistance of the Lipetsk State Pedagogical University.

REFERENCES 1. Romanovskij, V.I. (1932). Sur une classe d’equations integrales lineares. Acta Mathematica, vol. 59, 99-208.

2. Kantorovich, L.V., Akilov, G.P. (1984). Functional analysis. Nauka, Moscow.

446

Paper submitted: 27.07.2018.

Paper accepted: 14.09.2018.

This is an open access article distributed under the CC BY-NC-ND 4.0 terms and conditions.

Journal of Applied Engineering Science Vol. 16, No. 3, 2018 ISSN 1451-4117