Journal of Mechanical Engineering 2013 12 by Darko Svetak

http://www.sv-jme.eu

59 (2013) 12

Since 1955

Contents

Papers

707

Rok Simič, Mitjan Kalin: Comparison of Alcohol and Fatty Acid Adsorption on Hydrogenated DLC Coatings Studied by AFM and Tribological Tests

719

Wei Tan, Xiaoan Chen, Shaojiang Dong: A New Method For Machinery Fault Diagnoses Based On an Optimal Multiscale Morphological Filter

735

Manoj Modi, Gopal Agarwal: Powder-Mixed Electro-Discharge Diamond Surface Grinding Process: Modelling, Comparative Analysis and Multi-Output Optimisation Using Weighted Principal Components Analysis

748

Branislav Sredanovic, Gordana Globocki-Lakic, Djordje Cica, Davorin Kramar: Influence of Different Cooling and Lubrication Techniques on Material Machinability in Machining

755

Jafar Arashmehr, Gholam Hossein Rahimi, Seyed Fazel Rasouli: An Experimental and Numerical Investigation of a Grid Composite Cylindrical Shell Subjected to Transverse Loading

763

XiaoYang Jiao, GuoJun Liu, JianFang Liu, Xinbo Li, XiaoLun Liu, Song Lu: Research on Levitation Coupled with Standing Wave Levitation and Electromagnetic Levitation

Journal of Mechanical Engineering - Strojniški vestnik

725 Janez Benedičič, Roman Žavbi, Jožef Duhovnik: Systematic Development of a Device for Bituminous Layer Application

12 year 2013 volume 59 no.

Strojniški vestnik Journal of Mechanical Engineering

Strojniški vestnik – Journal of Mechanical Engineering (SV-JME) Aim and Scope The international journal publishes original and (mini)review articles covering the concepts of materials science, mechanics, kinematics, thermodynamics, energy and environment, mechatronics and robotics, fluid mechanics, tribology, cybernetics, industrial engineering and structural analysis. The journal follows new trends and progress proven practice in the mechanical engineering and also in the closely related sciences as are electrical, civil and process engineering, medicine, microbiology, ecology, agriculture, transport systems, aviation, and others, thus creating a unique forum for interdisciplinary or multidisciplinary dialogue. The international conferences selected papers are welcome for publishing as a special issue of SV-JME with invited co-editor(s). Editor in Chief Vincenc Butala University of Ljubljana Faculty of Mechanical Engineering, Slovenia Technical Editor Pika Škraba University of Ljubljana Faculty of Mechanical Engineering, Slovenia Editorial Office University of Ljubljana (UL) Faculty of Mechanical Engineering SV-JME, Aškerčeva 6, SI-1000 Ljubljana, Slovenia Phone: 386-(0)1-4771 137 Fax: 386-(0)1-2518 567 E-mail: info@sv-jme.eu, http://www.sv-jme.eu Print DZS, printed in 440 copies

Founders and Publishers University of Ljubljana (UL), Faculty of Mechanical Engineering, Slovenia University of Maribor (UM), Faculty of Mechanical Engineering, Slovenia Association of Mechanical Engineers of Slovenia Chamber of Commerce and Industry of Slovenia, Metal Processing Industry Association President of Publishing Council Branko Širok, UL, Faculty of Mech. Engineering, Slovenia Vice-President of Publishing Council Jože Balič, UM, Faculty of Mech. Engineering, Slovenia Cover: Scanning electron microscope image of a silicon cantilever with a sharp tip, which is used for atomic force microscopy (AFM) in contact mode to determine the surface topography and related features. Surface coverage with the adsorbed molecules, which affect the tribological properties of the surfaces, can also be determined from the AFM topography measurements. Image Courtesy: Laboratory for Tribology and Interface NanoTechnology, Faculty of Mechanical Engineering, University of Ljubljana, Slovenia

International Editorial Board Koshi Adachi, Graduate School of Engineering,Tohoku University, Japan Bikramjit Basu, Indian Institute of Technology, Kanpur, India Anton Bergant, Litostroj Power, Slovenia Franci Čuš, UM, Faculty of Mech. Engineering, Slovenia Narendra B. Dahotre, University of Tennessee, Knoxville, USA Matija Fajdiga, UL, Faculty of Mech. Engineering, Slovenia Imre Felde, Obuda University, Faculty of Informatics, Hungary Jože Flašker, UM, Faculty of Mech. Engineering, Slovenia Bernard Franković, Faculty of Engineering Rijeka, Croatia Janez Grum, UL, Faculty of Mech. Engineering, Slovenia Imre Horvath, Delft University of Technology, Netherlands Julius Kaplunov, Brunel University, West London, UK Milan Kljajin, J.J. Strossmayer University of Osijek, Croatia Janez Kopač, UL, Faculty of Mech. Engineering, Slovenia Franc Kosel, UL, Faculty of Mech. Engineering, Slovenia Thomas Lübben, University of Bremen, Germany Janez Možina, UL, Faculty of Mech. Engineering, Slovenia Miroslav Plančak, University of Novi Sad, Serbia Brian Prasad, California Institute of Technology, Pasadena, USA Bernd Sauer, University of Kaiserlautern, Germany Brane Širok, UL, Faculty of Mech. Engineering, Slovenia Leopold Škerget, UM, Faculty of Mech. Engineering, Slovenia George E. Totten, Portland State University, USA Nikos C. Tsourveloudis, Technical University of Crete, Greece Toma Udiljak, University of Zagreb, Croatia Arkady Voloshin, Lehigh University, Bethlehem, USA General information Strojniški vestnik – Journal of Mechanical Engineering is published in 11 issues per year (July and August is a double issue). Institutional prices include print & online access: institutional subscription price and foreign subscription €100,00 (the price of a single issue is €10,00); general public subscription and student subscription €50,00 (the price of a single issue is €5,00). Prices are exclusive of tax. Delivery is included in the price. The recipient is responsible for paying any import duties or taxes. Legal title passes to the customer on dispatch by our distributor. Single issues from current and recent volumes are available at the current single-issue price. To order the journal, please complete the form on our website. For submissions, subscriptions and all other information please visit: http://en.sv-jme.eu/. You can advertise on the inner and outer side of the back cover of the magazine. The authors of the published papers are invited to send photos or pictures with short explanation for cover content. We would like to thank the reviewers who have taken part in the peerreview process.

ISSN 0039-2480 © 2013 Strojniški vestnik - Journal of Mechanical Engineering. All rights reserved. SV-JME is indexed / abstracted in: SCI-Expanded, Compendex, Inspec, ProQuest-CSA, SCOPUS, TEMA. The list of the remaining bases, in which SV-JME is indexed, is available on the website.

The journal is subsidized by Slovenian Research Agency. Strojniški vestnik - Journal of Mechanical Engineering is also available on http://www.sv-jme.eu, where you access also to papers’ supplements, such as simulations, etc.

Instructions for Authors All manuscripts must be in English. Pages should be numbered sequentially. The maximum length of contributions is 10 pages. Longer contributions will only be accepted if authors provide justification in a cover letter. Short manuscripts should be less than 4 pages. For full instructions see the Authors Guideline section on the journal’s website: http://en.sv-jme.eu/. Please note that file size limit at the journal’s website is 8Mb. Announcement: The authors are kindly invited to submitt the paper through our web site: http://ojs.sv-jme.eu. Please note that file size limit at the journal’s website is 8Mb. The Author is also able to accompany the paper with Supplementary Files in the form of Cover Letter, data sets, research instruments, source texts, etc. The Author is able to track the submission through the editorial process - as well as participate in the copyediting and proofreading of submissions accepted for publication - by logging in, and using the username and password provided. Please provide a cover letter stating the following information about the submitted paper: 1. Paper title, list of authors and affiliations. 2. The type of your paper: original scientific paper (1.01), review scientific paper (1.02) or short scientific paper (1.03). 3. A declaration that your paper is unpublished work, not considered elsewhere for publication. 4. State the value of the paper or its practical, theoretical and scientific implications. What is new in the paper with respect to the state-of-the-art in the published papers? 5. We kindly ask you to suggest at least two reviewers for your paper and give us their names and contact information (email). Every manuscript submitted to the SV-JME undergoes the course of the peer-review process. 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. - 6 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 square brackets [1] and collected together in a reference list at the end of the manuscript. Units - standard SI symbols and abbreviations should be used. Symbols for physical quantities in the text should be written in italics (e.g. v, T, n, etc.). Symbols for units that consist of letters should be in plain text (e.g. ms-1, K, min, mm, etc.) 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. However, graphs and line drawings should be prepared as vector images, e.g. CDR, AI. When labeling axes, physical quantities, e.g. t, v, m, etc. should be used whenever possible to minimize the need to label the axes in two languages. Multi-curve graphs should have individual curves marked with a symbol. The meaning of the symbol should be explained in the figure caption. 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. In addition to the physical quantity, e.g. t (in italics), units (normal text), should be added in square brackets. 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 A reference list must be included using the following information as a guide. Only cited text references are included. Each reference is referred to in the text by a number enclosed in a square bracket (i.e., [3] or [2] to [6] for more references). No reference to the author is necessary. References must be numbered and ordered according to where they are first mentioned in the paper, not alphabetically. All references must be complete and accurate. All non-English or. non-German titles must be translated into English with the added note (in language) at the end of reference. Examples follow. Journal Papers: Surname 1, Initials, Surname 2, Initials (year). Title. Journal, volume, number, pages, DOI code. [1] Hackenschmidt, R., Alber-Laukant, B., Rieg, F. (2010). Simulating nonlinear materials under centrifugal forces by using intelligent crosslinked simulations. Strojniški vestnik - Journal of Mechanical Engineering, vol. 57, no. 7-8, p. 531-538, DOI:10.5545/sv-jme.2011.013. Journal titles should not be abbreviated. Note that journal title is set in italics. Please add DOI code when available and link it to the web site. Books: Surname 1, Initials, Surname 2, Initials (year). Title. Publisher, place of publication. [2] Groover, M.P. (2007). Fundamentals of Modern Manufacturing. John Wiley & Sons, Hoboken. Note that the title of the book is italicized. Chapters in Books: Surname 1, Initials, Surname 2, Initials (year). Chapter title. Editor(s) of book, book title. Publisher, place of publication, pages. [3] Carbone, G., Ceccarelli, M. (2005). Legged robotic systems. Kordić, V., Lazinica, A., Merdan, M. (Eds.), Cutting Edge Robotics. Pro literatur Verlag, Mammendorf, p. 553-576. Proceedings Papers: Surname 1, Initials, Surname 2, Initials (year). Paper title. Proceedings title, pages. [4] Štefanić, N., Martinčević-Mikić, S., Tošanović, N. (2009). Applied Lean System in Process Industry. MOTSP 2009 Conference Proceedings, p. 422-427. Standards: Standard-Code (year). Title. Organisation. Place. [5] ISO/DIS 16000-6.2:2002. Indoor Air – Part 6: Determination of Volatile Organic Compounds in Indoor and Chamber Air by Active Sampling on TENAX TA Sorbent, Thermal Desorption and Gas Chromatography using MSD/FID. International Organization for Standardization. Geneva. www pages: Surname, Initials or Company name. Title, from http://address, date of access. [6] Rockwell Automation. Arena, from http://www.arenasimulation.com, accessed on 2009-09-07. EXTENDED ABSTRACT By the time the paper is accepted for publishing, the authors are requested to send the extended abstract (approx. one A4 page or 3.500 to 4.000 characters). The instructions for writing the extended abstract are published on the web page http://www.sv-jme.eu/ information-for-authors/. COPYRIGHT Authors submitting a manuscript do so on the understanding that the work has not been published before, is not being considered for publication elsewhere and has been read and approved by all authors. The submission of the manuscript by the authors means that the authors automatically agree to transfer copyright to SV-JME and when the manuscript is accepted for publication. All accepted manuscripts must be accompanied by a Copyright Transfer Agreement, which should be sent to the editor. The work should be original by the authors and not be published elsewhere in any language without the written consent of the publisher. The proof will be sent to the author showing the final layout of the article. Proof correction must be minimal and fast. Thus it is essential that manuscripts are accurate when submitted. Authors can track the status of their accepted articles on http://en.svjme.eu/. PUBLICATION FEE For all articles authors will be asked to pay a publication fee prior to the article appearing in the journal. However, this fee only needs to be paid after the article has been accepted for publishing. The fee is 300.00 EUR (for articles with maximum of 10 pages), 20.00 EUR for each addition page. Additional costs for a color page is 90.00 EUR.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12 Contents

Contents Strojniški vestnik - Journal of Mechanical Engineering volume 59, (2013), number 12 Ljubljana, December 2013 ISSN 0039-2480 Published monthly

Papers Rok Simič, Mitjan Kalin: Comparison of Alcohol and Fatty Acid Adsorption on Hydrogenated DLC Coatings Studied by AFM and Tribological Tests Wei Tan, Xiaoan Chen, Shaojiang Dong: A New Method For Machinery Fault Diagnoses Based On an Optimal Multiscale Morphological Filter Janez Benedičič, Roman Žavbi, Jožef Duhovnik: Systematic Development of a Device for Bituminous Layer Application Manoj Modi, Gopal Agarwal: Powder-Mixed Electro-Discharge Diamond Surface Grinding Process: Modelling, Comparative Analysis and Multi-Output Optimisation Using Weighted Principal Components Analysis Branislav Sredanovic, Gordana Globocki-Lakic, Djordje Cica, Davorin Kramar: Influence of Different Cooling and Lubrication Techniques on Material Machinability in Machining Jafar Arashmehr, Gholam Hossein Rahimi, Seyed Fazel Rasouli: An Experimental and Numerical Investigation of a Grid Composite Cylindrical Shell Subjected to Transverse Loading XiaoYang Jiao, GuoJun Liu, JianFang Liu, Xinbo Li, XiaoLun Liu, Song Lu: Research on Levitation Coupled with Standing Wave Levitation and Electromagnetic Levitation

707 719 725 735 748 755 763

Received for review: 2013-05-20 Received revised form: 2013-07-22 Accepted for publication: 2013-09-27

Comparison of Alcohol and Fatty Acid Adsorption on Hydrogenated DLC Coatings Studied by AFM and Tribological Tests Simič, R. – Kalin, M. Rok Simič – Mitjan Kalin*

University of Ljubljana, Faculty of Mechanical Engineering, Slovenia Adsorption on hydrogenated diamond-like carbon (DLC) coatings with respect to boundary lubrication has been studied only very rarely, even though such adsorption has proved to be one of the main boundary-lubricating mechanisms for metal contacts. For this reason, we used atomic force microscopy (AFM) and tribotests to study the ability of hexadecanol and hexadecanoic acid to adsorb onto DLC and affect the tribological properties of DLC coatings, where steel was used as a reference. We have shown that alcohols and fatty acids can adsorb onto the DLC under static conditions. However, under dynamic tribocontact conditions, although alcohol and fatty acid molecules help to decrease the wear of the coatings, they are inefficient in reducing the friction. In all the experiments the fatty acid proved to have superior adsorption abilities compared to the corresponding alcohol. Based on our results and the existing literature, tentative adsorption mechanisms that include an environmental species effect, a temperature effect and a tribochemical effect are discussed for DLC surfaces. Keywords: DLC, AFM, fatty acid, alcohol, adsorption, tribology

0 INTRODUCTION Low-friction, low-wear and anti-adhesion properties are characteristics of diamond-like carbon (DLC) coatings, making them very useful for the protection of the contact surfaces of mechanical systems, see [1] to [3]. Since the contact surfaces of various mechanical systems are usually lubricated, an investigation of the effects of lubricants and additives on DLC coatings is crucial for the improvement and adaptation of the lubrication for DLC coatings. It is known that DLC coatings are, in general, chemically very stable and inert towards external species under static conditions [4]. On the other hand, under dynamic tribocontact conditions DLC surfaces can interact with opposing surfaces and with the gaseous molecules present in their surroundings [4]. Hydrogen, either atmospheric or originating from the coating itself, can passivate the surface’s dangling bonds and cause H-termination, which results in very low friction [5] to [7]. It was shown that in the case of water lubrication DLC can also tribochemically react with water molecules, where besides hydrogen passivation, OH groups are also formed on the DLC surface, causing OHtermination [8]. Although H-termination seems to be preferable to OH-termination due to the absence of van der Waals forces between the opposing surfaces, the OH-termination is assumed to play a significant role in reducing the friction and wear of the DLC contacts by preventing the strong adhesive bonds between the non-passivated surfaces in the contact [4]. The formation of an OH-terminated carbon surface was also reported by other authors [9], who

used GMO as an additive on the ta-C surfaces. Hydroxyl groups from the GMO were assumed to tribochemically react with the friction-activated ta-C surfaces causing ultra-low friction. The hydroxylation of the surfaces was also suggested in another study where glycerol was used as an additive [10]. Here, the authors proposed that hydroxylated surfaces also enhance the intercalation and adsorption of the surrounding alcohol molecules via hydrogen bonding. The tribochemical reactions of DLC surfaces were also confirmed with relatively complex boundarylubricating additives, like ZDDP and MoDTC [11] to [13]. It was suggested that additive species interact with the dangling bonds of the friction-activated carbon surfaces, forming low-shear MoS2 sheets on the surface. Nevertheless, the base oils also affect the performance of the DLC coatings [14] to [17]. Increasing the oil’s viscosity was found to decrease the wear of the DLC contacts in boundary-lubricating conditions, although the friction increased [15] and [16]. An improved wear performance due to the polarbased adsorption was also reported as the degree of unsaturation and the polarity of the base oil was increased [17]. Although the understanding of the lubrication of DLC coatings has improved drastically in the past decade, a systematic investigation of the adsorption of some polar groups like hydroxyl and carboxyl is still necessary. Since the chemisorption of polar molecules proved to be one of the efficient boundary-lubrication mechanisms for steel surfaces, it is very relevant to study the effect of such molecules on DLC coatings.

*Corr. Author’s Address: University of Ljubljana, Faculty of Mechanical Engineering, Aškerčeva 6, 1000 Ljubljana, Slovenia, mitjan.kalin@tint.fs.uni-lj.si

707

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

In this work we have focused on the adsorption ability of alcohols and fatty acids on DLC because these molecules have proved to be efficient in improving the tribological properties of steel contacts. To examine the adsorption ability on the nanoscale and in the absence of any mechanical and tribological effects the experiments were performed under controlled, static conditions using an atomic force microscope (AFM). The DLC and steel surfaces were exposed to various concentrations of hexadecanol and hexadecanoic acid in hexadecane, and the surface coverage of the adsorbed polar molecules was measured using the AFM. Tribological tests using boundary-lubricated conditions were also performed to correlate the wear and friction behaviors with the AFM results.

Germany). For the AFM analyses the hexadecanol and hexadecanoic acid molecules were introduced into the hexadecane (Merck KGaA, Darmstadt, Germany) in various concentrations from 2 to 20 mmol/l. For the tribological tests the polar molecules were mixed into a polyalphaolephin PAO6 base oil (Neste Oil, Espoo, Finland) with a kinematic viscosity of 30 mm2/s at 40°C, again with concentrations from 2 to 20 mmol/l. It was assumed that the adsorption from the PAO and hexadecane is relatively similar, despite the solvents having different structures [18]. Tribotests with pure PAO6 oil were also made to serve as a reference.

1 EXPERIMENTAL 1.1 Materials Steel (AISI 52100/DIN 100Cr6) discs of 24-mm diameter were used in this study. The hardness of the discs was 850 HV, as measured with a microhardness tester (Leitz Miniload, Wild Leitz GmbH, Germany). The discs were polished to an average surface roughness of Ra = 3 to 4 nm for the AFM analyses and Ra = 30 to 40 nm for the tribological tests, as measured using a CP-II AFM (Veeco, USA). A very low surface roughness in the case of the AFM analyses was necessary in order to observe the adsorbed species on the surface. Standard 100Cr6 steel bearing balls with a surface roughness Ra of 10 to 15 nm were used as a counter body in the ball-on-disc tribological tests. Half of the steel discs and balls were coated with 1.5 µm of hydrogenated amorphous diamondlike carbon (a-C:H, 30% H). Two interlayers – a Ti interlayer and a hydrogenated amorphous diamondlike carbon layer containing Si and O (a-C:H:Si:O) – were deposited between the coating and the steel substrate for better adhesion of the coating. All the layers were deposited using a hybrid PVD/PACVD process (Sulzer Sorevi SAS, Limoges, France). The DLC sample composition is presented in Fig. 1. The coating deposition only had a minor effect on the surface roughness. The surface roughness Ra of the steel and the DLC-coated samples is presented in Table 1. The polar molecules selected for the adsorption analyses in this work were straight-chain hydrocarbons with 16 carbon atoms and a hydroxyl (OH) or carboxyl (COOH) end-group, i.e., hexadecanol and hexadecanoic acid (Merck KGaA, Darmstadt, 708

Fig. 1. Schematic cross-section of the DLC-coated 100Cr6 steel samples

Table 1. Surface roughness Ra of the steel and DLC-coated samples; only the discs were used for the AFM adsorption analysis Steel disc DLC disc Steel ball DLC ball AFM 3 ± 0.4 nm 4 ± 0.2 nm / / Analysis Tribotests 30 ± 5 nm 40 ± 5 nm 10 ± 0.2 nm 15 ± 0.3 nm

b) Ai A

Fig. 2. Schematic representation of a) the AFM; and b) an image of the surface, where the adsorbed molecules appear in the shape of small bright spots; the A denotes the whole scanned area, while Ai is the projected surface area of a single spot

1.2 AFM Analysis Prior to the tests, the samples were cleaned by sonication in ethanol for 1 min and then left to dry in air. The analysis of the samples prior to their exposure to the solutions revealed no polishing or cleaning residues on the surfaces. Moreover, analyses of the DLC were performed after the coating deposition, where no polishing residues could be present due to industrially defined cleaning and subsequent coating deposition. Moreover, after exposing the samples

Simič, R. – Kalin, M.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

to pure hexadecane (without alcohol or fatty acid molecules) and subsequent cleaning the AFM data revealed no surface debris or aggregates. For the final analysis equally sized drops of mixtures were spread over the surfaces. Then, one set of samples was left at 25 °C for 10 minutes and the other at 80 °C for 20 minutes to allow the adsorption of the hexadecanoic acid molecules onto the surfaces. Afterwards, the samples were cleaned by sonication in ethanol for one minute to remove the unbound molecules, and then left to dry in air. Samples were then analyzed using the AFM in contact mode at a constant load, which proved to be a suitable technique for the imaging of adsorbed layers in our previous research. About 10 to 20 scans of 50 µm × 50 µm were collected from various locations on a single sample to ensure statistically representative results. The analyses were made using the ProScan 1.8 Data Acquisition image-analyzing software associated with the AFM. The topographic images were used to determine the average surface coverage with the remaining adsorbed molecules, which appeared in groups in the shape of small islands that were scattered across the surface, Fig. 2. The surface coverage (SC) was calculated as follows:

∑ SC =

N i =1

, (1)

where N is the number of adsorption spots, A is the whole scanned area, and Ai is the projected surface area of one adsorption spot, Fig. 2. 1.3 Tribological Tests The tribological tests were performed using a CETR UMT tribometer (CETR, CA, USA) with a ballon-flat reciprocating test geometry. Self-mated contacts were always used. In this way the chemical reactions between the DLC and the additives could be observed, without the interference from the steel counter body, which affects the reaction products and befogs the actual reactivity of the DLC. The tests were performed at 25 °C and at 80 °C. The normal load was set to 10 N, corresponding to 1 GPa of maximum initial Hertzian contact pressure. The tests with a 6.8-mm stroke consisted of 7350 cycles to reach a total sliding distance of 100 m. The average relative contact velocity was set to 0.01 m/s, so the boundary-lubricating conditions were ensured. The characteristic coefficient of friction (COF) for a specific test was determined as the average coefficient of friction over the last 3000 cycles. Each experiment was repeated several times to ensure statistically

relevant results. An optical microscope equipped with a CCD camera (DS-fi1, Nikon Corp., Tokyo, Japan) was used to verify that the coating did not wear through and to measure the diameters of the wear scars on the balls. The wear volume was subsequently calculated. The wear volume on the discs, however, was usually below the limits of the technique used and was not measured. 2 RESULTS 2.1 AFM Examples of the topographic images of the surfaces after their exposure to the solutions of alcohol or fatty acid in the PAO and cleaning are shown in Fig. 3. The surfaces were partially covered with islands of adsorbed molecules of the alcohol or fatty acid that were retained on the surfaces after the ultrasonic cleaning. Dispersive intermolecular forces caused the molecules to merge into groups and appear in the images as nanoscopic bright spots. The total areal size of these spots was found to depend on the concentration, temperature, type of additive and type of surface. This is discussed in detail later on. On the other hand, for the height of the spots, which was typically from 10 to 25 nm, no such obvious correlations to the parameters studied in this work were found. Therefore, the height of the spots was assumed to depend primarily on the cohesive intermolecular forces between the molecules in the mixtures rather than on their interactions with the substrates. The measured surface coverage with adsorbed molecules as a function of their concentration at 25 °C is presented in the graph in the Fig. 4a. The surface coverage was increasing monotonously with the increasing concentrations for both additives and on both surfaces. It is clear that the surface coverage of the steel with alcohol molecules (Steel+OH) was the lowest across the whole range of tested concentrations. The measured surface coverage was 0.24% at 2 mmol/l and was increasing monotonously up to 0.64% at 20 mmol/l. Considering the assumed nonreactivity of the DLC, it was surprising to measure a higher surface coverage of alcohol molecules on the DLC surface (DLC+OH) compared to the steel. The surface coverage on the DLC was 0.47% at 2 mmol/l and was increasing up to 0.92% at 20 mmol/l for the alcohol molecules. As expected, due to the higher polarity of the fatty acid molecule compared to the corresponding alcohol molecule, we measured a higher surface coverage when the fatty acid was used

Comparison of Alcohol and Fatty Acid Adsorption on Hydrogenated DLC Coatings Studied by AFM and Tribological Tests

709

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

Concentration 7 mmol/l

20 mmol/l

2 mmol/l

10 μm

COOH

Steel

COOH

DLC

10 μm

Fig. 3. AFM topographic images of the DLC and steel surfaces after their exposure to various solutions of alcohol (OH) or fatty acid (COOH) and cleaning

instead of the alcohol. This was true for the DLC and for the steel surface. Again, the surface coverage was increasing monotonously with increasing concentration for both surfaces. In contrast to the adsorption of the alcohol, the surface coverage with 710

fatty acid molecules was mostly higher on the steel (Steel+COOH) than on the DLC (DLC+COOH). And although the surface coverage at lower concentrations of the fatty acid was relatively similar for the steel and DLC, the surface coverage for the steel was increasing

Simič, R. – Kalin, M.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

Surface coverage [%]

3.0

DLC+COOH DLC+OH Steel+COOH Steel+OH

2.5

Surface coverage [%]

2.0 1.5 1.0

3.0

DLC+COOH DLC+OH Steel+COOH Steel+OH

2.5 2.0 1.5 1.0 0.5

0.5

0.0

0.0 0

10 15 Concentration [mmol/l]

Fig. 4. Surface coverage with adsorbed alcohol (OH) or fatty acid (COOH) molecules for DLC and steel surfaces at a) 25 °C and b) 80 °C as measured with the AFM

2.2 Tribotests The coefficients of friction (COFs) and ball-wear volumes that were measured for each additive concentration are presented in Figs. 5 and 6. Fig. 5a shows the coefficient of friction for a steel contact

0.16 0.14

Pure PAO

0.12 0.10

COF

0.08 0.06 0.04 Steel+OH Steel+COOH

0.02 0.00 0

5 10 15 Concentration [mmol/l]

Ball wear [10-5 mm3]

1.25% at 2 mmol/l, then it increased a lot to 1.93% at 5 mmol/l, and above 5 mmol/l it was decreasing slowly to a value of 1.49% at 20 mmol/l. As for the steel, the surface coverage with fatty acid at 80 °C was 1.59% at 2 mmol/l, then it increased greatly to 2.35% at 5 mmol/l, and above 5 mmol/l it was decreasing gradually to a value of 1.99% at 20 mmol/l. Therefore, the adsorption of fatty acid on steel (Steel+COOH) showed similar behavior as on the DLC; however, higher values of the surface coverage were measured on the steel than on the DLC. This is a similar relation to the one that was observed at 25 °C, Fig. 4a.

Steel+OH Steel+COOH

50 40 30 20

Pure PAO

with a significantly higher rate than for the DLC. At 20 mmol/l the surface coverage with fatty acid molecules was 1.37% on the DLC and 2.64% on the steel surface. When the temperature was set to 80 °C the surface coverage increased a great deal at low concentrations for both additives and on both surfaces, Fig. 4b, compared to the corresponding coverage at 25 °C, Fig. 4a. At 80 °C, the surface coverage with alcohol molecules on the steel surface (Steel+OH) was 0.44% at 2 mmol/l and 0.69% at 20 mmol/l. The alcohol molecules again adsorbed better on the DLC than on the steel. The surface coverage with alcohol molecules on the DLC (DLC+OH) was 0.96% at 2 mmol/l and a surprisingly high 2.28% at 20 mmol. The fatty acid molecules tended to absorb better on the DLC (DLC+COOH) than the alcohol molecules, similar to the situation at 25 °C. At 80 °C the surface coverage with fatty acid molecules on the DLC was

10 0

5 7 15 Concentration [mmol/l]

Fig. 5. a) Coefficient of friction (COF) and b) ball wear for the steel contact lubricated by PAO oil with added different concentrations of hexadecanol (OH) or hexadecanoic acid (COOH) Comparison of Alcohol and Fatty Acid Adsorption on Hydrogenated DLC Coatings Studied by AFM and Tribological Tests

711

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

DLC+OH DLC+COOH

0.14 0.12 0.10

COF

Pure PAO

0.08 0.06 0.04 0.02 0.00 0

5 10 15 Concentration [mmol/l]

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

DLC+OH DLC+COOH

Pure PAO

0.16

Ball wear [10-5 mm3]

5 7 15 Concentration [mmol/l]

Fig. 6. a) Coefficient of friction (COF) and b) ball wear for the DLC contact lubricated by PAO oil with added different concentrations of hexadecanol (OH) or hexadecanoic acid (COOH)

lubricated with various concentrations of alcohol or fatty acid in the PAO base oil. The coefficient of friction with just the base oil being used (Fig. 5a) was approximately 0.15. When alcohol was added, the coefficient of friction did not change for additive concentrations up to 5 mmol/l. At 7 mmol/l the coefficient of friction decreased to a value of 0.13, and was gradually decreasing with an increasing alcohol concentration. At 20 mmol/l the coefficient of friction reached its lowest value of 0.11, which is about 30% lower than with the base oil alone. On the other hand, even the smallest addition of hexadecanoic acid at a concentration of just 2 mmol/l decreased the COF by 30%, to a value of about 0.1. Increasing the concentration from 2 to 20 mmol/l had only a little effect on the COF; however, a monotonous decrease of the COF with higher concentrations was observed. At 20 mmol/l the COF for steel reached its lowest value of 0.094. A similar trend as with the coefficient of friction was observed in the wear behavior for the steel contact, Fig. 5b. The ball-wear volume for the base oil alone was 4.6 × 10-4 mm3, Fig. 5b. When the alcohol was added at 2 mmol/l, the average wear volume did not decrease; it even increased to 5.1 × 10-4 mm3. When higher concentrations of alcohol were used, however, the wear began to decrease, reaching the lowest value of around 1.0 × 10-4 mm3 at the highest two concentrations. This is about 80 % lower than with the base oil alone. The results show that hexadecanol in higher concentrations decreased the coefficient of friction and wear (Fig. 5), while it increased the surface coverage of the steel surfaces (see Fig. 4). On the other hand, the addition of hexadecanoic acid at a concentration of just 2 mmol/l decreased the wear by 93%, to a value of 3.34 × 105 mm3. Increasing the concentration from 2 to 20 mmol/l had relatively little 712

effect on the wear; however, a continuous decrease in the wear for higher fatty acid concentrations was observed. At 20 mmol/l the wear of the steel ball was only 1.70 × 10-5 mm3. The results show that the presence of fatty acid, even in a concentration as low as 2 mmol/l, can drastically reduce the friction and wear. The fatty acid also resulted in a lower coefficient of friction and wear of the steel contacts than with the alcohol. This result correlates very well with the data obtained with the AFM (Fig. 4), where the fatty acid led to much better adsorption properties for the steel than the alcohol. Fig. 6 shows the coefficient of friction (COF) and the ball wear for the DLC/DLC contact at an environmental temperature of 25 °C. Notice the different scales of the wear for the DLC (Fig. 6b) and the steel (Fig. 5b). The coefficient of friction for the DLC contact (Fig. 6a) using base oil alone was around 0.059, which is approximately 60% lower compared to the steel (Fig. 5a). The addition of hexadecanol or hexadecanoic acid did not affect the coefficient of friction of the DLC significantly. The measured values of the coefficient of friction varied between 0.056 and 0.060 for all the concentrations; the additive being either alcohol or fatty acid. Although there were no significant changes in the coefficient of friction for the DLC, a decrease in the wear is clearly visible when either the alcohol or fatty acid was added to the base oil, Fig. 6. The ball-wear volume was approximately 1.5 × 10-5 mm3 when only the base oil was used, Fig. 6b. When the alcohol was added in the smallest concentration of 2 mmol/l the wear did not change much; however, at 5 mmol/l of alcohol the wear volume decreased to a value of 1.30 × 105 mm3. A further increase in the concentrations monotonously decreased the wear volume to a value of 1.27 × 10-5 mm3 at the highest

Simič, R. – Kalin, M.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

concentration. This is about 15% lower wear than with the base oil alone. On the other hand, the addition of hexadecanoic acid at concentrations of 2 mmol/l and 5 mmol/l decreased the wear by 13 to 17 %, to values of 1.30 × 105 mm3 and 1.23 × 105 mm3, respectively. Increasing the concentration above 5 mmol/l did not influence the wear to any great extent, as it remained relatively constant, ranging from 1.23 × 105 mm3 to 1.28 × 105 mm3. Like with the steel contact, the fatty acid resulted in lower wear than the alcohol, especially at the lowest concentrations tested. This result correlates well with the AFM results (Fig. 4), where the fatty acid proved to have a better adsorption ability with respect to the DLC compared to the alcohol. Although the drop in wear is less obvious for the DLC than for the steel, it is important to consider the up to 30-times lower wear in the case of the DLC (Fig. 6b) compared to the steel (Fig. 5b), which makes any further reductions in the wear much more evident. 3 DISCUSSION 3.1 AFM An AFM analysis of the adsorption of hexadecanol and hexadecanoic acid on the DLC surfaces was performed, using steel surfaces as a reference. After the solutions of various concentrations had been left on the surfaces to react, the samples were cleaned, which was an important part of the work. Cleaning with an ethanol sonication tended to remove the non-bonded and weakly bonded molecules. Only the strongly bonded molecules remained on the surface, which is discussed in detail later on. In spite of this relatively severe cleaning, a clear correlation between the surface coverage and the molecule concentration could be observed. The obtained results therefore imply that the AFM can be used to analyze the adsorption ability of polar molecules on various surfaces. The monotonous increase of the surface coverage with an increasing concentration of the polar molecules at room temperature (Fig. 4a) could be explained by the theory that a larger number of active molecules present in the carrier fluid means that more binding sites on the surface can become occupied, which consequently manifests itself in an increased surface coverage. However, this holds only for low concentrations or a large number of available binding sites. Higher concentrations and/or a limited availability of free binding sites would result in the saturation of adsorption. Therefore, the surface coverage would level off at higher concentrations.

The heterogeneity of the affinity of the adsorption sites and the low adsorption ability of the molecules might well be the reason for the leveling off of the surface-coverage curves. In the case of hexadecanol and steel (Fig. 4a), the low surface coverage and the leveling off above the concentration of 15 mmol/l indicate the weak adsorption ability of the long-chain alcohol molecules with respect to the steel in static conditions. However, the same alcohol molecules seem to have a slightly better adsorption ability when in combination with the DLC surfaces. Since DLC coatings are known to be rather nonreactive towards surrounding species under static conditions [4] this result was surprising. On the other hand, the fatty acids resulted in a higher surface coverage of the steel than the DLC, which proves the well-known good adsorption ability of the fatty acid with respect to steel surfaces. Moreover, the fact that the fatty acid resulted in a higher surface coverage on both surfaces compared to the alcohol proves the overall, much better adsorption ability of the fatty acids compared to the alcohol. This very much agrees with several complementary studies, see [19] and [20]. Although the 100Cr6 steel and the a-C:H coating possess relatively similar total surface energies, the polar component of the surface energy of the steel is much larger, i.e., almost double, compared to the polar component of the a-C:H, which indicates the better overall adsorption ability of steel with respect to the polar molecules compared to the hydrogenated DLC [21]. This agrees well with the results of this study, where the surface coverage of the steel with fatty acid molecules was much greater compared to the DLC. In the case of hexadecanol, however, the differences in adsorption between the steel and the DLC are not so obvious, which may be due to the lower polarity of the alcohol compared to the fatty acid. Namely, the polarity of the alcohols may be too low to reliably differentiate the behavior based on the surface-energy theorem. At 80 °C the relations between the surface coverage for both the additives and the surfaces remained the same, which proves that the above statements also hold true at elevated temperatures. However, by elevating the temperature the surface coverage increased drastically for the lowest concentrations in all cases, Fig. 4. This could be due to the better mobility of the molecules at higher temperatures, and as a result, more molecules hit the surface and adsorb. The affinity of the binding sites may change as well. The leveling off in the case of the hexadecanoic acid, where the surface coverage reached its highest values, may be caused by a limited

Comparison of Alcohol and Fatty Acid Adsorption on Hydrogenated DLC Coatings Studied by AFM and Tribological Tests

713

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

and relatively small number of available adsorption sites on the surface. The slowly decreasing trend of the surface coverage on both surfaces in the case the hexadecanoic acid, above the concentration of 5 mmol/l, might be a consequence of the micelle structures that can be formed from polar acid molecules at higher concentrations. 3.2 Tribotests Tribological tests were performed to compare the friction and wear behaviors with the AFM results, and to investigate the molecules’ ability to retain on the surface and to act as a potential boundary-lubricating agent. Although a minimum concentration (2 to 5 mmol/l) of alcohol molecules was necessary to achieve significant reductions in the friction and wear, their further decreasing with an increasing concentration is in agreement with the well-known fact that alcohols help to decrease the friction and wear of steel contacts [22] to [24]. Namely, higher concentrations enable a more complete adsorbed film, which protects the surface asperities from direct mechanical contact, passivates the worn and freshly exposed surfaces and therefore prevents any strong adhesion. Decreasing the COF and wear with the increasing hexadecanol concentration, therefore, combines well with the AFM results, Figs. 4. In contrast to the hexadecanol, the hexadecanoic acid in a concentration as low as 2 mmol/l was sufficient to form an efficient protective film on the surface. Nevertheless, it was already proven that fatty acids readily form adsorbed layers and help to improve the tribological properties of metals, see [18] and [25] to [27]. However, the dependence of the tribological performance on the acid (or alcohol) concentration at 25 °C does not directly resemble the surface coverage measured with the AFM at 25 °C (Fig. 4a). The reason for this is the increased contact temperature and wear. The temperature in the tribocontacts may rise well above the surrounding temperature due to the frictional heat [28]. The increased contact temperature, if not too high, usually enhances the adsorption. This is why the tribological behavior of steel (Fig. 5) correlates much better with the AFM results obtained at increased temperature (Fig. 4b) than at room temperature, Fig. 4a. In addition to the frictional heat, the wear causes the removal of the passivating surface layers, which creates new dangling bonds on the surface. This is an important phenomenon, in addition to temperature, that distinguishes the tribocontact from the static conditions. 714

The fact that the coefficient of friction of the DLC/DLC contact was not affected by the presence of the hexadecanol or hexadecanoic acid molecules (Fig. 6a) does not necessarily imply the absence of adsorption. It has been shown [5] that hydrogenated DLC surfaces, like in our case, possess inherently lowfriction properties due to the passivation of the surface dangling bonds by the hydrogen from the coating, which prevents the formation of strong adhesion bonds between the opposing surfaces. Passivating the surfaces with the polar molecules may therefore have no noticeable influence on the friction behavior of the coating, since the shear between the adsorbed hydrocarbon topcoats does not differ from the shear between the bare hydrogenated DLCs. This seems to be true at 25 °C as well as at 80 °C. In contrast to the friction, the wear of the DLC decreased in the presence of the hexadecanol and hexadecanoic acid molecules (Fig. 6b). Similar results, where the wear decreased in the presence of alcohol, i.e., pentanol, were also reported in another study with steel/DLC contacts [29]. According to the obtained results, we presume that a film of adsorbed polar molecules acts as a thin, highly viscous layer, which prevents direct mechanical contacts between the surface asperities and thus reduces the wear. The fact that the acid always resulted in lower wear of the DLC compared to the alcohol, irrespective of the concentration used, again proves the superior adsorption capabilities of the acid compared to the alcohol. Moreover, trends in the wear reduction on DLC in the presence of the alcohol and acid seem to be similar as on the steel. Namely, the acid resulted in the greatest reduction of wear at the lowest concentration and caused only slight changes at higher concentrations. The alcohol, on the other hand, caused no noticeable changes at 2 mmol/l, but it resulted in a monotonous decrease in the wear above 2 mmol/l. These trends appear similar to those of the steel, despite the overall lower effect that both molecules had on the DLC compared to the steel. However, a much lower degree of wear of the DLC compared to the steel, which makes any further improvements in the case of the DLC very difficult, has to be considered here. Nevertheless, the wear behavior of the DLC in the presence of the polar molecules also correlates well with the AFM data. This is especially true for the surface coverage obtained at 80 °C (Fig. 4b), which is due to the same reasons described earlier in the text for the steel. Besides wear, rubbing may also trigger a tribo-emission, which can occur at the surfaces of DLC, see [30] and [31]. Tribo-emitted electrons, ions

Simič, R. – Kalin, M.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

and photons can ionize molecules, allowing them to chemisorb or even decompose, see [32] and [33]. The emission of charged particles, on the other hand, leaves charged areas on the surface, which then serve as new bonding sites. The wear and tribo-emission can thus affect the adsorption of polar molecules and befog the effect of the environmental temperature on adsorption. Accordingly, the tribocontact appears to enhance the chemisorption, which appears to be a preferable mechanism in the tribological tests. On the other hand, in the static AFM conditions, where the formation of dangling bonds is less probable and limited to higher temperatures, hydrogen bonding seems to be a more plausible mechanism for the molecule adsorption.

3.3 On the Adsorption Mechanisms 3.3.1 Static Conditions Similar to steel, amorphous carbon coatings that are exposed to ambient air and moisture also possess a certain amount of oxides and hydroxides on the surface, see [34] and [35]. The amorphous structure and oxidation cause these surfaces to be heterogeneous in the sense that they contain exposed adsorption sites with a range of affinities for interactions with polar end-groups [36]. The surfaces of the amorphous carbon substrates therefore contain sp2 and sp3 bonded carbon atoms, whereas the surface bonds are passivated by hydrogen, oxides and hydroxides. Fig. 7 shows schematic diagrams illustrating several tentative adsorption mechanisms for alcohols and fatty acids in different environmental and contact conditions. In the

Fig. 7. Tentative adsorption mechanisms for alcohol and fatty acid molecules onto DLC; alcohol a) and fatty acid b) molecules physisorb by the hydrogen bonding to the surface oxides and hydroxides; in the static conditions the alcohol c) and fatty acid d) molecules may chemisorb to an unsaturated sp2 system; during the dynamic tribocontact conditions chemisorption of the alcohols e) and fatty acids f) to the surface carbon atoms is greatly enhanced Comparison of Alcohol and Fatty Acid Adsorption on Hydrogenated DLC Coatings Studied by AFM and Tribological Tests

715

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

static conditions, the alcohol or fatty acid molecules close to the surface physically adsorb by hydrogen bonding to surface oxides and hydroxides, Figs. 7a and 7b. Since the physical bonding is relatively weak (~ 20 kJ/mol) it be can easily disrupted by the ambient thermal energy or removed by solvent rinsing [35]. As proposed by Kasai [35] and [37], alcohol molecules can also chemisorb to the carbon surface, where chemical bonding occurs when the hydrogen atom from the hydroxyl endgroup migrates to the dangling bond inside the amorphous carbon structure, while the remaining alkoxy group interacts with the unsaturated system at the surface, Figs 7c. We assume that the fatty acids can chemisorb in a similar way, Fig 7d. According to Kasai [35] and [37], such an adsorption process is slow at room temperature, but it can be accelerated by elevating the temperature. This agrees well with our results, where the adsorption of both alcohol and fatty acid molecules, the alcohol and acid, on DLC was greatly promoted at all concentrations when the temperature was increased from 25 to 80 °C. Even though ultrasonic cleaning may cause the removal of a large fraction of physisorbed molecules, some of them may still persist on the surface and contribute to the total surface coverage. Their exact share of the total surface coverage, however, cannot be determined from the technique used in this work. 3.3.2 Dynamic Conditions The rubbing during a tribological contact of DLC surfaces causes the removal of the surface oxides, hydroxides and even carbon atoms, which results in the wear of the material. The consequential absence of the surface passivating layers creates new dangling bonds on the freshly exposed carbon atoms. These dangling bonds tend to passivate themselves by reacting immediately with the surrounding species, like free hydrogen, oxygen and, to some extent, also polar alcohol or fatty acid molecules. The alcohol and fatty acid molecules chemisorb to the exposed dangling bonds by the attachment of the oxygen from the hydroxyl/carboxyl group onto the exposed surface carbon atoms. The hydrogen that gets cut off from the hydroxyl/carboxyl group during the chemisorption can passivate the remaining dangling bonds on the surface or gets removed from the contact as a H2 molecule. The alkyl chain of the alcohol therefore attaches to the surface by creating an ether bond (C-O-C), while the fatty acid presumably forms ester bonds. The ions, electrons and photons may be generated in the tribological contacts of the hydrogenated DLCs 716

[30] to [32], causing the formation of a triboplasma and tribocharging. A triboplasma at the DLC surface can cause tribochemical reactions, similar to those proposed for steel by Kajdas in the NIRAM approach [38]. All these phenomena contribute greatly to the chemisorption of the alcohol and fatty acid molecules on the DLC surfaces in the contact, Figs. 7e and 7f. Since the physisorption is easily disrupted due to the relatively weak van der Waals forces, it can only have a minor effect on the friction and wear. The physisorbed molecules are presumably unable to sustain these boundary-lubrication conditions and such tribological improvements would not be observed. We therefore assume that it is the chemisorption that causes the significant improvements in the triboperformance, although both physisorption and chemisorption may be present in the contact. The adsorption mechanisms for the hydrogenated DLC are in a way similar to the mechanisms of adsorption onto steel. However, the bonding energies between the polar molecules and the steel (iron-alkoxide/carboxylate bond) are presumably very different from the energies between the polar molecules and the DLC (ether/ester-type bond). Therefore, the ability of the additive molecules to adsorb on the DLC surface and thus affect the friction and wear may differ from the ability of the alcohol molecules adsorbed onto steel. 4 CONCLUSIONS In this work we showed that AFM can be used successfully to detect islands of adsorbed molecules on steel and DLC. We showed that polar molecules can adsorb on the DLC also under static conditions, although the DLC is assumed to be rather nonreactive. The analysis of the surface coverage revealed that fatty acid possesses an overall superior adsorption ability compared to the related alcohol molecules. Comparing the results for the steel and the DLC revealed that fatty acid adsorbs better on the steel than on the DLC, while the opposite holds for the alcohol molecules. The presence of the fatty acid or alcohol molecules in the lubricant can also reduce the wear of the DLC coatings, which proves the adsorption ability of these molecules. The adsorbed alcohol and fatty acid molecules protected the surface asperities from direct mechanical contacts and thus helped to reduce the wear. At low concentrations (2 to 5 mmol/l) fatty acid resulted in lower wear compared to the alcohol, which proved the better adsorption ability of the acid. Above 7 mmol/l both molecules

Simič, R. – Kalin, M.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

resulted in similar reductions in the wear, which – in contrast to the differences in the static AFM tests – is a consequence of the better film formation due to the enhanced adsorption on the tribo-activated surfaces. The fact that the adsorption of molecules had no significant influence on the coefficient of friction can be explained by the inherent low-friction properties of the hydrogenated DLC itself. To sum up, the results of this work showed that long-chain alcohols and fatty acids adsorb onto DLC and can therefore serve as potential green boundary-lubrication agents for DLC. 5 ACKNOWLEDGMENTS The authors greatly appreciate the generous support of the Taiho Kogyo Tribology Research Foundation, Japan, as well as of F. Meunier from Sulzer Sorevi for providing the coatings. The authors also wish to thank J. Jelenc (Nanotul, Slovenia), whose help with the AFM analyses contributed greatly to this research. 6 REFERENCES [1] Robertson, J. (2002). Diamond-like amorphous carbon. Materials Science and Engineering: R: Reports, vol. 37, no. 4-6, p. 129-281, DOI:10.1016/S0927796X(02)00005-0. [2] Erdemir, A. (2004). Diamond-like carbon films. Tribology of Mechanical Systems, ASME Press, New York, p. 139-156, DOI:10.1115/1.802094.ch8. [3] Neville, A., Morina, A., Haque, T., Voong, M. (2007). Compatibility between tribological surfaces and lubricant additives—How friction and wear reduction can be controlled by surface/lube synergies. Tribology International, vol. 40, no. 10-12, p. 1680-1695, DOI:10.1016/j.triboint.2007.01.019. [4] Erdemir, A., Donnet, C. (2006). Tribology of diamondlike carbon films: recent progress and future prospects. Journal of Physics D: Applied Physics, vol. 39, R311-R327, DOI:10.1088/0022-3727/39/18/R01. [5] Fontaine, J., Donnet, C., Grill, A., LeMogne, T. (2001). Tribochemistry between hydrogen and diamondlike carbon films. Surface and Coatings Technology, vol. 146-147, p. 286-291, DOI:10.1016/S02578972(01)01398-6. [6] Erdemir, A. (2001). The role of hydrogen in tribological properties of diamond-like carbon films. Surface and Coatings Technology, vol. 146-147, p. 292-297, DOI:10.1016/S0257-8972(01)01417-7. [7] Li, H., Xu, T., Wang, C., Chen, J., Zhou, H., Liu, H. (2007). Tribochemical effects on the friction and wear behaviors of a-C:H and a-C films in different environment. Tribology International, vol. 40, no. 1, p. 132-138, DOI:10.1016/j.triboint.2006.03.007. [8] Wu, X., Ohana, T., Tanaka, A., Kubo, T., Nanao, H., Minami, I., Mori, S. (2008). Tribochemical

investigation of DLC coating in water using stable isotopic tracers. Applied Surface Science, vol. 254, no. 11, p. 3397-3402, DOI:10.1016/j.apsusc.2007.11.024. [9] Kano, M., Yasuda, Y., Okamoto, Y., Mabuchi, Y., Hamada, T., Ueno, T., Ye, J., Konishi, S., Takeshima, S., Martin, J.M., de Barros Bouchet, M.I., Le Mogne, T. (2005). Ultralow friction of DLC in presence of glycerol mono-oleate (GMO). Tribology Letters, vol. 18, no. 2, p. 245-251, DOI:10.1007/s11249-004-27494. [10] Matta, C., Joly-Pottuz, L., de Barros Bouchet, M.I., Martin, J.M., Kano, M., Zhang, Q., Goddard III, W.A. (2008). Superlubricity and tribochemistry of polyhydric alcohols. Physical Review B, vol 78, no. 8, p. 085436-1 - 085436-8, DOI:10.1103/PhysRevB.78.085436. [11] de Barros Bouchet, M.I., Martin, J.M., Le-Mogne, T., Vacher, B. (2005). Boundary lubrication mechanisms of carbon coatings by MoDTC and ZDDP additives. Tribology International, vol. 38, no. 3, p. 257-264, DOI:10.1016/j.triboint.2004.08.009. [12] Equey, S., Roos, S., Mueller, S., Hauert, R., Spencer, N.D., Crockett, R. (2008). Tribofilm formation from ZnDTP on diamond-like carbon. Wear, vol. 264, no. 3-4, p. 316-321, DOI:10.1016/j.wear.2007.03.012. [13] Topolovec-Miklozic, K., Lockwood, F., Spikes, H. (2008). Behaviour of boundary lubricating additives on DLC coatings. Wear, vol. 256, no. 11-12, p. 1893-1901, DOI:10.1016/j.wear.2008.04.051. [14] Kalin, M., Roman, E., Vižintin, J. (2007). The effect of temperature on the tribological mechanisms and reactivity of hydrogenated, amorphous diamondlike carbon coatings under oil-lubricated conditions. Thin Solid Films, vol. 515, no. 7, p. 3644-3652, DOI:10.1016/j.tsf.2006.09.049. [15] Velkavrh, I., Kalin, M., Vižintin, J. (2009). The influence of viscosity on the friction in lubricated DLC contacts at various sliding velocities. Tribology International, vol. 42, no. 11-12, p. 1752-1757, DOI:10.1016/j.triboint.2009.04.022. [16] Velkavrh, I., Kalin, M. (2012). Comparison of the effects of the lubricant-molecule chain length and the viscosity on the friction and wear of diamond-likecarbon coatings and steel. Tribology International, vol. 50, p. 57-65, DOI:10.1016/j.triboint.2012.01.008. [17] Kalin, M., Vižintin, J., Vercammen, K., Barriga, J., Arnšek, A. (2006). The lubrication of DLC coatings with mineral and biodegradable oils having different polar and saturation characteristics. Surface and Coatings Technology, vol. 200, no. 14-15, p. 45154522, DOI:10.1016/j.surfcoat.2005.03.016. [18] Lundgren, S. M., Persson, K., Mueller, G., Kronberg, B., Clarke, J., Chtaib, M., Claesson, P.M. (2007). Unsaturated Fatty Acids in Alkane Solution: Adsorption to Steel Surfaces. Langmuir, vol. 23, no. 21, p. 1059810602, DOI:10.1021/la700909v. [19] Kalin, M., Simič, R., Hirayama, T., Geue, T., Korelis, P. (2014). Neutron-reflectometry study of alcohol adsorption on various DLC coatings. Applied Surface

Comparison of Alcohol and Fatty Acid Adsorption on Hydrogenated DLC Coatings Studied by AFM and Tribological Tests

717

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 707-718

Science, vol. 288, p. 405-410, DOI:10.1016/j. apsusc.2013.10.047. [20] Simič, R., Kalin, M., Hirayama, T., Korelis, P., Geue, T. (2013). Fatty acid adsorption on several DLC coatings studied by neutron reflectometry. Tribology Letters, (in Press), DOI:10.1007/s11249-013-0257-0. [21] Kalin, M., Polajnar, M. (2013). The correlation between the surface energy, the contact angle and the spreading parameter, and their relevance for the wetting behaviour of DLC with lubricating oils. Tribology International, vol. 66, p. 225-233, DOI:10.1016/j. triboint.2013.05.007. [22] Hardy, W., Doubleday, I. (1922). Boundary lubrication. The paraffin series. Proceedings of the Royal Society of London A, vol. 100, p. 550-574, DOI:10.1098/ rspa.1922.0017. [23] Hu, Y., Liu, W. (1998). Tribological properties of alcohols as lubricating additives for aluminum-onsteel contact. Wear, vol. 218, no. 2, p. 244-249, DOI: 10.1016/S0043-1648(98)00162-8. [24] Przedlacki, M. Kajdas, C. (2006). Tribochemistry of fluorinated fluids hydroxyl groups on steel and aluminum surfaces. Tribology Transactions, vol. 49, no. 2, p. 202-214, DOI:10.1080/05698190500544676. [25] Raman, A., Gawalt, E.S. (2007). Self-assembled monolayers of alkanoic acids on the native oxide surface of SS316L by solution deposition. Langmuir, vol. 23, no. 5, p. 2284-2288, DOI:10.1021/la063089g. [26] Sahoo, R.R., Biswas, S.K. (2009). Frictional response of fatty acids on steel. Journal of Colloid and Interface Science, vol. 333, no. 2, p. 707-718, DOI:10.1016/j. jcis.2009.01.046. [27] Bowden, F.P., Gregory, J.N., Tabor, D. (1945). Lubrication of metal surfaces by fatty acids. Nature, vol. 156, p. 97-101, DOI:10.1038/156097a0. [28] Kalin, M. (2004), Influence of flash temperatures on the tribological behaviour in low-speed sliding: a review. Materials Science and Engineering: A, vol. 374, no. 1-2, p. 390-397, DOI:10.1016/j.msea.2004.03.031. [29] Marino, M.J., Hsiao, E., Bradley, L.C., Eryilmaz, O.L., Erdemir, A., Kim, S.H. (2011). Is ultra-low friction needed to prevent wear of diamond-like carbon (DLC)? An alcohol vapor lubrication study for stainless steel/

718

DLC interface. Tribology Letters, vol. 42, p. 285-291, DOI:10.1007/s11249-011-9771-0. [30] Nakayama, K. (2004). Triboemission of electrons, ions, and photons from diamond-like carbon films and generation of tribomicroplasma. Surface & Coatings Technology, vol. 188-189, p. 599-604, DOI: 10.1016/j. surfcoat.2004.07.103. [31] Matta, C., Eryilmaz, O.L., de Barros Bouchet, M.I., Erdemir, A., Martin, J.M., Nakayama, K. (2009). On the possible role of triboplasma in friction and wear of diamond-like carbon films in hydrogen-containing environments. Journal of Physics D: Applied Physics, vol. 42, no. 7, p. 075307, DOI:10.1088/00223727/42/7/075307. [32] Nakayama, K., Martin, J.M. (2006). Tribochemical reactions at and in the vicinity of a sliding contact. Wear, vol. 261, no. 3-4, p. 235-240, DOI:10.1016/j. wear.2005.10.012. [33] Kajdas, C. (2005). Importance of the triboemission process for tribochemical reaction. Tribology International, vol. 38, no. 3, p. 337-353, DOI:10.1016/j. triboint.2004.08.017. [34] Yang, M., Marino, M.J., Bojan, V.J., Eryilmaz, O.L., Erdemir, A., Kim, S.H. (2011). Quantification of oxygenated species on a diamond-like carbon (DLC) surface. Applied Surface Science, vol. 257, no. 17, p. 7633-7638, DOI:10.1016/j.apsusc.2011.03.152. [35] Kasai, P.H., Shimizu, T. (2012). Bonding of hard disk lubricants with OH-bearing end groups. Tribology Letters, vol. 46, p. 43-47, DOI:10.1007/s11249-0129916-9. [36] Shukla, N., Gellman, A.J., Gui, J. (2000). The interaction of CF3CH2OH and (CF3CF2)2O with amorphous carbon films. Langmuir, vol. 16, p. 65626568, DOI:10.1021/la0004496. [37] Kasai, P.H. (2002). Carbon overcoat: Structure and bonding of Z-DOL. Tribology Letters, vol. 13, no. 3, p. 155-166, DOI:10.1023/A:1020101007436. [38] Kajdas, C. (1987). About an anionic-radical concept of the lubrication mechanism of alcohols. Wear, vol. 116, no. 2, p. 167-180, DOI:10.1016/0043-1648(87)902316.

Simič, R. – Kalin, M.

Received for review: 2013-01-06 Received revised form: 2013-05-20 Accepted for publication: 2013-09-25

A New Method For Machinery Fault Diagnoses Based On an Optimal Multiscale Morphological Filter Tan, W. – Chen, X.A – Dong, S.J. Wei Tan1,2,* – Xiaoan Chen1 – Shaojiang Dong1

2 Key

1 State Key Laboratory of Mechanical Transmission, Chongqing University, China Laboratory of Advanced Manufacturing Technology for Automobile Parts, Chongqing University of Technology, China

In order to effectively eliminate the noise and extract the impulse components in the vibration signals, a new method based on an optimal multiscale morphological filter is proposed. In this method, firstly, the average of the closing and opening operator is used to construct the morphological filter, then the multiscale morphological filters’ structure elements (SEs) are optimized and selected using a particle swarm optimization algorithm (PSO). The noise in the original signal is filtered by the multiscale morphological filter. The proposed method was evaluated by simulated signals and bearing fault signals. The results show that the method can effectively filter the noise and extract the impulse characteristics of the vibration signals, which demonstrate the effectiveness of the proposed method. Keywords: multiscale morphological filter, structure element, particle swarm optimization algorithm, noise reduction

0 INTRODUCTION Rolling bearings are widely used in rotating machinery. Unexpected bearing failures can cause unscheduled downtime and loss. Therefore, it is very important to diagnose the fault in the bearing. However, even useful features in the bearing fault signals are often filled with strong background noise, which prevents the detection of bearing faults, so an effective signal processing method is desired to provide more information about the fault. Recently, the mathematical morphological filter has been introduced into the fault diagnosis of machinery [1]. It can decompose the original signal into several physical parts according to certain geometric characteristics, thus overcoming the drawbacks of other signal processing methods such as the Fast Fourier transform (FFT) method, which cannot be used to the nonlinear and non stationary signal, the wavelet transform method, which has to choose the basic wavelet and needs a long time to compute [2], and the empirical mode decomposition (EMD) method, which has the problem of mode mixing [3]. There are some studies of morphology analysis in one-dimensional (1-D) signals; a flat structuring element (SE) with a length around 0.6 times the pulse repetition period was used to demodulate the fault signals [1]. An open-closing and close-opening combined morphological filter to de-noise the vibration signal of rotating machinery was proposed [4]. Zhang, et al. [5] introduced an approach based on the morphological filter to extract the features of the signal from a faulty gear according to the signal characteristics. Li and Xiao [6] introduced pattern

classification based on a 1-D adaptive rank-order morphological filter. However, those studies required prior knowledge of the signals and a fixed SE in single-scale morphology analysis. For a special signal, the characteristic features may be presented in multiscales. In order to extract these features, a multiscale morphology analysis is required. Multiscale morphology analysis can be used to extract morphological features of different scales and it is independent of prior knowledge when selecting SEs. Hence, it is often more feasible to remove the noise using a multiscale morphological filter [7]. The rest of this paper is organized as follows. In Section 1, the theory behind the morphological filter is introduced, and a multiscale morphological filter is constructed; the SEs are optimized by PSO. Sections 2 discuss the presented method in detail. Section 3 applies the proposed method to the simulated signal and Section 4 applies the proposed method to the bearing fault experiment. Some conclusions are given in Section 5. 1 MORPHOLOGICAL FILTER ANALYSIS 1.1 The Fundamental Theory of Morphological Filter The morphological filter requires less computational time than other traditional signal processing methods. By constantly moving the SE to match the signal, feature extracting and de-noising can be achieved. For a signal x(n) defined by Df = 0, 1, ..., N–1 boundary, the SE g(n) is a discrete function defined by Dg = 0, 1, ..., M–1 boundary, where N ≥ M. The

*Corr. Author’s Address: The State Key Laboratory of Mechanical Transmission, Chongqing University, No. 174 Shazheng street, Chongqing, China, tanwv@163.com

719

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 719-724

yλ ( x) = λ y ( x / λ ) = λ / 2 ( x / λ ) • g + ( x / λ )  g  =

transformation consists of four basic operations: erosion, dilation, opening and closing:

( x ⊕ g )( n ) = max  x ( n − m ) + g ( n ) , (1)

( x ⊗ g )( n ) = max  x ( n + m ) − g ( n ) , (2)

( x  g )( n ) = ( x ⊗ g ⊕ g )( n ) , (3)

( x • g )( n ) = ( x ⊕ g ⊗ g )( n ). (4)

In the formula ⊕, ⊗, and • symbols corresponds to the dilation, erosion, opening and closing operation. 1.2 Design of the Multiscale Morphology Filter Multiscale morphological filters were first presented by Maragos [7]. In this paper, we defined T to denote the morphological operator as proposed by Maragos. Based on multiscale analysis, we can define:

= λ / 2 ( x / λ ) ⊕ g ⊗ g + ( x / λ ) ⊗ g ⊕ g  = = λ / 2[ x ⊕ λ g ⊗ λ g + x ⊗ λ g ⊕ λ g ],

where the λ is the scale parameter. If λ = 1, the size of g was set to 3×3. If λ = 2, the size of 2g was set to 5×5. So, the size of the ig was set to (2i + 1) × (2i + 1) which according to the length of the noise period [8]. Because the triangle has good stability and good symmetry, the triangular shape SE was selected as an effective filtering window in this paper. The amplitudes of the SEs were optimized through PSO in this paper. 1.3 Optimization of the Morphological Filter The PSO was chosen to optimize the SE of the morphological filter through the following formula [9]:

( x ⊗ g )λ = λ ( x / λ ) ⊗ g  = x ⊗ λ g , (6)

( x ⊕ g )λ = λ ( x / λ ) ⊕ g  = x ⊕ λ g , (7)

(

)

vij ( t + 1) = wvij (t ) + c1r1 j pij (t ) − xij (t ) +

Tλ ( x ) = λT ( x / λ ) , (5)

where {Tλ λ > 0, λ ∈ N } . Similarly, multiscale erosion and dilation can be defined as:

(9)

(

)

+c2 r2 j pg j (t ) − xij (t ) ,

(10)

xij (t + 1) = xij (t ) + vij (t + 1), (11)

where the subscript i represents the ith particle. j represents the j-dimensional. The subscript t represents the t generation. vij(t) is the velocity of the ith particle in the tth iteration; xij(t) is the position of the ith particle; pij(t) is the pbest position of the ith particle; pg j is the gbest position (pbest represents the local optimum of the particles, gbest represents the overall situation optimum of the particles); w represents the inertia weight. c1, c2 are learning factors. r1 ~ U(0,1), r2 ~ U(0,1) represent two independent random functions.

where λ g = g ⊕ g ⊕ ... ⊕ g ( λ − 1) times. In morphological operations, opening and closing operations have different processing performances. An opening operation could restrain the positive impulse and keep the negative impulse, while the closing operation will have inverse function. So, in practical applications, the relative morphological operation should be selected to correspond to the 1.4. Determination of the Fitness Function processing aim. However, it is sometimes it is difficult to get transcendental knowledge of practical positive The signal-to-noise ratio (SNR) is selected as the and negative impulses; the general situation is that fitness function: the positive and negative impulses are contained in practical data at the same time. Therefore, it is p ( w) 2 = RSN ( x) 10 = log , p ( w) Y ( w) , (12) necessary to construct a morphological filtering N ( w) algorithm with a combination of open and closed operations. In this article, the nonlinear filter was where RSN(x) is the SNR function of the system, y(x) is constructed in the form: the output of the filter, Y(w) is the power spectrum of the output signal; N(w) is the power spectrum of the y ( x) = 1 / 2 [ x • g + x  g ] , (8) noise that is removed from the original signal when where y(x) is the filtered signal. So the nonlinear filter the terminal interaction time is reached, to obtain the can be expressed as: SNR value. 720

Tan, W. – Chen, X.A – Dong, S.J.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 719-724

The process of optimizing the SE based on the PSO is given below: 1. At the beginning of the optimization process, randomly initialize positions and velocities, pbest and gbest of the particles; 2. Set the parameters of SEs equal to the parameters of the particles’ positions; 3. Use the SEs to construct the morphological filters; 4. Use the morphological filters to deal with the signal and get the filtered signal y(x), then calculate the current fitness value R(xi) of each particle using Eq. (12); 5. Use the pbest to construct the morphological filter and calculate the fitness value R(xp), then compare the R(xp) with R(xi). If R(xi) is greater than R(xp), then set the parameters of particle xi to the pbest; 6. Use the gbest to construct the morphological filter and calculate the fitness value R(xg) and compare the R(xg) with R(xg). If R(xp) is greater than R(xg), then set the parameters of pbest to the gbest; 7. For each particle i in the swarm, calculate positions xi+1, velocities vi+1 using Eqs. (10) and (11); 8. While the termination conditions are not met, return to step 3; 9. End loop.

y = ∑ ωi yi , (13) i =k

where [k,n] is the range of λ, ωi is the weight of different λi. In this paper, the ωi took the mean value, ωi = 1/(n-k). 3 THE VALIDATED THROUGH SIMULATION DATA A simulated signal is built to validate the proposed method. It is defined as:

x(t) = x1(t) + x2(t) +x3(t) .

(14)

The signal x1(t) = cos(2π 30t) + cos(2π 50t) is shown in Fig.1a; the impulsive signal x2(t) is a typical series of exponentially decaying pulses used to simulate the impulsive signal (the repetition period is 0.0625 s and the impulsive function in one period is e–5t sin(10πt) as shown in Fig. 1b; x3(t) is the Gaussian white noise with a standard deviation of 0.5. The composite signal x(t) is shown in Fig.1c.

2 PROCEDURES OF MULTISCALE MORPHOLOGY ANALYSIS FOR FAULT DIAGNOSIS The procedure for using a multiscale morphological filter for fault diagnosis is as follows: 1. Select a multiscale SE λg, in this paper the scales are selected from 3, 5 and 7 where the unit is one sampling point. (The size of the SE is subject to the length of the noise period [8]) 2. For the scale of λi, the PSO is used to optimize the amplitudes of the SE; the SNR function is used as the fitness function. 3. Then, the SE with an optimized length is used to perform the multiscale morphological filter operation on the original signal and impulsive components are extracted so that the background noise can be better restrained. 4. Change the scale to λi+1, repeat steps (2 to 4); 5. Through a weight operation the final results of the multiscale morphology analysis can be obtained:

Fig. 1. The signal Waveform graph

The parameters of the PSO for optimizing the SE were set as the original position x(t) = 0 and velocity v(t) = 0, pbest = 0, gbest = 0, the population scale m = 20, the terminal interaction time tmax = 200, the inertia weight w = 0.5, c1 = c2 = 1.2, the search space dimension d = 3. The optimal SEs were obtained as {0, 0.0006, 0}, {0, 0.1115, 0.2229, 0.1115, 0}, {0, 0.1484, 0.2968, 0.4452, 0.2968, 0.1484}. Fig. 2 is the FFT spectrum of the simulated signal, the impulsive frequency was laid in the strong harmonic frequencies (30 and 50 Hz). The impulsive frequency may be covered by the noise if it is strong or the impulsive signal is weak. The purpose of the

A New Method For Machinery Fault Diagnoses Based On an Optimal Multiscale Morphological Filter

721

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 719-724

simulated experiment is to extract the impulsive features at 16, 32 and 48 Hz, to suppress the harmonic features at 30 and 50 Hz, and remove the white noise feature.

The FFT spectrums of the simulated signal are shown in Figs. 3 and 4. In Fig. 3b, obvious impulsive features, i.e. 16, 32 and 48 Hz, are presented. Compared to Fig. 4b, there are some background signals, such as at 16 and 32 Hz, in Fig. 3b. This shows that the optimal multiscale morphology analysis has better performance in impulsive features extraction and noise reduction than the traditional morphology filter. 4 EXPERIMENTAL VALIDATION

Fig. 2. The FFT spectrum of the simulated signal

The results achieved using the original singlescale morphological filter employed by Nikolaou [1] (the structure element g = {0, 0, 0}), and the method proposed in this paper are shown and compared in Fig. 3. Since the SEs were constructed according to the characters of the signal, the entire impulsive signal can be extracted from the background noise, which cannot be carried out by single-scale morphology analysis.

The effectiveness of the proposed optimal multiscale morphology analysis method was evaluated using the vibration data measured in our lab. In the experimental setup detailed in Fig. 5, the power is provided by an electrical motor. To measure the vibrations three accelerometers were mounted on the square housing of the analyzed bearing. The outer and inner races have a fault created by electro-discharge machining as shown in Figs. 6 and 7. The fault size is 0.007 inches in diameter and 0.011 inches in depth for ball, inner and outer races. The motor speed is 1721 RPM (28.7 Hz), the inner defect frequency was calculated to be 156 Hz and the outer defect frequency 103 Hz.

Fig. 3. The FFT spectrums of the simulated signal

Fig. 5. Experimental setup

Fig. 4. The FFT spectrums of the simulated signal Fig. 6. Roller bearings with outer race fault

722

Tan, W. – Chen, X.A – Dong, S.J.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 719-724

Fig. 7. Roller bearings with inner race fault

Data were collected at 12,000 samples/second, the number of sampling points is 12k. Roller bearings with inner and outerrace faults are used for the analysis. Their time domain waveform and FFT spectrum are shown in Figs. 8 and 10. However, due to the noise disturbance, it was difficult to obtain useful information from the analysis.

Fig. 9. Outer race fault vibration signal analyzed using an optimal multiscale morphology filter

Fig. 10. Inner race fault signal waveform and FFT spectrum

Fig. 8. Outer race fault signal waveform and FFT spectrum

For the outer race fault signal, the parameters of the PSO for optimizing the SE were set such that the original position was x(t) = 0 and the velocity was v(t) = 0, pbest = 0, gbest = 0, the population scale was m = 20, the terminal interaction time was tmax = 300, the inertia weight was w = 0.5, c1 = c2 = 1.2, and the search space dimension was d = 3. The optimal SEs obtained were {0, 0.001, 0}, {0, 0.1003, 0.3211, 0.1003,0}, {0, 0.1156, 0.2701, 0.5423, 0.2701, 0.1156, 0}. The vibration signal was analyzed using an optimal multiscale morphology filter. Fig. 9b shows that the outer defect frequency of 103 Hz together with its second and third harmonics, i.e. 206 and 309 Hz, side frequency and modulation frequency (28.7 Hz) were all clearly detected. There is a good match between the expected features of the FFT spectrum and the actual fault features associated with the roller bearing with the outer race fault.

For the inner race fault signal, the parameters of the PSO for optimizing the SE were set such that the original position was x(t) = 0 and velocity was v(t) = 0, pbest = 0, gbest = 0, the population scale was m = 20, the terminal interaction time was tmax = 300, the inertia weight w = 0.5, c1 = c2 = 1.2, and the search space dimension was d = 3. The optimal SEs obtained were {0, 0.0006, 0},{0, 0.0703, 0.1211, 0.0703, 0},{0, 0.0156, 0.1301, 0.2351, 0.1301, 0.0156, 0}.

Fig. 11. Inner race fault vibration signal analyzed by optimal multiscale morphology filter

A New Method For Machinery Fault Diagnoses Based On an Optimal Multiscale Morphological Filter

723

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 719-724

Fig. 11b shows that the inner defect frequency 156 Hz together with its second and third harmonics, i.e., 313 and 467.9 Hz, and side frequencies (156 ± 28.7 Hz) are prominent. The modulation frequencies 28.7 Hz (the frequency of rotor rotating) are also very clear. This reveals that the modified approach is effective in detecting faults in the bearing. 5 CONCLUSIONS In this paper, an optimal multiscale morphology analysis based fault diagnosis approach was proposed. Compared with the traditional morphological filter, two improvements were made in the proposed method. Firstly, the morphological operation was decided with an average combination of multiscale open-closing and close-opening operation, so that not only impulses could be extracted but also the noise could be removed . Then, the structure elements were selected and optimized using aparticle swarm optimization algorithm, which has features similar to the object signal. A step-by-step procedure was also defined to illustrate how the proposed approach can be applied. The validation results show that the proposed approach is more effective and robust in extracting impulsive features than the traditional single-scale morphology analysis. Moreover, since the SE is optimized by the PSO according to the signal, the optimized multiscale morphological filter is beneficial in improving the accuracy of mechanical fault diagnosis. However, as the redundancy increases, the complexity of the computation increases as well. This is one of the main shortcomings of the proposed transform, which should be explored in the future. 6 ACKNOWLEDGMENT We would like to thank the anonymous reviewers and editors for their valuable comments and suggestions. We also thank the Science and Technology Commission of Chongqing (Grant No. cstc2012jcsfjfzhX0027), China.

724

7 REFERENCES [1] Nikolaou, N.G., Antoniadis, I.A. (2003). Application of morphological operators as envelope extractors for impulsive-type periodic signals. Mechanical Systems and Signal Processing, vol. 17, no. 6, p. 1147-1162, DOI:10.1006/mssp.2002.1576. [2] Su, W.S., Wang, F.T., Zhu, H. (2010). Rolling element bearing faults diagnosis based on optimal Morlet wavelet filter and autocorrelation enhancement. Mechanical Systems and Signal Processing, vol. 24, no. 5, p. 1458-1472, DOI:10.1016/j.ymssp.2009.11.011. [3] Cheng, J.S., Yu, D.J., Yang, Y.(2006). A fault diagnosis approach for roller bearings based on EMD method and AR model. Mechanical Systems and Signal Processing, vol. 20, no. 2, p. 350-362, DOI:10.1016/j. ymssp.2004.11.002. [4] Hu, A., Tang, G., An, L. (2006). De-noising technique for vibration signals of rotating machinery based on mathematical morphology filter. Chinese Journal of Mechanical Engineering, vol. 42, no. 4, p. 127-130, DOI:10.3901/JME.2006.04.127. [5] Zhang, L., Yang, D., Xu, J., Chen, Z. (2007). Approach to extracting gear fault feature based on mathematical morphological filtering. Chinese Journal of Mechanical Engineering, vol. 43, no. 2, p. 71-75, DOI:10.3901/ JME.2007.02.071. [6] Li, H., Xiao, D. (2012).Fault diagnosis using pattern classification based on one-dimensional adaptive rank-order morphological filter. Journal of Process Control, vol. 22, no. 2, p. 436-449, DOI:10.1016/j. jprocont.2011.12.005. [7] Maragos, P. (1989). Pattern spectrum and multiscale shape representation. IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 11, no. 7, p. 701-716, DOI:10.1109/34.192465. [8] Ji, T.Y., Lu, Z., Wu, Q.H. (2007). Optimal soft morphological filter for periodic noise removal using a particle swarm optimiser with passive congregation. Signal Processing, vol. 87, no. 11, p. 2799-2809. DOI:10.1016/j.sigpro.2007.05.024. [9] Zhang, Z.Y., Zhang, H.Y. (2004). Principal manifolds and nonlinear dimensionality reduction via tangent space alignment. Journal of Shanghai University (English Edition), vol. 8, no. 4, p. 406-424, DOI:10.1007/s11741-004-0051-1

Tan, W. – Chen, X.A – Dong, S.J.

Received for review: 2013-04-18 Received revised form: 2013-09-12 Accepted for publication: 2013-09-25

Systematic Development of a Device for Bituminous Layer Application Benedičič, J. – Žavbi, R. – Duhovnik, J. Janez Benedičič – Roman Žavbi* – Jožef Duhovnik

University of Ljubljana, Faculty of Mechanical Engineering, Slovenia A method has been developed that enables a systematic approach to searching for opportunities to design new products by taking account of the relevant business environments. The process of searching for opportunities involves the already recognized social, economic, technological and legislative factors (SETL) through which outside influences are measured, and the use of the SETL factors makes it even more systematic. The industrial application of this method as described below has resulted in an innovative product that has already increased existing productivity by a factor of 3, as well as knowledge transfer. The new product has also expanded a developing product family in the company’s existing product portfolio. Keywords: methodology, industrial case, competitive products, productivity, academic-industrial team, training of product developers

0 INTRODUCTION There is no product development without proper knowledge of technical processes, as products are those that lead to individual transformations in the course of technical processes. A technical process is a set of transformations that change an unsatisfactory situation into a satisfactory one [1] and [2]. In general, individual transformations within a technical process can be accomplished by more or less simple products, as well as by physiological machines (humans or animals). A school case of a technical process is e.g. timber harvesting with transformations that involve felling of trees, cutting of branches and tree trunks, hauling of wood to the forest road, and its loading onto an appropriate vehicle to transport it to the sawmill. Each transformation requires a more or less simple product (or physiological machine), e.g. an axe, motor saw, horse, tractor, cable lift, or truck. Individual simple products can also be combined into more complex ones and forest harvester is an example of that. The implementation of technical processes is associated with productivity. In their living and working environment, people strive for maximum productivity, and this is one of the most important economic parameters involved in the implementation of technical processes. The paper describes one application of the opportunity search method to find new products in the early phase (i.e. at the fuzzy front end) of product development that would assist in the implementation of various technical processes in construction. It was done by a mixed industrial-academic team consisting of the representatives of a Slovene metal processing company and the LECAD Laboratory (Faculty of

Mechanical Engineering, University of Ljubljana). Role of teamwork in product development is extensively described in [3]. The above mentioned company has already been oriented into new product development for the field of construction. The automation of work processes in construction thus served as a starting point and the company’s intention was to develop products or product families to facilitate the implementation of technical processes in construction. The said product family was thus supposed to be based on common technological (similar manufacturing technology) and economic (identical or related sales routes and customers) starting points. Many technical processes in construction (e.g. bricklaying, installation of panelling, application of coatings, laying of hydro insulation, binding of reinforcements) are still done manually using simple products (trowels, hammers, brushes, burners, tongs, etc.). As a result, the productivity of these processes is low [4]. In many cases, the above-mentioned products also have harmful effects on the health of workers who use them. Due to constant exposure, occupational diseases may develop and have permanent consequences for the users (e.g. carpal tunnel syndrome in workers exposed to repetitive movements of the wrist, as in the case of reinforcement binding, repetitive and forceful gripping [5] and [6]). Because of a lack of staff and insufficient knowledge about systematic approaches to product development, the company could otherwise consciously try to expand its product portfolio via trial-and-error. By doing so, it could potentially put its own future at risk, but unfortunately such deficits are characteristic of many SMEs. This method, however,

*Corr. Author’s Address: University of Ljubljana, Faculty of Mechanical Engineering, Aškerčeva 6, 1000 Ljubljana, Slovenia, roman.zavbi@@lecad.fs.uni-lj.si

725

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 725-734

brings a systematic approach to the early phase of the development process. Laurie et al. [7] has attached great importance to systematic discovering of new products and opportunities. Companies have no time to wait for a member of the development team to come up with a great idea. The development staff needs to look for opportunities constantly and systematically. The systematic approach is also supported by Hayek [8], who said that general knowledge cannot be a source of new wealth because its broad distribution among potential imitators is antithetical to excess profits. Conversely, specific knowledge holds the potential for wealth generation as a function of its limited, asymmetric distribution. The findings of Fiet et al. [9] provide initial evidence that constrained searches in domains in which an aspirant already possesses specific knowledge yield a higher probability of success than random, unconstrained searches. This paper is of qualitative character and its basic purpose is to present an industrial application of the opportunity search method in detail and to demonstrate the usefulness of a systematic approach/method. The authors have already presented the applicability of this method for various branches of industry [10] and [11]. 1 LITERATURE REVIEW In our analysis of engineering design and development design processes available to date (e.g. [12] to [14]), none of those occurring before 1984 apparently included the opportunity search process, either as an integral part or as a separate process. However, recently the term ‘integrated product development’ has appeared, in which some authors also include the opportunity search process into the initial phase of product development. Andreasen and Hein [2] made a large step forward with their integrated product development (IPD), in which they included both a multidisciplinary approach and an initial phase of the development process during which ideas are created and tasks can be defined on their basis. It is essential for the development process to be initiated by searching for market needs. This is an entirely new approach compared to the “classical” development process. The author states that the search can be systematic, but they do not clearly define or describe such systematics. Bhave [15] also proposed a model of integrated product development, in which he places great emphasis on the recognition of opportunities, but the model does not include any systematic search 726

for information or generation of ideas. Rather, ideas are thought to appear randomly and are also highly dependent on the individual. All decisions are left to the individual and his ‘feeling’. The new concept development (NCD) model [16] contains a phase of opportunity identification, but the identification method itself is not defined, as only an opportunity is recognized. Within its element called the ‘analysis of opportunities’, this model suitably defines the need for collecting information. It is one of rare models that also takes into account the legislation and patents as very important factors. The company management is also present as a central figure that provides management and supervision. Integrated new product development (iNPD) [17] is one of rare processes that is very well structured, phase by phase. The objectives, results and methods are clearly defined. Although the opportunity search process is well defined, the opportunity search phase itself is deemed less important than a thorough clarification and understanding of opportunities. Ardichvili et al. [18] were among the first to emphasize the significance of the cyclic nature of the design process and type of opportunity, but they did not properly define the decision to repeat opportunity development cycles. It is also interesting that they do not advocate a systematic approach and states that investments in the development and marketing of opportunities are recovered sooner if an opportunity is the result of sudden discovery and not of systematic search. By saying this, they partially negates the method, because in their model the development of opportunities is done via predefined steps. In the development of new products, Crawford and Di Benedetto [19] very progressively emphasize the importance of both the market and the technology available within the company. They systematically and efficiently divide the search for opportunities and new products into four basic directions, addressing both external and internal factors. They also take into account the legislation. The process they present, involves a generalized search for predefined external and internal sources. The company management and its owners do not participate in the decision-making process, but their plans do serve as a source for the opportunity search. The Stage-Gate process and the supplemented NexGen Stage-Gate process are both integrated product development processes [20]. The steps within individual phases are not clearly defined and the lower degree of process systemization reduces the probability of success. In a technology driven (TD) variant of the Stage-Gate process, cooper emphasizes

Benedičič, J. – Žavbi, R. – Duhovnik, J.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 725-734

the importance of idea generation from several different sources, but does not define more detailed procedures to search for ideas or to generate them. Laurie’s new growth platform (NGP) [7] includes searching for opportunities and defines the relevant opportunity search area, which had not been so explicitly stated by anyone before him. He builds the process from above down, taking into account a weakly systematic structure. In the process itself, he does not define the phase of idea generation, but prefers to emphasize the significance of the team as the element that identifies and develops an opportunity, i.e. the relevant platform.

promising area and proceed with the opportunity search process.

2 SEARCH FOR OPPORTUNITIES-THE METHOD The method is generally suitable for small as well as large companies. It is of utmost importance for company management to recognize the need for systematic development of new products and for it to be willing to invest in the training of its staff and the implementation of systematic development methods. The method was developed within the scope of doctoral studies [21]. Opportunity search is divided into four steps (Fig. 1): • Step 1: boundary conditions for opportunity search; • Step 2: recognition of opportunities; • Step 3: elimination of irrelevant opportunities; • Step 4: opportunity analysis and ranking. 2.1 Step 1 The first step of our method actually answers the problems encountered by the companies with which we collaborate and deficiencies in the initial part of existing processes and opportunity search methods (Fig. 1). Step one results in a selected subarea and a motto for opportunity search on the basis of each company’s characteristic features, its relevant markets, and the general market trends. Step one is among the most important ones, because the choice of the opportunities search area for new products can significantly influence a company’s business operations and its future orientations. By selecting the opportunities search area and focusing on it, the possibility to discover opportunities in other areas is reduced. However, when faced with failure in the selected area, one can always go back to the first step and choose the next promising area. In this way, it is not necessary to repeat the entire method. Instead, the product developer can choose the second most

Fig. 1. Graphical representation of the SETL method (adapted from [10])

In our case, the company formed a mixed development team and defined the starting points for development work: • In searching for processes, emphasis should be on economic compatibility with existing products from the designed product family; • Emphasis should also be on technological compatibility with existing products from the designed product family; • Suitable work processes in construction should be found that could be automated in line with the relevant economic and technological factors. The predefined starting points are extremely important for directing the development team and its individual members. The degree of work process automation was also taken into account because processes having a lower level of automation (or those which are done manually as was the case in our project) show greater potential for improving productivity.

Systematic Development of a Device for Bituminous Layer Application

727

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 725-734

2.2 Step 2

2.2.2 Social Factor

In step two of the SETL method, opportunities are recognized. In most other processes and methods, the opportunity search process can usually begin only when an opportunity is recognized (Fig. 1). The second step in the SETL method involves intensive data and information gathering about a particular area, based on four recognized influential factors: social, economic, technological and legislative. We look for opportunity carriers that are represented by wishes, fantasies, work processes, trends and reference products in a particular area or subarea. An opportunity is recognized if opportunity carriers, together with the company’s characteristic features, yield a positive financial value. Step two is a cyclic one, whereby each cycle adds to the volume of information and thus deepens one’s understanding of the particular area or subareas. This step allows more detailed work and breakdown of subareas into smaller, more specific parts.

The social factor is becoming an increasingly important part of decision-making related to product purchases. It represents our way of life and our local culture. It has developed over centuries under the influence of constant economic, social and political changes. The social factor thus also involves all those influences that cannot be related to the legislation, technology and economic category of a product. The key influences within the social factor are as follows: • demographics (married, single, family size, children, ageing of the population etc.); • social security (social issues, health protection, unemployment, etc.); • environment; • cultural aspects; • fashion trends (visual and emotional impact of fashion trends); • spare time (vacations, hobbies, entertainment, etc.); • politics (impact of politics, lobbying, etc.).

2.2.1 SETL Factors

2.2.3 Economic Factor

The SETL factors are four mutually complementary factors that assist in finding information about a particular area or subareas and arranging the obtained information [7]. The SETL factors structurally capture all the information about an area, opportunity or product. They capture all influences on opportunity recognition and can vary slightly, depending on the search area. The SETL factors are similar to the catalogues used in some development methods [18]. They guide and lead the user in his/her search for information. They provide a guarantee that all vital information will be gathered. The more diverse the sources of data and information, the greater the volume of various pieces of information, and this enhances the probability of diverse recognized opportunities. Gathering of information and data from primary sources is difficult as it requires a lot of direct personal communication, which in turn means that more human resources are needed. At the beginning of the opportunity recognition step, secondary sources constitute the first supply of data and information. They do not require personal communication and are mostly accessible without major problems. Later on, primary sources play the most important role, because the biggest volume of information and data can be expected from direct conversations with users, specialists and others, as well as from observing of work processes. 728

With the economic factor, one tries to define the economic potential of an area, opportunity or product. When the already recognized opportunities are analyzed, it is possible to obtain quite exact economic information on any particular opportunity. However, while opportunities are still searched for or the decision to enter a specific area or branch is still being made, our thinking and gathered information are very abstract. In order to be able to define the economic prospects of an area and subsequently also of an opportunity, concrete economic and financial data and indicators are required. However, since it is very difficult to do so before opportunities have been recognized, one can resort to using so-called reference products for a certain area or branch. They are those that are important for the company for which one is trying to find promising areas or opportunities. Therefore, they are typical representatives of the area and match the company’s development orientations. When looking for information, the following directions can be helpful: • size of the area; • current and expected growth of the selected area; • current and future purchasing power of the user; • added value of the reference products; • financial analysis of the work processes; • productivity; • financial effect of any related products.

Benedičič, J. – Žavbi, R. – Duhovnik, J.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 725-734

•

2.2.4 Technological Factor Good knowledge of modern technology and past scientific discoveries, and consequently also of future technologies, is of vital importance for the opportunity search process. These factors, all of which are included in the technological factor, represent an important source of new opportunities. There are many cases when it is exactly through the use of new technologies that new and beneficial uses of existing products are found. When looking for information, the following directions can be helpful: • the technology of work processes; • the technology of reference product manufacture; • the technology of reference product application; • patents; • new technologies; • environmentally appropriate technologies; • use of natural resources.

• • • • • •

2.2.5 Legislative Factor The regulatory issues involving the manufacture, sale and application of products are becoming ever stricter, and this will make the legislative factor that much more important in the future. Laws can encourage a trend in a particular area and vice versa, i.e. trends can also dictate the adoption of new legislation. Furthermore, new discoveries and radical innovations can give rise to the adoption of specific rules and laws. One should always be aware that with radical innovations one may also set new market standards. When looking for information, the following directions can be helpful: • EU Regulations and directives; • legislation of the state in whose markets one wishes to be present; • legislation of the states that serve as models to the one in whose markets one wishes to be present (legislative trends); • rules and regulations; • standards. The rest of this subsection presents a case involving the acquisition and categorizing of information for the process of Ibitol (bituminous layer) application: • • •

Social factors: Desire for simple use of the machine; Possibility for workers to use the machine in a comfortable, upright posture; Elimination of repetitive movements and unnatural positions with reduced total number of necessary movements;

•

Possibility of more precise dosage of the coating (kg/m2) to reduce its consumption as well as its environmental impact. Economic factors: Efficient work process; Value of existing comparable machines: Manual application with a roller (i.e. the value corresponds to a worker’s salary); Effective machine use: Up to 5 hours a day per person (worker’s self-assessment); Average machine capacity: Up to 60 m2 per hour (worker’s self-assessment and measured value); Sales potential: Estimated 75%. Technological factors: Work process description: Prior to application, the surface should be dry (permitted moisture of the base ≤ 4% weight) and dust-free, there should be no remaining large particles from the previous base and also no loose particles. The coating material supplied from a tank (which should be heated if necessary) on the upgrade should be uniformly spread across the base by using a (motorized) roller. A variant with a heated tank for preheating of the coating material is also possible. The objective of the work process is to ensure a uniform coating of desired thickness across the entire surface area. Estimate of the necessary power for performing the work process: P = 50 W for moving the machine and P = 50 W for coating application, plus the power of the gas heater for preheating the coating material. Review of existing tools for work process implementation: Application of coatings in construction is done primarily by manual means. Machines for applying coatings by spraying are also used, but their applicability is limited or nil for materials of higher viscosity such as e.g. bitumen. Advantages of manual application: lower initial tool costs, simplicity, no machine maintenance. Disadvantages: low productivity, greater staff workload, lower coat uniformity. Current trends in the area of the analyzed process: The area of automated application of coatings is still quite undeveloped, because in most cases application is done manually using rollers. New technological solutions can be expected.

Legislative factors: • Regulations on movable pressurized equipment (Official Gazette of the RS, No. 18/2004,

Systematic Development of a Device for Bituminous Layer Application

729

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 725-734

138/2006) (economic activities: technical regulations and measurements, No. 98) • Regulations on the permitted amount of toxic solvents in coatings are stricter each year. In our case, the following work processes suitable for automation were identified as potential opportunities for new product development: • floor sanding, • drying of the base, • application of coatings, • application of industrial flooring or paving, • transport of small loads within a construction site, • application of glue onto a prepared surface, • cutting of concrete base, • removal of indoor flooring, • grinding of concrete base, • fixing of road surface cracks, • maintenance of indoor floors, • fixing of hydro insulation, • road surface grooving to prevent hydroplaning, • application of road markings. 2.3 Step 3 In this step, the unsuitable opportunities are eliminated according to the criteria determined in collaboration with the company management, which should enable a clear distinction between suitable and unsuitable opportunities. An opportunity that has been designated as unsuitable according to any criterion should be eliminated unless the company management disagrees despite the assessment. Note that all four factors should be taken into consideration. This step also allows the possibility of Eureka ideas. During step two, the team identified several different work processes. A decision had to be made which of them were promising and suitable, and these had to be analysed in greater detail. The elimination method was used, i.e. if the work process did not fulfil any of the set criteria, it was eliminated. There were two exclusion criteria: • Sales potential assessed from the available information (YES/NO); • Economic and technical congruity of process automation with the directions of the product family (use in construction for performing currently remaining manual work processes, smaller product dimensions) (YES/NO). • Work processes which fulfilled both criteria: • floor sanding, • drying of the base, • application of coatings, • application onto a prepared surface, 730

• • •

fixing of road surface cracks, fixing of hydro insulation, application of road markings. The degree of automation depends on the solution and on the social, economic, technological and legislative indicators. 2.4 Step 4 Most processes and opportunity search methods (Fig. 1) also include a detailed analysis of each opportunity, which in fact helps the product developer understand the true potential of the opportunity. It is also necessary to answer the question of whether the recognized opportunities are indeed realistic or it was only believed so due to insufficient information. The procedure results in the suitability ranking of opportunities and in their detailed description, which can already lead to some requirements for a new product or service. In step four, the opportunities are analysed up to the point at which the company is able to include them in its development projects. If one wants to further analyse individual opportunities and establish their suitability, it is necessary to carry out a detailed analysis on the basis of the SETL factors. Close attention should be paid to primary sources because in order to confirm an opportunity carrier, a more personal contact with the users is required. 2.4.1 Ranking of Opportunities Ranking of opportunities is one of the most important decision-making phases in the further product development process. The range of data on each opportunity is sufficiently broad to enable proper ranking of the opportunity within the development project portfolio. In our case, the team conducted an in-depth analysis of each of the selected processes. The main emphasis was on primary sources. Three interviews with insulation experts were conducted, for both asphalt surfaces and roofs, and this yielded additional information on individual processes. One important piece of information was that legislative requirements regarding the solvent content of coating materials become stricter every year. Therefore, some foreign companies have started producing environmentally friendly hydro insulation coatings. These have a disadvantage, however, in terms of lower adherence to the base and the consequentially longer application process. On the other hand, the said disadvantage can also be an advantage to us, as an automated application process

Benedičič, J. – Žavbi, R. – Duhovnik, J.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 725-734

(or a product used to perform it more quickly) could lead to a higher quality of the coating and an increased productivity. The opportunity rank was created using the AHP method based on the comparison of opportunities [22] and [23]. The application of coatings was selected as the most promising process for automation, and fixing of hydro insulation ranked second. The company management decided to proceed with the development of a new product for the implementation of the coating application process. Table 1. Parameters of SETL method implementation in the discussed company

Step 1 Step 2 Step 3 Step 4 Sum

Duration Workload Number Number of Time [weeks] per team of core extended consumption member team team per step [hours] members members [hours] 3 15 3 0 45 10 60 3 2 320 1 3 3 0 9 8 51 3 2 255 22 129 629

•

Regardless of the application method, the consumption of the coating material per surface area amounted to 0.3 l/m2 , which is in agreement with the manufacturer’s data that states coating material consumption of between 0.3 and 0.5 l/ m2 ; • Because of the influence of dust and surface tension, it was characteristic of all the application methods that the coating material did not properly adhere to the base without being rubbed in; • By far the most appropriate application method was to pour the coating and then spread it across the surface with circular brush strokes (Fig. 2a); • The worst tested work process for insulation material application was pressing it in with a sponge. Fig. 2b above clearly shows that in spite of a completely wet sponge pressing of the insulation material yields extremely poor surface coverage. Based on the test results, the working principle of pouring the insulation material and its circular spreading was selected for continued work.

3 FURTHER SYSTEMATIC DEVELOPMENT The recognized opportunity (in our case the coating application process) served as the basis for continued development of a product for process implementation. The search for the opportunity (four steps) lasted a total of 22 weeks (Table 1). The team members faced the greatest time constraints in the second and fourth step. Generally this depends on the level of difficulty of the selected area and the necessary analyses for evaluating the suitability of individual opportunities. In pursuing the goal of high-quality product development, a systematic approach to the development process is essential also in further development steps [1], [13] and [14]. The key functions of a product for the implementation of a process for applying hydro insulation coatings are coating application and transport of the coating material from the storage to the site. The tests were done according to a predefined protocol. Since the coating is pressed onto dust (that is present on the surface even after cleaning, which usually occurs before coating application) on concrete surfaces, pouring of the insulation material and its spreading with a brush in a circular motion proved to be the most effective working principle for coating application (Fig. 2a). Analysis of the test results showed the following:

b) Fig. 2. Testing of the working principle for a) circular spreading of the hydro insulation material with a brush and b) pressing it with a sponge

Systematic Development of a Device for Bituminous Layer Application

731

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 725-734

Later on, appropriate testing of simple prototypes (material, geometry and geometric distribution of bristles, system for transport of the coating material, etc.) was performed and the results were used in continued product development for the implementation of the application process for hydro insulation coatings (Fig. 3). After the conceptual design, testing and selection of working principles for all partial product functions, a complete prototype was designed by combining all partial functions of the machine for application of hydro insulation coatings. Systematic improvement of the product continued to the phase in which the company management decided to manufacture a test batch (Fig. 4). A patent [24] was granted for the machine in 2011 and its prototype was presented at an innovation forum in 2010 [25].

4 IMPACT OF THE DEVELOPED PRODUCT ON THE APPLICATION OF HYDRO INSULATION COATING The success of the entire procedure, i.e. searching for opportunities and further systematic development process, has been proven (strictly speaking, a properly controlled formal experiment should have been performed to formally prove the utility of the presented method) by the characteristics of the developed machine, which partially automates the process for the application of hydro insulation coatings. Both processes (the manual one and the partially automated one) were performed, measured and analysed. The testing was done under the same conditions, i.e. above all on the same surface, using the same type of coating and at the same ambient and coating temperatures. The surface quality affects the speed of application as much as 30%, assuming the same required quality of the application. The result of the process for applying the hydro insulation coating using the developed machine was manifested directly as increased productivity of the process (Table 2). Table 2. Time dynamics of the application process Preparation [min/day] Manual Using the developed machine

Fig. 3. Partial prototype at initial testing

Fig. 4. Machine for application of insulation coatings

732

15 20

Container Application Cleaning exchange [m2/hour] [min/day] [min/exchange] 2 60 3 1.5

180

Taking into account a 10-hour work day, the effects of preparation, container exchange and cleaning can be considered negligible (approx. 30 minutes per 10 hours of effective application), as the application time is significantly longer than the duration of all other accompanying tasks. It is evident that the productivity of machine application increases by a factor of 3 compared to manual application. A partially automated process therefore has a significant advantage also because it reduces repetitive movements of the hands and thus lessens workplace injuries. However, for a product to succeed, the technical aspect is not enough. The owner or the investor also has to see the financial benefits of process automation, based on which he/she will be willing to finance the investment and attain some market advantage over his/her competitors. In calculating the economic benefit, the gross price per work hour of EUR 7.05 was taken into account. Table 3 shows the calculated savings of machine application with respect to the defined productivity. Return on investment in equipment and the costs of

Benedičič, J. – Žavbi, R. – Duhovnik, J.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 725-734

reproduction materials could be measured as savings compared to manual work. These in turn represent the direct economic benefit for the machine’s owner who provides the services of coating application. Such a machine with included maintenance for 50,000 m2 of coated area costs about € 3,200. Compared to manual application (see Table 3), € 0.059 (pessimistic estimate) or € 0.079 (optimistic estimate) of labour costs per applied square meter are saved. If the machine costs € 3,200 with included maintenance (i.e. machine price of € 3,000 and maintenance costs for 50,000 m2), the investment is recovered after coating 54,237 m2 (pessimistic estimate) or 40,506 m2 (optimistic estimate). Each year, an area of about 30,000 m2 can be coated by an employee. This means that the related investment would be recovered within one to two years. Table 3. Savings with machine application

Labour costs [€/m2] 0.118

Savings compared to manual work [€/m2] /

120

0.059

180

0.039

0.079

Productivity [m2/hour] Manual application Machine application (pessimistic scenario) Machine application (optimistic scenario)

5 CONCLUSIONS The opportunity search method used for the development of new products provides systematic searching and provide traceability and iterability to the opportunity search process. Compared to other methods and processes, this method covers all phases of opportunity search, from a company’s decision that it needs a new product, up to its inclusion in current development projects. The initial part represents an important contribution of this method to the systematic analysis of development processes. When deciding on new product development, one should not immediately focus on opportunities. An analysis of the specific entrepreneurial environment and global trends should be done first and then a decision should be made on where to search for opportunities. This would ensure a systematic approach as well as the traceability and iterativity of steps in the initial part of the method. The methods recognised so far mostly start directly by searching for opportunities without prior selection of an area or subarea. Process tracking is not performed because the input parameters are missing. With the

introduction of Step 1, our method enables traceability from the very start. The systematics and traceability of opportunity search are also ensured through the use of social, economic and legislative factors, which is the second major difference compared to other methods and processes. Completely and thoroughly defined SETL factors ensure systematic gathering of data and information and provide for a continual interweave of the design process with the specific entrepreneurial environment. The method’s complexity level is suitable for its simple introduction in companies without the need for major tutorials. Training takes place in parallel with its use. Despite its simplicity, the method has yielded good results, as has been demonstrated by the presented case and by other examples [10] and [11]. The application of this method and the systematic product development process has also resulted in three patents so far [24], [26] and [27]. We are well aware that the application of this method is supported only by few examples, which is not enough to validate the method. In another words, while it appears that the opportunity search method was a success for the companies (which applied it), the results cannot be generalized. However, the primary goal of this article is dissemination of the method and its successful application(s), which makes it possible for other teams to use the method. 6 ACKNOWLEDGEMENT This work was performed in collaboration with NIKO d.o.o., Železniki, and co-funded by the Ministry of Education, Science, Culture and Sport of the Republic of Slovenia (Contract No. P2-0256). 7 REFERENCES [1] Hubka, V., Eder, E.W. (1988). Theory of Technical Systems. Springer-Verlag, Berlin, DOI:10.1007/978-3642-52121-8. [2] Andreasen, M.M., Hein, L. (2000). Integrated Product Development, reprint. Technical University of Denmark, Institute for Product Development, Lyngby. [3] Rihar, L., Kušar, J., Gorenc, S., Starbek, M. (2012). Teamwork in the Simultaneous Product Realisation. Strojniški vestnik - Journal of Mechanical Engineering, vol. 58, no. 9, p. 534-544, DOI:10.5545/svjme.2012.420. [4] Rivas, R.A., Borcherding, J.D., Gonzales, V. Alarcon, L.F. (2011). Analysis of factors influencing productivity using craftsmen questionnaires: case study in a Chilean construction company. Journal of Construction

Systematic Development of a Device for Bituminous Layer Application

733

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 725-734

Engineering and Management, vol. 137, no. 4, p. 312320, DOI:10.1061/(ASCE)CO.1943-7862.0000274. [5] Barcenilla, A., March, L.M., Chen, J.S. Sambrook, J.S. (2012). Carpal tunnel syndrome and its relationship to occupation: A meta-analysis. Rheumatology, vol. 51, no. 2, p. 250-261, DOI:10.1093/rheumatology/ker108. [6] Palmer, K.T., Harris, E.C., Coggon, D. (2007). Carpal tunnel syndrome and its relation to occupation: A systematic literature review. Occupational Medicine, vol. 57, no. 1, p. 57-66, DOI:10.1136/oem.64.12.787-a. [7] Laurie, D.L., Doz, Y.L., Sheer, C.P. (2006). Creating new growth platforms. Harvard Business Review, vol. 84, no. 5, p. 80-90. [8] Hayek, F.A. (1945). The use of knowledge in society. American Economic Review, vol. 35, no. 4, p. 519-530. [9] Fiet, J.O., Norton Jr., W.I., Van Clouse, G.H. (2007). Systematic search as a source of technical innovation: An empirical test. Journal of Engineering and Technology Management, vol. 24, no. 4, p. 329-346, DOI:10.1016/j.jengtecman.2007.09.001. [10] Žavbi, R., Benedičič, J., Duhovnik, J. (2010). Use of mixed academic-industrial teams for new product development: delivering educational and industrial value. International Journal of Engineering Education, vol. 26, no. 1, p. 178-194. [11] Benedičič, J., Duhovnik, J., Žavbi, R. (2011). Innovations for future development of farms: A case study of the implementation of an opportunity search method on a farm. Transactions ASABE, vol. 54, no. 2, p. 743-752. [12] Hubka, V. (1976). Theorie der Konstruktionsprozesse. Springer-Verlag, Berlin, DOI:10.1007/978-3-64281035-0. [13] Cross, N. (2000). Engineering Design Methods: Strategies for Product Design. J. Wiley & Sons, Chicester. [14] VDI 2221 (1993). Beuth Verlag, Berlin. [15] Bhave, M.P. (1994). A process model of entrepreneurial venture creation. Journal of Business Venturing, vol. 9, no. 3, p. 223-242, DOI:10.1016/0883-9026(94)900310. [16] Koen, P., Ajamian, G., Burkart, R., Clamen, A., Davidson, J., D’Amore, R., Elkins, C., Herald, K., Incorvia, M., Johnson, A., Karol, R., Seibert, R., Slavejkov, A., Wagner, K. (2001). Providing clarity and a common language to the “Fuzzy Front End”.

734

Research Technology Management, vol. 44, no. 2, p. 46-55. [17] Cagan, J., Vogel, C.M. (2002). Creating Breakthrough Products: Innovation from Product Planning to Program Approval. Financial Times Prentice Hall, Upper Saddle River. [18] Ardichvili, A., Cardozo, R., Ray, S. (2003). A theory of entrepreneurial opportunity identification and development. Journal of Business Venturing, vol. 18, no. 1, p. 105-123., DOI:10.1016/S08839026(01)00068-4. [19] Crawford, C.M., Di Benedetto, C.A. (2003). New Products Management. McGraw-Hill, Irwin. [20] Cooper, R.G. (2006). Formula for Success in New Product Development, Working Paper No. 23, p. 1-10, from: http://www.stage-gate.com/knowledge.php accessed 2013-04-10. [21] Benedičič, J. (2007). Searching for Opportunities for New Product Development, Ph.D. Thesis. University of Ljubljana, Faculty of Mechanical Engineering, Ljubljana. [22] Saaty, T.L. (1988). Multicriteria Decision Making The Analytic Hierarchy Process, 2nd ed., University of Pittsburgh, Pittsburgh. [23] Nang-Fei, P. (2008). Fuzzy AHP approach for selecting the suitable bridge construction method. Automation in Construction, vol. 17, no. 8, p. 958-965, DOI:10.1016/j. autcon.2008.03.005. [24] Benedičič, J., Duhovnik, J., Potočnik, S. Sedej, L., Žavbi, R. (2010). Device for Application of Bituminous Layer on at Least Approximately Horizontal Base. Patent No. SI 23364 A, Slovenian Intellectual Property Office, Ljubljana. [25] Catalogue, Public Agency of the Republic of Slovenia for Entrepreneurship and Foreign Investments, (2010), from http://www.foruminovacij.si/sfi/kataloginovacij/?year=2010, accessed on 2013-03-28. [26] Benedičič, J., Duhovnik, J., Žavbi, R., Slak, A., Podgornik, A. Poljanec, D. (2010). Device for Applying a Bitumen Hydroinsulation Band. Patent No. SI 22928 A, Slovenian Intellectual Property Office, Ljubljana. [27] Potočnik, S., Benedičič, J., Čufar, R., Marenk, S., Bertoncelj, T., Žavbi, R. (2010). Appliance for Strengthening Wire Ties between Reinforcing Elements. Patent No. SI 22929 A, Slovenian Intellectual Property Office, Ljubljana.

Benedičič, J. – Žavbi, R. – Duhovnik, J.

Received for review: 2013-04-18 Received revised form: 2013-07-16 Accepted for publication: 2013-08-20

Powder-Mixed Electro-Discharge Diamond Surface Grinding Process: Modelling, Comparative Analysis and Multi-Output Optimisation Using Weighted Principal Components Analysis Modi, M. - Agarwal, G. Manoj Modi* - Gopal Agarwal

Malaviya National Institute of Technology, Mechanical Engineering Department, India Powder-mixed electro-discharge diamond surface grinding (PMEDDSG) is an efficient process for shaping hard materials, such as Ti-6Al4V. Modelling, comparative analysis to study the behaviour of input variables against the responses for both processes, determination of optimum combination of parameters, and studying the effect of input variables on white recast layer thickness and finally on different surface generation during the PMEDDSG processing of Ti-6Al-4V are reported in this paper. The response surface methodology was used to develop the mathematical models of both the responses. Thirty-one experiments were performed on the PMEDDSG set-up without powder-mixed dielectric fluid. Another thirty-two experiments were performed on the PMEDDSG set-up with aluminium powder-mixed dielectric fluid. The current, pulse-duration, wheel-speed, duty-cycle and powder-concentration (considered only as an input variable with the powder-mixed dielectric fluid) were taken as input parameters. The material-removal-rate (MRR) and average-surface-roughness (Ra) were measured as responses in both the process. The weighted principal components (WPC) analysis has been applied to find an optimum setting of PMEDDSG process parameters during multi-output optimisation. A total of 18 experiments were performed according to Taguchi L18 orthogonal-array on the PMEDDSG set-up. The optimum combination suggested by the WPC method was tested to obtain the optimum values of MRR and Ra. The scanning electron microscopy images showed that the surface view and white recast layer thickness of machined workpieces are largely influenced by ampere-current, pulse-duration, duty-cycle, wheel-speed and powder-concentration. Keywords: powder-mixed electro-discharge diamond surface grinding (PMEDDSG), electro-discharge diamond surface grinding (EDDSG), weighted principal components (WPC), Ti-6Al-4V

0 INTRODUCTION Powder-mixed electro-discharge diamond surface grinding (PMEDDSG) is an efficient process for shaping hard materials, such as Ti-6Al-4V. It is widely used in many applications, including aerospace industries and medical implants. This process is a combination of conventional surface grinding with electrical spark machining with the presence of powder in dielectric fluid. In this paper, a PMEDDSG set-up has been designed and fabricated in house. Chen et al. [1] experimented with electrical discharge machining of Ti-6Al-4V. They explained that the removal of material in distilled water occurs due to melting, vaporization and crack propagation, whereas in kerosene, it is due to melting and vaporization. They compared the material removal rate (MRR) and electrode wear ratio in the dielectric fluid and reported the effect of the dielectric fluid on the properties of the EDM-generated surface. Choudhury [2] conducted experiments on the EDDG of high-speed steel. He described how the MRR and grinding forces affected by input machining variables, i.e. current (I), voltage (V), pulse duration (Ton) and duty cycle (DC). Koshy et al. [3] performed experiments on the EDDG of high-speed steel and explained that how the grinding force and the rate of material removal are influenced by ampere-current and the wheel revolution per *Corr. Author’s Address: MNIT, Jaipur, Rajasthan, India, manojmnitjaipur1@gmail.com

minute. Koshy et al. [4] carried out experimentation on electric discharge diamond grinding of WC-Co and explained that how the grinding force and rate of material removal was influenced by amperecurrent and the on-length of the pulse. Kansal et al. [5] explained that MRR and surface roughness are affected by mixing silicon powder in the kerosene of the electric discharge machine. They selected EN-31 as a workpiece material and copper as an electrode material. Habib [6] formulated the models for MRR, the wear ratio of electrode, the gap size and surface roughness by using response surface methodology (RSM). The input machining variables were I, V and SiC % in aluminium. He performed the experiments on electrical discharge machining with an Al/Sic workpiece and copper as an electrode. Hewidy et al. [7] developed models using RSM in the wire EDM of Inconel 601 material. The input machining variables are I, duty factor, wire tension and water pressure whereas the responses are MRR, wear ratio and surface roughness. He reported that the RSM method has the advantage of interpreting the effect of each machining variable on the output response. Lin and Lin [8] applied the combined grey relational analysis-orthogonal array approach for the multi-output optimization of process parameters in electrical discharge machining of SKD11 alloy steel and reported that multi-output optimization problems 735

StrojniĹĄki vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

can be simplified with this approach. Pearson offered principal component analysis (PCA), which was later developed by Hotelling [9]. Su and Tong [10] and Jiju [11] offered an easy and step-by-step solution based on PCA for multi-output optimisation with principal component analysis methodology. Liao [12] explained that there are some shortcomings in the PCA methodology, and used weighted principal components (WPC) method to overcome them. He suggested that the WPC approach reduces the complexity of the engineerâ&#x20AC;&#x2122;s judgement in comparison to the Taguchi approach. Agarwal and Modi [13] showed that effect of current and speed on the MRR in the EDDG process and reported that the MRR increases with increase in current and speed. In this research paper, an experimental study of surface grinding in the EDDG process (with and without powder-mixed dielectric fluid) has been made. The objective of this research is to develop mathematical models of the responses, a comparative analysis to study the behaviour of input variables against the responses for both the processes, to determine the optimum combination of parameters, and studying the effect of input variables on the white recast layer thickness and on different surface generation during the PMEDDSG processing of Ti6Al-4V. In this research work, RSM has been applied to develop the mathematical models of responses. These models were used for understanding and predicting the behaviour of input machining variables over the responses. For this purpose, thirty-one and another thirty-two experiments were conducted on an inhouse-designed and fabricated PMEDDSG set-up with and without aluminium powder-mixed dielectric fluid. Minitab software is used to develop the mathematical models of both the responses. Optimal combinations of machining variables are calculated within the boundary of formulated mathematical models. An optimum combination of machining variables for both the responses is not suggested by this approach. To overcome this problem, weighted principal components analysis has been applied to determine an optimum setting of powder-mixed electro-discharge diamond surface grinding (PMEDDSG) process parameters during multi-output optimisation. A total of eighteen experiments were performed according to the Taguchi L18 orthogonal array on the PMEDDSG set-up. Analysis of variance (ANOVA) has been applied to determine the contribution percentage (C [%]) of different parameters during the PMEDDSG process. 736

1 OBSERVATIONAL METHODOLOGY All the experiments have been conducted on an inhouse-designed and fabricated PMEDDSG set-up with and without powder-mixed dielectric fluid. For this purpose, a special attachment has been designed and fabricated inside the main tank of the EDM machine: a separate acrylic transparent bathtub-like container (called the machining tank, capacity = 36 litres of dielectric fluid) was fixed on the machine table with the help of clamping bolts. In this set-up, a separate pump and stirrer assembly is fixed in the side wall of the machining tank. The pump circulates the powder-mixed dielectric fluid in the inter-electrode gap (IEG) and stirrer is used to mix the powder particles uniformly with dielectric fluid. The details of PMEDDSG set-up are depicted in Fig. 1a. The relative motion between bronze-diamond wheel and work-piece is achieved by reciprocating the machine table through an automatic feed arrangement. The lead screw of EDM machine table was driven by a reversible AC synchronous motor to obtain the relative motion between the workpiece and bronzediamond grinding wheel. The workpiece speed is 0.38 cm/s. The wheel speed is selected according to plan of experimentation (see Tables 3 and 6). The mechanism for removing metal in the PMEDDSG process is due to the occurrence of spark discharge, series discharge and grinding action by the diamond grains of the bronzediamond wheel over the workpiece. Whereas, in the EDDSG process, it is due to the occurrence of spark discharge and grinding action by the diamond grains of the bronze-diamond wheel over the workpiece. Series discharge occurred in the PMEDDSG process due to the presence of powder particles between the wheel and workpiece, i.e. inter-electrode gap (IEG). These particles of powder in IEG become energised and accelerated by the applied electric field and finally move in a zigzag fashion. These electrically conductive powder particles boost the breakdown in the IEG and increase the space between the wheel and workpiece. After that, the particles arranged themselves in chain-structures and finally bridged the gap between the wheel and workpiece. This type of bridging decreases the insulating strength of the dielectric. Therefore, a short circuit occurs, which is responsible for an explosion in the space between the wheel and workpiece, and series discharge finally occurs in the IEG. At the same time, few secondary discharges also occur between the side corner edge of the grinding wheel and the workpiece. The mechanism of removal of metal in the PMEDDSG process is shown in Fig. 1b. The details of the bronze-diamond

Modi, M. - Agarwal, G.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

wheel are depicted in Table 1. Ti-6Al-4V has been taken as the workpiece material for experimentation. It is flat and rectangular; its size is (60×10×20 mm). The advantage of this hybrid process is that it reduces the thickness of recast layer by the grinding action due to abrasion. Table 1. Details of bronze-diamond grinding wheel Abrasive Diameter Thickness Bond Material

Diamond 100 mm 10 mm Bronze

Concentration Bore Depth of abrasive Grit size

75% 32 mm 5 mm 80/100

MRR =

(Wb m − Wa m ) ×1000 t (Wb m − Wa m ) ×1000 t×ρ

, (1a) , (1b)

where Wb m and Wa m are the weights of the workpeace before and after machining, t is the machining time (45 min) and ρ is the density (4.37 gm/cm³) of workpiece material. The hardness of Ti-6Al-4V material is 32 HRC. The weight of the workpiece was measured using a high precision electronic balance (WENSAR HPB-310 model) before and after the experiment, after which the difference in weight was determined. Finally, the MRR is determined by dividing this difference in the weight of the workpiece by the machining time. Ra is measured with a Surtronic-25 Taylor Hobson surface roughness tester at a cut-off value of 0.8 mm. A digital tachometer was used for grinding wheel speed measurement (revolutions per minute (RPM)). 2 RESPONSE SURFACE MODELLING

Fig. 1a) Schematic diagram of in-house-designed and fabricated powder-mixed electro-discharge diamond surface grinding set-up

Response surface modelling (RSM) is a combination of statistics and a mathematics approach. This approach is used to obtain the relationship between input variables and output responses. All experiments have been conducted according to central composite rotary design (CCRD). The response surface in general is expressed by the Eq. (2).

i =1

Y = a0 + ∑ ai xi + ∑ aii xi 2 + ∑ aij xi x j , (2) j >i

where, Y are output responses, a0 coefficients for the free terms, ai coefficients of linear terms, aii coefficients of quadratic terms and aij coefficients for the interaction terms. Thirty-one experiments were performed on an electro-discharge diamond surface grinding set-up. Table 2. Value of machining variables for EDDSG process Fig. 1b) Details of mechanism of removal of material in powdermixed electro-discharge diamond surface grinding process

Eqs. 1a and b are used to calculate the MRR (mg/min) and MRR (mm³/min) for each machining process.

Coded value of machining variables at different levels according to Parameters Symbol Unit CCRD plan -2 -1 0 +1 +2 Current I [A] 1 3 5 7 9 Pulse on-time Ton [µs] 40 70 100 130 160 Wheel Speed WS [RPM] 350 450 550 650 750 Duty Cycle DC [%] 0.59 0.63 0.67 0.71 0.75

Powder-Mixed Electro-Discharge Diamond Surface Grinding Process: Modelling, Comparative Analysis and Multi-Output Optimisation ...

737

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

The value of machining variables involved in this experimental work is depicted in Table 2. The experimental central composite rotary design plan and the responses for the EDDSG process are depicted in Table 3.

Table 4. ANOVA for MRR and Ra models without powder-mixed dielectric fluid

Experiment No.

Table 3. Experimental plan for CCRD and responses for EDDSG process

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

Coded value of machining variables at different levels according to CCRD plan Pulse Wheel Duty Current on-time speed cycle (I) (Ton) (WS) (DC) 0 0 0 2 0 0 0 -2 0 0 -2 0 0 0 0 0 0 2 0 0 -1 -1 -1 -1 -1 1 1 -1 -1 1 -1 -1 1 -1 1 1 1 1 1 -1 1 -1 -1 -1 2 0 0 0 0 0 0 0 -1 1 1 1 -1 -1 -1 1 -1 -1 1 1 0 0 0 0 1 -1 -1 1 0 0 0 0 -2 0 0 0 -1 1 -1 1 -1 -1 1 -1 1 1 -1 -1 0 0 2 0 1 -1 1 -1 1 1 -1 1 0 -2 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0

MRR [mm³/min] Exp. 0.322 0.440 0.268 0.352 0.502 0.232 0.410 0.306 0.496 0.798 0.440 0.712 0.356 0.348 0.164 0.266 0.308 0.384 0.424 0.240 0.252 0.306 0.532 0.450 0.534 0.488 0.298 0.592 0.380 0.368 0.360

Eq. (3) 0.312 0.448 0.270 0.380 0.490 0.212 0.432 0.322 0.494 0.672 0.452 0.732 0.380 0.364 0.144 0.254 0.380 0.384 0.380 0.252 0.254 0.322 0.562 0.490 0.562 0.494 0.270 0.604 0.380 0.380 0.380

Model Residual Error Total

5.08 3.13 4.62 3.97 4.13 3.14 2.96 3.28 4.60 3.94 3.97 4.97 3.92 4.17 4.23 3.61 4.12 5.31 4.00 2.78 4.55 2.51 4.38 3.67 3.33 5.81 3.25 5.19 3.99 3.91 4.07

Eq. (4) 5.19 2.87 4.75 4.03 4.22 2.96 2.88 3.40 4.64 3.92 4.00 5.07 4.03 4.04 4.12 3.60 4.03 5.16 4.03 2.99 4.56 2.44 4.44 3.71 3.48 5.60 3.34 5.08 4.03 4.03 4.03

The mathematical models for MRR and Ra are obtained via Minitab 14 software after eliminating the non-significant terms for the EDDSG process, and are depicted by Eqs. (3) and (4). Using Eqs. (3) and (4), the responses are calculated and are shown in Table 3, in the columns MRR (Eq. (3)) and Ra (Eq. (4)). MRR = 0.38 + 0.12 I + 0.055Ton + 0.055WS − − 0.034 DC + 0.028 I 2 , (3) 738

Source

Ra [µm] Exp.

Ra = 4.03 + 0.52 I + 0.22 Ton − 0.26WS + + 0.58 DC − 0.062 Ton2 + 0.052WS 2 . (4)

Source Model Residual Error Total

DOF 5 25 30

SS 0.55 0.032 0.58

DOF 6 24 30

SS 17.5 0.41 17.9

For MRR Model MS F 0.11 86.5 0.0012

P <0.0001

For Ra Model MS F 2.92 171.3 0.017

P < 0.0001

The ANOVA results are depicted in Table 4. It has been determined that the p-value is less than 0.05 for both models. Thus, it is concluded that the model equations of both responses are significant in describing the relationship between the input parameters and output responses. Another thirty-two experiments were performed on the PMEDDSG setup with aluminium powder. The value of machining variables involved in this experimental work is depicted in Table 5. The experimental CCRD plan and responses for PMEDDSG process are depicted in Table 6. Table 5. Value of all machining variables for PMEDDSG process Parameters

Symbol

Current Pulse on-time Wheel Speed Duty Cycle Al Powder Conc.

I Ton WS DC PC

Coded value of machining variables at different levels according to CCRD plan -2 -1 0 +1 +2 [A] 1 3 5 7 9 [µs] 40 70 100 130 160 [RPM] 350 450 550 650 750 [%] 0.59 0.63 0.67 0.71 0.75 [gm/l] 1.5 2 2.5 3 3.5 Unit

Similarly, the mathematical models for MRR and Ra are obtained with the aid of Minitab 14 software after eliminating the non-significant terms for PMEDDSG process (see Eqs. (5) and (6)). Using Eqs. (5) and (6), the responses are calculated and are shown in Table 6, in the columns MRR (Eq. (5)) and Ra (Eq. (6)).

MRRAl = 0.51 + 0.16 I + 0.070 Ton + 0.067WS − − 0.044 DC + 0.022 PC + 0.022 I 2 , (5)

RaAl = 3.88 + 0.49 I + 0.20 Ton − 0.25WS + 0.54 DC + (6) +0.14 PC − 0.078Ton2 + 0.0072 PC 2 .

Modi, M. - Agarwal, G.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

Coded value of machining variables at different levels according to CCRD plan

Exp.

Eq. (6)

Ra [µm]

-2

0.272 0.278

2.69

2.90

-1

0.440 0.443

3.80

3.73

3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

0 -1 0 1 -1 0 0 0 0 0 1 1 -1 -1 0 -1 -1 1 0 -1 0 0 1 1 2 1 0 0 0 1

0 -1 0 1 1 0 0 0 0 0 -1 1 -1 1 2 1 -1 1 0 -1 0 0 -1 1 0 -1 -2 0 0 -1

0 -1 0 -1 -1 0 -2 0 0 0 -1 -1 1 -1 0 1 -1 1 0 1 0 2 -1 1 0 1 0 0 0 1

0 1 0 1 1 0 0 0 0 0 1 -1 -1 -1 0 -1 -1 1 0 1 0 0 -1 -1 0 1 0 2 -2 -1

0 -1 0 -1 1 0 0 2 0 -2 1 1 -1 -1 0 1 1 1 0 1 0 0 -1 -1 0 -1 0 0 0 1

0.480 0.206 0.490 0.614 0.330 0.550 0.336 0.650 0.510 0.460 0.482 0.840 0.398 0.384 0.630 0.540 0.294 0.760 0.530 0.360 0.480 0.660 0.580 0.880 0.921 0.630 0.374 0.405 0.552 0.678

3.84 3.93 3.82 5.46 4.41 3.76 4.45 4.73 3.96 3.58 5.15 4.27 2.54 3.02 3.98 2.87 3.06 5.02 3.77 3.50 3.95 3.53 3.69 3.65 4.80 4.28 3.13 4.88 3.01 3.23

3.88 3.83 3.88 5.21 4.51 3.88 4.38 4.44 3.88 3.88 5.09 4.41 2.25 3.15 3.96 2.93 3.03 4.99 3.88 3.61 3.88 3.38 3.73 3.63 4.86 4.31 3.16 4.96 2.80 3.51

Exp.

Eq. (5)

MRR [mm³/min]

Pulse ontime (Ton) Wheel speed (WS) Duty cycle (DC) Powder Conc.

Current (I)

Experiment No.

Table 6. Experimental plan for CCRD and responses for PMEDDSG Process

0.510 0.169 0.510 0.629 0.353 0.510 0.376 0.554 0.510 0.466 0.533 0.761 0.391 0.397 0.650 0.575 0.301 0.807 0.510 0.347 0.510 0.644 0.577 0.851 0.918 0.623 0.370 0.422 0.598 0.755

Table 7. ANOVA for MRR and Ra models with powder-mixed dielectric fluid Source Model Residual Error Total Source Model Residual Error Total

DOF 6 25 31

SS 0.90 0.039 0.94

DOF 7 24 31

SS 16.1 0.65 16.7

For MRR Model MS F 0.15 95.72 0.0015

P < 0.0001

For Ra Model MS F 2.31 84.8 0.027

P < 0.0001

The ANOVA results are depicted in Table 7. It is found that the p-value is less than 0.05 for both models. Therefore, it is concluded that model equations of both the responses are significant in describing the relationship between the input parameters and output responses. 3 ANALYSIS OF OUTCOMES AND INTERPRETATIONS (FROM MATHEMATICAL MODELS OF MRR AND Ra WITH AND WITHOUT POWDER) The analysis of outcomes is summarized in Table 8, which also includes the interpretations of Figs. 2 to 9. With reference to Figs. 2 to 5, the MRR is greater in the powder-mixed dielectric fluid as compared to the dielectric fluid without it. The reason for this is the occurrence of the series discharge in the IEG. These series discharges have been taken place (in addition to the spark discharge and grinding action in the EDDSG process) in the IEG due to the presence of aluminium powder in the dielectric fluid. With reference to Figs. 6 to 9, the Ra is less in powder-mixed dielectric fluid as compared to without powder-mixed dielectric fluid. The presence of aluminium powder in the dielectric fluid modified the plasma channel, i.e. it became larger and wider. The spark energy is equally distributed between the powder particles. Hence, the density of spark discharge is reduced, which is responsible for the formation of smaller cavities on the machined workpiece. The optimum setting of different machining variables for the PMEDDSG and EDDSG processes depends on maximizing the MRR and minimizing the Ra; it is depicted in Table 9. These values were calculated within the boundary of formulated mathematical models. The Taguchi method is commonly used for single response optimization; it does not suggest an optimum combination of machining variables for both the responses. There are significant complications involved in multiresponse optimisation as compared to single response optimisation, because the higher log S value of one response (i.e. MRR) may correspond to a lower log S value for another response (i.e. Ra). Hence, an overall evaluation of log S values is needed for the multi-output optimisation. In order to overcome this problem, a Weighted Principal Components Analysis has been applied to find an optimum setting of PMEDDSG process parameters during multi-output optimisation. In Tables 8 and 9, it is shown that the Al powder-mixed EDDSG process yields better results (i.e. higher MRR and lower Ra) as compared to the

Powder-Mixed Electro-Discharge Diamond Surface Grinding Process: Modelling, Comparative Analysis and Multi-Output Optimisation ...

739

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

Table 8. Analysis of outcomes and interpretations of Figs. 2 to 9 Fig. Continuous/ Condition Number Dotted Line (with/without powder) 2

Continuous line Dotted line Continuous line Dotted line

With Powder Without Powder

Reason for Trend 1 in Fig.

Trend 2 in Fig.

MRR rises with increase in current.

More spark energy is developed due to the rising trend in current.

MRR rises with fall in duty cycle.

–||–

MRR rises with increase in pulse on time.

Ton is increasing continuously. Thus, sufficient time is available for conduction.

–||–

MRR rises with increase in wheel speed.

More spark energy is dissipated with the increase in wheel speed. It happens because the rate of current-flow in the grinding area rises.

With Powder Without Powder

Continuous line

With Powder

Dotted line

Without Powder

Continuous line

With Powder

Dotted line

Without Powder

Continuous line

With Powder

–||–

Dotted line

Without Powder

Continuous line

With Powder

Dotted line

Without Powder

Continuous line

With Powder

Trend 1 in Fig.

Dotted line Continuous line

Without Powder

Dotted line

Without Powder

With Powder

MRR rises Flushing increases and MRR rises with with increase More spark energy is developed due to gap width decrease due increase in in wheel the rising trend in current. to rise in wheel speed. current. speed. More spark energy is developed due to the Toff is decreasing continuously to obtain Ra rises with Ra rises with rising trend in current. the rising trend in duty cycle. Thus, a increase in increase in duty Hence, wider size short time is available for flushing and current. cycle. cavities are developed deionization of the dielectric. on work surface. Ton is increasing continuously. So, sufficient on time is available for Ra rises with conduction of heat into workpiece. MRR –||– –||– increase in increases due to softening of workpiece pulse on time. and hence Ra increases due to the formation of wider cavities. All eroded particles from IEG are carried Ra decreases away by the good flushing. This is –||– –||– with increase in achieved with increase in wheel speed. wheel speed. The adherence of eroded particles on work surface is decreased. Ra decreases Flushing increases Ra rises with with increase with increase in wheel More spark energy is developed due to increase in in wheel speed and decreases the rising trend in current. current. speed. the Ra.

EDDSG process during the machining of Ti-6Al4V. Hence, we have applied the weighted principal components approach to find an optimum setting of the PMEDDSG process parameters during multioutput optimization with Al powder-mixed dielectric fluid. After that, SEM analysis was performed on Al powder-mixed EDDSG machined workpieces.

Table 9. Optimal setting of PMEDDSG and EDDSG variables Process parameter MRR [mm³/min] Ra [µm] Process parameter MRR [mm³/min] Ra [µm]

PMEDDSG Process Ton WS DC 130 645 0.63 70 650 0.65 0.883 2.54 EDDSG Process WS DC I Ton 7 130 645 0.63 3 70 650 0.65 0.665 2.71 I 7 3

PC 2.93 2.4

#I / #D [%]

4 OPTIMIZATION THROUGH WEIGHTED PRINCIPAL COMPONENTS (WPC) ANALYSIS

24.6 (#I) 6.71 (#D)

The Principal Components Analysis (PCA) method was first proposed by Pearson and modified by Hotelling [9]. The short comes associated with the PCA method have been rectified by WPC method proposed by Liao [12]. In this method, total principal

#I: improvement in MRR and #D: decrement in Ra in PMEDDSG process as compared to EDDSG process

740

Reason for Trend 2 in Fig. Toff is increasing continuously to get the falling trend in duty cycle. Thus, sufficient time is available for flushing and deionization of the dielectric.

Modi, M. - Agarwal, G.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

Fig. 2. Influence of current on MRR for different duty cycle (WS = 550 RPM, Ton = 100 µs, PC = 2.5 gm/l)

Fig. 3. Influence of current on MRR for different pulse on-time (WS =550 RPM, DC = 0.67, PC= 2.5 gm/l)

Fig. 4. Influence of current on MRR for different wheel speed (Ton = 100 µs, DC = 0.67, PC= 2.5 gm/l)

Fig. 6. Influence of current on Ra for different duty cycle (WS =550 RPM, Ton = 100 µs, PC= 2.5 gm/l)

Fig. 7. Influence of current on Ra for different pulse on-time (WS =550 RPM, DC = 0.67, PC= 2.5 gm/l)

Fig. 8. Influence of current on Ra for different wheel speed (Ton = 100 µs, DC = 0.67, PC=2.5 gm/l)

Fig. 5. Influence of wheel speed on MRR for different current (Ton = 100 µs, DC = 0.67, PC= 2.5 gm/l)

Fig. 9. Influence of wheel speed on Ra for different current (Ton = 100 µs, DC = 0.67, PC = 2.5 gm/l)

components are used. The detailed procedure of this method is depicted in Fig. 10. The various parameters

and their levels are depicted in Table 10. MOPI is depicted in Table 11.

Powder-Mixed Electro-Discharge Diamond Surface Grinding Process: Modelling, Comparative Analysis and Multi-Output Optimisation ...

741

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

For ANOVA analysis, Minitab software is used. From Table 12, it is observed that current (> 95% CL) and powder concentration (≤ 90% CL) are the significant parameters. Ton, WS and DC are the non-significant parameters. It is also observed that percentage contribution C [%] of various process parameters is powder concentration (19%), pulse on-time (1.5%), wheel speed (1.5%), duty cycle (7%) and current (61%) for the responses under the

Table 10. Input variables and their levels for aluminium powdermixed dielectric fluid Symbol Parameters PC I Ton WS DC

Unit

Al Powder Concentration [gm/l] Current [A] Pulse on-time [μs] Wheel Speed [RPM] Duty Cycle [%]

Level 1 1.5 3 100 450 0.61

Level 2 3 7 150 650 0.69

Level 3 11 200 850 0.77

Table 11. L18 orthogonal array, output responses, normalised log S value and MOPI Exp. No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

PC 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2

I 1 1 1 2 2 2 3 3 3 1 1 1 2 2 2 3 3 3

Control Factor Ton WS 1 1 2 2 3 3 1 1 2 2 3 3 1 2 2 3 3 1 1 3 2 1 3 2 1 2 2 3 3 1 1 3 2 1 3 2

DC 1 2 3 2 3 1 1 2 3 3 1 2 3 1 2 2 3 1

Responses MRR [mm³/min] Ra [µm] 0.521 2.21 0.540 2.60 0.660 3.20 0.790 3.40 0.880 4.01 0.659 4.65 0.956 4.15 0.980 4.63 0.835 5.80 0.780 3.60 0.625 4.70 0.726 4.98 1.025 4.35 0.880 5.80 0.760 5.40 0.915 4.70 1.035 5.90 0.840 6.54

log S value MRR Ra -5.66 -6.89 -5.35 -8.30 -3.61 -10.10 -2.05 -10.63 -1.11 -12.06 -3.62 -13.35 -0.39 -12.36 -0.18 -13.31 -1.57 -15.27 -2.16 -11.82 -4.08 -14.15 -2.78 -13.94 0.21 -12.77 -1.11 -15.27 -2.38 -14.65 -0.77 -13.44 0.30 -16.31 -1.51 -15.42

Normalized log S value MRR Ra 0.00 1.00 0.05 0.85 0.34 0.66 0.61 0.60 0.76 0.45 0.34 0.31 0.88 0.42 0.92 0.32 0.69 0.11 0.59 0.48 0.27 0.23 0.48 0.25 0.99 0.38 0.76 0.11 0.55 0.18 0.82 0.30 1.00 0.00 0.70 0.09

MOPI -0.45 -0.35 -0.05 0.16 0.34 0.10 0.44 0.51 0.44 0.20 0.08 0.23 0.53 0.49 0.31 0.44 0.71 0.45

Table 12. Average MOPI values at different factor levels and ANOVA table for MOPI Average MOPI values at different factor levels Level 1 Level 2 Level 3 Max-Min 0.03 0.60 0.57 -0.05 0.32 0.495 0.545 0.21 0.29 0.24 0.08 0.20 0.27 0.28 0.08 0.18 0.21 0.35 0.17

Symbol

DF 1 2 2 2 2 6 17

PC I Ton WS DC Error Total

Table 13. Display the explained variation and eigenvector of each principal component Principal Component PC1 PC2

742

Eigen values 1.6384 0.3616

Explained variation 81.9 18.1

Cumulative variation 81.9 100

Eigen vector [0.707 -0.707] [0.707 0.707]

SS 0.2983 0.9564 0.019 0.020 0.1035 0.1425 1.5675

ANOVA table for MOPI MS F 0.2983 3.76 0.4782 11.74 0.009 0.09 0.010 0.09 0.0517 0.53

P 0.070 0.001 0.914 0.910 0.599

C [%] 19 61 1.5 1.5 7 10 100

Table 14. Result of the confirmation test Factor Level

Initial process parameters

PC1I1Ton1WS1DC1 MRR [mm³/min] 0.521 Ra [µm] 2.21 Overall MOPI grade –0.45 Improvement in MOPI is 0.93.

Modi, M. - Agarwal, G.

Optimal process parameters (WPC) approach PC2I3Ton2WS3DC3 1.055 4.76 0.48

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

Start



First step is to calculate the Signal to Noise ratio (aij) of MRR (according to higher is better, βij = −10 log  1 (according to lower is better, βij = −10 log  1

m

∑o t =1

2 ijt



1  ) and R a 2  ijt  

∑ m o t =1

  ) criterion, where m is number of trials and oijt is a value of output. 

Normalize the Signal to Noise ratio of removal rate of material and average surface roughness in zero to one value according to math. min max min formula X = β ij − β i , β i = minimum of βi1, βi2, ..., βpj, and β i = maximum of βi1, βi2, ..., βpj, where Xij is normalised value. ij max β i − β imin

After the normalising the values Minitab software is used for Principal Component Analysis (depicted in Table 13). The relation between Responses and Principal Components are given by PCn = ∑ i =1 ani xij , ∑ i =1 ani2 = 1; where Xij is a normalised value of ith response at jth trial for i = 1, 2,..., l; n = 1, 2, ..., k; P is a number of resonses; ani Eigen Vector and k(k≤p) components to determine the p

variance in p responses. PC1 = 0.707 X MRR – 0.707 X Ra and PC2 = 0.707 X MRR + 0.707 X Ra. Finally calculate the MOPI = ∑ n =1 wn PCn = 0.819 X PC1 + 0.181 X PC2, where wn is weight of nth principal components, MOPI = Multi Output Performance Index. k

End

Fig. 10. Flow chart to calculate MOPI for WPC approach

multi-output optimization (maximization of MRR and minimization of Ra) in the PMEDDSG of Ti-6Al4V. The optimum combination of different factors is PC2I3Ton2WS3DC3, and its confirmation experimental results are depicted in Table 14. 5 ANALYSIS OF PMEDDSG-GENERATED SURFACE The machined surface generated through the PMEDDSG process is subject to many changes in the form of surface cavities and pits. To investigate the effect of various input parameters i.e. current, pulse-duration, duty cycle, wheel speed and powder concentration on the PMEDDSG-generated surfaces, electron scanning microscopy (SEM) is used at 500× magnification. To prepare the specimens for SEM analysis, emery paper (3/0 and 4/0) were used to grind the produced machined surfaces; after that, these surfaces were cleaned and shined by polishing and finally etched with K-AGENT (combination of eighty-eight percent pure water, ten percent nitric acid and two percent hydrofluoric acid) for 50 seconds. The effect of current and pulse-duration on PMEDDSG of Ti-6Al-4V (with Al powder-mixed dielectric fluid) generated surfaces are depicted in Figs. 11a and b and 12a and b, respectively. It was observed from all images that the surface cavities become wider and bigger as the current or pulse-

duration increases. Increases in current or pulseduration lead to increase in spark energy, which leads to the evacuation of more material in molten form from work surfaces. Hence, increases in current or pulse-duration is the reason for the formation of wider and bigger sizes cavities on work surface. The effect of duty cycle on the PMEDDSG of Ti-6Al-4V generated surface (with Al powdermixed dielectric fluid) is depicted in Figs. 13a and b. It was observed from both images that surface cavities are wider and bigger with increase in duty cycle. Pulse on-duration remains constant throughout the experiment and the duty cycle increases due to decreased pulse off-time. This means less time is available for dielectric flushing to sweep the ejected particles from the IEG and for the deionization of the dielectric. Therefore, a continuous spark takes place and Ra increases. The effect of wheel speed on PMEDDSG of Ti6Al-4V generated surfaces (with Al powder-mixed dielectric fluid) is depicted in Figs. 14a and b. It was observed from both images that the surface cavities are narrower and smaller with increase in wheel speed. Ra decreases with increase in wheel speed. The increased in-wheel speed is the reason for the efficient dielectric flushing in the IEG, which reduces the deposition of re-solidified ejected particles on a worksurface and also reduces the possibilities of arcing.

Powder-Mixed Electro-Discharge Diamond Surface Grinding Process: Modelling, Comparative Analysis and Multi-Output Optimisation ...

743

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

b) Fig. 11. Influence of current on PMEDDSG-generated surface; a) I = 2 A, b) I = 10 A (WS = 550 RPM, DC = 0.80, Ton = 300 µs, PC = 5 gm/l of Al Powder)

b) Fig. 13. Influence of duty cycle on PMEDDSG-generated surface; a) DC = 0.70, b) DC = 0.75 (WS = 550 RPM, I = 6 A, Ton = 100 µs, PC = 5 gm/l of Al Powder)

b) Fig. 12. Influence of pulse-duration on PMEDDSG-generated surface; a) Ton = 100 µs, b) Ton = 200 µs (WS = 550 RPM, DC = 0.80, I = 6 A, PC = 5 gm/l of Al Powder)

b) Fig. 14. Influence of wheel speed on PMEDDSG-generated surface; a) WS = 650 RPM, b) WS = 750 RPM (I = 6 A, DC = 0.80, Ton = 100 µs, PC = 5 gm/l of Al Powder)

Hence, increase in wheel speed is the reason for the formation of narrower and smaller sizes cavities on work surface. The effect of powder on the PMEDDSG of Ti-6Al-4V generated surface (with Al and without powder-mixed dielectric fluid) is depicted in Figs. 15a and b. It was observed from both images that surface cavities are smaller and smooth due to the presence of Al powder in the IEG as compared to without powder. This happened because the added powder in the dielectric fluid modified the plasma channel, i.e. it becomes larger and wider, and the spark energy is

equally distributed between the powder particles. Hence, the electrical density of spark discharge is reduced, which is responsible for the production of smaller cavities on work-surface. It is also observed that number of cavities on the machined surfaces increases with increase in pulseduration, duty cycle and current. Similarly, the number of cavities on the machined surfaces decreases with increase in wheel speed. A few cracks are developed on the machined surface when the stress induced by the PMEDDSG process exceeds the material tensile strength. A few pockmarks are developed on the

744

Modi, M. - Agarwal, G.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

b) Fig. 17. Influence of pulse on-duration on PMEDDSG-generated surface; a) Ton = 200 µs, t = 30 µm, b) Ton = 300 µs, t = 41 µm (I = 6 A, DC = 0.80, WS = 550 RPM, PC = 5 gm/l of Al powder)

Fig. 15. Influence of presence and absence of powder on PMEDDSG-generated surface; a) with Al powder, PC = 5 gm/l of Al powder, b) without powder (WS = 550 RPM, I = 2 A, DC = 0.80, Ton = 300 µs)

b) Fig. 16. Influence of current on PMEDDSG-generated surface; a) I = 6 A, t = 30 µm, b) I = 10 A, t = 49 µm (Ton = 200 µs, DC = 0.80, WS = 550 RPM, PC = 5 gm/l of Al powder)

b) Fig. 18. Influence of wheel speed on PMEDDSG-generated surface; a) WS = 550 RPM, t = 30 µm, b) WS = 750 RPM, t = 24 µm (I = 6 A, Ton = 200 µs, DC = 0.80, PC = 5 gm/l of Al powder)

machined surface due to escaping of trapped gases from the re-solidified material.

increases. The increased value of current is the reason for increase in the energy of each spark and hence raises the MRR. In addition, the eroded materials from the work surface are swept away by the dielectric and the greater volume of un-flushed molten substances in the IEG (inter electrode gap) have undergone the resolidification process and are finally deposited on the upper surface of the workpiece. From Figs. 17a and b, it is observed that WRLT increases with increase in the pulse on-length. The increased value of the pulse on-length is the reason

6 ANALYSIS OF WHITE RECAST LAYER THICKNESS (WRLT) If the molten material from work surface is not flushed out quickly, it will re-solidify due to the cooling effect of the dielectric fluid and adhere to the machined surface. This deposited layer is called the white recast layer. From Figs. 16a and b, it was observed that white recast layer thickness (WRLT) increases as the current

Powder-Mixed Electro-Discharge Diamond Surface Grinding Process: Modelling, Comparative Analysis and Multi-Output Optimisation ...

745

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

for greater thermal softening of the workpiece through heat conduction and hence increases the MRR. In addition, the eroded materials from the work surface are swept away by the dielectric and a greater volume of un-flushed molten substances in the IEG have undergone the re-solidification process and are finally deposited on the upper surface of the workpiece. From the Figs. 18a and b, it is observed that WRLT decreases with increase in wheel speed. The increase in wheel speed is the reason for efficient dielectric flushing in the IEG, which reduces the deposition of re-solidified ejected particles on a worksurface and also reduces the possibilities of arcing.

b) Fig. 19. Influence of duty cycle on PMEDDSG-generated surface; a) duty cycle = 0.80, t = 30 µm, b) duty cycle = 0.85, t = 39 µm (I = 6 A, Ton = 200 µs, WS = 550 RPM, PC = 5 gm/l of Al powder)

From the Figs. 19a and b, it is observed that WRLT increases with increasing the duty cycle. Pulse on-duration remains constant throughout the experiment and duty cycle increases due to decreased pulse off-time. This means less time is available for dielectric flushing to sweep the ejected particles from the IEG and for the deionization of the dielectric. Thus, continuous sparking takes place and a greater volume of un-flushed molten substances in IEG have undergone the re-solidification process and are finally deposited on the upper surface of the workpiece. 7 CONCLUSIONS The following conclusions can be drawn, based on the analysis of outcomes, interpretations of Figs. 2 to 9, the optimization approach, and analysis of SEM images: 746

1. MRR is greater in powder-mixed dielectric fluid in comparison to dielectric fluid without it, during the PMEDDSG of Ti-6Al-4V. Some authors, i.e. Choudhury et al. [2], Koshy et al. [3] and [4], obtained the same results in cut-off grinding mode with different workpiece materials. 2. Ra is less in powder-mixed dielectric fluid in comparison to dielectric fluid without it, during the PMEDDSG of Ti-6Al-4V. 3. The PMEDDSG is an efficient process for shaping the hard material. This process has the advantage of reducing the recast layer thickness by the grinding action due to abrasion. 4. The highest MRR is achieved when current, pulse on-time and wheel speed are at peak levels. Similarly, The highest MRR is achieved when the duty cycle is at the lowest level. 5. The highest Ra is achieved when current, pulse on-time and duty cycle are at peak levels. Similarly, The highest Ra is achieved when wheel speed is at the lowest level. 6. The best results (higher MRR and lower Ra) are obtained in the PMEDDSG of Ti-6Al-4V with Al powder-mixed dielectric fluid. 7. Surface cavities are wider and bigger with the increase in current, pulse-duration and duty cycle. Similarly, surface cavities are narrower and smaller with the increase in wheel-speed. 8. Surface cavities are smaller and smoother due to the presence of aluminium powder in comparison to dielectric fluid without it; this is confirmed from studying the SEM images. These images were captured from specimens/workpieces operated with Al powder (powder concentration = 5 gm/l) and without aluminium powder (powder concentration = zero gm/l) during PMEDDSG process of Ti-6Al-4V. 9. White-recast layer thickness is increased by increase in current, pulse on-length and duty cycle. Similarly, white-recast layer thickness is decreased by increase in wheel speed. 10. It has been observed that the number of cavities on the machined surfaces increases with increase in pulse-duration, duty cycle and current. It has also been found that number of cavities on the machined surfaces decreases with increase in wheel speed. A few cracks are developed on the machined surface when the stress induced by the PMEDDSG process exceeds the material tensile strength. A few pockmarks are made on the machined surface due to escaping of trapped gases from the re-solidified material.

Modi, M. - Agarwal, G.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 735-747

11. It is observed that based on WPC approach the improvement in MRR from 0.521 to 1.055 mm³/ min, deterioration in Ra from 2.21 to 4.76 µm and overall improvement in MOPI is 0.93 have been found during the multi-output optimization in PMEDDSG of Ti-6Al-4V with Al powder-mixed dielectric fluid. These results are obtained when the optimum process parameter combination (PC2I3Ton2WS3DC3) results were compared with an initial process parameter combination (PC1I1Ton1WS1DC1) results. 12. It is observed through ANOVA analysis in the WPC approach that the percentage contribution of various process parameters is powder concentration (19%), pulse on-time (1.5%), wheel speed (1.5%), duty cycle (7%) and current (61%) for the responses under the multioutput optimization (maximization of MRR and minimization of Ra) in the PMEDDSG of Ti-6Al4V. The most significant parameters that affect the PMEDDSG process are current and powder concentration. 8 NOMENCLATURE C [%] Contribution percentage CL Confidence level DC Duty cycle DOF Degree of freedom EDM Electrical discharge machining EDDG Electro-discharge diamond grinding EDDSG Electro-discharge diamond surface grinding IEG Inter electrode gap F F ratio K-AGENT Kroll’s agent Log S Value S/N Ratio SS / MS Sum of squares / Mean squares Ton [µs] Pulse on-time / Pulse-duration / POT WP / WOP With powder / Without powder WS Wheel speed 9 REFERENCES [1] Chen, S.L., Yan, B.H., Huang, F.Y. (1999). Influence of kerosene and distilled water as dielectrics on the electric discharge machining characteristics of Ti– 6A1–4V. Journal of Materials Processing Technology, vol. 87, no. 1-3, p. 107–111, DOI:10.1016/S09240136(98)00340-9. [2] Choudhury, S.K., Jain, V.K., Gupta, M. (1999). Electrical discharge diamond grinding of highspeed steel. Machining Science and Technology:

An International Journal, vol. 3, no.1, p. 91-105, DOI:10.1080/10940349908945685. [3] Koshy, P., Jain, V.K.,Lal, G.K. (1996). Mechanism of material removal in electrical discharge diamond grinding. International Journal of Machine Tools and Manufacture, vol. 36, no.10, p. 1173-1185, DOI:10.1016/0890-6955(95)00103-4. [4] Koshy, P., Jain, V.K., Lal, G.K. (1997). Grinding of cemented carbide with electrical spark assistance. Journal of Materials Processing Technology, vol. 72, no. 1, p. 61-68, DOI:10.1016/S0924-0136(97)00130-1. [5] Kansal, H.K., Sehijipal S., Kumar, P. (2005). Parametric optimization of powder mixed electrical discharge machining by response surface methodology. Journal of Materials Processing Technology, vol.169, no. 3, p. 427-436, DOI:10.1016/j.jmatprotec.2005.03.028. [6] Habib, S.S. (2009). Study of the parameters in electric discharge machining through response surface methodology approach. Applied Mathematical Modelling, vol. 33, no. 12, p. 4397-4407, DOI:10.1016/j.apm.2009.03.021. [7] Hewidy, M.S., El-Taweel, T.A., El-Safty, M.F. (2005). Modelling the machining parameters of wire electrical discharge machining of Inconel 601 using RSM, Journal of Materials Processing Technology, vol. 169, no. 2, p. 328-336, DOI:10.1016/j.jmatprotec.2005.04.078. [8] Lin, J.L., Lin, C.L. (2002). The use of orthogonal array with grey relational analysis to optimize the electrical discharge machining process with multiple performance characteristics. International Journal of Machine Tools & Manufacture, vol. 42, no. 2, p. 237244, DOI:10.1016/S0890-6955(01)00107-9. [9] Hotelling, H. (1933). Analysis of a complex of statistical variables into principal components. Journal of Educational Psychology, vol. 24, no. 6, p. 417, DOI:10.1037/h0071325. [10] Su, C.T., Tong, L.I. (1997). Multi-response robust design by principal component analysis, Total Quality Management, vol. 8, no. 6, p. 409-416, DOI:10.1080/0954412979415. [11] Antony, J. (2000). Multi-response optimization in industrial experiments using Tagchi’s quality loss function and principal component analysis, Quality and Reliability Engineering International, vol. 16, no. 1, p. 3-16, DOI:10.1002/(SICI)10991638(200001/02)16:1<3::AID-QRE276>3.0.CO;2-W. [12] Liao, H.-C. (2006). Multi-response optimization using weighted principal component, International Journal of Advanced Manufacturing Technology, vol. 27, p. 720-725, DOI:10.1007/s00170-004-2248-7. [13] Modi, M., Agarwal, G. (2012). Design, Development & Experimental Investigation of Electro-Discharge Diamond Surface Grinding of Ti-6Al-4V. International Journal of Advanced Materials Research, vol. 418420, p. 1478-1481, DOI:10.4028/www.scientific.net/ AMR.418-420.1478.

Powder-Mixed Electro-Discharge Diamond Surface Grinding Process: Modelling, Comparative Analysis and Multi-Output Optimisation ...

747

Received for review: 2013-02-22 Received revised form: 2013-05-15 Accepted for publication: 2013-09-05

Influence of Different Cooling and Lubrication Techniques on Material Machinability in Machining Sredanovic, B. – Globocki-Lakic, G. – Cica, D. – Kramar, D. Branislav Sredanovic1,* – Gordana Globocki-Lakic1 – Djordje Cica1 – Davorin Kramar2 1 University

of Banja Luka, Faculty of Mechanical Engineering, Bosnia and Herzegovina of Ljubljana, Faculty of Mechanical Engineering, Slovenia

2 University

In this paper, a novel approach to the definition of universal machinability is presented. The machinability model is based on analysing the vector of the cutting process performance. The machinability of C45E steel was analysed and evaluated according to the developed machinability definition. As the machinability criteria, cutting force, intensity of tool wear and surface roughness were used. Analysis of machinability was performed using different cooling and lubrication conditions: conventional flooding, minimum quantity lubrication (MQL) and high-pressure jet-assisted machining (HPJAM). Technological parameters were adjusted to the semi-finish regime, with the use of the highest parameter values possible. During the research, cutting forces, chip shapes, tool wear and surface roughness were monitored and measured. The influence of different cooling and lubrication techniques and the relationships between process performance indicators on C45E steel machinability are analysed. Keywords: machinability model, high-pressure jet assisted machining, cooling and lubrication

0 INTRODUCTION Material machinability is defined as the ease of material machining with different tools and conditions within economic regimes [1]. As a general definition of material machinability, the following is used: the greatest values of machinability are with those materials with which the smallest cutting forces, vibrations, tool wear and surface roughness processing occur [1]. The goal of machinability testing is finding the optimal regime of material processing and the effective management of the cutting process. The improvement of existing technologies and the developments of new cutting technologies through experimental research are ongoing tasks for researchers. Analysing machinability is a complicated task because of the large number of influencing factors. Such factors on the material machinability in cutting are the parameters that significantly affect the ease of processing: workpiece material, tool material and geometry, tool machine, processing type, cooling and lubrication fluids properties and delivery, machining condition, etc. The basic functions of cutting fluid are the reduction of the friction, the reduction of generated heat and the improvement of chip evacuation [2]. The positive effects on machinability are reflected in increasing tool life, reducing energy consumption, improving of the surface roughness, etc. The effect of cooling and lubricating depends on fluid flow parameters, fluid characteristics and delivery technique. 748

The necessity to machine using less harmful cutting fluids has prompted many researchers to investigate the use of minimum quantity lubrication (MQL). In MQL, lubrication is obtained via the lubricant, while a minimum cooling action is achieved by the pressurised air that reaches the tool/workpiece interface [3]. In MQL, however, secondary characteristics are important. These include their safety properties in environmental pollution and human contact as well as properties in biodegradability, oxidation and storage stability [4].

Fig. 1. Jet setup in HPJAM

Compared to conventional cooling, the concept of high-pressure jet-assisted machining (HPJAM) is to inject a high pressure jet of cutting fluid in the cutting zone. The jet is normally directed to the cutting edge at a low angle directly between the rake face and the chip (Fig. 1). HPJAM is beginning established as a method for a substantial increase of removal rate and productivity in machining of advanced materials such as hardened steels used for moulds and constructions,

*Corr. Author’s Address: University of Banjaluka, Faculty of Mechanical Engineering, Stepe Stepanovica 75, 78000 Banja Luka, Bosnia and Herzegovina, e-mail: sredanovic@gmail.com

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 748-754

Ti-based alloys used in gas turbines and in the aerospace industry. The advantage of this method is reduction of cutting temperature, reduction of tool wear, improvement of chip breakability, etc. [5] and [6]. In this paper, the influence of different cooling and lubrication techniques in turning process with coated carbide tool on steel C45E machinability is analyzed. 1 METHOD FOR UNIVERSAL MATERIAL MACHINABILITY DEFFINIG Material machinability is defined according to the main criteria. These criteria are the output parameters of the cutting process: tool wear, surface roughness, cutting forces, chip shape, etc., which depend on the input-cutting parameters [1]. Based on analyses of previous approaches of machinability definition, it can be concluded that machinability is defined by one criterion only. Machinability is defined via machinability indexes as the relative estimate of machinability, without units [1]. Some material may have considerable machinability according to one criterion, but exceptionally poor machinability when the other criteria are considered. A complete definition of machinability or absolute machinability is based on the consideration of many criteria simultaneously [1]. In [7], the author described Boulger’s specific test for machinability estimation based on process economy. In that paper, Lee’s machinability evaluation based on the thermodynamic number was described. Enache et al. [8] developed a “global model” of relative machinability, which contained more criteria. Rao and Gandhi [9] used the matrix and graph theory in a model developed for absolute machinability defining, termed the “matrix model”. In this paper, a novel method of machinability definition, based on vector analysis, is introduced. It considers several machinability criteria [10] and requires a three-axes rectangular coordinate system. Each axis in the system represents an aforementioned criterion (Fig. 2). The criterion values (Ki) are set on the corresponding axis. Criteria whose increasing values decrease the machinability are taken as inputs values in the model. Each axis has a corresponding unit vector ( kiko ) whose length corresponds to the unit value of criterion value (infinite length). The sum of unit vectors yields the control vector ( C ), whose direction determines the best uniformity of machinability criteria or the best machinability (Fig. 2).

Fig. 2. Machinability vector definition

Based on the criterion values and the corresponding unit vectors, the criterion vector that lies in the appropriate coordinate axis is formed as a scalar product of the unit vector and criterion vector:

K iKO = kiko ⋅ K i . (1)

The machinability vector is the sum of all individual criteria vectors:

n M j = ∑ K iKO . (2) i =1

In general, the material whose machinability vector has less intensity and closes smaller angles with a control vector has better machinability [11]. The size of the parallelogram area formed by the machinability vector and its control vector is related to the machinability of materials (Fig. 3.). It can be concluded: material with a smaller parallelogram surface area has better machinability in compare to material with a larger parallelogram surface area.

Fig. 3. Area of the parallelogram related to machinability of materials j and j+1

Influence of Different Cooling and Lubrication Techniques on Material Machinability in Machining

749

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 748-754

The size of parallelogram area can be expressed through the value of vector product between the machinability vector and control vector:

Pj = C × M j = C ⋅ M j ⋅ sin ∠ C , M j . (3)

(

)

It can be concluded that universal machinability (Mu) is inversely proportional to the area of parallelogram formed by the machinability vector and control vector. Based on established relationships, on Eqs. (3) and (4), it can be written in its final form: M u = Pj−1 =

Fig. 4. Experimental setup

 ( K + K 2 + K3 ) =  K12 + K 22 + K 32 ⋅ 1 − 12  K1 + K 22 + K 32 

(

)

−1

  . (4)  

In Eq. (4), as criteria values can be take the measured or calculated values. This values depend on input parameters, on example cutting force is F = f (ap, f, vc). Eq. (4) has two parts: the first part represents the influence of criteria values and the second part shows the influence of criteria values uniformity on machinability.

With the MQL technique, an external spray system with a was used. Pressure of pMQL = 0.3 MPa and a flow rate of QMQL = 0.0005 l·min-1 was set. The mixed jet was directed in the cutting zone with its nozzle located Hn = 30 mm away from tool tip, at an angle of νMQL = 90° of the cutting edge and at an angle of ψMQL = 30° from clearance face (Fig. 5).

2 EXPERIMENTAL SETUP In the experimental research, longitudinal turning operations with different types of cooling and lubrication techniques were used: conventional flooding, MQL and HPJAM. Experiments were conducted on a universal Boehringer lathe with the following characteristics: power of P = 8 kW, maximum spindle speed n = 2240 rev·min-1 and feed f = 1.6 mm·rev-1. An SNMG 1204 08 NMX carbide cutting tool for semi-machining was used. The clearance angle was 10°, and the rake angle 0°; the radius of the tool tip was 0.8 mm with a simple chip breaker. A PSDN 2525 M12 tool holder with an inclination angle of 45° was chosen (Fig. 4.). The workpiece material used in experimental research was C45E carbon steel. Experimental research was performed on a workpiece with dimensions of 150×250 mm. The tensile straight of this material is σ = 820 N·mm-², the module of elasticity is E = 2·10³ MPa, and the hardness of steel is 45 HRC. All experiments (over 108 experimental measurements with repeating) were carried out using cutting fluid with a 3% emulsion of vegetable oil. In conventional flooding, a free fall from above is used. A pressure of pCON = 0.3 MPa and flow rate of QCON = 2.0 l·min-1 was applied. 750

Fig. 5. MQL nozzle setup

In HPJAM, a conventional lathe was fitted with a Hammelmann high pressure plunger pump. The pressure was set at pHP = 50 MPa and the flow rate at QHP = 2.0 l·min-1. A standard sapphire orifice with a diameter of dHP = 0.4 mm (commonly used in waterjet cutting applications) was installed at a distance of Hn = 30 mm from the tool cutting edge in order to assure its use in the core zone of the jet and to avoid variations in the diameter of the jet and the radial distribution of the pressure. The jet was directed to the cutting edge (νHP = 90°), and ψHP = 30° from clearance face at a low angle (about 5 to 6°) with the tool rake face (Fig. 1). With the goal of achieving an effective material removal process, the cutting speeds and feeds were set higher than recommended. Turning was investigated according to the following input process parameters and levels:

Sredanovic, B. – Globocki-Lakic, G. – Cica, D. – Kramar, D.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 748-754

• • •

Depth of cut ap [mm]: 1.5, 2.0, 2.5. Feed f [mm·rev-1]: 0.224, 0.280, 0.355, 0.400. Cutting speed vc [m·min-1]: 210, 320, 400.

conditions for different depths of cut and feed rates is shown.

Fig. 6. Flank tool wear and chip shape (ap = 2 mm, f = 0.28 mm·rev-1, vc=320 m·min-1) for: a) conventional, b) MQL and c) HPJAM

During the experimental research, crucial process output parameters that describe cutting processes and material machinability were monitored: main cutting force (Fc), feed cutting force (Ff) and passive cutting force (Fp); flank tool wear land width (VBB) and intensity of tool wear (IVB); as the surface roughness parameters: arithmetic average height (Ra) and maximum height of the profile (Rz); and chip shape. 3 ANALYSIS OF EXPERIMENTAL RESULTS The best chip shape was obtained via HPJAM (Fig. 6) according the ISO classification [1]. According to its colour, it can be concluded that the impact of generated heat was reduced. Chip shape (conventional and MQL) was similar and was acceptable. Flank tool wear in the conventional and MQL at the end of tool life was extremely poor, as could be observed in the chip shape. In Fig. 7, a comparison of the three cutting force components values for turning in conventional flooding, MQL and HPJAM cooling and lubrication

Fig. 7. Values of cutting forces for different cooling and lubrication conditions in machining of C45E with vc = 320 [m·min-1]

From the diagrams (Fig. 7), it can be concluded that the values of the main cutting forces for all three cooling and lubrication condition are almost equal. The values of the feed and passive forces were the smallest in the case of HPJAM. In comparison, the values of the aforementioned forces in the MQL were smaller than in the conventional technique. The differences are greater for smaller cutting depths. Fig. 8 shows the results of tool wear on flank faces for different cooling and lubrication techniques. MQL and conventional techniques show almost the same values of tool wear, about T = 8 min, while

Influence of Different Cooling and Lubrication Techniques on Material Machinability in Machining

751

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 748-754

HPJAM had a four-times longer tool life. Other than this tool wear in HPJAM, conditions are uniform, while in conventional flooding and in MQL conditions some notch wear is recognized (Fig. 6). This is a result of intensive lubrication between the tool rake face and chip, chip deforming and evacuation in the case of HPJAM.

Fig. 8. Tool wear for different cooling and lubrication conditions

was analysed. Furthermore, a comparison with Enache’s global model [8] and Rao-Ghandi’s matrix model [9] and a comparison for different parameter input combinations were performed. In theory of machining, machinability evaluation values are relative values and there are no unit for it.

Fig. 10. Comparison of different machinability models for C45E (ap = 2 mm, f = 0.28 mm·rev-1, vc = 320 m·min-1)

In Fig. 10, a comparison of different material machinability evaluation models is shown. For calculation of machinability evaluation values, cutting force, intensity of flank tool wear and surface roughness were used. The order tendency of input combinations for the new model corresponds to previously developed models. It can be concluded that the output of the newly developed model correspond to the outputs of the matrix and the global model [11].

Fig. 9. Surface roughness parameters for different cooling and lubrication conditions

Values of surface roughness for different cooling and lubrication techniques are shown in Fig. 9. For conventional lubrication, surface finish is better than in case of other two techniques used, but surface roughness in MQL and HPJAM is also acceptable. Predictive regression models for monitored and measured machinability criteria values were developed, and a database was formed. The aforementioned prediction regression models take into account the cutting input parameters: workpiece material, depth of cut, feed, cutting speed, tool geometry, cooling and lubrication type, etc. [11]. According to the new theory described in this paper (the “vector model”), material machinability 752

Fig. 11. Comparison of different machinability models for C45E (f = 0.28 mm·rev-1, vc = 320 m·min-1)

Sredanovic, B. – Globocki-Lakic, G. – Cica, D. – Kramar, D.

Strojni拧ki vestnik - Journal of Mechanical Engineering 59(2013)12, 748-754

The machinability evaluation for combinations of investigated inputs is shown in Figs. 11 and 12. With increased feed and depth of cut, increasingly poorer machinability was calculated. The reason for this the highest values of cutting force, tool wear and surface roughness. HPJAM yields the best machinability during machining with the same cutting parameters, because of lower cutting force values and lower flank tool wear values.

Fig. 12. Machinability evaluation values for different conditions in machining of C45E (ap = 2 mm, vc = 320 m路min-1)

Fig. 13. Machinability evaluation values for different conditions in machining of C45E (vc = 320 m路min-1)

As shown by the experiments and calculations, turning with conventional flooding at a depth of cut of ap = 1.5 mm, MQL at a depth of cut of ap = 2 mm and with HPJAM at a depth of cut of ap = 2.5 mm has approximately the same universal machinability (Fig. 13). Based on the values of the developed model of universal machinability, it can be concluded that the

highest values of the process parameters can be used in HPJAM. The same conclusions about the machinability evaluation values for the studied cooling and lubrication techniques using the various criteria of machinability, such as cutting energy, productivity, chip shape, material removal rate (MMR), etc. can be achieved [12]. 4 CONCLUSIONS Machinability is an extremely significant factor in industry. It can be estimated by output cutting parameters / machinability criteria, which depend on input cutting parameters. Cutting force, intensity of flank tool wear and parameters of surface roughness were used as the machinability criteria. The newly developed machinability evaluation model shows considerable congruence with previously developed models [8] and [9]. Based on experimental research and using the new model, the machinability of different cooling lubrication techniques can be calculated. The analysis shows that turning with HPJAM provides the best machinability. Experiments were performed in the use of higher technological parameters values than in previous research and recommendation [5] and [6]. Based on experimental studies, it can be concluded that the tool life during HPJAM turning is four times higher than with MQL and conventional techniques. The same surface roughness was obtained in all three cooling and lubrication techniques. In HPJAM, feed and passive cutting forces are 5 to 10% lower than with MQL and conventional flooding. Moreover, HPJAM has significant contributions for chip breakability. Based on these facts, we can conclude that HPJAM yields the best results. The operational area for the investigated tool-workpiece combination can be expanded approximately 35 to 40% in comparison with conventional technology. This expansion is in the increasing of depth of cut, feed and cutting speed. Moreover, a study shown that MQL, as an ecological technology, has slightly better results than conventional lubrication technique. In the future research, developing of machinability evaluation model in terms of using greater a number of machinability criteria will be investigated. Also, properly inputs and its description in novel model will be investigated. Based on the experimental research, database of technology parameters will be developed.

Influence of Different Cooling and Lubrication Techniques on Material Machinability in Machining

753

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 748-754

5 ACKNOWLEDGMENTS This research is the result of a bilateral project BIBA/12-13-001 between Slovenia and Bosnia and Herzegovina with collaboration between Laboratory for Cutting from Faculty of Mechanical Engineering in Ljubljana and Laboratory for Cutting Technology from Faculty of Mechanical Engineering in Banja Luka. 6 REFERENCES [1] Grzesik, W. (2008). Advanced Machining Processes of Metallic Materials: Theory, Modelling and Application. Elsevier, Amsterdam. [2] Kopač, J. (2002). Cutting forces and their influence on the economics of machining. Strojniški vestnik Journal of Mechanical Engineering, vol. 48, no. 3, p. 121-132. [3] Dhar, N.R., Ahmed, M.T., Islam, S (2007). An experimental investigation on effect of minimum quantity lubrication (MQL) in machining AISI 1040 steel. International Journal of Machine Tools & Manufacture, vol. 47, p. 748-753, DOI:10.1016/j. ijmachtools.2006.09.017. [4] Leppert, T. (2011). Effect of cooling and lubrication conditions on surface topography and turning process of C45 steel. International Journal of Machine Tools & Manufacture, vol. 51, p. 120-126, DOI:10.1016/j. ijmachtools.2010.11.001. [5] Courbon, C., Kramar, D., Krajnik, P., Pusavec, F., Rech, J., Kopac, J. (2009). Investigation of machining performance in high-pressure jet assisted turning of Inconel 718: an experimental study. International

754

Journal of Machine Tools & Manufacture, vol. 49, no. 14, p. 1114-1125, DOI:10.1016/j. ijmachtools.2009.07.010. [6] Kramar, D., Kopac, J. (2009). High pressure cooling in the machining of hard-to-machine materials. Strojniški vestnik - Journal of Mechanical Engineering, vol. 55, no. 11, p. 685-694. [7] Theile, E.W., Kuding, K.J., Murphy, D.W., Soloway, G., Duffin, B. (1990). Comparative machinability of brasses, steel and aluminum alloy: CDA’s universal machinability index, Publication of CDA, New York. [8] Enache, S., Strâjescu, E., Opran, C., Minciu, C., Zamfirache M. (1995). Mathematical model for the establishment of material machinability. Annals of CIRP, vol. 44, no. 1, p. 79-82, DOI:10.1016/S00078506(07)62279-3. [9] Rao, R.V., Gandhi, O.P. (2002). Diagraph and matrix methods for machinability evaluation of works material. International Journal of Machine Tools & Manufacture, vol. 42, no. 2, p. 321-330. [10] Sredanovic, B., Globocki, L.G. (2010). Quality monitoring of production systems and processes in form of vector of power. Proceedings of 9th International Scientific and Practical Conference: Research, Development and Application High Technologies in Industry, Saint Petersburg, p. 418-420. [11] Globocki, L.G., Sredanovic, B., Nedic, B., Cica, Dj., Catic, D. (2011). Development of mathematical model of universal material machinability. Journal of the Balkan Tribological Association, vol. 17, no. 4, p. 501511. [12] Kramar, D., Sredanović, B., Globočki, L. G., Kopač, J. (2012). Contribution to material machinability definition. Journal of Production Engineering, vol. 15, no. 2, p. 27-32.

Sredanovic, B. – Globocki-Lakic, G. – Cica, D. – Kramar, D.

Received for review: 2013-03-26 Received revised form: 2013-05-07 Accepted for publication: 2013-05-29

An Experimental and Numerical Investigation of a Grid Composite Cylindrical Shell Subjected to Transverse Loading Arashmehr, J. – Rahimi, G.H. – Seyed Fazel Rasouli, S.F. Jafar Arashmehr – Gholam Hossein Rahimi – Seyed Fazel Rasouli Tarbiat Modares University, Department of Mechanical Engineering, Iran This paper presents an experimental and numerical study of a stiffened composite cylindrical shell with clamped-free boundary condition under transverse end loading. Electrical strain gauges were employed to measure the strains. Likewise a finite element analysis has been used to validate the experimental results. The finite element method (FEM) software used was ANSYS11. Fairly good agreements were observed between the numerical and the measured results. In addition, a comparison between the stiffened and unstiffened composite shell was carried out. It was found that the intersection of two stiffeners has an important effect in reducing Von Mises stress in the shell. Failure analysis based on Tsai–Wu failure theory was also performed. It was concluded that stiffening of the shell by helical stiffeners increases the load capacity considerably. Keywords: grid composite structure, transverse loading, stress analysis, finite element analysis

0 INTRODUCTION Composite materials are desirable in lightweight structures due to their high specific stiffness and strength. These materials have great importance in the aerospace and aircraft industries. The main reasons for the success of composite materials use in the aerospace industry are the high performance versus weight ratios that it is possible to achieve using these materials and the necessity of lowering the weight of the aircraft as much as possible in order to enhance its efficiency [1]. Use of stiffeners can lead to higher load resistance of a composite structure without much increase in weight [2]. The grid composite structure is made of stiffeners and skin and the stiffeners can be made of other materials like metal alloys [3]. The stiffening structures are either on the inner or the outer or both sides of the skin. Recent research on grid composite structures has tended to concentrate on the axial loading while the structures are exposed to transverse loadings as well as axial loadings. This investigation attempts to study the grid composite cylindrical shell under transverse loadings, both experimentally and numerically. In recent years, a series of studies on the grid composite cylindrical shell have been carried out by Rahimi et. al. [4] to [6]. In the present study, a stress analysis of the structure under a clamped-free boundary condition subjected to transverse end load has been carried out both experimentally and numerically. Following this, a finite element modeling was utilized to verify the experimental result. To our knowledge, no investigation of this kind has been previously reported.

1 EXPERIMENTAL STUDY 1.1 Fabrication Process The specimen in this study was fabricated using a specially designed filament winding machine designed and built in Tarbiat Modares University as shown in Fig. 1. This machine executes a continuous procedure to fabricate the specimen in which the ribs are first made, the skin is immediately wound around it, and then the specimen is cured. This process ensures a solid fusion of the ribs to the skin. E-glass fibers and a room-temperature-curing resin epoxy matrix were used for the fabrication. The nominal mechanical properties of the material are presented in Table 1 [5].

Fig. 1. Specimen fabrication using a filament winding process [4] Table 1. Nominal material properties Material properties Longitudinal modulus Transverse modulus Transverse modulus Shear modulus Poisson’s ratio

Symbol E11 E22 E33 G12 ν12

Unit [GPa] [GPa] [GPa] [GPa] [-]

*Corr. Author’s Address: Tarbiat Modares University, Department of Mechanical Engineering, PO box 14115-143, Tehran, Iran, arashmehr@alum.sharif.edu

E-glass/Epoxy 36 5.8 5.8 3.22 0.3

755

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 755-762

The geometrical properties of the stiffened composite cylindrical shell are shown in Table 2: Table 2. Geometrical properties of the stiffened composite cylindrical shell Geometrical properties Length Diameter Number of ribs Cross section’s dimensions of ribs Thickness of shell

Value 31 cm 14 cm 6 6×6 mm 0.8 mm

Fig. 3. Sketch of the specimen mounted on the fixture

The ribs form a 30 degrees angle with the axis of revolution of the cylindrical shell. It should be noted that the shape of the cross section of the stiffeners can change the results [7]. The fabricated specimen is shown in Fig. 2.

Fig. 4. Sketch of the bonding of the strain gauges on the top of the specimen

Fig. 2. A sample of the fabricated specimens

1.2 Testing An L-shaped steel fixture was designed to mount the specimen on the testing machine. The specimen was installed in a circular groove created in the fixture. The steel sheet attached to the free end of the specimen provided distribution of the load onto the circumference of the edge of the shell and prevented local damage due to concentrated load. The stiffened shell was tested under clamped-free boundary conditions, with the load transversely exerted on the free end. Fig. 3 shows a sketch of the test. For strain measurement, 120 Ω strain gauges were employed [8]. Each gage must be bonded carefully to the structure. The strain gauges were attached to the top face of the specimen at four points (Three strain gauges at each point) in a rosette arrangement as shown in Fig. 4. 756

An INSTRON test machine was used to apply the load in a displacement-controlled procedure with a rate of 1 mm/min [9]. Fig. 5 shows the stiffened composite shell during the test. The loading was continued until the structure failed from the clamped end. It should be mentioned that a marker was drawn under the load cell and on the steel sheet to see if there was any rotation in the shell or not. It was observed after the test that the marker did not have any horizontal displacement with respect to the load cell, which means that no rotation was found.

Fig. 5. The specimen under loading

Arashmehr, J. – Rahimi, G.H. – Seyed Fazel Rasouli, S.F.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 755-762

2. FINITE ELEMENT MODELING

3 RESULTS

ANSYS 11.0 software was employed to make a 3-D model of the structure [10]. The intersections of the ribs were modeled as distinct volumes with twice the fiber volume fraction as the ribs due to accumulation of fibers in these areas during the fabrication process. The stiffeners were meshed using 20-node layered solid elements, SOLID191. Since the fibers are unidirectional within the ribs, the number of layers would not make any difference to the results. SHELL99 was used to mesh the shell [7]. It was found to be adequate to assume the shell to be a 20-ply laminate after a divergence test. The nodes of the shell and the rib elements within a tolerance of 0.6 millimeters were coupled together in order to stick the shell and the ribs together. The boundary conditions, including clamped-free and transverse end loading, were applied to the model. It should be noted that the weight of the steel sheet attached to the free end (2 kg), was considered in the FEM analysis. The created FEM model is presented in Fig. 6.

The obtained results in 400 N of the experimental and numerical studies are shown in Table 3. The number of each strain gauge is shown in Fig. 4. There is a good accordance between the results for both approaches. Fig. 7 shows the variations in the displacement of the free end versus the applied transverse end load obtained in the experiment. Fig. 7 indicates 1700 N of load capacity for this structure under the mentioned boundary conditions. The behaviour of this structure changes linearly until approximately 1400 N. After a 7 mm deflection of the free end, the first reduction in the load variation is observed. This is the time that buckling or collapse takes place in the structure. In the failure analysis part of this paper, it will be illustrated that the collapse in the composite shell occurs in this point. Following this, the load increases again up to the maximum load capacity.

Fig. 7. Deflection of the free end versus the applied load

In this experiment εx , εy , were measured. By using Eq. (1), the γxy is obtained: Fig. 6. The FEM Model

ε (θ ) = ε x cos 2 θ + ε y sin 2 θ + γ xy sin θ cosθ , (1)

therefore

Static analysis was conducted to extract the strains and stresses along the top face of the stiffened shell.

γ xy =

ε (θ ) − ε x cos 2 θ − ε y sin 2 θ . (2) sin θ cosθ

Table 3. Result of strains in experimental and numerical studies Number of strain gauge Experimental (micro strain) FEM (micro strain) Difference [%]

1 867 729 19

2 56 46 17

3 19 14 26

4 1412 1208 16

5 92 69 25

6 103 79 23

7 149 122 18

8 -171 -133 22

9 -232 -187 24

An Experimental and Numerical Investigation of a Grid Composite Cylindrical Shell Subjected to Transverse Loading

10 3022 2387 21

11 -283 -232 18

12 18 14 22

757

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 755-762

3.1 Stress Distribution Considering that the strain gauges show the strains in the outer layer of the shell [11], and the fact that the shell is wound at a 75 degrees angle, a coordinate system transformation should be done prior to using stress-strain relations. This is shown in Fig. 8.

linear zone. The stresses in the experimental study are obtained from Eq. (3). 3.2 Experimental versus Finite Element Results A comparison of the variation in σx versus the distance from the free end between the experimental and finite element results is shown in Fig. 9:

Fig. 8. Global and material coordinate systems

σ x   Q11 Q12    σ y  = Q12 Q22 τ xy  Q Q    16 26

Q16   ε x    Q26   ε y  . (3)  Q66  γ xy 

Where the transformed reduced stiffness Qij is given in terms of reduced stiffness Qij in Eq. (4). Q11 = n Q11 + 2m n ( Q12 + 2Q66 ) + m Q22 , 4

(

)

Q12 = n 2 m 2 ( Q11 + Q22 − 4Q66 ) + m 4 + n 4 Q12 ,

(

)

Q16 = mn n 2 ( Q11 − Q12 − 2Q66 ) + nm3 ( Q12 − Q22 + 2Q66 ) ,

Q22 = m 4Q11 + 2m 2 n 2 ( Q12 + 2Q66 ) + n 4Q22 , (4) Q26 = m3 ( n ( Q11 − Q12 − 2Q66 ) ) + mn3 ( Q12 − Q22 + 2Q66 ) ,

(

)

Q66 = n 2 m 2 ( Q11 − 2Q12 − 2Q66 + Q22 ) + n 2 + m 2 Q66 , where m = sin θ , n = cos θ and θ is the angle of the fibers and the axis of revolution of the shell. For the orthotropic lamina, Qij are: E1 ν E Q11 = , Q12 = 12 2 = Q21 , 1 −ν 12ν 21 1 −ν 12ν 21 (5) E2 Q22 = , Q66 = G12 . 1 −ν 12ν 21 It should be noted that the stresses were calculated at 400 N. Looking back at the load-displacement graph, it is clear that at this load the structure is in the 758

Fig. 9. σx - Finite element vs. experimental result comparison

Fig. 9 shows that at a distance of 23 cm from the free end of the cylinder which is the intersection point of the two ribs, the σx considerably decreases in comparison with other points. It was also found that the values of the σy , τxy decrease noticeably at the intersection point of the two stiffeners as well as at σx . The von Mises equivalent stress for plane-stress is obtained from Eq. (5) and the values of this stress for experimental and numerical analyses are compared in Fig. 10.

σ vonMises =

(σ x − σ y ) 2 + σ x 2 + σ y 2 + 6τ xy 2 2

. (6)

It is evident from the graph that the equivalent von Mises stress declines significantly at the intersection point of the stiffeners. The deformed shape of the structure in the FEM analysis indicates that the decrease in the stresses at the abovementioned point is due to the deformation in the stiffened shell under the transverse end load. Fig. 11 shows the large scale deformed shape of the specimen.

Arashmehr, J. – Rahimi, G.H. – Seyed Fazel Rasouli, S.F.

Strojniﾅ｡ki vestnik - Journal of Mechanical Engineering 59(2013)12, 755-762

To determine the effect of stiffening of a cylindrical composite shell, the stiffened composite shell is compared to an unstiffened composite shell. 3.3 Stiffened Shell vs. Unstiffened Shell Result

Fig. 10. ﾏプonMises - Finite element vs. experimental result comparison

The elements on the top face of the structure are expected to be in tension. As shown in Fig. 11, the deformation near the junction of two stiffeners on the top of the structure is remarkably less than that seen in other parts of the structure. In addition the contour of stress distribution in the stiffened shell illustrates a significant decrease in von Mises stress at this point (Fig. 12).

In this section, the obtained FE results are compared to the results of an unstiffened shell under the same boundary conditions. Therefore, a finite element model of an unstiffened composite shell with the same material properties was built [13]. Fig. 13 shows a comparison of the variations in the ﾏプonMises between the stiffened shell and the unstiffened shell under the same boundary conditions. It is concluded that the greatest difference in von Mises stress is at a distance 23 cm from the free end of the structure. This is where the junction of the two stiffeners is located in the stiffened shell. The above graph shows lower values of von Mises stress in the stiffened shell compared to the unstiffened shell.

Fig. 13. ﾏプonMises - stiffened vs. unstiffened composite shell

4 FAILURE ANALYSIS Fig. 11. Deformation of the structure

Fig. 12. von Mises stress distribution

Grid composite structures have different failure modes that strongly depend on the type of loading and boundary conditions. The stiffened composite shell may fail due to some mechanism such as collapse of the stiffeners or skin. It was observed from the experiments that the stiffened composite cylindrical shell would finally fail in the clamped end. This failure mode occurs in the shell. Fig. 14 shows the failure of the structure. In addition, no other failure modes in other parts of the stiffeners or the shell were observed. However, other failure modes like collapse of the stiffener are

An Experimental and Numerical Investigation of a Grid Composite Cylindrical Shell Subjected to Transverse Loading

759

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 755-762

possible in other types of loading, such as local loads [14]. The Tsai–Wu failure theory failure theory is used for the failure analysis [15]. Two methods were applied to the failure analysis. In the first approach, the experimental results near the clamped end and the Tsai–Wu failure theory are employed. The values of σx , σy and τxy are obtained from Eq. (7). These stresses are in a global coordinate system, which is needed to calculate them in the material coordinate system.

H1 =

H2 =

T 1 ult

(σ )

−

, H11 =

C 1 ult

(σ )

1 , (σ ) (σ 1C )ult T 1 ult

1 1 1 − , H 22 = , (σ 2T )ult (σ 2C )ult (σ 2T )ult (σ 2C )ult

H 6 = 0, H12 = −

1 1 , T C 2 (σ 1 )ult (σ 1 )ult (σ 2T )ult (σ 2C )ult

H 66 =

1 . (10) (τ 12 ) 2ult

The ultimate strengths of the e-glass fibers is presented in Table 4: [17] Table 4. Properties of unidirectional e-glass fibers

Fig . 14. Failure of the structure in the clamped end under a transverse end load

The global and local stresses in an angle lamina are related to each other through the angle of the lamina, θ: σ x  σ 1  σ  = T σ  ,  2  [ ]  y  (7) τ xy  τ 12   

where [T] is defined as:  cos 2 θ  [T ] =  sin 2 θ  − sinθ cosθ 

sin θ cos 2 θ sinθ cosθ

2 sinθ cosθ   −2 sinθ cosθ  , (8) cos 2 θ − sin 2 θ 

where θ = 75°. Therefore the results of point 4 which is near the clamped end are used for the calculation of the required stresses in the material coordinate system. The Tsai–Wu failure theory is defined as: H1σ 1 + H 2σ 2 + H 6τ 12 + H11σ 12 +

+ H 22σ 2 2 + H 66τ 12 2 + 2 H12σ 1σ 22 < 1, (9)

where [16] 760

Parameter

Unit

Value

[MPa]

1080

[MPa]

620

[MPa]

128

[MPa]

[-]

0.028

[-]

0.005

(σ1 ult (σ2T)ult (σ1C)ult (σ2C)ult (τ12)ult (ε1T)ult (ε2T)ult

Table 5. Failure analysis results Tsai–Wu failure theory index FEM EXP FEM EXP FEM EXP FEM EXP 400 6.4 5.24 13.17 10.53 0.26 0.2 0.222 0.196 800 12.81 10.76 26.53 21.75 0.486 0.383 0.495 0.460 1450 22.92 19.71 47.4 39.34 0.78 0.631 1.021 0.952 Load [N]

σ1

[MPa]

σ2

[MPa]

τ12

[MPa]

In the second step, finite element failure analysis is carried out based on the Tsai–Wu failure theory [18]. The failure criterion was defined using the ultimate strengths in Table 4. Both methods were carried out for three different loads and the results are presented in Table 5. The results show a good agreement between both methods. It should be noted that FEM analysis also shows that the stiffened composite cylindrical shell will fail at the clamped edge as observed in the experiments. The failure load of the structure was found to be 1465 N in the experimental methods and 1437 N in the FEM failure analysis. On the other hand in the Load-Displacement graph (Fig. 7), at 1394 N the load decreases temporarily and

Arashmehr, J. – Rahimi, G.H. – Seyed Fazel Rasouli, S.F.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 755-762

the increases again. This is where the failure occurs in the clamped end of the stiffened shell. Afterwards, FEM failure analysis was carried out for an unstiffened shell using the same boundary conditions. The result clarified that the unstiffened shell fails in the clamped end similarly to the stiffened shell. The failure load for the unstiffened composite shell was found to be 1215 N, which is about 17% lower than the load capacity of the stiffened composite shell. 5 CONCLUSION In this study, a stress analysis for a stiffened composite cylindrical shell under clamped-free boundary conditions subjected to a transverse end load was carried out using experimental and numerical methods. A failure analysis for the structure was also performed and the results were compared to the unstiffened composite shell. The most important conclusions are listed below: • The effect of the intersection of the stiffeners can be used in some parts of the structure where a high level of stress is expected. For example this effect would be useful around areas where stress concentration exists. • Under this load and these boundary conditions, the most important concern is related to the areas near the clamped end. In other words, the failure probability at this point is far higher than other points in the structure. This structure also protected against other failure modes including separation between the ribs and skin or collapse of the ribs. • The failure analysis showed a 17% higher load capacity in the stiffened composite shell compared to the unstiffened shell, which is undoubtedly a useful point in the design of these structures. 6 APPENDIX Specimen Fabrication. The process of fabrication begins with laying the e-glass fibers within the predesigned grooves in the mandrel until reaching the desired number of fiber threads within a single groove, which was 100 threads for the specimens in this paper. The fibers are therefore uni-directional along the grooves (or eventually the stiffeners). After the grooves of the ribs were filled with the desired volume of fibers and resin, the shell was immediately wound as the mandrel rotated and also reciprocated and a stationary filler fed the fibers. The rotational

and reciprocal speed of the mandrel was adjusted to obtain 72 degree fiber angles with respect to the axis of rotation of the cylinder. The process cannot define a definite number of layers for the fibers; therefore, the number of layers in the FEA was defined by a convergence test, which means that the number of layers were increased until the results reached an approximately constant point (20 layers for the shell and 1 layer for the fibers, since they are unidirectional). The fiber volume fraction was designed to be approximately 30%. It should be mentioned that at the intersection of the stiffeners there are twice as many fibers as other sections of the stiffeners, which was modeled by doubling the volume fraction in the FEA model. The final geometrical properties are given in Table 2. 7 REFERENCES [1] Jaunky, N., Knight, N., Ambur, D. (1998). Optimal design of general stiffened composite circular cylinders for global buckling with strength constraints. Journal of Composite Structures, vol. 41, no. 3-4, p. 243-252, DOI:10.1016/S0263-8223(98)00020-8. [2] Wodesenbet, N., Kidane, S. (2003). Optimization for buckling loads of Grid stiffened composite panels. Journal of Composite structures, vol. 60, no. 2, p. 159169, DOI:10.1016/S0263-8223(02)00315-X. [3] Huo, S., Yuan, Z., Wang, F., Yue, Z., Liu, Y. (2013). Effect of static aeroelasticity on composite wing characteristics under different flight attitudes. Journal of Central South University,vol. 20, no. 2, p. 312-317, DOI:10.1007/s11771-013-1489-8. [4] Yazdani, M., Rahimi, G.H. (2009). An experimental investigation into the Buckling of GFRP stiffened shells under axial loading. Academic Journal, vol. 4, no. 9, p. 914-920. [5] Rahimi, G.H., Yazdani, M. (2010). The effects of helical ribs’ number and grid types on the buckling of thin-walled GFRP-stiffened shells under axial loading. Journal of Reinforced Plastics and Composites, vol. 29, no. 17, p. 2568-2575, DOI:10.1177/0731684409355202. [6] Yazdani, M., Rahimi, G.H. (2011). The behavior of GFRP-stiffened and -unstiffened shells under cyclic axial loading and unloading. Journal of Reinforced Plastics and Composites, vol. 30, no. 5, p. 440-445, DOI:10.1177/0731684411398537. [7] Rahimi, G.H., Zandi, M., Rasouli, S.F. (2013). Analysis of the effect of stiffener profile on buckling strength in composite isogrid stiffened shell under axial loading. Journal of Aerospace Science and Technology, vol. 24, no. 1, p. 198-203, DOI:10.1016/j.ast.2011.11.007. [8] Arashmehr, J., Rahimi, G.H. (2012). Experimental investigation of clamped stiffened composite cylindrical shells under end load. Experimental Solid

An Experimental and Numerical Investigation of a Grid Composite Cylindrical Shell Subjected to Transverse Loading

761

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 755-762

Mechanic and Dynamics Conference (xmech 2012), Tehran. [9] Zheng, Z., Wang, J., Mi, L. (2010). Preparation of new tissue engineering bone-CPC/PLGA composite and its application in animal bone defects. Journal of Central South University, vol. 17, p. 202-210, DOI:10.1007/ s11771-010-0031-5. [10] Movaghghar, A., Ivanovich Lvov, G. (2012). Theoretical and experimental study of fatigue strength of plain woven glass/epoxy composite. Strojniški vestnik - Journal of Mechanical Engineering, vol. 58, no. 3, p. 175-182, DOI:10.5545/sv-jme.2011.135. [11] Li, G. (2007). Experimental study of hybrid composite cylinders. Composite Structures, vol. 78, no. 2, p. 170181, DOI:10.1016/j.compstruct.2005.08.028. [12] Jones, R. (1999). Machromechanical behavior of a lamina a lamina. Mechanics of Composite Materials, Taylor&Francis, Portland, p. 75-77. [13] Arashmehr, J., Rahimi, G.H., Rasouli, S.F. (2012). Numerical and experimental stress analysis of stiffened cylindrical composite shell under transverse end load.

762

Journal of World Academy of Science, Engineering and Technology, vol. 67, p. 1298-1302. [14] Arashmehr, J., Rahimi, G.H. (2012). Investigation of grid composite cylindrical shell under local load. 11th Conference of Iranian Aerospace Society, Tehran. [15] Ilić, I., Petrovic, Z., Maksimović, M. (2012). Computation method in failure analysis of mechanically fastened joints at layered composites. Strojniški vestnik – Journal of Mechanical Engineering, vol. 58, no. 9, p. 553-559, DOI:10.5545/sv-jme.2011.157. [16] Kaw, A. (2006). Micromechanical analysis of a lamina. Mechanics of Composite Materials, Taylor & Francis, Portland, p. 137-158. [17] Daniel, I. (1994). Elastic behavior of unidirectional lamina, Engineering Mechanic of Composite Materials, Oxford University, New York, p. 34 [18] Jadhav, P., Mantena, P. (2007). Parametric optimization of grid-stiffened composite panels for maximizing their performance under transverse loading. Composite Structures, vol. 77, p. 353-363, DOI:10.1016/j. compstruct.2005.07.015.

Arashmehr, J. – Rahimi, G.H. – Seyed Fazel Rasouli, S.F.

Received for review: 2013-03-11 Received revised form: 2013-06-13 Accepted for publication: 2013-08-23

Research on Levitation Coupled with Standing Wave Levitation and Electromagnetic Levitation Jiao, X.Y. – Liu, G.J. – Liu, J.F. – Li, X.B. – Liu, X.L. – Lu, S. XiaoYang Jiao1 – GuoJun Liu1 – JianFang Liu1 – Xinbo Li2,* – XiaoLun Liu1 – Song Lu1 1 Jilin

University, College of Mechanical Science and Engineering, China University, School of Communication Engineering, China

2 Jilin

In order to solve the problem caused by metal materials' inability to be cooled without contact with other materials after being heated by electromagnetic levitation, a new method is proposed: using a standing wave levitator to levitate the melted metal. The standing wave levitator adopts a concave spherical surface on the emitter and the reflector. Using ANSYS software, the transducer and the standing wave fields were simulated. Based on the simulation, the distribution and the maximum acoustic pressure with different radii of the concave spherical surface on the emitter and the reflector can be obtained, from which the optimal radius was determined. Based on the optimisation, a prototype of a standing wave levitation device was designed and manufactured. Levitation experiments for light and heavy specimens were carried out. It is shown that steel balls can be levitated stably when the distance between the emitter and the reflector is two times that of the wavelength. Next, the standing wave levitator was used in an attempt to levitate a steel ball of 5 mm in diameter after being non-contact heated by electromagnetic levitation. The results show that the method utilising a standing wave levitator to levitate and cool the metal materials after being non-contact heated by the electromagnetic levitation is feasible at this preliminary state. Keywords: standing wave levitation, ANSYS simulation, electromagnetic levitation, non-contact cooling

0 INTRODUCTION Outer space environments can provide experimental conditions for high vacuum, non-contact and microgravity testing, which is appropriate for material solidification to study various kinds of fluid phenomenon [1] and [2]. However, it is far too costly for most researchers to perform space experiments. Consequently, many methods and technologies have been developed to simulate the outer space environment on earth, including acoustic levitation [3], electromagnetic levitation [4], aerodynamic levitation [5], electrostatic levitation [6], optical levitation [7], magnetic levitation [8] and superconducting magnetic levitation [9]. Among these methods and technologies, acoustic levitation has its own distinct advantages, such as good stability, simple construction, no special requirements on the levitated materials, etc. Therefore, research on acoustic levitation is attracting the interest of increasing numbers of researchers. The principle of acoustic levitation is based on acoustic radiation pressure generated in a highly intense ultrasonic acoustic field, which produces a levitation force to overcome the suspended sample’s gravity. In linear acoustics, the sound pressure varies periodically with time and its mean time in a cycle time is zero. However, in highly intense acoustic fields, the nonlinear effect becomes much more significant. In essence, the gravitational force is counteracted by a steady-state acoustic radiation force, as a result of the nonlinear effect in the laminar flow around the levitated specimen when its size is much smaller than

the wavelength of the ultrasound [10]. The levitation force is so large that common solids and liquids of densities thousands of times greater than air can be suspended in air [11]. There is no additional effect on the sample levitated by acoustic levitation force, which in principle can levitate any substance. In areas such as micro-machine technology, biotechnology, and the processing of new kinds of material, the research and application of acoustic levitation has become increasingly important [12]. Acoustic levitation is classified as near-field levitation and standing wave levitation according to the mechanism of ultrasonic levitation. The levitation height of the specimen suspended by the near-field is in the scale of micrometres, and its levitation force is so large that it can reach tens of kilograms. Furthermore, near-field levitation has been successfully applied to transport objects and rotate rotors without contact [13]. Unlike near-field levitation, a reflector is a part of a standing wave levitation device. The sound waves reflected by the reflector interfere with the forward waves produced by the emitting surface, and standing waves are generated when the distance between the emitting surface and the reflecting surface is an integer multiple of half a wavelength. Compared with near-field levitation, standing wave levitation has more levitation space. However, its levitation force becomes much smaller, and the diameter of the levitated sample can be reduced to millimetre scale. Electromagnetic levitation technology not only levitates metal samples, but can also heat and melt

*Corr. Author’s Address: Jilin University, School of Communication Engineering, Renmin Road 5988, Changchun, China, xb_li@jlu.edu.cn

763

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 763-771

them, preventing other materials contacting with the samples. Therefore, pure metal materials can be obtained. However, the power supply applied to the electromagnetic coils must be turned off when the heating is not needed, and the metal sample needs to cool down. The levitation force on the metal sample also disappears, and the levitated samples are in contact with other materials, which also contaminates them. Several methods have been proposed to solve this problem, such as filling space around the electromagnetic coils with cool inert gases [14], using high drop tubes [15], cold-crucible induction furnaces [16] and so on. This paper proposes a new kind of method using a standing wave levitator to suspend the metal after heated by the electromagnetic levitation. When the metal sample is levitated and heated by the electromagnetic coils, little or no force from the standing wave levitator is applied on the sample. When the power supply of the electromagnetic coils is turned off, the gravity of the metal is entirely overcome by the standing wave levitation force. The standing wave levitator is usually made by steel, and it also can be easily induced and heated by the electromagnetic coils. Therefore, it is necessary to improve the levitation force and increase the distance between the emitter and the reflector for the standing wave levitator. In order to improve the levitation force of standing wave levitation, researchers throughout the world have made many significant attempts. Barmatz [17] designed and analysed different resonance chambers of rectangular, spherical and cylindrical geometries. Magill [18] filled the resonance chamber with high pressure gas, and a tungsten carbide sample with a density of 15 g/cm3 was successfully levitated. Lee [19] arranged multiple transducers in arrays, which levitated a gold particle. Xie and Wei [20] and [21] designed standing wave levitation device with reflector of concave spherical surface, and an Iridium sphere was stably suspended in the standing wave field. Different kinds of liquids were also experimented with as reflectors [22]. A gas stream was used to balance the gravity of the levitating particle whose stability was controlled via a three-axis acoustic levitator [23] and [24]. A piezoelectric transducer with a concave emitting surface and a concave reflector was presented [25] and [26], which increased the lateral forces and reduced the lateral oscillations of the levitated object significantly, compared with the traditional single-axis acoustic levitator. In order to further optimise the acoustic levitator and increase the distance between emitter and 764

reflector, the emitter and the reflector with different concave radii were analysed via ANSYS, and the pressures of the standing wave field under different parameters are compared in this paper. Based on the optimisation results, the prototype of the levitator with a concave spherical surface on the emitter and reflector was designed and manufactured. In the lab, levitation experiments were carried out using foam and steel balls. Finally, the standing wave levitator was used to suspend the metal materials after being non-contact heated by the electromagnetic coils. 1 THEORETICAL ANALYSIS OF STRUCTURAL MODE 1.1 Structural Mode Based on the principle of focused ultrasound, a concave spherical reflecting surface can improve the levitation force of a standing wave levitator and decrease wave diffusion during the delivery process. Therefore, the emitter and reflector of the standing wave levitator are designed as concave spherical surfaces, as is shown in Fig. 1. For the sake of simplification during analysis, an assumption is made that the diameter of the concave spherical reflecting surface is the same as that of the emitting surface.

Fig. 1. Structural model

Steel balls can be stably levitated and heated in the electromagnetic coils without contact. The schematic diagram of the coils is shown in Fig. 2. The direction of the electric current in the upper stabilising coils is in opposition to that in the lower levitated coils. The upper coils play the main role of keeping the levitated sample stable and the lower coils provide the levitating force. Generally, electric current with a frequency of tens of thousands hertz flows through the coils, producing alternating electromagnetic fields in the space coated by the coils. Metal samples can

Jiao, X.Y. – Liu, G.J. – Liu, J.F. – Li, X.B. – Liu, X.L. – Lu, S.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 763-771

be induced by the alternating electromagnetic field, which generates induced eddies on the surface of the metal sample. In reverse, the induced eddy acts with the electromagnetic field producing the levitation force on the metal sample, and the levitated force can overcome the gravity of the metal sample. At the same time, there is significant heating effect on the metal sample with the induced eddy. Thus, the metal sample can be levitated and heated without contact with other materials. However, when heating the metal sample is completed and needs to be cooled, the electric current flowing through the coils must be cut off. The heating effect disappears while the levitated force also vanishes, and the metal sample will drop downward. Therefore, the metal sample with a high temperature comes in touch with other materials that may cause contamination. In this paper, a standing wave levitator was used in an attempt to suspend the metal sample after it was levitated and heated by the electromagnetic coils. The schematic diagram is shown in Fig. 3.

1.2 Vibration Analysis of Concave Spherical Emitting Surface A sandwich-type structure is generally used in a piezoelectric transducer, which produces a stretching vibration in the axial direction. When the emitting surface is planar, the emitter vibration excites the air, and planar waves are then formed. After reaching the reflecting surface, the planar waves will be reflected. Finally, the planar standing waves come into being when the forward travelling waves are interfered with the reflecting waves. In order to study the vibration model of the concave spherical emitting surface when the piezoelectric transducers produces stretching vibrations in the axial direction, the model analysis of the transducer and emitter were carried out using ANSYS software. The vibration models of the concave spherical emitting surface are shown in Fig. 4. Only half of the transducer and emitter was selected during analysis, due to its symmetrical structure. The simulation results of Fig. 4 show that the vibration direction of the emitting surface points to the centre position of the spherical surface, and the vibration displacement of the edge is larger than that of the middle. Thus, it can be approximately seen that spherical waves are produced when air is excited by a concave spherical emitting surface. It can be deduced that diffusion to the surrounding space is largely reduced, and the acoustic pressure in the levitating space is enhanced, which will improve the levitation force.

Fig. 2. Schematic diagram of electromagnetic levitation

Fig. 3. Schematic diagram of levitation-coupled standing wave levitation and electromagnetic levitation

c) Fig. 4. Vibration model of the concave spherical emitting surface; a) initial model of the emitter , b) emitter model when stretched at the middle, c) emitter model when it was stretched at end

Research on Levitation Coupled with Standing Wave Levitation and Electromagnetic Levitation

765

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 763-771

2 OPTIMISATION OF THE CONCAVE SPHERICAL SURFACE As is known, the larger the acoustic pressure, the larger the levitation force. When other parameters are fixed, standing wave pressure differs according to the radii of the concave spherical emitting surface and the reflecting surface. Based on the above conclusion, the piezoelectric transducer, the concave spherical emitting surface and reflecting surface are simulated using ANSYS software, and the pressure value of the standing wave is obtained. By comparing the different acoustic pressure under different radii, the optimal radius of the concave spherical surface can be obtained. 2.1 ANSYS Simulation of the Standing Wave Levitator The designed vibration frequency of the piezoelectric transducer is 20 kHz, of which the emitting surface is considered as a plane. The resonant frequency of the transducer will shift when the emitting surface adopts a concave spherical structure. Thus, it is necessary to obtain the actual vibration frequency through the harmonic analysis method for the piezoelectric transducer and the emitter. Fig. 5 shows dimensions schematic of piezoelectric transducer.

In the ANSYS model, zero displacement boundary conditions are applied to the nodes at the top of the mechanical amplifier, and the electrical boundary conditions of the 1 V are applied to the piezoelectric materials. The piezoelectric materials are simulated using a PLANE 13 piezoelectric couplefield element, which has displacement and voltage degrees of freedom. An acoustic fluid element, FLUID 29, is used to simulate the air region, which has a pressure degree of freedom. FLUID 29 also can simulate the interaction between the transducer and the air region, which has displacement and pressure degrees of freedom. The air region edge adopts a noreflection boundary condition and its elements are FLUID 129. A harmonic analysis is used to simulate the process and phenomenon for the sphere standing wave with the ANSYS software. Firstly, when the radius of the emitting surface is a specific value, the resonant frequency is confirmed after analysing the curve whose amplitude varies with a vibration frequency. Then, the wavelength in the air is calculated, and the distance of the emitting surface and the reflecting surface are adjusted to 1.5 the times of the wavelength. Using ANSYS software again to simulate, the distribution and the maximum pressure of the standing wave in the resonant frequency are obtained. For other emitting surfaces of different radii, the maximum acoustic pressure can be obtained in the same way. Fig. 6 shows the distribution of the standing wave field when the radius R is 27 mm.

Fig. 6. Standing wave pressure distribution (R = 27 mm, the distance between the emitter and the reflector is 1.5 times of the wavelength)

Fig. 5. Dimensions schematic

766

Fig. 6 shows not only the distribution of the standing wave sound field, but also the maximum acoustic pressure. When R is set as 27 mm, the maximum pressure is 1.18×10-9 Pa (only 1 V voltage

Jiao, X.Y. – Liu, G.J. – Liu, J.F. – Li, X.B. – Liu, X.L. – Lu, S.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 763-771

was acted on the piezoelectric chips on the ANSYS simulation). The transducer, the emitting surface and the reflecting surface of different radii (from R = 21 to 32 mm) have all been simulated one after another with the space R of 1 mm. Boundary conditions and the input voltage are identical in any radius condition. Fig. 7 shows the curve of the maximum acoustic pressure with different radii of the concave spheres. It can be seen that when R equals 27 mm, the acoustic pressure reaches the maximum, which shows that levitation ability is the strongest under this condition.

some acoustic waves, which leads to lower acoustic energy compared to the standing wave field without the coils. Therefore, it can be shown that the maximum acoustic pressure of Fig. 8 is lower than that of Fig. 6. In the levitation experiment coupled with the standing wave levitation and the electromagnetic levitation, a greater voltage should be input to the standing wave transducer.

Fig. 7. The max acoustic pressure changes with radius

Fig. 8. Distribution of the acoustic field affected by the electromagnetic coils

2.2 ANSYS Simulation of the Effect on the Standing Waves Field by the Coils

3 STANDING WAVE LEVITATION EXPERIMENT

If the diameter of the electromagnetic coils is too much bigger than that of the levitated specimen, it will lead to unstable levitation. Therefore the electromagnetic coils can only stay inside the standing wave fields space, instead of outside the fields. The electromagnetic coils are between the acoustic levitation emitter and reflector and will disturb the standing wave field. The standing waves will have reflections, interferences, etc. at the coils. ANSYS software continues to be used to simulate the acoustic field affected by the coils. The simulated model is similar to above, except for the coils between the acoustic levitation emitter and reflector. Zero displacement boundary conditions are applied to the edge of the coils. The air edge contacting the coils is simulated by an acoustic fluid element, FLUID 29, which has displacement and a pressure degree of freedom. Other boundary conditions are similar to above. Fig. 8 shows the simulation results. Compared with the results shown in Fig. 6, the acoustic field obtains some difference when the electromagnetic coils are added between the emitter and the reflector. A strong acoustic field also exists in the space that is surrounded by the electromagnetic coils. Meanwhile, the coils reflect

3.1 Standing Wave Levitation Device According to the optimisation results via the ANSYS software, a standing wave levitator was designed and manufactured (Fig. 9). The piezoelectric transducer is fastened with the adjusting screw at the fixed base, which can be adjusted up and down. The ultrasonic frequency power supply sends an AC signal to the piezoelectric transducer. By turning the adjustment for the reflector, the reflector can be moved up or down, and then the resonant distance between the emitter and the reflector is adjusted.

Fig. 9. Ultrasound standing wave levitation device; 1 ultrasonicfrequency power supply, 2 piezoelectric transducer, 3 adjusting screw, 4 amplitude transformer, 5 emitter, 6 reflector, 7 adjustment for the reflector and 8 fixed base

Research on Levitation Coupled with Standing Wave Levitation and Electromagnetic Levitation

767

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 763-771

3.2 Sample Levitation Experiment Light samples similar to foam balls were levitated in the standing wave levitator in order to determine the probable node positions. In Fig. 10, when there are 3, 4 or 6 levitation positions, the light levitated samples are distributed with equal space between them. The samples nearby the emitter are off the axial line and others are on the axial line. Several items can be levitated in any position of the annular nodes. As a result of the annular potential dispersing power of the standing waves, it can only levitate some light materials. A high density specimen, such as a steel sphere, cannot be levitated in the annular potential, as shown in Fig. 11, which can only levitate steel spheres at other positions. When the distance is near 34.9 mm, i.e. about twice the wavelength, two steel balls (diameter 5 mm) can be levitated at the two positions near the reflector or the middle locations, as shown in Figs. 11a and b. However, the same steel sphere cannot be levitated at the rest position at the same time. The levitation state of the two steel balls is steady. However, owning to the smaller restoring force, the levitation can be easily destroyed by exterior disturbances. Moreover, as the distance becomes longer, the restoring force and stability of the levitated sample will become worse. In Fig. 11c, three steel balls of 3mm diameter can be suspended at three positions near the reflector at the same time when the distance between the emitter and the reflector is twice the wavelength. The levitator presented by Andrade et al. [25] and Baer et al. [26] can levitate three steel spheres of 2.5 mm diameter when the distance between the emitter and the reflector is 1.5 times that of the wavelength. The levitating number of the same sample is a powerful reference that can be used as an evaluation method for the working stability and levitation ability of a standing wave levitation device. The levitation force and stability of the standing wave device presented in this paper are observably improved.

768

Fig. 10. Light specimen levitated by standing wave

Fig. 11. Steel spheres levitated by standing wave

When the distance continues to increase, no steel sphere can be levitated at any position of the standing wave field. The reason is that too many waves spread outside of the levitation space when the distance increases, which weakens the power of standing wave. If a resonance tube is used in the test, the results will be better. However, it is difficult to place specimen in the standing wave field. 3.3 Electromagnetic Levitation Experiment Coupled with Standing Wave Levitation In the coupled levitation, the distance between the emitter and the reflector of the standing wave levitator is adjusted to two times that of the wavelength and a steel ball of 5 mm diameter can be stably levitated at the second levitation position near the reflector. The axial line of the electromagnetic coils should coincide with that of the standing wave levitator. The levitated position of the steel ball in the electromagnetic coils should be coincident with the second levitated position near the reflector in the standing wave levitator, as shown ‘*’ in Fig. 12. The relationship between the levitated force and the distance between the emitter and the reflector is in accordance with the sine equation [27]. As shown in Fig. 12, when the steel ball is levitated and heated by the electromagnetic coils, the steel ball is seated at the second node of the standing wave levitator and the force on it from the standing waves is zero or extremely little. When the heating of the steel ball is completed, the electric current input to the electromagnetic coils is decreased. The electromagnetic force on the steel ball will lessen, and the ball will start to move downward gradually. At the same time, the acoustic force on the steel ball from the standing wave levitator will become larger until the electric current of the electromagnetic coils is entirely cut off and the steel ball is levitated by the standing wave levitator alone. In this way, the steel ball is cooled to room temperature without contact with any other materials. In the process of

Jiao, X.Y. – Liu, G.J. – Liu, J.F. – Li, X.B. – Liu, X.L. – Lu, S.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 763-771

transferring the levitated force on the steel ball from the electromagnetic force to acoustic force, the drop distance of the sample is about 1/8 of the wavelength (2.2 mm) and the motion is smooth and steady. The photo of the levitation coupled the standing wave levitation and the electromagnetic levitation is shown in Fig. 13.

the resonance condition will be weakened, and the levitation force on the levitated sample will become correspondingly smaller. Therefore, the levitated position will also move downward. Thus, when the downward motion of the steel ball is observed, it immediately shows a decrease in the distance between the emitter and the reflector. Simultaneously, enough sound pressure must be kept. In this way, the standing wave field can be balanced at an approximation resonance condition [28]. In this method, the steel ball at a higher temperature was cooled to room temperature without contact with any other materials. It is shown that the standing waves can be used to suspend the metal sample that has been levitated and heated by the electromagnetic coils. 4 CONCLUSIONS

Fig. 12. Schematic diagram of the sample position transformation in coupled levitation; ‘*’ is the position where the acoustic levitated force is zero. ‘A’ is possible positions where the metal ball is levitated by the acoustic levitated force alone

Fig. 13. Photo of the levitation coupled the standing wave levitation and the electromagnetic levitation

When the steel ball, at a higher temperature, is levitated and cooled in the standing wave levitator, it is extremely important to adjust the resonant acoustic field. Since the change of the temperature leads to the shift of the wavelength, the distance between the emitter and the reflector needs to be adjusted to accommodate the resonance condition. However, it is difficult to adjust the distance in time by hand. In fact, the distance between the emitter and the reflector can be set a little smaller. When the temperature decreases,

With the purpose of further optimisation of the acoustic levitator and the levitation via coupling standing wave levitation and electromagnetic levitation, the emitter and the reflector with different concave radii were analysed with ANSYS software. The formation of the standing waves was simulated. The distribution and the maximum acoustic pressure were ascertained with different radii of the concave spherical surface on the emitter and the reflector. The acoustic pressure of different levitators was compared, and the optimal radius R = 27 mm for this levitator was determined. Based on the optimisation, a standing wave levitator was manufactured. With this device, levitation experiments for light and heavy specimens were carried out. Many types of light foams could be simultaneously levitated at discrete node positions; the foam levitated near the emitter deviated from the axial line. However, steel balls could not be levitated at the node position near the emitter. When the distance between the emitter and the reflector equalled twice that of the wavelength about 34.9 mm, three steel balls of 3mm diameter could be simultaneously levitated in three disparate node positions, excluding the position near the emitter. Compared with other standing wave levitation devices, the levitation capability of the levitator presented in this paper was considerably enhanced. Following that, an attempt was made to use the standing wave levitator to levitate a steel ball of 5 mm diameter after non-contact heating via electromagnetic levitation. Though the heated steel ball was not particularly stable when it was cooled, the coupled levitation experiment could be successful if the distance between the emitter and the reflector was

Research on Levitation Coupled with Standing Wave Levitation and Electromagnetic Levitation

769

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 763-771

suitably adjusted. The results show that it is feasible preliminarily by using standing wave levitator to levitate and cool the metal materials after non-contact heating via the electromagnetic levitation. 5 ACKNOWLEDGMENT This work was supported by the National Natural Science Foundation of China (Grant No. 51075181). 6 REFERENCE [1] Gao, J.R., Cao, C.D., Wei, B. (1999). Containerless processing of material by acoustic levitation. Advances in Space Research, vol. 24, no. 10, p. 1293-1297, DOI:10.1016/S0273-1177(99)00736-X. [2] Urevc, J., Koc, P., Stok, B. (2011). Characterization of material parameters used in the mathematical modelling of arc welding and heat treatment processes Transactions of Famena, , vol. 35, no. 4, p. 1-14. [3] Brandt, E.H. (2001). Suspended by sound. Nature, vol. 413, no. 6855, p. 474-475, DOI:10.1038/35097192. [4] Tsukada, T., Sugioka, K., Tomoya, T., Fukuyama, H., Kobatake, H. (2009). Effect of static magnetic field on a thermal conductivity measurement of a molten droplet using an electromagnetic levitation technique. International Journal of Heat and Mass Transfer, vol. 52, no. 21-22, p. 5152-5157, DOI:10.1016/j.ijheatmass transfer.2009.04.020. [5] Hennet, L., Cristiglio, V., Kozaily, J., Pozdnyakova, I., Fischer, H E., Bytchkov, A., Drewitt, J.W.E., Leydier, M., Thiaudiere, D., Gruner, S., Brassamin, S., Zanghi, D., Cuello, G.J., Koza, M., Magazu, S., Greaves, G.N., Price, D.L. (2011). Aerodynamic levitation and laser heating: Application at synchrotron and neutron sources. The European Physical Journal - Special Topics, vol. 196, no. 1, p. 151-165, DOI:10.1140/epjst/ e2011-01425-0. [6] Kordel, T., Holland-Moritz, D., Yang, F., Peters, J., Unruh, T., Hansen, T., Meyer, A. (2011). Neutron scattering experiments on liquid droplets using electrostatic levitation. Physical Review B, vol. 83, no. 10, p. 104205:1-9, DOI:10.1103/PhysRevB.83.104205. [7] Kohira, M., Isomura, A., Magome, N., Mukai, S., Yoshikawa, K. (2005). Optical levitation of a droplet under a linear increase in gravitational acceleration. Chemical Physics Letters, vol. 414, no. 4-6, p. 389-392, DOI:10.1016/j.cplett.2005.08.083. [8] Geim, A.K., Simon, M.D., Boamfa, M.I., Hefinger, L.O. (1999). Magnetic levitation at your fingertips. Nature, vol. 400, no. 6742, p. 323-324, DOI:10.1038/22444. [9] Liu, J.H., Wang, Q.L., Yan, L.G. (2009). Analysis of force characteristics of a superconducting ball in a given magnetic field. Physica C: Superconductivity, vol. 469, no. 13, p. 756-759, DOI:10.1016/j.physc. 2009.04.006.

770

[10] Brandt, E.H. (1989). Levitation in Physics. Science, vol. 243, no. 4889, p. 349-355, DOI:10.1126/ science.243.4889.349. [11] Xie, W.J., Cao, C.D., Lu, Y.J., Wei, B. (2002). Levitation of Iridium and Liquid Mercury by Ultrasound. Physical Review Letters, vol. 89, no. 10, p. 104304:1-4, DOI:10.1103/PhysRevLett.89.104304. [12] Kozuka, T., Yasui, K., Tuziuti, T., Towata, A., Lida, Y. (2007). Noncontact Acoustic Maniqulation in Air. Japanese Journal of Applied Physics, vol. 46, no. 7B, p. 4948-4950, DOI:10.1143/JJAP.46.4948. [13] Ryuto, Y., Manabu, A., Hideki, T., Takehiro, T. (2011). Novel transfer method using near-field acoustic levitation and its application. Japanese Journal of Applied Physics, vol. 50, no. 7, p. 07HE29:1-5, DOI:10.1143/JJAP.50.07HE29. [14] Malekzadeh, M., Halali, M. (2011). Production of silver nanoparticles by electromagnetic levitation gas condensation. Chemical Engineering Journal, vol. 168, no. 1, p. 441-445, DOI:10.1016/j.cej.2010.12.081. [15] Cadirli, E., Herlach, D.M., Volkmanb, T. (2010). Characterization of rapidly solidified Ni-Si and Co-Al eutectic alloys in drop tube. Journal of Non-Crystalline Solids, vol. 356, no. 9-10, p. 461-466, DOI:10.1016 /j.jnoncrysol.2009.12.019. [16] Takeuchi, M., Arai, Y., Kase, T., Nakajima, Y. (2013). Corrosion study of a highly durable electrolyzer based on cold crucible technique for pyrochemical reprocessing of spent nuclear oxide fuel. Journal of Nuclear Materials, vol. 432, no. 1-3, p. 35-41, DOI:10.1016/j.jnucmat.2012.07.048. [17] Barmatz, M., Collas, P. (1985). Acoustic radiation potential on a sphere in plane, cylindrical, and spherical standing wave fields. The Journal of the Acoustic Society of America, vol. 77, no. 3, p. 928-945, DOI:10.1121/1.392061. [18] Magill, J., Capone, F., Beukers, R., Werner, P., Ohse, W. (1987). Pulsed laser heating of acoustically levitated microspheres under pressure. High Temperatures- High Pressures, vol. 19, no. 5, p. 461-471. [19] Lee, M.C., Feng, I. (1982). Acoustic levitating apparatus for submillimeter samples. Review of Scientific Instrument, vol. 53, no. 6, p. 854-859, DOI:10.1016/1.1137068. [20] Xie, W.J., Wei, B. (2002). Dependence of acoustic levitation capabilities on geometric parameters. Physical Review E, vol. 66, no. 2, p. 026605:1-11, DOI:10.1103/PhysRevE.66.026605. [21] Xie, W.J., Cao, C.D., Lu, Y.J., Wei, B. (2002). Levitation of Iridium and Liquid Mercury by Ultrasound. Physical Review E, vol. 89, no. 10, p. 104304:1-4.. [22] Hong, Z.Y., Xie, W.J., Wei, B. (2011). Acoustic levitation with self-adaptive flexible reflectors. Review of Scientific Instrument, vol. 82, no. 7, p. 074904:1-5, DOI:10.1063/1.3610652. [23] Weber, J.K.R., Hampton, D.S., Merkley, D.R., Rey, C.A., Zatarski, M.M., Nordine, P.C. (1994). Aeroacoustic levitation: A method for containerless liquid-

Jiao, X.Y. – Liu, G.J. – Liu, J.F. – Li, X.B. – Liu, X.L. – Lu, S.

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, 763-771

phase processing at high temperatures. Review of Scientific Instrument, vol. 65, no. 2, p. 456-465, DOI:10.1063/1.1145157. [24] Stephens, T.L., Budwig, R.S. (2007). Three-axis acoustic device for levitation of droplets in an open gas stream and its application to examine sulfur dioxide absorption by water droplets. Review of Scientific Instrument, vol. 78, no. 1, p. 014901:1-8, DOI:10.1063 /1.2424454. [25] Andrade, M.A.B., Buiochi, F., Adamowski, J.C. (2010). Finite element analysis and optimization of a single-axis acoustic levitator. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 57, no. 2, p. 469-479, DOI:10.1109/TUFFC.2010. 1427.

[26] Baer, S., Andrade, M.A.B., Esen, C., Adamowski, J.C., Schweiger, G., Ostendorf, A. (2011). Analysis of the particle stability in a new designed ultrasonic levitation device. Review of Scientific Instruments, vol. 82, no. 10, p. 105111:1-7, DOI:10.1063/ 1.3652976. [27] King, L.V. (1934). On the acoustic radiation pressure on spheres. Proceedings of the Royal Society of London. Series A, Mathematical and Physical Sciences, vol. 147, no. 861, p. 212-240, DOI:10.1098/rspa. 1934.0215. [28] Xie, W.J., Wei, B. (2001). Space environment simulation for material processing by acoustic levitation. Chinese Physics Letters, vol. 18, no. 1, p. 68-70, DOI:10.1088/0256-307X/18/1/324.

Research on Levitation Coupled with Standing Wave Levitation and Electromagnetic Levitation

771

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12 Vsebina

Vsebina Strojniški vestnik - Journal of Mechanical Engineering letnik 59, (2013), številka 12 Ljubljana, december 2013 ISSN 0039-2480 Izhaja mesečno

Rok Simič, Mitjan Kalin: Primerjava adsorpcije alkohola in maščobne kisline na hidrogenirani DLC prevleki s pomočjo AFM analize in triboloških testov Wei Tan, Xiaoan Chen, Shaojiang Dong: Nova metoda za diagnostiko napak strojev na osnovi optimalnega večskalnega morfološkega filtra Janez Benedičič, Roman Žavbi, Jožef Duhovnik: Sistematični razvoj stroja za nanašanje ibitola Manoj Modi, Gopal Agarwal: Kombinirani postopek diamantnega površinskega brušenja in elektroerozijske obdelave v dielektriku z dodanim prahom: modeliranje, primerjalna analiza in večizhodna optimizacija z analizo po metodi uteženih glavnih komponent Branislav Sredanovic, Gordana Globocki-Lakic, Djordje Cica, Davorin Kramar: Vpliv različnih tehnik hlajenja in mazanja na obdelovalnost materiala pri odrezavanju Jafar Arashmehr, Gholam Hossein Rahimi, Seyed Fazel Rasouli: Eksperimentalna in numerična preiskava kompozitne cilindrične lupine z ojačitvenimi rebri pod transverzalno obremenitvijo XiaoYang Jiao, GuoJun Liu, JianFang Liu, Xinbo Li, XiaoLun Liu, Song Lu: Raziskava sklopitve levitacije na stojno valovanje in elektromagnetne levitacije Osebne vesti Doktorske disertacije, specialistično delo, diplomske naloge

SI 148 SI 149 SI 150 SI 151 SI 152 SI 153 SI 154

SI 155

Prejeto v recenzijo: 2013-05-20 Prejeto popravljeno: 2013-07-22 Odobreno za objavo: 2013-09-27

Primerjava adsorpcije alkohola in maščobne kisline na hidrogenirani DLC prevleki s pomočjo AFM analize in triboloških testov Simič, R. – Kalin, M. Rok Simič – Mitjan Kalin*

Univerza v Ljubljani, Fakulteta za strojništvo, Slovenija

V kontaktih mehanskih sistemov prihaja do trenja in obrabe, kar povzroča energijske izgube, znižuje funkcionalnost sistemov in povečuje stroške vzdrževanja. Uporaba prevlek iz diamantu podobnega ogljika (ang: Diamond-Like Carbon – DLC), ki so poznane po svojih nizko-tornih lastnostih in odpornosti na obrabo, lahko znatno zniža energijske izgube in podaljša življenjsko dobo ključnih mehanskih komponent. Ker so kontakti mehanskih sistemov večinoma mazani, to pomeni mazanje tudi DLC prevlek, zato je preučevanje vpliva maziv in aditivov na učinkovitost mazanja DLC prevlek ključnega pomena za izboljšanje in prilagoditev mazanja. Nizko-torne lastnosti DLC prevlek naj bi delno izvirale iz njihove kemijske stabilnosti in splošne nereaktivnosti. Omejene interakcije z molekulami v okolici pa zavirajo možnosti izboljšanja mazalnih lastnosti DLC prevlek. Kljub domnevni nereaktivnosti, pa so prisotnost interakcij in izboljšanje triboloških lastnosti DLC prevlek dokazali v kar nekaj primerih, a splošni mehanizmi delovanja maziv in njihovih aditivov še niso poznani. Namreč, šele temeljito poznavanje osnovnih mehanizmov delovanja maziv in aditivov na DLC prevlekah omogoča optimizacijo mazanja omenjenih prevlek. V tem delu smo se osredotočili na interakcije med DLC prevleko ter alkoholi in maščobnimi kislinami, za katere je znano, da se adsorbirajo na površino jekla in tako učinkovito izboljšajo tribološke lastnosti jeklenih kontaktov. Sposobnost adsorpcije omenjenih molekul na DLC površino smo pod kontroliranimi pogoji preučili z uporabo mikroskopa na atomsko silo (ang: Atomic Force Microscope - AFM). DLC in jeklene površine, ki smo jih uporabili za primerjavo, smo izpostavili različnim raztopinam (2 do 20 mmol/l) heksadekanola ali heksadekanojske kisline v heksadekanu. Z AFM metodo smo po ultrazvočnem čiščenju površin izmerili delež površine, ki so ga zasedale adsorbirane molekule. Z enakimi koncentracijami heksadekanola in heksadekanojske kisline v baznem olju smo opravili tudi tribološke teste za kontakte DLC/DLC in jeklo/jeklo pri pogojih mejnega mazanja. Rezultate trenja in obrabe smo primerjali z izmerjenimi vrednostmi zasedenosti površine iz AFM analize. S povečevanjem koncentracije polarnih molekul se je zasedenost površine DLC prevleke in jekla pričakovano povečevala. S tem smo dokazali, da AFM lahko uporabimo za zaznavanje skupkov adsorbiranih molekul na površini. Poleg tega smo pokazali, da se polarne molekule lahko adsorbirajo na površino hidrogeniranih DLC prevlek tudi ob odsotnosti tribološkega kontakta in da je zasedenost površine DLC prevleke podobna kot na jeklu. Pokazali smo tudi, da se z večanjem koncentracije polarnih molekul v mazivu in večanjem zasedenosti površine hkrati izboljšujejo tudi tribološke lastnosti DLC površin in jekla. Ob prisotnosti polarnih molekul se je obraba DLC prevleke znižala za kar 15 do 20%, medtem, ko je bil vpliv polarnih molekul na trenje zaradi naravnih nizko-tornih lastnosti DLC prevlek neizrazit. V splošnem so se maščobne kisline izkazale kot bolj učinkovite od alkoholov, saj so tvorile bolj celovito adsorbirano plast in učinkoviteje znižale obrabo DLC prevlek. V tej raziskavi smo torej pokazali, da se polarne molekule lahko adsorbirajo na DLC površino in da med zasedenostjo površine, ki je merilo za količino adsorbiranih molekul na površini, in obrabo DLC prevlek obstaja pomembna povezava, ki do sedaj še ni bila eksplicitno objavljena v literaturi. Pokazali smo, da so alkoholi in maščobne kisline potencialni kandidati za ključne komponente aditivov, ki bi jih lahko uporabili v kombinaciji z DLC prevlekami in hkrati ustrezajo najnovejšim okoljevarstvenim smernicam. Predstavili smo tudi možne mehanizme adsorpcije alkoholov in maščobnih kislin na DLC prevlekah, ki vključujejo vpliv okolja, temperature in tribološkega kontakta. Ključne besede: DLC, AFM, maščobne kisline, alkoholi, adsorpcija, tribologija

SI 148

*Naslov avtorja za dopisovanje: *Naslov avtorja za dopisovanje: Univerza v Ljubljani, Fakulteta za strojništvo, Aškerčeva 6, Ljubljana, Slovenija, mitjan.kalin@tint.fs.uni-lj.si

Prejeto v recenzijo: 2013-01-06 Prejeto popravljeno: 2013-05-20 Odobreno za objavo: 2013-09-25

Nova metoda za diagnostiko napak strojev na osnovi optimalnega večskalnega morfološkega filtra Tan, W. – Chen, X.A – Dong, S.J. Wei Tan1,2,* – Xiaoan Chen1 – Shaojiang Dong1 1 Univerza v Chongqingu, Državni laboratorij za mehanske prenose, Kitajska 2 Laboratorij za napredne proizvodne tehnologije za atomobilske dele, Univerza za tehnologijo v Chongqing, Kitajska

Kotalni ležaji so zelo razširjen element rotacijskih strojev. Nenadna odpoved ležajev lahko povzroči nenačrtovane zastoje in izgubo, iz česar izhaja velik pomen diagnostike napak ležajev. Za diagnosticiranje napak ležajev se najpogosteje uporabljajo vibracijski preskusi. Signal napake kotalnega ležaja pa je kompleksen signal in njegovim uporabnim delom se pogosto pridružuje tudi močan šum ozadja, ki preprečuje odkrivanje napak ležaja. Zato je zaželena učinkovita metoda za obdelavo signala, ki razkriva več informacij o napaki. Analiza vibracijskega signala temelji na strukturnih parametrih ležaja, delovnih pogojih in lastni frekvenci komponent ležaja. Fourierjeva transformacija, valčna transformacija in časovno-frekvenčna analiza razkrijejo lastne frekvence komponent karakteristike frekvenčne modulacije napake. Pri posebnih signalih odpovedi ležajev pa se lahko značilne lastnosti skrivajo na več skalah. Fourierjeva transformacija ne more v celoti popisati časovno spremenljivih lastnosti nestacionarnih signalov, valčna transformacija pa ima učinek aliasa v pasovih, učinek uhajanja itd. Rezultati zato zlasti pri diagnostiki odpovedi notranjega obroča in kotalnega ležaja niso zadovoljivi. Za izluščenje teh značilnosti je potrebna večskalna morfološka analiza. Večskalna morfološka analiza omogoča razkrivanje morfoloških značilnosti na različnih skalah in ni odvisna od predhodnega znanja pri izbiri strukturnih elementov. Večskalni morfološki filter je zato primernejši za odstranitev šuma. Predlagana je nova metoda na osnovi optimalnega večskalnega morfološkega filtra za učinkovito odpravo šuma in izločitev impulznih komponent v vibracijskih signalih. Pri tej metodi se najprej uporabi povprečje zapiralnega in odpiralnega operatorja za izgradnjo morfološkega filtra. Elementi strukture večskalnih morfoloških filtrov so optimizirani in izbrani z optimizacijskim algoritmom z rojem delcev. Šum v izvornem signalu je filtriran z večskalnim morfološkim filtrom. Predlagana metoda je bila ovrednotena s simuliranimi signali in s signali odpovedi ležajev. Rezultati kažejo, da lahko metoda učinkovito filtrira šum in izloči impulzne značilnosti vibracijskih signalov, s čimer je ponazorjena učinkovitost predlagane metode. Predlagana metoda ima dve izboljšavi v primerjavi s tradicionalnim morfološkim filtrom. Morfološka operacija je bila določena s povprečno kombinacijo večskalne odpiralno-zapiralne in zapiralno-odpiralne operacije, zato se lahko ne le izluščijo impulzi, ampak se tudi zgladi šum. Nato so bili izbrani strukturni elementi in optimizirani po algoritmu optimizacije z rojem delcem, ki ima podobne lastnosti kot signal objekta. Opredeljeni so tudi koraki postopka za ponazoritev uporabe predlaganega pristopa. Rezultati validacije kažejo, da je predlagani pristop učinkovitejši in robustnejši od tradicionalne enoskalne morfološke analize pri razkivanju impulznih značilnosti. Ker je strukturni element optimiziran po metodi PSO z ozirom na signal, optimizirani večskalni morfološki filter tudi izboljšuje natančnost diagnostike mehanskih napak. Pri uporabi optimalnega večskalnega morfološkega filtra se s povečevanjem redundance poveča tudi zahtevnost izračunov. To je ena glavnih pomanjkljivosti predlagane transformacije, ki bi jo bilo treba v prihodnje še dodatno raziskati. Analiza oblike signala napake na osnovi optimalnega večskalnega morfološkega filtra in opis različnih morfoloških značilnosti signala napake omogočata identifikacijo napak. Večskalna morfološka analiza je preprosta operacija, primerna za računalniško strojno opremo. Rezultati normaliziranih izračunov so lahko enostavni in jasno razločujejo razne značilnosti napak za ugotavljanje stanj in diagnostiko napak. Večskalna morfološka analiza je torej nov pristop k identifikaciji in klasifikaciji napak rotacijskih strojev. Ključne besede: večskalni morfološki filter, strukturni element, optimizacijski algoritem z rojem delcev, zmanjšanje šuma, razkrivanje značilnosti, diagnostika napak kotalnih ležajev

*Naslov avtorja za dopisovanje: Laboratorij za napredne proizvodne tehnologije za atomobilske dele, Univerza za tehnologijo v Chongqingu, No. 174 Shazheng street, Chongqing, Kitajska, tanwv@163.com

SI 149

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, SI 150 Prejeto v recenzijo: 2013-04-18 © 2013 Strojniški vestnik. Vse pravice pridržane. Prejeto popravljeno: 2013-09-12 Odobreno za objavo: 2013-09-25

Sistematični razvoj stroja za nanašanje ibitola Benedičič, J. – Žavbi, R. – Duhovnik, J. Janez Benedičič – Roman Žavbi* – Jožef Duhovnik Univerza v Ljubljani, Fakulteta za strojništvo, Slovenija

Avtomatizacija procesov v življenjskih in delovnih okoljih postaja stalnica razvoja procesov. Eden izmed bistvenih ciljev avtomatizacije je povečanje produktivnosti, ki je eden izmed pomembnejših ekonomskih parametrov pri načrtovanju procesov. Vsekakor se podjetja za avtomatizacijo procesov odločajo predvsem na podlagi neposrednih ali posrednih ekonomskih učinkov. Ročno izvajan ali nezadostno avtomatiziran delovni proces pa je lahko tudi priložnost za nove izdelke ali storitve. Vsekakor pa je potrebno proces najprej prepoznati kot potencialno zanimiv za avtomatizacijo. Članek opisuje sistematičen pristop in metodološko razvit preskriptivni model procesa iskanja priložnosti za nove izdelke, ki je bil uporabljen za prepoznavanja delovnih procesov primernih za določeno stopnjo avtomatizacije v gradbeništvu. Smiselnost avtomatizacije ročnih delovnih procesov v gradbeništvu se ovrednoti na podlagi ekonomskega, socialnega, tehnološkega in zakonodajnega dejavnika. S posameznim dejavnikom opišemo določene značilnosti in informacije, ki nam omogočajo prepoznati določeno izboljšavo procesa kot potencialno razvojno priložnost. Uporaba posameznih dejavnikov pri prepoznavanju priložnosti v avtomatizaciji ročnih delovnih procesov v gradbeništvu pomeni temelj sistematičnega pristopa. Sistematično iskanje možnosti za avtomatizacijo ročnih delovnih procesov v gradbeništvu omogoči le-te prepoznati kot poslovne priložnosti. Te se v razvojnem procesu oblikujejo v rešitve, ki dosežejo ekonomsko in tehnološko primerno stopnjo avtomatizacije ročnega delovnega procesa. V razvojnem procesu je bila v začetnem delu uporabljena metoda SETZ (Socialni, Ekonomski, Tehnološki in Zakonodajni dejavnik). Eden izmed bistvenih elementov metode je zbiranje primarnih in sekundarnih informacij z ustreznim vodenjem in usmeritvijo. Informacije so strukturirane s pomočjo socialnega, ekonomskega, tehnološkega in zakonodajnega dejavnika, kar omogoča sistematičen pristop in učinkovitejše prepoznavanje priložnosti. Pravilnost odločitve o stopnji avtomatizacije nam potrdi trg, na kar pa vpliva še veliko drugih dejavnikov, kakor so razvoj, razvojna ekipa, načrt trženja in samo trženje. Tako kot za vsako drugo dejavnost, tudi za gradbeništvo veljajo specifični pogoji komuniciranja pri iskanju informacij, testiranju in trženju. V prvem delu članka so opisane tudi časovne značilnosti uporabe metode SETZ. Akademsko-industrijska razvojna ekipa (Laboratorij LECAD in NIKO d.o.o., Železniki) je uporabila opisan sistematični pristop s katerim je bilo prepoznano, da na trgu obstaja potreba po avtomatizaciji oziroma delni avtomatizaciji nanašanja ibitola. Omenjena razvojna ekipa je razvila tudi stroj za nanašanje ibitola. Izvedeni so bili testi delnih prototipov ter testi funkcionalnosti in ekonomičnosti končnega fizičnega prototipa. Razvidno je, da se produktivnost delno avtomatiziranega procesa poveča do 3 – krat v primerjavi z ročnim procesom. Delno avtomatiziran proces ima bistveno prednost tudi v zmanjšanju enakih in ponavljajočih gibanjih delavca in s tem zmanjšanju delovnih poškodb. Uspešnost celotnega procesa, tako iskanja priložnosti, kakor tudi nadaljnjega sistematičnega razvojnega novega izdelka temelječega na prepoznani priložnosti, nedvomno dokazujejo značilnosti razvitega stroja, ki delno avtomatizira proces nanašanja hidroizolacijskih premazov. Z uporabo metode SETZ, je mogoče zagotoviti sistematičnost v procesu iskanja priložnosti in povečati verjetnost uspeha razvoja izdelka, kar je bilo prikazano v industrijskih primerih opisanih v drugih publikacijah. Ključne besede: metodologija, industrijska uporaba, delovni proces, iskanje priložnosti, razvoj novega izdelka, inovacija, produktivnost

SI 150

*Naslov avtorja za dopisovanje: Univerza v Ljubljani, Fakulteta za strojništvo, Aškerčeva 6, 1000 Ljubljana, Slovenija, roman.zavbi@@lecad.fs.uni-lj.si

Prejeto v recenzijo: 2013-04-18 Prejeto popravljeno: 2013-07-16 Odobreno za objavo: 2013-08-20

Kombinirani postopek diamantnega površinskega brušenja in elektroerozijske obdelave v dielektriku z dodanim prahom: modeliranje, primerjalna analiza in večizhodna optimizacija z analizo po metodi uteženih glavnih komponent Modi, M. - Agarwal, G. Manoj Modi* - Gopal Agarwal

Nacionalni institut za tehnologijo Malaviya, Oddelek za strojništvo, Indija

Kombinirani postopek diamantnega površinskega brušenja in elektroerozijske obdelave v dielektriku z dodanim prahom (PMEDDSG) je učinkovit postopek za oblikovanje trdih materialov kot je Ti-6Al-4V. Ta hibridni postopek je kombinacija konvencionalnega površinskega brušenja in elektroerozijske obdelave z dodanim prahom v dielektrični tekočini. Brušenje trdega materiala z diamantnim brusom zahteva razmeroma veliko silo, brusilni kolut pa ostane glaziran. Pri elektroerozijski obdelavi trdega materiala se neodstranjeni material zaradi učinka hlajenja dielektrične tekočine spet strdi in oblikuje izjemno trdo resolidificirano plast, v obdelani površini pa ostanejo tudi toplotne napetosti. Kot odgovor na te težave je bil zasnovan in razvit postopek PMEDDSG, pri katerem se oba procesa dogajata istočasno. Hibridni postopek zmanjša sile pri brušenju s toplotnim mehčanjem površine obdelovanca, odpravlja glaziranje koluta z mehčanjem veziva, odpravlja zabijanje koluta z iskrami, odpravlja resolidificirano plast z brusilnim učinkom, zmanjšuje preostale napetosti z abrazijo ter oblikuje gladko površino obdelovanca zaradi enakomerne porazdelitve isker na delcih prahu. Novi hibridni proces odpravlja slabosti konvencionalnih postopkov obdelave trdih materialov, kot so neprimerna kakovost površine, krajša življenjska doba orodja in večji stroški. Pri tej izvedbi je uporaba procesa omejena na obdelavo ravnih površin in samo na električno prevodne materiale. Članek poroča o modeliranju in primerjalni analizi razmerja med vhodnimi spremenljivkami in izhodi obeh procesov, določanju optimalne kombinacije parametrov in preučitvi vpliva vhodnih spremenljivk na debelino bele plasti, kakor tudi o oblikovanju različnih površin pri obdelavi materiala Ti-6Al-4V po postopku PMEDDSG. Za razvoj matematičnega modela obeh odgovorov je bila uporabljena metodologija odzivne površine, ki daje veliko informacij iz majhnega števila eksperimentov za manjšo porabo časa in nižje stroške. Na razvitih matematičnih modelih je bila uporabljena analiza variance, ki pokaže, ali so modeli uporabni za opisovanje razmerja med vhodnimi parametri in izhodnimi odgovori. Opravljenih je bilo 31 poskusov PMEDDSG brez prahu v dielektrični tekočini. Nato je bilo opravljenih še 32 eksperimentov PMEDDSG s primešanim aluminijevim prahom v dielektrični tekočini. Za vhodne parametre so bili vzeti tok, trajanje impulza, hitrost koluta, delovni cikel in koncentracija prahu (le-ta je bila uporabljena kot vhodna spremenljivka samo pri dielektrični tekočini, mešani s prahom). Kot odgovor sta bila pri obeh postopkih merjeni stopnja odvzema materiala (MRR) in povprečna površinska hrapavost (Ra). Analiza rezultatov je pokazala, da na MRR in Ra v veliki meri vplivajo tok, trajanje impulza, delovni cikel, hitrost koluta in koncentracija prahu. PMEDDSG obdelava materiala Ti-6Al-4V s prahom v dielektrični tekočini daje boljše rezultate kot obdelava brez prahu. Optimalna nastavitev parametrov postopka PMEDDSG pri večizhodni optimizaciji je bila ugotovljena z analizo po metodi uteženih glavnih komponent (WPC). Opravljenih je bilo skupno 18 eksperimentov s PMEDDSG po ortogonalnem polju Taguchi L18. Optimalna kombinacija, ki izhaja iz metode WPC, je bila eksperimentalno preskušena za določitev optimalnih vrednosti MRR in Ra ter za ponazoritev učinkovitosti metode WPC. Pristop WPC uporablja preproste računske postopke in je zato uporaben za inženirsko prakso. Posnetki vrstične elektronske mikroskopije so pokazali, da na stanje površine in na debelino bele plasti obdelovanca v veliki meri vplivajo tok, trajanje impulza, delovni cikel, hitrost koluta in koncentracija prahu. Članek podaja tudi informacije za razumevanje mehanizma odstranjevanja kovine pri postopku PMEDDSG. Ključne besede: kombinirani postopek diamantnega površinskega brušenja in elektroerozijske obdelave v dielektriku z dodanim prahom (PMEDDSG), kombinirana elektroerozijska obdelava in diamantno površinsko brušenje (EDDSG), utežene glavne komponente (WPC), analiza variance (ANOVA), Ti-6Al-4V

*Naslov avtorja za dopisovanje: Nacionalni institut za tehnologijo Malaviya, Oddelek za strojništvo, Jaipur, Rajasthan, Indija, manojmnitjaipur1@gmail.com

SI 151

Prejeto v recenzijo: 2013-02-22 Prejeto popravljeno: 2013-05-15 Odobreno za objavo: 2013-09-05

Vpliv različnih tehnik hlajenja in mazanja na obdelovalnost materiala pri odrezavanju Sredanovic, B. – Globocki-Lakic, G. – Cica, D. – Kramar, D. Branislav Sredanovic1,* – Gordana Globocki-Lakic1 – Djordje Cica1 – Davorin Kramar2 1 Univerza

v Banja Luki, Fakulteta za strojništvo, Bosna in Hercegovina

2 Univerza

v Ljubljani, Fakulteta za strojništvo, Slovenija

V prispevku je predstavljen nov pristop k definiciji univerzalne obdelovalnosti. Pojem obdelovalnost materiala v splošnem razumemo kot sposobnost materiala, ki ga je mogoče obdelovati in se v najširšem smislu navezuje na vse načine mehanske obdelave, torej pokriva tudi obdelavo z odrezavanjem. Taka definicija je sicer na videz povsem jasna, vendar jo je težko podrobneje določiti, zlasti pa meriti. Kljub vsemu je odrezovalnost zelo pomembna lastnost materialov, ki bi jo morali upoštevati že konstrukterji, v kolikor želimo izdelovati izdelke z zahtevano kakovostjo čim bolj ekonomično. Pri ožji definiciji je bolj obdelovalen tisti material, ki ga lahko obdelujemo, pri katerem je obstojni čas orodja daljši, pri katerem nastajajo manjše odrezovalne sile, pri katerem dobimo boljšo kakovost obdelane površine, pri katerem dobimo ugodnejšo obliko odrezkov in/ali pri katerem dosežemo večjo natančnost obdelave. Vsako od naštetih stališč predstavlja pomembno okoliščino pri določanju obdelovalnosti za določen material. Različni avtorji so s številnimi preizkusi skušali dobiti stalne vrednosti ali določiti parametre, s katerimi bi lahko ugotavljali in določevali univerzalno obdelovalnost posameznih materialov, vendar doslej še ni bilo mogoče določiti zakonitosti medsebojnih vplivov materiala orodja, obdelovanca in obrabe orodja. Če jo želimo izraziti številčno, lahko uporabimo naslednje kriterije: obraba orodja, obstojnost orodja, oblika odrezkov, kakovost obdelane površine in natančnost obdelave in/ali velikost rezalnih sil. Pri določanju obdelovalnosti materiala povzroča veliko težav tudi to, da osnovni kriteriji obdelovalnosti pri različnih primerih obdelave niso enako pomembni. Tako je pri grobem struženju najpomembnejši čas obstojnosti orodja, pri finem struženju pa kakovost obdelane površine, pri delu na avtomatih pa oblika odrezkov. Zato lahko govorimo le o lastnostih glede na posamezne kriterije, pa še te v večini lahko samo primerjamo. Predstavljeni model obdelovalnosti temelji na analizi vektorja učinkovitosti procesa rezanje. To je nov pristop k definiciji obdelovalnosti s katero se že vrsto let ukvarjajo znanstveniki in raziskovalci širom sveta. Kot merila oz. kriteriji obdelovalnosti so uporabljeni: rezalna sila, intenzivnost obrabe orodja in hrapavost. Metoda tako omogoča primerjavo obdelovalnosti različnih materialov na osnovi več kriterijev hkrati. Analiza obdelovalnosti in potrditev razvitega modela je bila izvedena pri različnih hladilno-mazalnih pogojih, in sicer: za konvencionalno oblivanje, minimalno količino maziva (MQL-minimum quantity lubrication) in za visokotlačno odrezavanje (HPJAM-high pressure jet assisted machining). Obdelovalnost jekla C45E pri različnih pogojih hlajenja in mazanja smo analizirali in ovrednotili glede na razviti model univerzalne obdelavnosti. Tehnološki parametri so bili prilagojeni semi-končni obdelavi, z uporabo najvišjih vrednosti parametrov mogoče. Med raziskavami so bile za potrebe določitve univerzalne obdelovalnosti pri različnih načinih mazanja in hlajenja merjene rezalne sile in hrapavost obdelane površine, spremljala pa se je tudi oblika odrezkov in obraba orodja. Vpliv različnih tehnik hlajenje in mazanje na kazalnike uspešnosti procesa za jeklo C45E je potrdil razvit model obdelovalnosti. Ključne besede: model obdelovalnosti, visokotlačno odrezavanje, hlajenje in mazanje, hlajenje z minimalno količino hladilno-mazalnega sredstva, struženje, obdelovalnost jekla C45E

SI 152

*Naslov avtorja za dopisovanje: Univerza v Banja Luki, Fakulteta za strojništvo, Stepe Stepanovica 75, 78000 Banja Luka, Bosna in Hercegovina, e-mail: sredanovic@gmail.com

Prejeto v recenzijo: 2013-03-26 Prejeto popravljeno: 2013-05-07 Odobreno za objavo: 2013-05-29

Eksperimentalna in numerična preiskava kompozitne cilindrične lupine z ojačitvenimi rebri pod transverzalno obremenitvijo Arashmehr, J. – Rahimi, G.H. – Seyed Fazel Rasouli, S.F. Jafar Arashmehr – Gholam Hossein Rahimi – Seyed Fazel Rasouli Univerza Tarbiat Modares, Oddelek za strojništvo, Iran

Glavni namen tega članka je eksperimentalna in numerična preiskava ojačene kompozitne cilindrične lupine z robnim pogojem nevpetega konca pod transverzalno končno obremenitvijo. Študija je obsegala izdelavo preskušancev, preskušanje in analizo po metodi končnih elementov. Preskušanci za raziskavo so bili izdelani s posebnim strojem za navijanje filamentov. Konstrukcija iz fiberglasa ima vijačna rebra in cilindrično lupino. Preskušanec je bil vpet v stroj za preskušanje s posebnim jeklenim vpenjalom v obliki črke L. Za merjenje deformacij so bili uporabljeni merilni lističi, pritrjeni na več mestih na zgornji površini konstrukcije. Obremenjevanje je bilo izvedeno s strojem INSTRON po postopku nadzorovanega premika. 3D-model konstrukcije je bil pripravljen v programski opremi ANSYS 11.0. Ojačitve so bile mrežene z 20-vozliščnimi polnimi plastnimi elementi, za mreženje lupine pa je bil uporabljen element SHELL99. Robni pogoj je bil nevpet konec pod transverzalno obremenitvijo. Napetosti in deformacije zgornje površine ojačene kompozitne lupine so bile ugotovljene s statično analizo. Eksperimentalno ugotovljene deformacije se razmeroma dobro ujemajo z rezultati numerične študije. Merilni lističi merijo deformacije v zunanji plasti lupine, ki je navita pod kotom 75°, zato je bila pred uporabo odvisnosti med napetostmi in deformacijami opravljena transformacija koordinatnega sistema. Nato so bile izračunane napetosti s transformirano reducirano togostjo. Ugotovljeno je bilo, da se napetosti v cilindrični lupini občutno zmanjšajo v točki sečišča dveh ojačitev. Za določitev učinka ojačitev cilindrične kompozitne lupine je bila ojačena kompozitna lupina primerjana z neojačeno kompozitno lupino z enakimi robnimi in obremenitvenimi pogoji. Med obema vrstama cilindrične lupine je bila ugotovljena občutna razlika von Misesovih napetosti v točki sečišča dveh ojačitev. Rezultati kažejo tudi manjše vrednosti von Misesovih napetosti v ojačeni lupini v primerjavi z neojačeno lupino. Ti rezultati so lahko v pomoč pri snovanju podobnih konstrukcij. Nato je bila opravljena analiza odpovedi na osnovi teorije odpovedi po Tsai-Wu-ju. Eksperimenti so pokazali, da ojačena kompozitna cilindrična lupina odpove na vpetem koncu in do odpovedi pride v lupini. Ugotovljen ni bil noben drug način odpovedi drugih delov konstrukcije, npr. zlom ojačitev ali ločitev reber od lupine. Pri analizi napak sta bili uporabljeni dve metodi: pri prvem pristopu so bili uporabljeni rezultati eksperimentov blizu vpetega konca in teorija odpovedi po Tsai-Wu-ju. Izračunane so bile napetosti v koordinatnem sistemu materiala z globalnimi in lokalnimi napetostmi v navitem laminatu. Rezultati eksperimentov v točki blizu vpetega konca so bili uporabljeni za izračun potrebnih napetosti v koordinatnem sistemu materiala. Za teorijo odpovedi po TsaiWu-ju je bila uporabljena natezna trdnost steklenih vlaken. Po isti teoriji odpovedi je bila izvedena tudi analiza odpovedi s končnimi elementi. Rezultati obeh metod se dobro ujemajo. Nato je bila opravljena analiza odpovedi po MKE za neojačeno lupino z enakim robnim pogojem. Neojačena lupina podobno kot ojačena lupina odpove na vpetem koncu, nosilnost neojačene kompozitne lupine pa je za 17 % nižja od nosilnosti ojačene kompozitne lupine. Omeniti je treba, da so bile nedavne raziskave kompozitnih konstrukcij z ojačitvenimi rebri usmerjene predvsem v aksialne obremenitve, medtem ko so takšne konstrukcije v industrijskih pogojih uporabe izpostavljene tako transverzalnim kot aksialnim obremenitvam. Gre za razmeroma novo raziskavo, ki vključuje eksperimentalni del in analizo po metodi končnih elementov. V obstoječi literaturi ni bilo mogoče najti podobnih raziskav konstrukcij z ojačitvenimi rebri. Ključne besede: kompozitna konstrukcija z ojačitvenimi rebri, transverzalne obremenitve, analiza napetosti, analiza po metodi končnih elementov, analiza odpovedi, merjenje deformacij

*Naslov avtorja za dopisovanje: Univerza Tarbiat Modares, Oddelek za strojništvo, p.p. 14115-143, Tehran, Iran, arashmehr@alum.sharif.edu

SI 153

Prejeto v recenzijo: 2013-03-11 Prejeto popravljeno: 2013-06-13 Odobreno za objavo: 2013-08-23

Raziskava sklopitve levitacije na stojno valovanje in elektromagnetne levitacije

Jiao, X.Y. – Liu, G.J. – Liu, J.F. – Li, X.B. – Liu, X.L. – Lu, S. XiaoYang Jiao1 – GuoJun Liu1 – JianFang Liu1 – Xinbo Li2,* – XiaoLun Liu1 – Song Lu1 1 Univerza

Jilin, Kolidž za strojništvo in tehniške vede, Changchun 130022, Kitajska 2 Univerza Jilin, Šola za komunikacijo in inženiring, Kitajska

Kovinskih materialov po brezkontaktnem segrevanju z magnetno levitacijo ni mogoče ohladiti brez stika z drugimi materiali. Kot rešitev tega problema je predlagana uporaba levitacijske naprave na stojno valovanje za lebdenje raztaljene kovine. Levitacijska naprava na stojno valovanje ima oddajnik in odbojnik s konkavno sferično površino. Pretvornik in polja stojna valovanja so bili simulirani s programsko opremo ANSYS. S simulacijo sta bila določena porazdelitev zvočnega tlaka in največji zvočni tlak pri različnih polmerih konkavne sferične površine na oddajniku in odbojniku za določitev optimalnega polmera. Med oddajnikom in odbojnikom akustične levitacijske naprave so elektromagnetne tuljave, ki motijo polje stojnega valovanja. Tuljave povzročajo odboje, interferenčne motnje ipd. Programska oprema ANSYS je bila uporabljena tudi za simulacijo zvočnega polja, na katerega vplivajo tuljave. Zvočno polje se spremeni, ko med oddajnik in odbojnik dodamo elektromagnetne tuljave. Tudi v prostoru, ki ga obdajajo elektromagnetne tuljave, je močno zvočno polje. Tuljave odbijejo del zvočnih valov, kar privede do zmanjšanja zvočne energije v primerjavi s poljem stojnega valovanja brez tuljav. Simulacija je pokazala, da je pri eksperimentu sklopljene levitacije na stojno valovanje in elektromagnetne levitacije potrebna večja vhodna napetost za pretvornik stojnega valovanja. Na podlagi optimizacije je bil zasnovan in izdelan prototip levitacijske naprave na stojno valovanje. Izvedeni so bili poskusi lebdenja z lažjim in težjim predmetom. Jeklene kroglice stabilno lebdijo, ko je razdalja med oddajnikom in odbojnikom enaka dvakratniku valovne dolžine. Nato je bila uporabljena levitacijska naprava na stojno valovanje za lebdenje jeklene kroglice premera 5 mm, ki je bila pred tem brezkontaktno segreta z elektromagnetno levitacijo. Ko jeklena kroglica pri višji temperaturi lebdi in se hladi v levitacijski napravi na stojno valovanje, je zelo pomembno, da se prilagodi resonančno zvočno polje. Sprememba temperature povzroči premik valovne dolžine in zato je treba uravnati razdaljo med oddajnikom in odbojnikom za doseganje resonance. Ročno nastavljanje razdalje je zamudno, razdaljo med oddajnikom in odbojnikom pa je pravzaprav mogoče nastaviti tudi na nekoliko manjšo vrednost. Z znižanjem temperature se oslabi resonanca in sila levitacije, ki drži lebdeči predmet, se ustrezno zmanjša. Zato se tudi položaj lebdenja predmeta nekoliko zniža. Ko je zabeleženo spuščanje jeklene kroglice, je zato treba takoj zmanjšati razdaljo med oddajnikom in odbojnikom, hkrati pa je treba tudi vzdrževati zadosten zvočni tlak. Polje stojnega valovanja je na ta način mogoče stabilizirati približno v pogojih resonance. Pri tej metodi se jeklena kroglica pri višji temperaturi ohladi na sobno temperaturo brez stika z drugimi materiali. Rezultati kažejo, da je z levitacijsko napravo na stojno valovanje možno lebdenje in ohlajanje kovinskih materialov po tem, ko so bili brezkontaktno segreti z elektromagnetno levitacijo. V tem članku je prvič predstavljena metoda za lebdenje raztaljene kovine v levitacijski napravi na stojno valovanje po segrevanju z elektromagnetno levitacijo. Opravljen je bil tudi eksperiment. Temperaturni vplivi lahko polje stojnega valovanja hitro uničijo, zato je akustična levitacija raztaljene kovine uspela šele po nekaj poskusih. V primeru hitrih temperaturnih sprememb je zato pomembno izboljšanje stabilnosti polja stojnega valovanja. Ogrevanje in ohlajevanje kovine je prav tako treba izvajati v atmosferi inertnega plina, ker pa avtorji članka niso razpolagali s potrebno opremo, so bili opisani eksperimenti opravljeni v zraku. V prihodnjih raziskavah bo treba razrešiti še vrsto drugih težav. Ključne besede: levitacija na stojno valovanje, simulacija ANSYS, elektromagnetna levitacija, brezkontaktno hlajenje

SI 154

*Naslov avtorja za dopisovanje: Univerza Jilin, Šola za komunikacijo in inženiring, Renmin Road 5988, Changchun, Kitajska, xb_li@jlu.edu.cn

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, SI 156-158 Osebne objave

Doktorske disertacije, specialistično delo, diplomske naloge

DOKTORSKE DISERTACIJE Na Fakulteti za strojništvo Univerze v Ljubljani so obranili svojo doktorsko disertacijo: ● dne 8. novembra 2013 Boris CENCIČ z naslovom: »Prenos in optodinamska karakterizacija laserskih bliskov z visoko vršno močjo in enakomernim profilom intenzitete« (mentor: prof. dr. Janez Možina); V nalogi obravnavamo širjenje laserskega snopa iz nestabilnega resonatorja z ravnim profilom intenzitete in visoko vršno močjo. Probleme pri širjenju zaradi uklona na robu snopa in termičnega fokusiranja laserske palice smo rešili s prenosom snopa s teleskopsko preslikavo. Na ta način smo ohranili razmere bližnjega polja blizu izhoda iz laserskega izvora tudi na razdalji nekaj metrov. Problem optičnega preboja v vmesnem gorišču smo rešili z vakuumskima celicama, ki smo ju vstavili v dva segmenta členaste roke. Teoretično smo obravnavali sipanje laserske svetlobe v tetovirani koži z metodo Monte Carlo. Laserski snop smo uporabili za študij odstranjevanje tetovaže in vivo na pacientih, ex vivo na svinjski koži in in vitro na kolagenskih kožnih nadomestkih. Z lasersko odklonsko sondo, fotodiodo in CCD kamero smo opazovali optodinamske in optične pojave, ki so posledica osvetlitve absorptivnega barvila v koži. Z odklonsko sondo smo določili prag za učinkovanje in prag za poškodbo kože, kar lahko uporabimo za spremljanje postopka v realnem času in in vivo. Pokazali smo, da pride po prvem laserskem blisku do strukturnih sprememb v koži, ki zmanjšajo učinek naslednjih bliskov na isto mesto. Dokazali smo povezavo in vivo med intenziteto sevanja plazme, ki nastane ob optičnem preboju na barvilu, in dolgoročno učinkovitostjo odstranjevanja tetovaže. S trajnimi plinskimi mehurčki v koži, ki nastanejo kot razpadni produkt barvila, smo pojasnili izboljšavo laserskega odstranjevanja tetovaže s predhodnim preluknjanjem kože; ● dne 12. novembra 2013 Boris KRŽAN z naslovom: »Raziskava delovanja gonil z zmanjšano količino tekočih maziv« (mentor: prof. dr. Jožef Vižintin); Povečevanje razmerja prenesene moči na enoto volumna je prepoznano kot eden izmed poglavitnih ciljev sodobnih zobniških gonil. Posledica zmanjševanja zunanjih dimenzij so večje kontaktne obremenitve in manj olja v sistemu, kar vodi do povišanih temperatur ter manjše debeline oljnega filma za ležaje in zobnike. V delu je izpostavljeno področje tehnologije zaščitnih prevlek na osnovi diamantu podobnega ogljika DLC. Na zobniškem preizkuševališču FZG smo izvedli preizkuse z zobniškimi dvojicami z in brez zaščitne prevleke. Preizkuse smo izvedli z zmanjšano količino olja v sistemu, zobnik je bil v oljno kopel potopljen za globino 3-kratne vrednosti modula. Zobniški par s

prevleko W-DLC in posebej prilagojeno formulacijo olja je dosegla višjo nosilnost v primerjavi s standardno zobniško dvojico in konvencionalnim oljem za gonila podobne viskoznosti. Približno 40 % nižja srednja aritmetična hrapavost zobnih bokov s prevleko je vzrok za 2-krat višjo vrednost parametra oljnega filma λ, v primerjavi s standardno kombinacijo. Zobniki z zaščitno prevleko so med obratovanjem podvrženi stalni obrabi, kjer se na površini generirajo majhni delci, površini v kontaktu pa se zgladita. Gladka površina zobnih bokov je tudi vzrok za odpornost na poškodbo jamičenja, ki preseže izračunano mejo za gonilo s polno polnitvijo olja, kot tudi obratovalno dobo standardnih zobniških dvojic s cementirano površino; ● dne 29. novembra 2013 Rok SIMIČ z naslovom: »Nanotribologija kemijsko in fizikalno aktivnih aditivov na prevlekah iz diamantu podobnega ogljika« (mentor: prof. dr. Mitjan Kalin); Doktorska naloga obravnava dve možnosti mazanja prevlek iz diamantu podobnega ogljika; s konvencionalnim načinom adsorpcije polarnih molekul kot kemijsko aktivnih aditivov ali z inovativnim pristopom mazanja z nanodelci kot fizikalno aktivnih aditivov. Ovrednotili smo vpliv obeh načinov mazanja na trenje in obrabo DLC prevlek. Mazanje z nanodelci se je izkazalo za neučinkovito, polarne molekule pa so znatno znižale obrabo DLC prevlek. Z različnimi površinsko občutljivimi metodami smo na DLC površinah na nanoskali raziskali adsorpcijske mehanizme maščobnih kislin in alkoholov ter razložili njihov vpliv v kontaktu DLC prevlek. Predlagali smo možne adsorpcijske mehanizme na DLC površinah, ki vključujejo vpliv okolja, temperature in tribološkega kontakta.

* Na Fakulteti za strojništvo Univerze v Mariboru so obranili svojo doktorsko disertacijo: ● dne 19. novembra 2013 Péter GÖNCZ z naslovom: »Računski model za določitev nosilnosti velikih aksialnih trirednih valjčnih ležajev« (mentor: prof. dr. Srečko Glodež); V predloženi doktorski disertaciji je prikazan računski model za določitev nosilnosti velikih aksialnih trirednih valjčnih ležajev. Jedro modela predstavlja matematično-analitični računski postopek za določitev porazdelitve notranjih kontaktnih sil pri obravnavanem tipu velikega aksialnega ležaja. Ta temelji na statičnem ravnovesju med zunanjo obremenitvijo ter notranjimi kontaktnimi silami med valjčki in tečinami ležajnih obročev. Računski model lahko upošteva vpliv poljubnega števila in geometrij valjčkov, nevzporednih relativnih pomikov ležajnih obročev in posledično neenakomerne kontaktne obremenitve valjčkov.

SI 155

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, SI 156-158

Omogočeno je upoštevanje poljubnih kombinacij aksialnih zračnosti ležaja ter vnaprej definiranih strukturnih deformiranosti ležajnih obročev. Rezultati računskih primerov za ta dva dejavnika so pokazali, da lahko bistveno zmanjšata nosilnost ležaja. Za praktično uporabnost je bil model implementiran v računalniško kodo. V okviru določitve statične nosilnosti površinsko zakaljenih tečin je bil analiziran vpliv geometrije valjčkov treh tipov: cilindričnega, logaritemskega in delno profiliranega. Za ta namen je bila izvedena numerična parametrična analiza kontakta valjčkov in segmenta tečine, ki je omogočala določitev vplivov izbranih kriterijev mejne nosilnosti, debeline zakaljene plasti tečine ter neenakomernih kontaktnih obremenitev valjčkov. S pomočjo Basquinove enačbe je bila računsko določena dinamična nosilnost površinsko zakaljene tečine ležaja v območju velikocikličnega utrujanja. Doba trajanja tečine je bila pri tem določena na dva načina: (i) z uporabo materialnih parametrov, določenih pri R = −1 ter upoštevanjem vpliva srednje primerjalne napetosti z Goodmanovo enačbo; (ii) z uporabo materialnih parametrov, določenih pri R = −∞. V omejenem obsegu je bila opravljena eksperimentalna določitev dinamične nosilnosti površinsko zakaljene tečine velikega kotalnega ležaja na preizkuševališču za kotalno-kontaktno utrujanje; ● dne 21. novembra 2013 Mihael VOLK z naslovom: »Stabilizacija procesa globokega vleka z obvladovanjem pridrževanja pločevine« (mentor: izr. prof. dr. Bojan Dolšak); Procesi preoblikovanja pločevine so že sami po sebi zelo kompleksni in predstavljajo ogromno izzivov tako orodjarjem pri načrtovanju preoblikovalnega orodja, kot tudi podjetjem, ki se ukvarjajo s proizvodnjo pločevinastih izdelkov. V zadnjem času pa so zahteve po izdelavi vedno bolj kompeksnih geometrijskih oblikah pločevinastih izdelkov, v uporabi pa je tudi vedno več novih materialov, ki imajo boljše mehanske lastnosti, vendar vse to na račun zmanjšanja sposobnosti preoblikovanja. Proces preoblikovanja je zato postal zahtevnejši, tehnološka okna pa zelo ozka. V doktorskem delu smo stabilizirali proces globokega vleka, ki je eden od najzahtevnejših proces preoblikovanja pločevine. V ta namen smo razvili pridrževalni sistem s segmentnimi vstavki, s katerimi smo povečali tehnološko okno in s tem povečali stabilnost procesa. Prav tako smo razvili sistem določevanja kakovosti vlečenih izdelkov, ki temelji na nadzoru toka materiala v orodje, razvili pa smo tudi optimizacijski sistem. Vsi sistemi skupaj sestavljajo sistem za optimizacijo proizvodnje pločevinastih izdelkov. V zadnji fazi smo vse sisteme verificirali tudi na primeru iz industrije. Rezultati so pokazali bistveno izboljšanje kakovosti vlečenih izdelkov, manj slabe kakovosti ter višjo stabilnost procesa preoblikovanja;

SI 156

● dne 22 novembra 2013 Darko FRIŠČIĆ z naslovom: »Analiza učinka biča v oskrbovalnih verigah z nivojskimi omejitvami« (mentor: izr. prof. dr. Borut Buchmeister);

Zadnja leta se gospodarske družbe in posamezna podjetja povezana v oskrbovalne verige soočajo z nestabilnimi razmerami, med katerimi je pomembnejše nihanje v potrebah kupcev. Ob upoštevanju nivojskih omejitev in različnih ovir pri prenosu informacij med posameznimi udeleženci v verigi, lahko nastanejo razlike v variancah naročil do dobaviteljev glede na povpraševanja kupcev. Velikost teh razlik pri premikanju po verigi navzdol ni vedno enaka in se odraža z nihanjem, ki je glavni pokazatelj obnašanja verige. Ta pojav je opredeljen kot učinek biča. Manjša kot je velikost učinka biča, manjše je nihanje in boljše je obnašanje verige. Takrat so tudi ostali rezultati verige boljši. V konceptu servisiranja kupcev iz zaloge plani proizvodnje in ostale aktivnosti temeljijo na napovedih povpraševanja kupcev. Da bi družbe in podjetja, vključena v oskrbovalno verigo, lahko imela vpogled v svoje potenciale v prihodnjih obdobjih, je bil v doktorski disertaciji analiziran učinek biča v oskrbovalnih verigah z nivojskimi omejitvami za prihodnje obdobje. V ta namen je bil zgrajen simulacijski model tročlenske verige s časovno vrsto sezonskega in desezoniranega značaja. V model je vgrajena skupna učinkovitost proizvodnih procesov (ang. Overall equipment effectiveness – OEE) in različna politika zalog. Simulacije različnih situacij so pokazale, da trenutni kazalec OEE in izbrana politika zalog vplivata na velikost učinka biča in s tem na obnašanje verige v prihodnje. Rezultati simulacij so potrdili naša pričakovanja, zato je model lahko uporaben tudi v širšem aplikativnem smislu; ● dne 28. novembra 2013 David POTOČNIK z naslovom: »Inteligentni računalniški sistem za podporo postopku konstruiranja orodij za hladno preoblikovanje pločevin« (mentor: izr. prof. dr. Miran Ulbin); Konstruiranje orodij za hladno preoblikovanje pločevin je kompleksen postopek, ki je domena bogatih izkušenj in znanja, na osnovi katerih lahko konstrukterji sprejemajo raznovrstne odločitve. Pomanjkanje izkušenih ter neučinkovito in počasno delovanje neizkušenih konstrukterjev so razlogi, da je ravno ta postopek ozko grlo celotnega razvojnega cikla pločevinskih izdelkov. Možen pristop k reševanju omenjene problematike se kaže v nadgradnji konvencionalnih CAD-sistemov z znanjem strokovnjakov. Navkljub mnogim poskusom razvoja tovrstnih inteligentnih sistemov za podporo postopku konstruiranja orodij za hladno preoblikovanje pločevin, pa je bilo zelo malo izmed njih usmerjenih v nadgradnjo omejenih zmogljivosti parametričnih 3D CAD-sistemov. To dejstvo je še posebej problematično, saj danes njihova uporaba predstavlja standard na področju konstruiranja orodij za hladno preoblikovanje pločevin. Pričujoče delo predstavlja novo metodologijo za vzpostavitev inteligentnega parametričnega 3D

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, SI 156-158

CAD-sistema, namenjenega celoviti podpori postopku konstruiranja orodij za hladno preoblikovanje pločevin. V ta namen je bila opravljena analiza tradicionalnega postopka konstruiranja orodij za hladno preoblikovanje pločevin ter sistematizirano kompleksno konstruktersko znanje, ki se navezuje na celoten razvojni cikel pločevinskega izdelka. Za razvoj sistema smo izbrali napredni 3D CAD-paket CATIA V5 R19, vključno z njegovim KBE-modulom Knowledgeware, ki ponuja bogat nabor funkcij za integracijo konstrukterskega znanja. Predlagan modulno zasnovan sistem je edinstven v smislu vzporednega povezovanja informacij, znanj in aktivnosti vseh razvojnih faz pločevinskega izdelka. Moduli omogočajo reševanje parametričnih omejitev, z dobljenimi rezultati pa lahko sistem avtomatsko izvrši konstruiranje preoblikovalnega orodja za veliko območje konstrukterskih problemov. Sistem prav tako omogoča svetovanje, preprečevanje napak ter izobraževanje neizkušenih konstrukterjev. Delovanje sistema so testirali tako izkušeni kot tudi neizkušeni konstrukterji znotraj podjetja Emo-orodjarna d. o. o. Na podlagi analize rezultatov lahko zaključimo, da sistem daje zelo kakovostne rezultate, bistveno prispeva k skrajšanju časa konstruiranja ter predstavlja učinkovit način za izobraževanje neizkušenih konstrukterjev; ● dne 29. novembra 2013 Miha PEVEC z naslovom: »Metoda napovedovanja nastanka termičnih razpok na torni površini avtomobilskih zavornih diskov« (mentor: prof. dr. Iztok Potrč); V doktorski disertaciji je razvita metoda za določitev dobe trajanja zavornih diskov zaradi termičnih razpok na torni površini. Po avtorjevem vedenju je to prvo delo, ki sistematično obravnava vsa področja, ki vplivajo na določitev dobe trajanja pri znanih obremenitvah. Tako razvita metodologija bo služila znanstvenikom ter zavornim inženirjem kot orodje pri optimizaciji geometrije zavornega diska ter materialov. Določitev dobe trajanja zavornih diskov je izredno multidisciplinarno raziskovalno področje, saj zajema znanje iz testiranj, materialov, mehanike, numeričnih metod, prenosa toplote ter dimenzioniranja na dobo trajanja. V doktorskem delu so najprej povzeti rezultati dosedanjih rezultatov iz posameznih področij, po kritični analizi dosedanjih del pa so izpostavljene opažene pomanjkljivosti, ki so v nadaljevanju podrobno obdelane. Za potrebe določitve natančnega temperaturnega polja je predstavljena numerična metoda za določitev faktorja prestopa toplote, ki upošteva realne pogoje med testiranji. Ugotovljeno je, da se rezultati predlagane metode razlikujejo z do sedaj objavljenimi analitičnimi ter numeričnimi metodami, ki ne morejo upoštevati vpliv vpihovanja hladilnega zraka med izvajanjem testa. Poleg tega je predstavljena tudi predlagana metoda za določitev emisijskega koeficienta sevanja zavornega diska med obratovanjem, katere rezultati se razlikujejo tudi za do 47 % od trenutno javno objavljenih. Opravljena je bila podrobna analiza temperaturno odvisnih fizikalnih lastnosti sive litine EN-GJL-250, kot so

toplotna prevodnost, specifična toplota ter temperaturna razteznost. Ugotovljeno je bilo, da se nekatere lastnosti razlikujejo od javno objavljenih tudi za do 48 %. S pomočjo numerične analize po metodi končnih elementov so bila določena temperaturna ter deformacijska polja med predstavljenim obremenitvenim ciklom. Za natančno določitev deformacijskega polja je bila opravljena podrobna analiza mehanskih lastnosti sive litine EN-GJL-250. Predstavljeni so rezultati nateznih in tlačnih preskusov pri sobni temperaturi ter povišanih temperaturah. Rezultati so statistično obdelani z Weibullovo verjetnostno teorijo. Na osnovi zahtev računskih modelov za izračun dobe trajanja po izbrani deformacijski metodi je bila v sklopu doktorske naloge opravljena podrobna analiza lastnosti utrujanja obravnavane sive litine. Nizkociklični preskusi utrujanja so bili izvedeni pri sobni temperaturi ter povišanih temperaturah. Določeni so bili parametri utrujanja, ki so potrebni za izračun dobe trajanja, ter izrisane deformacijske krivulje zdržljivosti. Na koncu je po izbranih hipotezah določena doba trajanja zavornega diska, ki je primerjana z rezultati eksperimenta. Eksperiment zajema pet obravnavanih zavornih diskov, ki so bili testirani po obravnavanem testu na odpornost proti utrujenostnim razpokam na zavornem dinamometru. Ugotovljeno je, da se rezultati eksperimenta ter predlagane metodologije odlično ujemajo, kar kaže na nujnost celovite obravnave problema, kot je to prikazano v predloženi doktorski disertaciji.

SPECIALISTIČNO DELO Na Fakulteti za strojništvo Univerze v Ljubljani je z uspehom zagovarjal svoje specialistično delo: ● dne 19. novembra 2013 Andrej BRAČUN z naslovom: »Raziskava vplivov na letalske nesreče športnih pilotov« (mentor: prof. dr. Jožef Vižintin).

DIPLOMSKE NALOGE Na Fakulteti za strojništvo Univerze v Ljubljani so pridobili naziv univerzitetni diplomirani inženir strojništva: dne 5. novembra 2013: Gregor JEVŠNIKAR z naslovom: »Energijska in eksergijska analiza delovanja enostavnega klimatskega sistema« (mentor: prof. dr. Vincenc Butala, somentor: doc. dr. Matjaž Prek); dne 27. novembra 2013: Blaž BLATNIK z naslovom: »Razvoj sistema za ročno po-prilagoditev vzglavnika za avtomobilski sedež s samodejno nastavitvijo višine vzglavnika« (mentor: izr. prof. dr. Jernej Klemenc); Luka PRISLAN z naslovom: »Brezžični uporabniški vmesnik za vodenje kirurškega orodja za vrtanje v kost« (mentor: prof. dr. Peter Butala); dne 28. novembra 2013:

SI 157

Strojniški vestnik - Journal of Mechanical Engineering 59(2013)12, SI 156-158

Luka STRNAD PETERCA z naslovom: »Kalibracija in analiza presluhov v posameznih oseh dinamometra« (mentor: doc. dr. Franci Pušavec, somentor: prof. dr. Janez Kopač). * Na Fakulteti za strojništvo Univerze v Ljubljani je pridobil naziv magister inženir strojništva: dne 27. novembra 2013: Matej CVETKO z naslovom: »Vpliv sekundarnega toka zraka na vzgonske karakteristike profila lopatice« (mentor: prof. dr. Branko Širok, somentor: doc. dr. Tom Bajcar); Janez KIDRIČ z naslovom: »Razvoj računalniške aplikacije za brezžični nadzor avtonomnega regulacijskega ventila za krmiljenje pretoka deponijskega plina« (mentor: doc. dr. Primož Podržaj); dne 28. novembra 2013: Mojca KOSEM z naslovom: »Sušilni stroj nove generacije« (mentor: prof. dr. Branko Širok, somentor: izr. prof. dr. Marko Hočevar); Damijan ZORKO z naslovom: »Konstruiranje spenjalnega spoja z objemko za aluminijaste profile« (mentor: prof. dr. Jožef Duhovnik, somentor: izr. prof. dr. Jože Tavčar); Nejc ZUPAN z naslovom: »Preprečevanje pojava drdranja pri odrezavanju s povečevanjem dinamične togosti sistema« (mentor: doc. dr. Franci Pušavec, somentor: prof. dr. Janez Kopač). * Na Fakulteti za strojništvo Univerze v Ljubljani je zagovarjal magistrsko delo v okviru Erasmus izmenjave: dne 5. novembra 2013: Jorge SANZ MUSTIELES z naslovom: »Sistem za kondicioniranje zraka s posrednim hlapilnim hlajenjem: analiza in numerično modeliranje / Air conditioning System by Indirect Evaporative Cooling: Analysis and Numerical Modeling« (mentor: prof. dr. Vincenc Butala, somentor: doc. dr. Matjaž Prek). * Na Fakulteti za strojništvo Univerze v Mariboru je pridobil naziv magister inženir strojništva: dne 27. novembra 2013: Primož ČERNIČ z naslovom: »Umerjanje optičnih in tipalnih trikoordinatnih merilnih naprav« (mentor: prof. dr. Bojan Ačko, somentor: doc. dr. Andrej Godina); Klemen GAJŠEK z naslovom: »Razvoj izdelka po principu sočasnega inženirstva« (mentor: prof. dr. Jože Balič, somentor: doc. dr. Iztok Palčič); Aleš ZEVNIK z naslovom: »Konstruiranje preme električnega avtomobila« (mentor: prof. dr. Srečko Glodež); dne 28. novembra 2013: Uroš PEŠAKOVIĆ z naslovom: »Vpliv različnih gostot in topologij mrež na izračunane karakteristike

SI 158

procesov zgorevanja pri GDI motorju« (mentor: prof. dr. Breda Kegl, somentor: prof. dr. Matjaž Hriberšek); Jure ŠTRUCL z naslovom: »Obravnava zaprtega stanja igle med posameznimi vbrizgi v dizelskem injektorju« (mentor: prof. dr. Leopold Škerget, somentor: dr. Zoran Žunič). * Na Fakulteti za strojništvo Univerze v Mariboru je pridobil naziv magister gospodarski inženir: dne 27. novembra 2013: Tomaž KOCIJANČIČ z naslovom: »Vpliv organizacijskih konceptov na inovativnost in uspešnost proizvodnih podjetij« (mentor: doc. dr. Iztok Palčič, somentor: prof. dr. Vojko Potočan). * Na Fakulteti za strojništvo Univerze v Ljubljani so pridobili naziv diplomirani inženir strojništva: dne 7. novembra 2013: Matjaž HATLAK z naslovom: »Primerjava enocevnega in dvocevnega ogrevalnega sistema« (mentor: prof. dr. Vincenc Butala); Danijel IVANČIČ z naslovom: »Hrapavost površine in bela plast pri potopni elektroeroziji« (mentor: doc. dr. Joško Valentinčič, somentor: prof. dr. Janez Grum); Jože NADU z naslovom: »Projektno vodenje izdelave modularnih enot« (mentor: izr. prof. dr. Janez Kušar, somentor: prof. dr. Marko Starbek); Miha PAPEŽ z naslovom: »Analiza elastičnega izravnavanja pri upogibanju v V-matrici ob različnih kotih krivljenja« (mentor: izr. prof. dr. Tomaž Pepelnjak). * Na Fakulteti za strojništvo Univerze v Mariboru so pridobili naziv diplomirani inženir strojništva:

dne 28. novembra 2013: Rudolf IMRE z naslovom: »Krmiljenje stroja za razrez cevnih segmentov« (mentor: doc. dr. Uroš Župerl); Sendi MIĆIĆ z naslovom: »Dvižna miza na jermenski pogon« (mentor: prof. dr. Iztok Potrč, somentor: izr. prof. dr. Tone Lerher). * Na Fakulteti za strojništvo Univerze v Mariboru so pridobili naziv diplomirani inženir strojništva (VS): dne 28. novembra 2013: Martin AVGUŠTIN z naslovom: »Sprememba konstrukcije kokile za litje na nizkotlačnem livnem stroju« (mentor: izr. prof. dr. Miran Ulbin); Blaž DOVNIK z naslovom: »Avtomatizacija centralnega mazalnega sistema na škarjah WT zdvajalec« (mentor: izr. prof. dr. Ivan Pahole).

http://www.sv-jme.eu

59 (2013) 12

Since 1955

Contents

Papers

707

Rok Simič, Mitjan Kalin: Comparison of Alcohol and Fatty Acid Adsorption on Hydrogenated DLC Coatings Studied by AFM and Tribological Tests

719

Wei Tan, Xiaoan Chen, Shaojiang Dong: A New Method For Machinery Fault Diagnoses Based On an Optimal Multiscale Morphological Filter

735

Manoj Modi, Gopal Agarwal: Powder-Mixed Electro-Discharge Diamond Surface Grinding Process: Modelling, Comparative Analysis and Multi-Output Optimisation Using Weighted Principal Components Analysis

748

Branislav Sredanovic, Gordana Globocki-Lakic, Djordje Cica, Davorin Kramar: Influence of Different Cooling and Lubrication Techniques on Material Machinability in Machining

755

Jafar Arashmehr, Gholam Hossein Rahimi, Seyed Fazel Rasouli: An Experimental and Numerical Investigation of a Grid Composite Cylindrical Shell Subjected to Transverse Loading

763

XiaoYang Jiao, GuoJun Liu, JianFang Liu, Xinbo Li, XiaoLun Liu, Song Lu: Research on Levitation Coupled with Standing Wave Levitation and Electromagnetic Levitation

Journal of Mechanical Engineering - Strojniški vestnik

725 Janez Benedičič, Roman Žavbi, Jožef Duhovnik: Systematic Development of a Device for Bituminous Layer Application

12 year 2013 volume 59 no.

Strojniški vestnik Journal of Mechanical Engineering