AFXZ Scientific Journal 1/2021

TECHNICKÁ UNIVERZITA VO ZVOLENE DREVÁRSKA FAKULTA

ACTA FACULTATIS XYLOLOGIAE ZVOLEN

VEDECKÝ ČASOPIS SCIENTIFIC JOURNAL

63 1/2021

Vedecký časopis Acta Facultatis Xylologiae Zvolen uverejňuje pôvodné recenzované vedecké práce z oblastí: štruktúra a vlastnosti dreva, procesy spracovania, obrábania, sušenia, modifikácie a ochrany dreva, termickej stability, horenia a protipožiarnej ochrany lignocelu-lózových materiálov, konštrukcie a dizajnu nábytku, drevených stavebných konštrukcií, ekonomiky a manažmentu drevospracujúceho priemyslu. Poskytuje priestor aj na prezentáciu názorov formou správ a recenzií kníh domácich a zahraničných autorov.

Scientific journal Acta Facultatis Xylologiae Zvolen publishes peer-reviewed scientific papers covering the fields of wood: structure and properties, wood processing, machining and drying, wood modification and preservation, thermal stability, burning and fire protection of lignocellulosic materials, furniture design and construction, wooden constructions, economics and management in wood processing industry. The journal is a platform for presenting reports and reviews of books of domestic and foreign authors.

VEDECKÝ ČASOPIS DREVÁRSKEJ FAKULTY, TECHNICKEJ UNIVERZITY VO ZVOLENE 63 1/2021 SCIENTIFIC JOURNAL OF THE FACULTY OF WOOD SCIENCES AND TECHNOLOGY, TECHNICAL UNIVERSITY IN ZVOLEN 63 1/2021 Redakcia (Publisher and Editor’s Office): Technická univerzity vo Zvolene (Technical university in Zvolen); TUZVO Drevárska fakulta (Faculty of Wood Sciences and Technology) T. G. Masaryka 2117/24, SK-960 01 Zvolen, Slovakia Redakčná rada (Editorial Board): Predseda (Chairman): prof. Ing. Ján Sedliačik, PhD., TUZVO (SK) Vedecký redaktor (Editor-in-Chief): prof. Ing. Ladislav Dzurenda, PhD., TUZVO (SK) Členovia (Members): prof. RNDr. František Kačík, PhD., TUZVO (SK) prof. RNDr. Danica Kačíková, MSc. PhD., TUZVO (SK) prof. Ing. Jozef Kúdela, CSc., TUZVO (SK) prof. Ing. Ladislav Reinprecht, CSc., TUZVO (SK) prof. Ing. Jozef Štefko. CSc., TUZVO (SK) doc. Ing. Pavol Joščák, CSc., TUZVO (SK) doc. Ing. Hubert Paluš, PhD., TUZVO (SK) Jazykový editor (Proofreader): Mgr. Žaneta Balážová, PhD. Technický redaktor (Production Editor): Antónia Malenká Medzinárodný poradný zbor (International Advisory Editorial Board): Bekhta Pavlo (Ukrainian Natl Forestry Univ, Ukraine), Deliiski Nencho (University of Forestry, Bulgaria), Jelačić Denis (Univ Zagreb, Croatia), Kasal Bohumi (Tech Univ Carolo Wilhelmina Braunschweig, Germany), Marchal Remy (Arts & Metiers ParisTech, France), Németh Róbert (Univ Sopron, Hungary), Niemz Peter (Bern Univ Appl Sci, Architecture Wood & Civil Engn, Switzerland), Orlowski Kazimierz A. (Gdansk Univ Technol, Poland), Pohleven Franc (Univ Ljubljana, Slovenia), Potůček František (Univ Pardubice, Czech Republic), Teischinger Alfréd (Univ Nat Resources & Life Sci, BOKU, Austria), Smardzewski Jerzy (Poznan Univ Life Sci, Poland), Šupín Mikuláš (Technical University Zvolen, Slovakia), Vlosky Richard P. (Louisiana State Univ, USA), Wimmer Rupert (Univ Nat Resources & Life Sci, Austria).

Vydala (Published by): Technická univerzita vo Zvolene, T. G. Masaryka 2117/24, 960 01 Zvolen, IČO 00397440, 2021 Náklad (Circulation) 150 výtlačkov, Rozsah (Pages) 178 strán, 15,59 AH, 15,71 VH Tlač (Printed by): Vydavateľstvo Technickej univerzity vo Zvolene Vydanie I. – marec 2021 Periodikum s periodicitou dvakrát ročne Evidenčné číslo: 3860/09 Acta Facultatis Xylologiae Zvolen je registrovaný v databázach (Indexed in): Web of Science, SCOPUS, ProQuest, AGRICOLA, Scientific Electronic Library (Russian Federation) Za vedeckú úroveň tejto publikácie zodpovedajú autori a recenzenti.

Rukopis neprešiel jazykovou úpravou Všetky práva vyhradené. Nijaká časť textu ani ilustrácie nemôžu byť použité na ďalšie šírenie akoukoľvek formou bez predchádzajúceho súhlasu autorov alebo vydavateľa.

© Copyright by Technical University in Zvolen, Slovak Republic.

ISSN 1336–3824

Obsah 1. BARBORA SLOVÁČKOVÁ: THERMAL CONDUCTIVITY OF SPRUCE, BEECH AND OAK HEARTWOOD DEGRADED WITH TRAMETES VERSICOLOR L. LLOYD ......................................................

5

2. RUSLAN R. SAFIN – ŠTEFAN BARCIK – EVGENY Y. RAZUMOV – PETR M. MAZURKIN – ALBINA V. SAFINA: MULTI-FACTOR MODELING OF DYNAMICS OF HARDWOOD DENSITY IN THE PROCESS OF THERMOMODIFICATION ..................................................

13

3. MICHAL DUDIAK: MODIFICATION OF MAPLE WOOD COLOUR DURING THE PROCESS OF THERMAL TREATMENT WITH SATURATED WATER STEAM ...................................................................

25

4. JOZEF KÚDELA – MICHAL ANDREJKO – OĽGA MIŠÍKOVÁ: WOOD SURFACE MORPHOLOGY ALTERATION INDUCED BY ENGRAVING WITH CO2 LASER UNDER DIFFERENT RASTER DENSITY VALUES .......................................................................................

35

5. NENCHO DELIISKI – LADISLAV DZURENDA – PETER NIEMZ – DIMITAR ANGELSKI – NATALIA TUMBARKOVA: COMPUTING THE 2D TEMPERATURE DISTRIBUTION IN LOGS STORED FOR A LONG TIME IN AN OPEN WAREHOUSE IN WINTER AND DURING SUBSEQUENT AUTOCLAVE STEAMING ...............................................

49

6. ZUZANA VIDHOLDOVÁ – DÁVID CIGLIAN – LADISLAV REINPRECHT: BONDING OF THE THERMALLY MODIFIED NORWAY SPRUCE WOOD WITH THE PUR AND PVAc ADHESIVES ...

63

7. DANIEL CHUCHAŁA – KAZIMIERZ A. ORŁOWSKI – GERHARD SINN – ALEKSANDRA KONOPKA: COMPARISON OF THE FRACTURE TOUGHNESS OF PINE WOOD DETERMINED ON THE BASIS OF ORTHOGONAL LINEAR CUTTING AND FRAME SAWING ..

75

8. GEORGI KOVATCHEV – VALENTIN ATANASOV: DETERMINATION OF VIBRATION DURING LONGITUDINAL MILLING OF WOOD-BASED MATERIALS ..............................................

85

9. STANISLAV KORCHAGIN – MIKHAIL CHERNYKH – VLADIMIR ŠTOLLMANN: PRODUCTION OF WOODEN HOLLOW SPIRAL BALUSTERS USING TURN-MILLING CNC MACHINES .......................

93

10. JACEK BARANSKI – PRZEMYSLAW DUDEK: CHIP SUCTION SYSTEM IN CIRCULAR SAWING MACHINE: EMPIRICAL RESEARCH AND COMPUTATIONAL FLUID DYNAMICS NUMERICAL SIMULATIONS .....................................................................................................

103

11. OLENA PINCHEVSKA – JÁN SEDLIAČIK – OLEXANDRA ZAVOROTNUK – ANDRIY SPIROCHKIN – IVAN HRABAR – ROSTISLAV OLIYNYK: DURABILITY OF KITCHEN FURNITURE MADE FROM MEDIUM-DENSITY FIBREBOARD (MDF) ......................

119

12. NADEŽDA LANGOVÁ – SYLVIA BLAŠKOVÁ – JOZEF GÁBORÍK – DENISA LIZOŇOVÁ – ANDREJ JUREK: MISMATCH BETWEEN THE ANTHROPOMETRIC PARAMETERS AND CLASSROOM FURNITURE IN THE SLOVAK PRIMARY SCHOOLS ............................

131

13. ELENA FARKAŠOVÁ – RENÉ BAĎURA: UNDERSTANDINGS OF DESIGN IN CIRCUMSTANCES OF HUMANITY .....................................

143

14. EMILIA GRZEGORZEWSKA – MARIANA SEDLIAČIKOVÁ – JÁN KALAFÚS: A COMPARISON OF THE IMPORTANCE OF THE FURNITURE MANUFACTURING IN EU COUNTRIES USING CLUSTER ANALYSIS AND HELLWIG'S METHOD ................................

151

15. ANDREA SUJOVÁ – ĽUBICA SIMANOVÁ: INFLUENCE OF ASPECTS OF CHANGE MANAGEMENT ON THE PERFORMANCE OF ENTERPRISES IN THE WOOD PROCESSING INDUSTRY ..............

165

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 5−12, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.01

THERMAL CONDUCTIVITY OF SPRUCE, BEECH AND OAK HEARTWOOD DEGRADED WITH TRAMETES VERSICOLOR L. LLOYD Barbora Slováčková ABSTRACT A new type of bio-based thermal insulation is proposed in this paper. This material is made of wood waste (from wood working manufacture or otherwise non-processable wood) which is intentionally degraded with a white rot fungus. Degradation of wood has an impact on its properties and this impact may be positive for re-purposing the use of such material.

The main focus is on thermal properties of this material. A decrease in the values of degraded wood thermal conductivity is described by a fraction of the thermal conductivity value of degraded wood to the value of thermal conductivity of undegraded wood. The fraction for spruce wood was 0.8 in longitudinal direction; 0.64 in radial direction and 0.54 in tangential direction. For beech wood, the fractions were 0.82 for longitudinal direction; 0.61 for radial direction and 0.69 for tangential direction. A decrease in thermal conductivity in oak heartwood was noticeable only in tangential direction; the fraction between thermal diffusivity values was 0.67. The values of thermal conductivity are in good agreement with thermal conductivity values of other bio-based thermal insulation materials.

Key words: degraded wood, thermal insulation, thermal conductivity, thermal diffusivity.

INTRODUCTION We spend a great amount of time inside buildings and it is therefore important that we feel good during this time. Building materials used for structures play an important role in this. Because the climate in our country has four seasons with fluctuating daily temperatures, it is important to choose a suitable thermal insulating material for these buildings. Conditions and technical requirements for thermal insulation of buildings are very strict and that is why we often choose materials which have the best thermal properties but their other properties related to changes in relative air humidity are not as good.

Thickness and thermal conductivity of a material are used to calculate the thermal insulation of a building. The values of thermal conductivity of the most frequently used insulation materials such as expanded Styrofoam (EPS), extruded Styrofoam or mineral wool are low, around 0.035 W·m−1·K−1 (according to technical documentation by manufacturer ISOVER, 2020). As it was already mentioned, manufacturing of these insulation materials requires a lot of energy and also chemicals. In the past, chlorine and fluor compounds were used as expanding agents in extruded Styrofoam. These compounds contribute to the greenhouse effect. Nowadays, carbon dioxide is used as the expanding agent (SVOBODA et al. 2005). Mineral fibers are made by melting a slag-basalt mixture at 5

very high temperatures. The fibers are then created by centrifugal spinning of the molten mixture at a temperature of 1350–1400°C. Melting of the basalt alone and of this mixture is extremely energy-consuming (SVOBODA et al. 2005).

Manufacturing of bio-based thermal insulations does not require the use of hazardous chemicals, nor high temperatures. The material is usually grinded into fine fibers, mixed with water and adhesives, filled into forms, eventually pressed and the excess water is evaporated by drying the material. Raw material used for bio-based insulations is waste from other manufactures like wood manufacture or agricultural production. The adhesives are usually natural too or they are harmless compounds. Adhesives used in bio-based thermal insulations are latex, resins or even lignin which is already contained in the raw material itself. Thermal insulation from wood fiber uses ammonium sulphate as fire retardant (STEICO flex038 technical documentation), though it is usually used as fertilizer.

One of the biggest advantages of bio-based thermal insulation is that the raw material is already highly hygroscopic. Hygroscopic materials have the capacity of adsorbing and desorbing water vapor which contributes to moderate extremes of humidity in indoor environments (OSANYINTOLA and SIMONSON 2006, PALUMBO et al. 2016, QIN et al. 2011, SIMONSON et al. 2004). The research by PALUMBO et al. (2018) confirmed, that natural materials absorb water immediately after beginning of the experiment. It is important to take this property into consideration as well since water vapors are present inside and outside of buildings all the time. Their movement can affect the performance of materials and foremost the comfortability of spending time in the building. Keeping air humidity at a certain, equilibrium level is important for a healthy living. Hence, natural movement of air humidity inside the material can become an important factor and an advantage in designing buildings.

Bio-based thermal insulations which are available on the market are made of wood fibers (STEICO) and recycled newspaper (blow-in cellulose insulation, STEICO and ISOCELL). Other proposed raw materials for bio-based thermal insulations are hay, corn pith and alginic acid (PALUMBO et al. 2018), coniferous tree needles and starch (MUIZNIECE et al. 2015, MUIZNIECE and BLUMBERGA 2016), coconut husk (VAN DAM 2004, ALAVEZRAMIREZ 2012, PANYAKAEW and FOTIOS 2011), wood saw dust (CETINER and SHEA 2018) and also mycelium (AMSTISLAVSKIJ et al. 2017, XING et al. 2018).

Another proposed bio-based insulation raw material is wood infested with a wood decaying fungus. Wood decaying fungi decrease the mass of the wood and the wood becomes more porous (REINPRECHT 2008). These properties – low mass and high porosity are significant for thermal insulations. The main focus of this work was to compare thermal conductivity values of degraded wood species to reference data of thermal conductivity values of undegraded wood species. Thermal conductivity values of degraded wood species were then compared to thermal conductivity values of other bio-based thermal insulations, available on the market and also experimental ones.

MATERIALS AND METHOD Three wood species were chosen for this experiment. The researched wood species are the most widespread species in Slovakia (Zelená správa za rok 2019 - MORAVČÍK et al. 2020).

From coniferous trees, it is spruce (Picea abies, L.). Among deciduous trees species it is beech (Fagus sylvatica, L.) and oak (Quercus petraea, Matt. Liebl.), its heartwood part was used for the experiment. The different wood species were also chosen to compare the effect of degradation on thermal properties. Spruce wood was provided by a wooden window frames manufacturer and beech and oak heartwood lumber was stored at the Department of Wood Science. The samples were cut and sanded to a size of 50 × 50 × 8 mm. The smallest 6

dimension of the sample was cut accordingly to each anatomical direction on wood. A total of 8 samples per anatomical direction per wood species were used for thermal properties measurement. The size of the samples was limited by the dimensions of the Kolle flasks.

These dimensions were also determined by preliminary calculations of the thermal field in wood according to literature (HRČKA and BABIAK 2017).

The intentional degradation was performed in the laboratory of Department of Wood Technology. Kolle flasks were sterilized in an autoclave prior to the experiment. A malt extract (prepared accordingly to STN EN 113) was poured into the Kolle flasks and it was inoculated with growing fungal cultures. The fungus was incubated until the surface of the malt extract was fully covered with mycelia growth. The samples were submerged in distilled water for 24 hours prior to the degradation experiment. Four or five samples were placed into one Kolle flask. Each sample was placed on a “U” shaped stainless steel support.

Duration of the intentional degradation was 6 months. This duration was chosen according to results from a similar experiment (SLOVÁČKOVÁ et al. 2018). After the time has passed, the samples were taken out of the Kolle flasks and cleaned off of visible mycelium remnants.

The samples were then submerged in distilled water. Submerging caused all air inside the sample to escape and this gradually stopped activity of the fungus. Wood decaying fungi need an air content of at least 5–20 % to be able to survive in wood (RYPÁČEK 1957). Each wood species was put into a separate container. The distilled water was changed gradually, once in every two weeks. The samples were kept in water until they reached a maximum moisture content. The maximum moisture content was checked regularly by weighing the samples.

After the samples reached the maximum moisture content, they were taken out of the distilled water and put into an air-conditioning chamber (A/C chamber, Binder, model KBF 780, Tuttlingen, Germany) at an air temperature of 20 ± 2°C and a relative air humidity of 60 ± 3%. When the samples reached the maximum moisture content, their dimensions were measured with a slide caliper (Mitutoyo Absolute Digimatic) and their masses were determined (laboratory scale RADWAG, Analytical Balances, model XA 60/20/X, accuracy 1·10−5 g).

Thermal properties were measured and calculated according to a method proposed by HRČKA and SLOVÁČKOVÁ (2019). Four samples with two heating foils and three thermocouples and a pyrometer were assembled in a fixed position (according to scheme 2; HRČKA and SLOVÁČKOVÁ 2019). Position of the thermocouples was fixed with a cellulose based scotch tape. The calculation was created in MS Excel Visual Basic for Applications and the Solver add-in program of MS Excel according to the equation proposed by HRČKA and BABIAK (2017).

The equation is solved as an 3-dimensional problem. Data needed to find the solution are gathered from measuring of the temperature change during a period of time (one hour per run) with the three thermocouples and the pyrometer. Thicknesses of the samples were also determined and their densities were calculated. The performance of the heating foil is also needed for the calculation, this was determined by Ohm´s equation. The final solutions of the equation are thermal conductivity, thermal diffusivity and specific heat capacity for all anatomical directions in wood. Because the measurement and evaluation of the results take a considerable amount of time, a follow-up experiment with undegraded wood was not performed. The results of thermal properties of undegraded wood done with this same method were published in other works (HRČKA and BABIAK 2017; SLOVÁČKOVÁ et al. 2018), so these data were used as reference data.

The 8 degraded samples were randomly divided into two sets, four samples each, for measuring thermal properties. Due to the level of degradation, only one full set of four samples was possible for some groups of samples (the second set of longitudinal and radial 7

direction groups of spruce wood contained only three samples). In these sets, the missing sample was always substituted by randomly choosing one sample from the full set.

RESULTS Medians of thermal conductivity values (λ), thermal diffusivity values (a) and specific heat capacity values (c) of the researched degraded wood species measured at their equilibrium moisture content reached at the air temperature of 20 ± 2°C and relative air humidity of 60 ± 3% are presented in Table 1. Equilibrium moisture contents reached at these conditions were 15.2% for degraded spruce wood, 15.0% for degraded beech wood and 14.9% for degraded oak heartwood. It must be noted, that these moisture contents were reached in the process of desorption and they are higher by approximately 2.5 % than moisture contents reached in the process of sorption. Densities of the degraded samples were 242.9 kg·m−3 for spruce wood, 375.4 kg·m−3 for beech wood and 523.8 kg·m−3 for oak heartwood. Thicknesses of the measured samples were 7.64 mm for spruce wood, 8.29 mm for beech wood and 7.78 mm for oak heartwood.

Tab. 1 Medians and quartiles Q1 and Q3 of thermal conductivity values, thermal diffusivity values, specific heat capacity for all anatomical directions of the researched wood species.

Wood species

Anatomical direction L

Degraded spruce

R T L

Degraded beech

R T L

Degraded oak heartwood

R T

Thermal conductivity [W·(m·K)-1] 0.28 (0.27 to 0.28) 0.09 (0.08 to 0.10) 0.07 (0.07 to 0.07) 0.31 (0.31 to 0.33) 0.14 (0.13 to 0.15) 0.11 (0.10 to 0.11) 0.35 (0.34 to 0.35) 0.20 (0.20 to 0.21) 0.14 (0.13 to 0.15)

Thermal diffusivity [m2·s-1] 6,97·10-7 (6.82·10-7 to 7.13·10-7) 2,50·10-7 (2.33·10-7 to 2.88·10-7) 2,15·10-7 (1.99·10-7 to 2.22·10-7) 6,69·10-7 (6.05·10-7 to 7.46·10-7) 2,75·10-7 (2.40·10-7 to 2.99·10-7) 2,13·10-7 (1.91·10-7 to 2.35·10-7) 4,81·10-7 (4.29·10-7 to 5.34·10-7) 2,65·10-7 (2.55·10-7 to 2.71·10-7) 2,46·10-7 (1.95·10-7 to 2.70·10-7)

Specific heat capacity [kJ·(kg·K)-1]

1.48 (1.43 to 1.55)

1.35 (1.24 to 1.40)

1.32 (1.23 to 1.39)

It is apparent, that the values of λ are the highest in longitudinal direction and the lowest in tangential direction. Degraded spruce wood reached the lowest values of λ from all researched wood species. It seems that the low density and high porosity of degraded spruce wood were significant factors influencing the values of λ.

The a values in transversal directions of degraded wood are in a similar range. The a value in longitudinal direction of degraded oak heartwood is lower than the a values in longitudinal directions in the other two wood species. Degraded spruce and beech wood reached a similar value of a in longitudinal direction. Thermal diffusivity is defined as the ratio of λ to the product of density and c. Hence, a conclusion on the density itself influencing the values of a is not possible to state without taking the influence of c and λ on thermal diffusivity into consideration as well. As it was similarly stated by GLASS and ZELINKA (2010), 8

conclusions regarding the thermal diffusivity variation with temperature and density are often based on calculating the effect of these variables on heat capacity and thermal conductivity.

All degraded wood species reached a value of specific heat capacity within a similar range despite having different densities. Specific heat capacity does not depend on wood species. It depends on temperature and moisture content of wood (GLASS and ZELINKA 2010).

Thermal conductivity values of degraded spruce wood are lower than in undegraded spruce wood. The λ values of undegraded spruce wood were 0.35 W·(m·K) −1 in the longitudinal direction; 0.14 W·(m·K)-1 in radial direction and 0.13 W·(m·K)-1 in tangential direction (SLOVÁČKOVÁ et al. 2018). These values were determined at an air temperature of 20 ± 2°C and relative air humidity of 65 ± 3%. λ values for degraded beech wood are also lower than in undegraded beech wood. The λ values for undegraded beech wood were (determined at an air temperature of 20 ± 2°C and a relative air humidity of 65%): 0.38 W·(m·K)-1 in the longitudinal direction; 0.23 W·(m·K)1 in radial direction and 0.16 W·(m·K)-1 in tangential direction (HRČKA and BABIAK 2017).

Hrčka and Babiak also stated a values of undegraded beech wood: 2.9·10-7 m2·s-1 in longitudinal direction; 1.7·10-7 m2·s-1 in radial direction and 1.2·10-7 m2·s-1 in tangential direction. These values are lower than a values of degraded beech wood which means that degraded beech wood reacts to temperature changes faster than undegraded wood.

Degraded oak heartwood has similar λ values as undegraded oak wood. The λ value in longitudinal direction was 0.348 W·(m·K)−1; 0.200 W·(m·K)−1 in radial direction and 0.21 W·(m·K)−1 in tangential direction (POŽGAJ et al. 1997). Only the λ value in tangential direction of degraded oak heartwood was lower compared to the λ value in tangential direction of undegraded oak wood. Varying porosity of the oak heartwood samples is suggested as one of the factors which caused a decrease in the value of λ only in tangential direction. The samples used for measuring thermal properties in tangential direction had a lower average porosity by approximately 6% than the samples used for measuring thermal properties in longitudinal and radial direction.

DISCUSSION λ values of various other bio-based thermal insulations are listed in table 2. The values of existing thermal insulations available on the market and also experimental bio-based thermal insulations are presented. All values were measured at similar conditions; an air temperature of 20 ± 2°C and relative air humidity of 50–65%.

Tab. 2 λ values of various bio-based thermal insulations. Thermal insulations available on the market are divided from experimental materials.

Type of insulation STEICO flex038 Blow-in cellulose (ISOCELL, STEICO) Coconut husk + bagasa Needles of coniferous trees Saw dust Corn pith and alginic acid Mycelium

Thermal conductivity [W·(m·K)-1] 0.038 0.038 0.048–0.068 0.0562–0.0654 0.0568–0.0629 0.042; 0.048 0.078–0.081 0.05–0.07

9

Source Technical documentation of STEICO flex038 product, 2020 Technical documentation of Blow-in cellulose product, 2020 PANYAKAEW, FOTIOS (2011) MUIZNIECE et al. (2015) CETINER and SHEA (2018) PALUMBO et al. (2018) XING et al. (2018) AMSTISLAVSKIJ (2017)

Based on the values listed in Table 2., it is possible to conclude that the λ values of bio-based thermal insulations are in a similar range. The λ value of degraded spruce wood in tangential direction comes the closest to λ values of other bio-based thermal insulations.

It is necessary to note, that the samples measured in the experiment in this paper were all solid whereas the materials listed in Table 2. were measured in an disintegrated state and made into a board.

Higher λ values may be caused by the influence of moisture on thermal properties of wood. Thermal conductivity value increases with increasing moisture content of wood (GLASS and ZELINKA 2010). Considering that raw materials for bio-based thermal insulations are highly hygroscopic and they absorb moisture immediately (PALUMBO et al. 2018), the fact that moisture content influences λ values must be taken in account.

Materials absorb water differently. Their moisture content at the same relative air humidity can differ. Inorganic materials are also able to absorb water. As it was proven in the research by Palumbo et al. (2018), extruded Styrofoam did not absorb water until relative air humidity reached 80% and then absorbed water abruptly, 4% at once. The absorption continued slowly until the material reached a 7% moisture content. The research by LI et al.

(2020) showed, that polyphenolic insulation material reached a moisture content of 6% at a relative air humidity of 65%. Polyurethane insulation material reached a moisture content of only 1.5% at the same relative air humidity.

The equilibrium moisture contents of the materials listed in Table 2 were: at a relative humidity of 65% - wood wool w = 8%; corn pith w = 10% (PALUMBO et al. 2018; saw dust w= 6.5–6.8% (CETINER and SHEA 2018). Equilibrium moisture contents of the samples used in our experiment reached at relative air humidity of 60% were: degraded spruce wood w = 15.2%; degraded beech wood w = 15.0% and degraded oak heartwood w = 14.9%. The experiment was performed in a desorption process, that is why the equilibrium moisture contents are slightly higher. Equilibrium moisture contents of the samples reached at the same relative air humidity but in the process of sorption were: 13.3% in degraded spruce wood; 12.3% in degraded beech wood and 11.9% in degraded oak heartwood. The difference in the moisture contents is small, but it may have had a small impact on the final values of degraded wood´s thermal properties presented in Table 1.

CONCLUSION The main focus of this work was thermal conductivity of wood degraded with the white rot fungus Trametes versicolor. The experiment was performed on three wood species – spruce, beech and oak heartwood. The values of λ in degraded spruce and beech wood conditioned at relative air humidity of 60 ± 5 % were lower than λ values of undegraded spruce and beech wood. In degraded oak heartwood, only the λ value in tangential direction was lower than the λ value of undegraded oak heartwood. To compare the decrease in thermal conductivity values, ratios of degraded and undegraded thermal conductivities are presented: longitudinal direction in spruce 0,8; radial direction 0,64; tangential direction 0,54.

The ratios for beech wood are as follows: λ ratio of degraded and undegraded wood for longitudinal direction 0,82; for radial direction 0,61 and 0,69 for tangential direction. The ratio for oak heartwood in tangential direction is 0,67.

The λ values of degraded spruce wood in tangential direction come close to λ values of experimental bio-based thermal insulations. It is important to note that values of thermal properties presented in this paper were measured in the process of desorption and the moisture content of the material was slightly higher than in the process of sorption which could have had an influence on the final values.

10

Bio-based thermal insulations are a suitable ecological alternative to inorganic thermal insulations. The λ values of bio-based thermal insulation are slightly higher than λ values of inorganic thermal insulations, but they have few advantages. The main advantage is, that the raw material for their production can be planted and raised and the manufacturing of these materials has a minimal impact on the environment.

REFERENCES ALAVEZ-RAMIREZ, R., CHIÑAS-CASTILLO, F., DOMINGUEZ-MORALES, V., J., ORTIZ-GURMAN, M. 2012. Thermal conductivity of coconut fibre filled ferrocement sandwich panels. In Construction and Building Materials, 37: 425–431.

AMSTISLAVSKI, P., YANG, Z., WHITE, M., D. 2017. United States Patent Application Publication; U. S. Patent and Trademark Office: Washington, DC, USA, 2017 CETINER, I., SHEA, A., D. 2018. Wood waste as an alternative thermal insulation for buildings. In Energy & Buildings, 168: 374–384 GLASS, S., V., ZELINKA, S., L. 2010. Chapter 4. Moisture relations and Physical Properties of Wood.

In: Wood Handbook – Wood as an engineering material, Forest Products Laboratory. 2010, Madison, Wisconsin.

HRČKA, R., BABIAK, M. 2017. Wood thermal properties. In Wood in civil engineering. Zagreb: InTech.

HRČKA, R., SLOVÁČKOVÁ, B. 2019. The Method of Wood Emmisivity Measurement. In Acta Facultatis Xylologiae Zvolen, 61(2): 17–24.

ISOCELL technical documentation, 2020. https://www.isocell.com/pdf/products/sk/D%C3%A1tov% C3%BD%20doklad%20k%20v%C3%BDrobku_Zellulose_SK.pdf [3.12.2020] ISOVER technical documentation, 2020. https://www.isover.sk/produkty/isover-woodsil [3.12.2020] LI., Y., SUN, Y., QIU, J., LIU, T., YANG, L., SHE, H. 2020. Moisture absorption characteristics and thermal insulation performance of thermal insulation materials for cold region tunnels. In Construction and Building Materials, 237: 117765 MORAVČÍK, M. and team of authors. 2020. Správa o lesnom hospodárstve v Slovenskej republike za rok 2019 – Zelená správa (Report about forestry in Slovakia in the year 2019 – Green report), Ministry of Agriculture and Rural Development of the Slovak Republic MUIZNIECE, I., BLUMBERGA, D., ANSONE, A. 2015. The use of coniferous greenery for heat insulation material production. In Energy Procedia, 72: 209–215.

MUIZNIECE, I., BLUMBERGA, D. 2016. Thermal conductivity of heat insulation material made from coniferous needles with potato starch binder. In Energy Procedia, 95: 324–329.

OSANYINTOLA, O., F., SIMONSON, C., J. 2006. Moisture buffering capacity of hygroscopic materials: experimental facilities and energy impact. In Energy Build, 38: 1270–1282.

PALUMBO, M., LACASTA, A., M., HOLCROFT, N., SHEA, A., WALKER, P. 2016. Determination of hygrothermal parameters of experimental and commercial bio-based insulation materials, Constr.

Buil. Mater., 124: 269–275.

PALUMBO, M., LACASTA, A., M., GIRALDO, M., P., HAURIE, L., CORREAL, E. 2018. Bio-based insulation materials and their hygrothermal performance in a building envelope system (ETICS). In Energy & Buildings, 174: 147−155.

PANYKAEW, S., FOTIOS, S. 2011. New thermal insulation boards made from coconut husk and bagasse. In Energy and Buildings, 43: 1732–1739.

POŽGAJ, A., CHOVANEC, D., KURJATKO, S., BABIAK, M. 1997. Štruktúra a vlastnosti dreva.

Bratislava: Príroda, a.s., ISBN 80-07-00960-4.

QIN, M., WALTON, G., BELARBI, R., ALLARD, F. 2011. Simulation of whole building coupled hygrothermal-airflow transfer in different climates, In Energy Convers. Manag., 52: 1470−1478.

REINPRECHT, L. 2008. Ochrana dreva., 1. vyd. Zvolen: Technická univerzita vo Zvolene, 2008. 453 S. ISBN 978-80-228-1863-6 RYPÁČEK, V. 1957. Biologie dřevokazných hub. Praha: Nakladatelství Československé Akademie Věd, 209 p.

11

SIMONSON C., J., SALAONVAARA M., OJANEN T. 2004. Heat and mass transfer between indoor air and a parmeable and hygroscopic building envelope: Part II – Verification and numerical studies. In J. Build. Phys. 28: 161–185.

SLOVÁČKOVÁ, B., VIDHOLDOVÁ, Z., HRČKA, R. 2018. Meranie koeficienta tepelnej vodivosti smrekového dreva degradovaného hubou Trametes versicolor. In Drevoznehodnocujúce huby, Zvolen: Technická univerzita vo Zvolene, ISBN 987-80-228-3134-5 STEICO flex038 technical documentation , 2020 https://www.steico.com/fileadmin/steico/content/pdf/Marketing/Czech/Products/STEICOflex_038_ cz_i.pdf [3.12.2020] STN EN 113: 1998. Ochranné prostriedky na drevo. Skúšobná metóda zisťovania ochrannej účinnosti proti drevokazným hubám Basidiomycetes. Zisťovanie hraníc účinnosti.

SVOBODA, Ľ., BAŽANTOVÁ, Z., MYŠKA, M., NOVÁK, J., TOBOLKA, Z., VÁVRA, R., VIMMROVÁ, A., VÝBORNÝ, J. 2005. Stavebné materiály. Bratislava: Jaga, ISBN 80-8076-014-4 VAN DAM, J., E., G., VAN DER OEVER, M., J., A., TEUNISSEN, W., KEIJSERS, E., R., P., PERALTA, A., G. 2004. Process for production of high density/high performance binderless boards from whole coconut husk. In Industrial Crops and Products, 19: 207–216.

XING, Y., BREWER, M., EL-GHARABAWY, H., GRIFFITH, G., JONES, P. 2018. Growing and testing mycelium bricks as building insulation materials. IOP Conf. Ser. Earth Environ. Sci. 2018, 121, 022032.

ACKNOWLEDGEMENTS This work was supported by the Slovak Research and Development Agency (Contract No. 16-0177) and the Internal Project Agency (contract No. 17/2020). The authors expresses gratitude to the Department of Wood Technology at the Technical University in Zvolen for their aid with sample preparation.

AUTHORS’ ADDRESSES Barbora Slováčková Technical University in Zvolen T. G. Masaryka 24 96001 Zvolen Slovakia

12

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 13−23, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.02

MULTI-FACTOR MODELING OF DYNAMICS OF HARDWOOD DENSITY IN THE PROCESS OF THERMOMODIFICATION Ruslan R. Safin − Štefan Barcik − Evgeny Y. Razumov − Petr M. Mazurkin − Albina V. Safina ABSTRACT The results of experiments and statistical modeling of changes in the density of hardwood over the cross section of the samples in the process of vacuum-contact heat treatment are given in the paper. The study of the layer-by-layer density of the samples was performed using a laboratory X-ray Density profile Analyzer - DPX300. In each series of experiments, the change in the density of wood samples was recorded at the different values of temperature and time of heat treatment, sample thickness and layer depth. Modeling the process by the identification method made it possible to obtain one-factor regularities, ranked according to an increase in the correlation coefficient. It was found that the main influence on the density change is exerted by such factors as processing time and temperature. Subsequent modeling made it possible to create a four-factor model of the dynamics of the density of samples during the heat treatment, which allows the prediction of the density profile of the material depending on the thickness of the material and the operating parameters of the process. The proposed model can be further used in the development of new operating parameters of the process of vacuum-conductive thermal modification of hardwoods in order to achieve a uniform degree of processing over the entire section of lumber.

Key words: thermal modification, hard wood, density, correlation coefficient.

INTRODUCTION The process of heat treatment of wood is being currently discussed and developed in Russia and abroad. It is of intense interest to many researchers, since this issue has a huge variety of development courses and a good prospect for in-depth study and the use in the future. Thermally treated wood surpasses untreated one in a number of indicators: it has improved performance characteristics (for example, such as biological stability, size and shape stability), has a decent appearance from light brown to dark aristocratic hues.

Heat treatment is gradually introduced into many wood processing technologies, such as production of massive thermally modified wood, composite materials and solid biofuels (BEHR et al. 2018, GALYAVETDINOV et al. 2016, SAFIN et al. 2015a, KHASANSHIN et al.

2016, SAFIN et al. 2015b). Chemical compositions are not used in the process of heat treatment; therefore, when the European Commission banned the use of chemically treated wood (2004), this type of modification of plant raw materials has become primary, both for producers and consumers of these products. The East European Plain ecosystem is 13

represented by taiga-broadleaved forests with spruce, pine, aspen, mountain ash, birch, oak, and other tree species. Hardwood species are of great interest, and their thermal modification allows the improved performance and improved decorative properties of wood.

Thermal modification is carried out at the processing temperatures from 180 to 2400С, while the wood undergoes thermochemical transformations causing a change in its physical properties. The feature of the thermal modification process is the change in properties throughout the volume of the material. The number of researchers claim (MÖTTÖNEN 2006, SALIN 2010, ESPINOZA et al. 2016, SANDOVAL-TORRES 2012, OUERTANI 2015) that the most suitable method for assessing the degree of heat treatment is the value describing the change in wood density. Still there are no theoretical and experimental data characterizing the change in the density of wood by layers in the process of thermal modification. However, the process of monitoring the changes in the density of the material by layers will ensure the uniformity of thermal modification of wood throughout the section of the workpiece, and, thereby, prevent possible negative phenomena in service. Thus the authors set the following objectives: - to investigate the change in the wood density of the main hardwood (aspen, birch and oak) over the cross section of the samples in the process of thermal modification; - to identify single-factor patterns and, according to an increase in the correlation coefficient, to identify the factors of the greatest influence on the change in the wood density in the process of heat treatment; - to conduct multi-factor process modeling with the use of an identification method to further control the process and ensure smooth heat treatment.

MATERIALS AND METHODS To research the process of thermal modification, the samples of aspen, birch and oak wood without wood defects with a cross section of 20 × 50 mm and the length of 250 mm were prepared. The samples of each tree species were taken from one radial sawing board, that helped ensure the same structure of growth rings along the tangential sections of the sample.

The bars underwent vacuum drying at 60 ºC for 5 days, which ensured a smooth moisture distribution over the cross section of the samples in the range between 5−6%.

All samples were weighed in order to determine the average density for the entire batch prior to the beginning of the experiment. The samples with a deviation of their density from the average value throughout the batch of not more than 1% were accepted for further experiments. The initial temperature of the samples was taken to 20 0С during their room storage.

48 samples of each tree species were selected to carry out the studies on the change in the density of wood during thermal modification. A number of experiments with varied parameters such as sample thickness, temperature and processing time were carried out using experimental equipment with diagram given in Fig. 1.

The equipment for the heat treatment of wood contains a chamber 1, which is connected to a vacuum line consisting of a vacuum pump 2 and a condenser 3. The supply of thermal energy to the treated timber 4 is carried out by means of a contact method using heat-supplying surfaces 5 and 6. They are perforated metal plates heated by filaments 7.

The plates are thermally insulated from the side opposite to the material processed with porous moisture-proof and breathable material 8. While conducting the experiment, the chamber is sealed with a cover 9. As a result of close contact of sample 4 with heating surfaces 5 and 6, heating of the sample can be observed. The heating temperature is controlled by a thermocouple 10 installed in the plate, a control electronic device 11 and a 14

control panel 12. The pressure in the chamber is recorded by a pressure gauge 13.

1 - chamber, 2 - vacuum pump, 3 - condenser, 4 - treated timber, 5,6 - heat-supplying surfaces, 7 - filaments, 8- porous heat insulator, 9 - camera cover, 10 - thermocouple, 11- control electronic device, 12 - control panel, 13 - pressure gauge.

Fig. 1 Experimental equipment of vacuum-contact heat treatment of wood.

In each series of experiments, four samples of a certain thickness (4, 12, or 20 mm) were simultaneously loaded into a chamber preheated to the required temperature, which were then subjected to the thermal modification under the same conditions, but at different temperatures (180, 200, 220, and 240 ºC). After a specified time interval (2, 3, 4, and 5 hours), one sample was taken out of the chamber and immediately placed in a desiccator for cooling without moisture gain.

The change in density over the thickness of the obtained samples was assessed with the XRAY Density Profile Analyzer - DPX300 (Fig. 2), which contains a radiation source (X-ray tube), a detector, and a data acquisition and processing unit.

a)

b)

Fig. 2. Measurement of sample density: a - appearance of the X-RAY Density Profile Analyzer - DPX300, b - sample in the equipment.

The non-contact technology allows scanning the samples under study by irradiation with primary X-ray radiation and, based on the response X-rays, registering the change in density over the cross section of the material with a step of 1 mm with a detector. The high sensitivity of the analyzer ensured small measurement errors, and the compatibility with Windows made it possible to obtain graphs of the density change during the measurement procedure.

RESULTS AND MODELLING An array of data was obtained as a result of the experiments carried out for each type of wood, which makes it possible to study the dynamics of wood density through multivariate modeling by the identification method (AHMED et al. 2008, WIECZOREK 2017, SERGIENKO 2014, ROFFEL et al. 2004).

15

The process of multi-factor modeling on the example of aspen is described in detail.

To do so, the data obtained within each series of experiments is combined in Table 1.

Tab. 1 Initial data for multi-factor modeling of the process of changing the density of the aspen samples in the process of thermal modification.

Item numb er 1 2 3 4 5 6 …

11 12 …

15 16 …

24 25 …

55 56 …

143 144 …

169 170 …

207 208

Sample thickness s , (mm) 4 4 4 4 4 4 …

4 4 …

4 4 …

12 12 …

12 12 …

20 20 …

20 20 …

20 20

Processing temperature t , (°С) 180 180 180 180 200 200 …

220 220 …

240 240 …

180 180 …

200 200 …

200 200 …

220 220 …

240 240

Wood Processing i -th depth Relative Absolute err. density time of layer 3  , (kg/m )  err.

3 , (%)  , (h) x , (mm)  , (kg/m ) 2 1 535 -3.51117 -0.66 2 2 536 -3.32837 -0.62 3 1 533 -3.59447 -0.67 3 2 535 -2.41177 -0.45 2 1 528 -4.80906 -0.91 2 2 529 -4.62631 -0.87 … … … … …

3 1 516 -2.19333 -0.43 3 2 517 -2.01053 -0.39 … … … … …

2 1 493 -4.63393 -0.94 2 2 495 -3.45113 -0.70 … … … …

3 2 536 -1.02577 -0.19 3 4 538 -0.226268 -0.04 4 4 …

5 5 …

4 4 …

5 5

4 6 …

1 2 …

4 6 …

8 10

519 521 526 528 …

520 522 499 500

0.247113 0.959613 …

-2.58409 -1.40135 …

1.42682 2.13934 …

0.789271 0.0647711

0.05 0.18 …

-0.49 -0.27 …

0.27 0.41 …

0.16 0.01

Similar tables of initial data for subsequent mathematical modeling of the process were obtained for the samples of birch and oak wood.

The experimental data were processed in the CurveExpert-1.40 software environment, which made it possible to obtain graphs based on the existing array of points and select the optimal dependences. The analysis of deviations of the obtained experimental data from the obtained model was carried out along with the selection. Further, the obtained regularities were ranked according to the increasing correlation coefficient.

One-factor patterns were identified at the first stage of modeling. The factors according to the growth of their influence on wood density according to the increase in the correlation coefficient were arranged.

In Figure 3, in the upper right corner are located: S – dispersion; r – correlation coefficient characterizing the levels of adequacy of the laws: above 0.7 – a strong relation between the factors; 0.5-0.7 – medium relation; 0.3-0.5 – weak relation (WIECZOREK 2017).

16

, kg/m3

, mm Fig. 3 The effect of thickness on the density of aspen samples.

The one-factor function of the effect of sample thickness on density with a correlation coefficient of 0.0088 is determined by the formula: (s) = 686.52127exp(−0.25658s 0.0025686 ) .

(1)

Thus, the aspen density decreases exponentially with average calculations with increasing sample thickness.

The data of graph 4 shows that when the correlation coefficient is 0.0969 (less than 0.3), the effect of the layer depth on the density of the samples becomes more obvious according to the law of exponential growth: ( х ) = 527.80274exp(0.00079932x1.18730 ) .

(2)

, kg/m3

, mm Fig. 4 The effect of layer depth on the density of aspen samples.

An increase in the density of the wood of the studied samples with an increase in the depth of the layer is observed.

The processing time is the next factor with an adequacy of 0.7199: () = 557.00000exp( −0.14383 0.43969 ) + .

(3) + 60.86110 0.55309 exp( −0.28278 0.12405 ) The indicated regularity includes two components. The first was the law of exponential death, and the second was the biotechnical law. The graph 5 shows that the impact of the first component was significant.

17

, kg/m3

, h

Fig. 5 The effect of the time of processing on the density of aspen samples.

And, finally, the highest correlation coefficient of 0.9836 has a factor in the influence of temperature on the change in density of wood samples (Fig. 6) according to the inverse Weibull law: ( t ) = 557.15069 − 0.37191exp(0.021100t ) .

(4) According to this formula, the exponential law of growth is subtracted from the initial density at growth rate 1. , kg/m3

t, 0C Fig. 6 The effect of temperature on the density of aspen samples.

Thus, single-factor patterns were determined, which were further ranked according to an increase in the correlation coefficient as a result of the modeling: 1) the function of the effect of sample thickness on density with a correlation coefficient of 0.0088; 2) the function of the effect of layer depth on density with a correlation coefficient of 0.0969; 3) the function of the effect of processing time on the density with a correlation coefficient of 0.7199; 4) the function of the effect of processing temperature on the density with a correlation coefficient of 0.9836.

Later we match the obtained data to get a multifactor model of the dynamics of the density of the hardwood sample in the process of heat treatment. Therefore, we take the absolute errors from model (1) and find their dependence on the next most important factor - the depth of the x layer. The values of the absolute errors  are given in Table 1.

s (х) = −557.55228exp(−5.49169x 0.076154 ) + 0.098879x1.71497 .

18

(5)

There was a slight increase in the correlation coefficient from 0.0969 (Fig. 4) to 0.1009. The modeling results are given in Figure 7. , kg/m3

, mm Fig. 7 The dependence of the absolute error of the thickness of the work material on the depth of the layer of the studied samples of aspen.

Then we find the dependence of the total absolute errors of the work material thickness and the layer depth on the next most important factor - the processing time  (Fig. 8).

 s,x () = 26.90364exp(−0.0163141.76265 ) − 1.30535  0.28704 exp(−0.089971 1.06196 )

.

(6)

, kg/m3

, h

Fig. 8 The dependence of the absolute error of the thickness and depth of the layer of the work material on the processing time.

Compared with Fig. 5, the correlation coefficient became 0.7277 instead of 0.7199 (in single-factor modeling).

Next, the dependence of absolute errors on the last factor-the processing temperature t is determined:  s,x , ( t ) = 2.86459 10−7 exp(8.41213t 0.15959 ) − 2.15207 10−9 t 4.59846 (7) .

Figure 9 shows the result of modeling this dependency.

19

, kg/m3

t, 0C Fig. 9 The dependence of the absolute error of the thickness, the depth of the layer and the processing time on the processing temperature.

According to calculations, a correlation coefficient of 0.9883 was obtained, which is slightly higher than the value in the single-factor model of 0.9836 (Fig. 6).

A four-factor model of the density dynamics of wood samples during thermo modification is presented in the form of a formula:

(s, х, , t) = (S) + s (х) + s,x () + s,x, (t) ,

(8)

where:

 (s ) - function of the influence of the sample thickness on the density according to the equation

(1);

 s (х ) - absolute accuracy of the sample thickness versus the layer depth of the samples under

study according to the equation (5);  s , x ( ) - absolute accuracy of the sample thickness and layer depth versus processing time according to the equation (6);  s , x , (t ) - absolute accuracy of the sample thickness, layer depth and processing time versus sample processing temperature according to the equation (7).

The relative error of the model at each point is determined by the formula:  = 100 / ф ,

(9)

where  ф - the actual values of the density of aspen according to Table 1. The maximum relative error of the four-factor model for aspen wood was 1.18%.

Figures 7-9 show small fluctuations confirming the vibrational nature of the process of thermal modification of wood (ZHOU et al. 2019, DEFO et al. 2004). Fluctuations occur due to frequent removal of the samples and rapid cooling to register the density profile.

Similarly, the dynamics of the density of birch and oak in the process of thermal modification were simulated.

When modeling the dynamics of the density of birch, we found that, unlike aspen and oak, the first factor with the lowest correlation coefficient is the depth of the layer, and the thickness of the work material goes second in terms of the adequacy rating of single-factor models. The highest correlation coefficient of 0.8845 has an effect of temperature on the change in density of hardwood samples.

The four-factor model for birch wood is given as a formula:

(x, s, , t) = (x) +  x (s) +  x ,s () +  x,s, (t) ,

20

(10)

where:

( x ) = 533.53307exp(0.00058880 x ) ,

(11)

 x (s) = 75.90639− 38.36827S 0.25332 ,

(12)

 x ,s () = 63.32358exp(6.6978 0.013788 ) − 51277.8183 0.093219 ,

(13)

 x ,s, ( t ) = 9.07283 10−6 exp(6.75709t 0.19455 ) − 1.97986 10−6 t 3.86043

.

(14)

a correlation coefficient of 0.6771 was obtained for the average constraint force between all factors, which is significantly less than the calculated above indicator. The maximum relative error of the four-factor model for birch was 13.39%.

The four-factor model for oak is presented as a formula:

 (s, х, , t) =  (S ) +  s ( х) +  s, x ( ) +  s, x, (t ) ,

(15)

(s) = 2026.66952exp(−0.99239s 0.0050216 ) ,

(16)

where:

 s ( x) = −281.48238exp(−1.8847610−5 x3.02178 ) + 276.03376x0.013825 ,

(17)

 s , x ( ) = 93.28925exp(0.27708 1.01788 ) + 84.02560 ,

(18)

 s,x , ( t ) = 2.37693exp(0.0014055t 1.61074 ) − 1.24980 10−30 t 14.05239 .

(19)

As a result of the simulation, a correlation coefficient of 0.7017 of a strong relation between all factors was obtined. The maximum relative error of the four-factor model for oak was 14.57%.

CONCLUSION The processes of thermal modification of wood are currently under study and of great interest both to manufacturers and to many researchers whose aim is to study and improve technologies. One of the important parameters for assessing the quality and degree of heat treatment of wood is the change in density over the cross section of the material. In this regard, monitoring the density drop of the material by layers will ensure the uniformity of thermal modification of wood throughout the entire section of the sample, thereby prevent possible negative phenomena in operation.

Numerous experimental studies of hardwood samples (aspen, birch and oak) of various thicknesses at different temperatures and processing times were carried out in order to study the dynamics of changes in wood density during heat treatment.

Multi-factor modeling of the process was carried out by the method of identification based on the results of the experimental studies. The obtained experimental data were processed in the CurveExpert-1.40 software environment, as a result of which the authors built graphical dependencies, performed regression analysis, and identified one-factor patterns, which were further ranked according to an increase in the correlation coefficient.

According to the rating of the adequacy of one-factor models for the considered tree species, strong factor relations with a correlation coefficient of not less than 0.7 were found under the influence of time and temperature of processing on the density of the samples under study.

The four-factor models of the dynamics of the density of aspen, birch and oak wood in the process of heat treatment were obtained based on the results of one-factor modeling by 21

the method of identification. Those models allow the prediction of the change in density depending on the thickness of the material and the parameters of the technological process.

The calculated correlation coefficients indicate a strong relation between all factors, while the greatest influence on the change in wood density is exerted by such factors as processing time and temperature. The maximum relative error of the presented models does not exceed 15%, which proves that they could be used in modeling the processes of vacuum-conductive thermal modification in order to optimize the operating parameters and ensure a uniform degree of processing over the entire section of lumber.

REFERENCES AHMED, R., HALL, T., WERNICK, P., ROBINSON, R., SHAH, M. 2008. Software process simulation modeling: A survey of practice. In Journal of Simulation, 2008, Volume 2, Issue 2.

BEHR, G., BOLLMUS, S., GELLERICH, A., MILITZ, H. 2018. Improvement of mechanical properties of thermally modified hardwood through melamine treatment. In Wood Material Science and Engineering, 2018, vol. 13, issue 5.

DEFO, M., FORTIN, Y., CLOUTIER, A. 2004. Modeling Superheated Steam Vacuum Drying of Wood.

In Drying Technology, 2004, Volume 22, Issue 10.

ESPINOZA, O., BOND, B. 2016. Vacuum Drying of Wood—State of the Art. In Curr Forestry Rep, 2016, Volume 2, Issue 4, pp 223–235.

GALYAVETDINOV, N.R., SAFIN, R.R., VORONIN, A.E. 2016. Analysis of physico-mechanical properties of composites based on polylactide and thermally modified wood fibers Materials Science Forum. In International Conference on Industrial Engineering, 2016, Volume 870, pp 202−206, Chelyabinsk; Russian Federation; 19 May 2016 to 20 May 2016.

KHASANSHIN, R.R., SAFIN R.R., RAZUMOV, E.Y. 2016. High Temperature Treatment of Birch Plywood in the Sparse Environment for the Creation of a Waterproof Construction Veneer. In Procedia Engineering, 2016, Volume 150, 2016, pp 1541−1546.

MÖTTÖNEN, V. 2006. Variation in Drying Behavior and Final Moisture Content of Wood during Conventional Low Temperature Drying and Vacuum Drying of Betula pendula Timber. In Wood Technology, 2006, Volume 24, Issue 11.

OUERTANI, S., KOUBAA, A., AZZOUZ, S. et al. 2015. Vacuum contact drying kinetics of Jack pine wood and its influence on mechanical properties: industrial applications. In Heat Mass Transfer, 2015, Volume 51, Issue 7, pp 1029–1039.

ROFFEL, B., BETLEM, B.H. 2004 Process Modeling and Identification. In Advanced Practical Process Control. Springer, Berlin, Heidelberg pp 45−71.

SAFIN R.R., KHASANSHIN, R.R., SHAIKHUTDINOVA, A.R., ZIATDINOV, R.R. 2015a. The technology for creating of decorative plywood with low formaldehyde emission. In IOP Conference Series: Materials Science and Engineering, 2015, Volume 93, Issue 1, National Research Tomsk Polytechnic UniversityTomsk; Russian Federation; 5 October 2015 to 9 October 2015.

SAFIN, R.R., SHAIKHUTDINOVA A.R., KHASANSHIN R.R., AKHUNOVA L.V., SAFINA A.V. 2015b.

Improving the energy efficiency of solid wood fuel. In International Multidisciplinary Scientific GeoConference Surveying Geology and Mining Ecology Management, SGEMVolume 1, Issue 4, 2015, Pages 315−322 15th International Multidisciplinary Scientific Geoconference and EXPO, SGEM 2015; Albena; Bulgaria; 18 June 2015 to 24 June 2015.

SALIN, J.-G. 2010. Problems and solutions in wood drying modelling: History and future. In Wood Material Science and Engineering, 2010, Volume 5, Issue 2.

SANDOVAL-TORRES, S., JOMAA, W., MARC, F. et al. 2012. Colour alteration and chemistry changes in oak wood (Quercus pedunculata Ehrh) during plain vacuum drying. In Wood Sci Technol, 2012, Volume 46, Issue 1–3, pp 177–191.

SERGIENKO I.V. 2014. Mathematical Modeling and Analysis of Complex Processes. In Topical Directions of Informatics. Springer Optimization and Its Applications, vol 78. Springer, New York, NY pp 153−188.

22

WIECZOREK, W. 2017. Identification Using Mathematical Modeling. In Grammatical Inference.

Studies in Computational Intelligence, Volume 673. Springer, Cham. pp 57−68.

ZHOU, Z., ZHANG, P., HUAI, B., HUANG, L. 2019. System Identification of Wood Drying Process Based on ARMAX Model. In Agricultural Sciences, Volume 10, 241−248.

AUTHORS‘ ADDRESSES

Ruslan Rushanovich Safin Kazan National Research Technological University 68 Karl Marx street Kazan 420015 Russian Federation cfaby@mail.ru Štefan Barcík Department of Manufacturing and Automation Technology Faculty of Technology Technical University in Zvolen T.G.Masaryka 24 SK-96001 Zvolen Slovak Republic barcik@tuzvo.sk Evgeny Yurevich Razumov Czech University of Life Sciences Prague Kamýcká 129, 16500 Praha 6 - Suchdol Czech Republik evgeny.razumov2011@yandex.ru Petr Matveevich Mazurkin Volga State University of Technology, 3 Lenin street Yoshkar-Ola 424000 Russian Federation kaf_po@mail.ru Albina Valerievna Safina Kazan National Research Technological University 68 Karl Marx street Kazan 420015 Russian Federation alb_saf@mail.ru

23

24

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 25−34, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.03

MODIFICATION OF MAPLE WOOD COLOUR DURING THE PROCESS OF THERMAL TREATMENT WITH SATURATED WATER STEAM Michal Dudiak ABSTRACT The technological process of colour modification and the associated parallel processes of changing the acidity and density of maple wood modified in pressure autoclave with saturated water steam is evaluated in the paper. Maple wood was thermally treatment at the temperatures of: tI = 105 ± 2.5 °C, tII = 125 ± 2.5 °C and tIII = 135 ± 2.5 °C for τ = 3, 6, 9 and 12 hours. Direct pH measurement of maple wood with the moisture content above the fiber saturation point (FSP) was performed using SI 600 pH meter with LanceFET + H puncture probe. The acidity values of maple wood ranged from pH = 5.1 in the case of native wood, and the acidity of the wood changed to pH = 3.3 due to the temperature and time of modification. The density of maple wood was determined using the RADWAG KIT 128 instruments. The total change in the density of wood during the modification process was 5.3 % lower than the density of native wood. Statistical evaluation of the measured results showed the correlations between the total colour difference and pH. Key words: maple wood, colour modification, acidity, wood density, saturated water steam.

INTRODUCTION Wet wood placed in the environment of hot water, saturated water steam, or saturated humid air is heated and its physical, mechanical, as well as chemical properties, are changed.

Thermal treatment of wood, besides physical and mechanical changes applied in the process of manufacturing veneers, plywood, bentwood furniture or pressed wood are accompanied with the changes in chemical properties and colour of the wood (KOLLMANN and GOTE 1968, TREBULA 1986, MOLNAR and TOLVAJ 2002, TOLVAJ et al. 2010, DZURENDA and ORLOWSKI 2011, DZURENDA 2013, BARANSKI et al. 2017).

In the past, colour changes when wood becoming darker during the steaming process were used to remove the undesirable colour differences between light-coloured sapwood and dark-coloured heartwood or to eliminate wood stain colours as a result of mold. In recent times, research into thermally modified wood has been focused on the issue of the colour change of specific tree species into more or less bright hues or wood imitation of domestic or exotic tree species (MOLNAR and TOLVAJ 2002, TOLVAJ et al. 2009, DZURENDA 2014, 2018a, b, c, BARCIK et al. 2015, BARANSKI et al. 2017).

The effect of heat on wet wood is also initiated by chemical changes in wood. The first chemical reactions include partial hydrolysis of hemicelluloses and extraction of watersoluble substances (FENGEL and WEGENER 1984, BUČKO 1995, SOLÁR 2004, 25

SUNDQVIST et al. 2006, SAMEŠOVÁ et al. 2018). Depending on the temperature and time of the activity of the hydrolysis products, e.g. acetic acid and formic acid, degradation of polysaccharides occurs. During the thermal treatment of wood, dehydration of pentoses to 2-furaldehyde as well as oxidation of carbohydrates also occurs. New chromophoric groups begin to form in lignin, causing the colour change of wood (FENGEL and WEGENER 1984, BUČKO 1995, HON and SHIRAISHI 2001, SOLÁR 2004, SUNDQVIST et al. 2006, GEFFERT et al. 2019).

The aim of this paper is to present changes in the properties of maple wood, such as density, acidity, and colour in the process of thermal treatment with saturated water steam in the temperature ranging from t = 105 °C to 135 °C and for the time from τ = 3 to 12 hours.

MATERIAL AND METHODS The wood of Acer pseudoplatanus in the form of blanks with dimensions: the thickness of 38 mm, the width of 90 mm and the length of 750 mm in 260 pieces were divided into 13 groups of 20 pieces in one group. The initial moisture content of wet maple wood was in the range between w = 57.2 ± 3.1 %. Blanks in the group 1 were not thermally treated. The other blanks were divided into 12 groups of 20 pieces and thermally treated with saturated water steam at the temperatures t = 105 °C, t = 125 °C, and t = 135 °C for 3, 6, 9, and 12 h. Thermal treatment of maple wood with saturated water steam was carried out in a pressure autoclave APDZ 240 (Himmasch AD, Haskovo, Bulgaria) installed in the company Sundermann s.r.o.

Banská Štiavnica (Slovakia).

The mode of thermal treatment in order to modify the colour of maple wood with saturated water steam is illustrated in Fig. 1. The conditions of thermal treatment of individual modes of colour modification are described in Tab. 1.

Fig. 1 Mode of colour modification of maple wood with saturated water steam.

The thermal process of maple wood colour modification was performed in the pressure autoclave at a higher saturated water steam pressure than atmospheric pressure. Saturated water steam temperatures in individual colour modes are given in Table 1. The temperatures tmax and tmin are the temperature intervals at which saturated water steam is fed into the autoclave to carry out the technological process. The temperature t4 is the temperature of the saturated water steam in the autoclave after reducing the water steam pressure in the autoclave to the atmospheric pressure to allow the safe opening of the pressure equipment and sampling after the time of thermal treatment.

26

Tab. 1 Modes of colour modification of maple wood with saturated water steam.

Temperature of saturated water steam Mode I Mode II Mode III

tmin

tmax

t4

102.5 122.5 132.5

107.5 127.5 137.5

100 100 100

Length of time the wood is exposed to colour modification τ1 = 3 h

τ2 = 6* h

τ3 = 9* h

τ4 = 12* h

Note: * After individual time of colour modification, exactly according to the determined diagram (course), 0.5 hour must be added to the given time of thermal modification of maple wood colour. This time between the individual sections of the modification serves to reduce the pressure for the safe opening of the autoclave and handling-selection of one group of modified blanks from the autoclave-closing and pressurization of the autoclave with saturated steam.

The moisture content of maple wood before entering the technological process of thermal treatment was determined by a random selection of 5 samples. Similarly, the moisture of thermally treated blanks was determined in individual modes and times after their selection from the autoclave and cooling to ambient temperature. The moisture content of wet maple wood was determined by the gravimetric method according to the standard STN EN 13183-1 (2003).

The acidity measurement of wet maple native and thermal treatment wood was performed using SI 600 pH meter with a functional Lance FET+H probe (Sentron, Roden, The Netherlands). Three measurements were performed on all samples in the middle of the thickness of the blank and 100 mm from the front of the blank. Using an accu drill (DeWalt DCD791NT, Germany), 12 mm diameter hole was made in the measuring point. The drilled wet sawdust was pushed back into the hole with a glass rod, where the LanceFET+H sensor head was inserted (GEFFERT et al. 2019). After about 60 seconds of stabilization, the pH value was read on SI 600 pH meter.

Dried colour-modified maple wood is used as a material for manufacturing the furniture, flooring, or interior tiling in the dry state. For this reason, the samples of untreated and thermally treated maple wood were dried in the convection hot air dryer KAD 1x6 (KATRES s.r.o. Czech Republic) with mild drying regime to the final moisture content w = 12 ± 0.5 % (DZURENDA 2020). Subsequently, the surface of dry blanks was machined using the FS 200 milling machine (BENET Trading, Kvasiny, Slovakia).

The colour of both thermal treatment and untreated maple wood in the CIE L*a*b* colour space was determined using Color Reader CR-10 colorimeter (Konica Minolta, Japan). The measurement was performed on the loading and side surfaces at a distance of 300 mm. A D65 light source with an illuminated area of 8 mm was used. The colour was evaluated based on the changes in the colour space CIE L*a*b* on the luminance coordinate L*, the red colour a*, the yellow colour b*, and the total colour difference ∆E*.

The value of the total colour difference is described with the equation: E  =

(L

 2

) ( 2

) ( 2

− L1 + a 2 − a 1 + b 2 − b1

)

2

(1)

where: L*1, a*1, b*1 are the values of the coordinates of the colour space of the dried milled thermally untreated maple wood surface, L*2, a*2, b*2 are the values of the coordinates of the colour space of the dried milled thermally treated maple wood surface.

Test samples for measuring density with the following dimensions were prepared from native and thermally treated blanks in individual modes and times: the thickness of 15 mm, the width of 50 mm, and the length of 100 mm. The produced samples were dried in the laboratory oven (MEMMERT UM110m, Germany) at the temperature of t = 103 ± 2 °C to constant weight. The density of maple wood in each mode and time of treatment was determined using the instrument RADWAG KIT 128.

27

DISCUSSION Physical properties of maple wood, such as wood moisture and acidity of wet wood before the heat treatment process, as well as wood density, colour space coordinates CIE L*a*b* in the dry state are given in Table 2.

Tab. 2 Measured values of density, colour space coordinates CIE L*a*b* of dry wood, moisture and acidity of wet maple wood.

Wood

Wood density ρ0 [kg∙m-3]

Dry maple wood Coordinate colour space CIE L*a*b* L* a* b*

Acer pseudoplatanus

600.1 ± 50.1

86.6 ± 1.2

5.9 ± 0.5

16.4 ± 0.5

Wet maple wood Moisture content Acidity w [%] pH 57.2 ± 3.1

5.1 ± 0.2

The measured values of the density of maple wood in the dry state are the average values of the density of healthy maple wood not damaged by fungi or molds. Similar values of maple wood density for the territory of Central Europe are given in the works: POŽGAJ et al. 1997, MAKOVÍNY 2010, KURJATKO et al. 2010. Based on the above statement, it is possible to mention the analyzed changes in the properties of maple wood achieved by thermal treatment in individual modes as representative (standard).

The results of laboratory work determining the density of thermally treated maple wood by individual modes in the dry state are given in Table 3.

Tab. 3 The densities of the dry thermally treated maple.

Temperature of saturated water steam tI = 105 ± 2.5 °C tII = 125 ± 2.5 °C tIII = 135 ± 2.5 °C

3 ρ0[kg/m3] 590.5 ± 27.8 587.9 ± 29.1 581.4 ± 21.9

Time of thermal modification of maple wood [h] 6 9 12 ρ0 [kg/m3] ρ0 [kg/m3] ρ0 [kg/m3] 586.4 ± 25.2 584.1 ± 29.3 582.9 ± 22.8 585.2 ± 22.6 580.4 ± 25.9 577.1 ± 20.9 577.1 ± 24.5 573.2 ± 21.0 568.4 ± 20.9

During the process of thermal modification of maple wood colour, weight loss occurs depending on the temperature of saturated water steam and the time of modification. The weight loss is mainly due to the process of hydrolysis of polysaccharides and extraction of water-soluble substances in wood (LAUROVÁ et al. 2004, SOLÁR 2014, VÝBOHOVÁ et al.

2018, GEFFERT et al. 2020). The reduction of the density of maple wood in the given modification process represents the value of by ρ0 ≤ 5.3 %. The changes in the density of colour-modified maple wood do not exceed the limits of the natural tolerance of maple wood densities, which is in the Central European area: ρ0 = 480-590-750 kg.m-3.

The moisture thermally treated maple wood for 3, 6, 9, 12 h measured after cooling the wood to ambient temperature is given in Table 4.

Tab. 4 Average values of maple wood moisture during the thermal treatment process Temperature of saturated water steam tI = 105 ± 2.5 °C tII = 125 ± 2.5 °C tIII = 135 ± 2.5 °C

3 w [%] 51.6 ± 1.9 52.3 ± 1.2 51.8 ± 1.6

Time of thermal modification of maple wood [h] 6 9 12 w [%] w [%] w [%] 51.2 ± 1.2 49.5 ± 1.0 48.9 ± 1.1 50.9 ± 1.4 50.2 ± 1.2 48.2 ± 1.1 51.0 ± 1.1 49.4 ± 1.1 47.9 ± 0.9

28

Reducing the maple wood moisture by ∆w ≈ 13.9 % resulting from the thermal treatment process is caused by evaporation of water from the wood into the environment in the autoclave during cooling to t = 100 °C. The source of heat for vaporization and evaporation of water from wood is the heat accumulated during the heating of wood to the required technological temperature (DZURENDA and DELIISKI 2000, DZURENDA 2018c).

Acidity values measured by the direct pH measurement method during the thermal treatment of maple wood at regular time intervals, after 3, 6, 9, 12 h are mentioned in Table 5.

Tab. 5 Measured average pH values of maple wood during the process of thermal treatment Temperature of saturated water steam tI = 105 ± 2.5 °C tII = 125 ± 2.5 °C tIII = 135 ± 2.5 °C

3 pH [-] 4.8 ± 0.3 4.1 ± 0.1 3.9 ± 0.1

Time of thermal modification of maple wood [h] 6 9 12 pH [-] pH [-] pH [-] 4.7 ± 0.1 4.6 ± 0.2 4.2 ± 0.2 3.9 ± 0.2 3.5 ± 0.3 3.3 ± 0.2 3.6 ± 0.2 3.3 ± 0.2 3.3 ± 0.1

The wet wood acidity in the case of most temperate tree species ranges from pH = 3.3 to 6.4 (SANDERMANN and ROTHKAMM 1956, IRLE 2012, SOLAR 2014, GEFFERT et al.

2019). The acidic reaction of most woody plants is caused by free acids and acidic groups found in an aqueous solution of dilute sugars, organic acids, and water-soluble inorganic substances fed by the root system to the tree in cell lumens (ČUDINOV 1968, BLAŽEJ et al.

1975, ZEVENHOVEN 2001, PŇAKOVIČ and DZURENDA 2015 ). The measured acidity values of wet untreated maple wood do not differ from the acidity values of maple wood with the moisture content above the saturated fiber point reported by the authors (SOLAR 2004, GEFFERT et al. 2019).

A decrease in the pH values of thermally modified maple wood by individual colour modification modes confirms the known fact about the course of hydrolysis of hemicelluloses in wet hardwood wood by the action of heat, as reported by MELCER et al.

1989, LAUROVÁ et al. 2004, SUNDQVIST et al. 2006, SAMEŠOVÁ et al. 2018, VÝBOHOVÁ et al. 2018, GEFFERT et al. 2019, DZURENDA et al. 2019).

The colour of dried, planed untreated maple wood and colour hues obtained during thermal treatment modes with saturated water steam is shown in Fig. 2. EPSON PERFECTION V850 PRO scanner with the quality of the created scan of maple wood colour samples of 1200 dpi was used in order to visually evaluate the colour change in the process of thermal treatment of maple wood by individual modes and for specific times.

Mode I t = 105 °C

29

Mode II t = 125 °C

Mode III t = 135 °C

Fig. 2 Changes in the colour of maple wood during the process of thermal treatment

A light white colour with yellow tinge of dry untreated maple wood was identified in the colour space CIE L*a*b* by the coordinates L* = 86.6 ± 1.2; a* = 5.9 ± 0.5; b* = 16.4 ± 0.5. The values given are comparable to the values of colour coordinates given for maple wood by the authors BABIAK et al. 2004, MEINTS et al. 2017.

The coordinate values describing the colour of the maple wood before and after the thermal treatment by the individual modes in the colour space CIE L*a*b* measured by colorimeter Color Reader CR-10 and the total colour difference ∆E* are given in Table 6.

Tab 6. Measured values of the coordinates L*, a*, b* in the colour space CIE L*a*b*, values of the total colour difference ∆E* of maple wood during the process of thermal modification.

Temperature of saturated water steam tI = 105 ± 2.5 °C

tII = 125 ± 2.5 °C

tIII = 135 ± 2.5 °C

Coordinates CIE L*a*b* L* a* b* ∆E* L* a* b* ∆E* L* a* b* ∆E*

Time of thermal modification of maple wood [h] 3 6 9 12 81.0 ± 0.4 79.1 ± 0.3 77.4 ± 0.6 76.2 ± 0.9 8.5 ± 0.3 8.9 ± 0.4 10.6 ± 0.6 10.9 ± 0.5 17.6 ± 0.5 17.7 ± 0.6 18.3 ± 0.6 18.9 ± 0.4 5.5 7.4 9.7 10.9 75.6 ± 0.6 73.0 ± 0.6 69.7 ± 0.8 69.3 ± 1.0 12.0 ± 0.4 12.0 ± 0.4 12.3 ± 0.5 12.4 ± 0.3 16.0 ± 0.5 16.1 ± 0.4 16.6 ± 0.5 17.1 ± 0.7 11.3 13.3 16.3 17.4 69.3 ± 0.8 65.3 ± 0.8 61.8 ± 0.7 59.1 ± 0.6 11.9 ± 0.5 11.5 ± 0.4 11.6 ± 0.2 11.8 ± 0.2 16.3 ± 0.3 17.0 ± 0.5 17.3 ± 0.4 17.8 ± 0.4 17.2 21.2 24.2 27.3

30

From a decrease in the value of lightness L0* = 86.6 of thermally untreated wood to the value of L1* = 76.2 of thermally treated maple wood with saturated water steam with a temperature tI = 105 ± 2.5 °C, to the value of L2* = 69.3 during the treatment of wood with water steam at the temperature of tII = 125 ± 2.5 °C, to the value of L3* = 59.1 at saturated water steam temperature tIII = 135 ± 2.5 °C, it is clear that increasing wood temperature results in the change in the brightness of the thermally treated maple wood, ∆L* increases and the wood darkens. A decrease in brightness is not uniform during the technological process of thermal modification of maple wood. A decrease in the brightness of thermally treated wet maple wood is in line with the knowledge about wood darkening in technological processes, such as wood steaming declared in the works of TOLVAJ et al. 2009, 2010, DZURENDA, 2018b.

Changes in the red chromatic coordinate a* have an increasing tendency. The red values of maple wood reach higher values during the modification depending on the temperature and time, while the modified wood acquires a redder tinge. The magnitudes of the changes in the red coordinate are significantly smaller compared to the changes in the luminosity coordinate. In the case of the yellow chromatic coordinate b*, the changes are slightly similar to the changes in brightness or changes of the colour coordinate a*.

The changes in the total colour difference ∆E* depending on the acidity of wet maple wood during the process of thermal treatment at the temperatures ranging between t = 105−135 °C and the time of the technological process τ = 3−12 h is shown in Fig. 3.

y = 100.6035 – 32.461∙x + 2.6799∙(x)2 R2 = 0.88

Fig. 3 Dependence of the total colour difference of maple wood ∆E* on the value of acidity of maple wood pH during the process of thermal treatment.

The dependence of the total colour difference of maple wood on the value of acidity is mathematically described by the equation: ∆E* = 100.6035 – 32.461∙pH + 2.6799∙(pH)2

(2)

where: pH - acidity value of wet maple wood.

The dependence of the total colour difference ∆E* on the change in acidity of maple wood in the thermal process is a suitable tool for evaluating the achieved colour change based on the pH of maple wood in the technological process.

31

CONCLUSION 1. The results of colour change, acidity and density of maple wood during the process of thermal modification of wood with saturated water steam at the temperatures: tI = 105 ± 2.5 °C, tII = 125 ± 2.5 °C and tIII = 135 ± 2.5 °C for τ = 3−12 hours are presented in the paper .

2. The wood colour changes as a result of the process of thermal modification. The change of the original white to light white-yellow colour, pale brown to dark brown colour was observed. More or less deep colour hues depending on the temperature and time of modification occured. The colour changes achieved during the modification process are described using the coordinates in the colour space CIE L*a*b*.

3. During the technological process of wood modification, the acidity of the wood changes, which is more pronounced at higher steam temperature and for a longer time. The change is from pH = 5.1 in the case of native wood to pH = 3.3 in mode III after 12 hours of modification. Simultaneously, with the change in the acidity of the wood, there is also a decrease in the total weight of the wood due to acid hydrolysis of the chemical components of the wood and extraction of water-soluble substances, which represents a total decrease in density by up to 5.3 %.

4. The dependence of the total colour change ∆E* of maple wood on the acidity of maple wood in the range between pH = 3.3 − 5.1 is described by Eq. 2.

REFERENCES BABIAK, M., KUBOVSKÝ, I., MAMOŇOVÁ, M. 2004. Farebný priestor vybraných domácich drevín In Interaction of wood with various Forms of Energy. Zvolen: Technical University of Zvolen. 113– 117.

BARAŃSKI, J., KLEMENT, I., VILKOVSKÁ, T., KONOPKA, A. 2017. High Temperature Drying Process of Beech Wood (Fagus sylvatica L.) with Different Zones of Sapwood and Red False Heartwood. In BioResources 12(1): 1861−1870. DOI:10.15376/biores.12.1.1761-1870.

BARCÍK, Š., GAŠPARÍK, M., RAZUMOV, E.Y. 2015. Effect of thermal modification on the colour changes of oak wood. In Wood Research. 60(3): 385−396.

BLAŽEJ, A., ŠUTÝ, L., KOŠÍK, M., KRKOŠKA, P., GOLIS, E. 1975. Chémia dreva. Bratislava: ALFA. BUČKO, J. 1995. Hydrolýzne procesy. Zvolen: Technical University of Zvolen. 116 p. ČUDINOV, B. S., STEPANOV, V. L. 1968. Phasenzusammensetzung der Wassers in gefrorenem Holz.

In Holztechnologie, 9(1): 14−18.

DZURENDA, L., DELIISKI, N. 2000. Analysis of moisture content changes in beech wood in the steaming process with saturated water steam. In Wood research 45(4): 1−8.

DZURENDA L., ORLOWSKI K. 2011. The effect of thermal modification of ash wood on granularity and homogeneity of sawdust in the sawing process on a sash gang saw PRW 15-M in view of its technological usefulness. In Drewno, 54(186): 27−37.

DZURENDA L. 2013. Modification of wood colour of Fagus sylvatica L. to a brown-pink shade caused by thermal treatment. In Wood research 58 (3): 475−482.

DZURENDA, L. 2014. Colouring of Beech Wood during Thermal Treatment using Saturated Water Steams. In Acta Facultatis Xylologiae Zvolen, 56(1): 13−22.

DZURENDA, L. 2018a. The Shades of Color of Quercus robur L. Wood Obtained through the Processes of Thermal Treatment with Saturated Water Vapor. In BioResouces 13(1): 1525−1533, doi: 10.1063/biores 13.1.1525-1533.

DZURENDA, L. 2018b. Hues of Acer platanoides L. resulting from processes of thermal treatment with saturated steam. In Drewno 61(202): 165−176.

32

DZURENDA, L. 2018c. The Effect of Moisture Content of Black Locust Wood on the Heating in the Saturated Water Steam during the Process of Colour Modification. In MATEC Web of Conferences 168, 06004. DOI: org/10.1051/matecconf/201816806004.

DZURENDA, L. 2020. Drying of Steaming Maple Timber in Drying Kilns, Preserving the Color Acquired by the Wood Steaming Process. In MATEC Web of Conferences 328, 04004. DOI: org/10.1051/matecconf/202032804004.

DZURENDA, L., GEFFERT, A., GEFFERTOVÁ, J., DUDIAK, M. 2020. Evaluation of the Process Thermal Treatment of Maple Wood Saturated Water Steam in Terms of Change of pH and Color of Wood. In BioResources 15(2): 2550−2559. DOI: 10.15376/biores.15.2.2500-2559.

FENGEL, D., WEGENER, G. 1989. Wood: Chemistry, Ultrastructure, Reactions. Walter de Gruyter: Berlin, Germany. 613 p.

GEFFERT, A., GEFFERTOVÁ, J., DUDIAK, M. 2019. Direct method of measuring the pH value of wood. In Forests 10(10), 852. DOI: 10.3390/f10100852.

GEFFERT, A., GEFFERTOVÁ, J., VÝBOHOVÁ E., DUDIAK, M. 2020. Impact of Steaming Mode on Chemical Characteristics and Color of Birch Wood. In Forests 11(4), 478. DOI: 10,3390/f11040478.

HON D.S.N., SHIRAISHI, N. 2001 Wood and cellulosic chemistry. 2nd edition. New York: MarcelDekker. 513 – 546.

IRLE, M. 2012. pH and why you need to know it. Wood Based Panels International 2012. Available online: http://www.wbpionline.com/features/ph-and-why-you-need-to-know-it/ (accesed on 20 August 2012).

KOLLMANN, F., GOTE, W. A. 1968. Principles of Wood Sciences and Technology. Vol. 1. Solid Wood, Berlin: Springer Verlag.

KURJATKO, S. et al. 2010. Parametre kvality dreva určujúce jeho finálne. Zvolen: Technical University of Zvolen. 352 p.

LAUROVÁ, M., MAMONOVÁ, M., KUČEROVÁ, V. 2004. Proces parciálnej hydrolýzy bukového dreva (Fagus sylvatica L.) parením a varením. Zvolen: Technical University of Zvolen. 59 p.

MAKOVÍNY, I. 2010. Úžitkové vlastnosti a použitie rôznych druhov dreva. Zvolen: Technical University of Zvolen. 104 p.

MEINTS, T., TEISCHINGER, A., STINGL, R., HASSMANNC. 2017. Wood colour of central European wood species: CIELAB characterisation and colour intensification. In Eur. J. Wood Prod., 75: 499−509. DOI: 10.1007/s00107-016-1108-0.

MELCER, I., MELCEROVÁ, A., SOLÁR, R., KAČÍK, F. 1989. Chemizmus hydrotermickej úpravy listnatých drevín. Zvolen: Technical University of Zvolen. 2/1989. 76 p.

MOLNAR, S., TOLVAJ, L. 2002. Colour homogenisation of different wood species by steaming. In Interaction of wood with various Forms of Energy. Zvolen: Technical university in Zvolen, 119−122.

POŽGAJ, A., CHOVANEC, D., KURJATKO, S., BABIAK, M. 1997. Štruktúra a vlastnosti dreva.

Bratislava: Príroda, a.s., 485 p. ISBN: 80-07-00960-4.

PŇAKOVIČ, Ľ., DZURENDA, L. 2015. Combustion characteristics of fallen fall leaves from ornamental trees in city and forest parks. In BioResources, 10(3): 5563−5572. DOI: 10.15376/biores.10.3.55635572.

SAMEŠOVÁ, D., DZURENDA, L., JURKOVIČ, P. 2018. Kontaminácia kondenzátu produktmi hydrolýzy a extrakcie z tepelného spracovania bukového a javorového dreva pri modifikácii farby dreva. In Chip and Chipless Woodworking Processes 2018. 11(1): 277–282.

SANDERMAN. W., ROTHKAMM.M. 1959. The determination of pH values of woods and their practical importance. In Holz Roh-Werkstoff, 17: 433−441.

SUNDQVIST, B., KARLSSON, O., WESTREMARK, U. 2006. Determination of formic-acid and acid concentrations formed during hydrothermal treatment of birch wood and its relation to color, strength and hardness. In Wood Sci Technol., 40(7): 549−561.

SOLÁR, R. 2004. Chémia dreva. Zvolen: Technical University in Zvolen, 102 p. ISBN: 80-228-1420-2.

STN EN 13183-1 (2003). Moisture content of a piece of sawn timber – Part1: Determination by oven dry method.

TOLVAJ, L., NEMETH, R., VARGA, D., MOLNAR, S. 2009. Colour homogenisation of beech wood by steam treatment. In Drewno. 52(181): 5−17.

33

TOLVAJ, L., MOLNAR, S., NEMETH, R., VARGA, D. 2010. Color modification of black locust depending on the steaming parameters. In Wood Research, 55(2): 81−88.

TREBULA, P. 1986. Sušenie a hydrotermická úprava. Zvolen: Technical University in Zvolen,255 p.

VÝBOHOVÁ E., GEFFERTOVÁ J., GEFFERT A. 2018. Impact of Steaming on the Chemical Composicion of Maple Wood. In BioResources. 13(3), 5862-5874. DOI: 10.15376/biores.13.3.5862-5874.

ZEVENHOVEN, M. 2001. Ash-forming Matter in Biomass Fuels, Åbo Akademi University, Turku, Finland.

ACKNOWLEDGMENT This experimental research was prepared within the grant project: APVV-17-0456 “Termická modifikácia dreva sýtou vodnou parou za účelom cielenej a stabilnej zmeny farby drevnej hmoty” as the result of work of author and the considerable assistance of the APVV agency.

AUTHOR ADDRESS Michal Dudiak Technical University in Zvolen T. G. Masaryka 24 960 01 Zvolen Slovak Republic xdudiak@is.tuzvo.sk

34

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 35–47, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.04

WOOD SURFACE MORPHOLOGY ALTERATION INDUCED BY ENGRAVING WITH CO2 LASER UNDER DIFFERENT RASTER DENSITY VALUES Jozef Kúdela - Michal Andrejko - Oľga Mišíková ABSTRACT The subject of this work was to study changes in surface morphology in three wood species differing in their structure, engraved with a CO2 laser under different raster density values. From the physical viewpoint, the morphology variation was assess based on the roughness parameters. Qualitative changes in the anatomical structure were inspected with the aid of light microscopy. Significant influences on roughness parameters have been confirmed for all the three factors acting during engraving (raster density, anatomical direction, wood species). Over the entire study range, increasing raster density caused increasing roughness parameters parallel to a well as perpendicular to the grain course.

Significantly higher roughness variation was recorded perpendicular to the grain course.

Significant influence of wood species on roughness parameters was explicitly confirmed in the case of roughness measuremrents perpendicular to the grain in radial direction. The most conspicuous changes were observed in spruce wood, the lowest in beech wood. The results of microscopical observations were effective for explaining the species-related differences in roughness in wood treated with a CO2 laser.

Key words: engraving, CO2 laser, spruce wood, beech wood, oak wood, morphology, roughness

INTRODUCTION Material treatment technology using a laser beam leads to very narrow cutting gaps.

Consequently, such technologies have found a wide range of use in material cutting and boring. In addition, there are emerging new potentials concerning the surface treatment of both metallic and non-metallic materials by engraving (PATEL et al. 2017, YANG et al. 2019, KUBOVSKÝ et al. 2020, YUNG et al. 2021). Engraving wood surface with a laser is done with the aim to alter the colour or morphology of the surface treated in this way. Wood treatment with a laser means a considerable benefit: for the given absorption coefficients it is possible to set the amount of the energy supplied by the laser beam onto the wood surface and thus to control the structure and properties of the treated surface (KAČÍK and KUBOVSKÝ 2011, VIDHOLDOVÁ et al 2017, KÚDELA et al. 2019, 2020).

Chemical, physical, and morphological alterations of the properties of engraved wood surface depend on the energy amount supplied by the laser beam onto the treated surface. This amount can be controlled through adjusting the laser power, the movement speed of the laser

35

head, the focal position, and the raster density (LIN et al. 2008, KUBOVSKÝ and KAČÍK 2013, HALLER et al. 2014, KUBOVSKÝ and KAČÍK 2014 GURAU et al. 2017, GURAU and PETRU 2018, Li et al. 2018, KÚDELA et al. 2020, ANDREJKO et al. 2020).

The energy concentrated in the laser beam and supplied to a specific area on the engraved surface is transformed to heat. At the very moment of contacting the wood surface, the great amount of heat concentrated within the laser beam with a very small diameter causes a very thin surface layer to sublimate immediately. Besides to the amount and concentration of the energy supplied, the thickness of the layer sublimed is dependent on the wood species, due to the inter-specific differences in wood structure and hardness. (ARAI and KAWASUMI 1980, BARCIKOWSKI et al. 2006, WUST et al. 2005, HALLER et al. 2014, DOLAN 2014, KÚDELA et al. 2020, ANDREJKO et al. 2020).

The microscopic observations show (HALLER et al. 2014, DOLAN 2014) that the wood surface treatment with a laser beam can reduce the wood roughness by melting but not carbonising the wood cells down to a depth of several micrometres. KÚDELA et al. (2019) report considerable morphological changes manifested through more roughness as late as under an irradiation dose of 75 J∙cm−2, primarily due to the carbonisation of the surface wood layer. GUO et al. (2021) used for ablating milled and ground surfaces a nano-second pulse laser. The engraving, however, may induce nn opposite effect (GURAU et al. 2017, GURAU and PETRU 2018, KÚDELA et al. 2020, ANDREJKO et al. 2020). The last cited works show that the wood surface roughness depends considerably on the laser power, head movement speed, raster density and other parameters.

KÚDELA et al. (2020), applying a CO2 laser on beech wood surface, observed moderately decreased roughness parameters in the fibre course direction at a low laser power (4 %). Their measurements resulted in finding lower values compared to the referential (non-treated with laser) ground specimens. This was caused by removal of released libriform fibres during the CO2 laser treatment process. The cells concerned were the ones maintained on the wood surface after grinding. For higher laser power (8 %), these authors report roughness increasing with increasing raster density. The roughness parameters values were considerably higher compared to the ground surface.

ANDREJKO et al. (2020), engraving oak wood surface, confirmed significant influence of all the factors studied (laser power, laser density, wood anatomical direction) on the roughness parameters Ra and Rz. The last cited work identified the major impact of the raster density. Over the whole study raster density range, the increasing raster density was followed by increasing roughness parameters, both parallel to and perpendicular to the grain course.

This paper also demonstrates that the oak wood surface roughness after CO2 laser treatment also depended on the heterogeneous oak wood structure as such. The trench after the laser beam was deeper in early wood with the major share of early vessels than in the late wood containing less vessels. This was reflected in more conspicuous roughness variation observed perpendicular to the grain.

Thus, the surface morphology of CO2 laser-treated wood depends on the laser technical parameters, on the irradiation methods and technology, and on the wood species used. The proper adjustment of parameters for a CO2 laser treatment offers possibilities to control the surface morphology of the treated wood. Wood surface morphology assessment needs to comprise both the anatomical and the physical aspects. The physical assessment is performed based on roughness and waviness parameters (GURAU 2013, GURAU and IRLE 2017, CZANADY and MAGOSS 2011 KÚDELA et al. 2018).

The aim of this work was to study CO2-laser-engraving-induced changes to surface morphology in three wood species differing in their structure. These changes were investigated under different raster densities. The surface geometry of the wood species concerned was

36

evaluated in two ways: quantitatively, based on roughness parameters and qualitatively, with the aid of light microscopy.

MATERIAL AND METHODS Wood surface engraving with a CO2 laser was performed on three wood species differing in their anatomical structure. The coniferous wood species were represented by spruce, the broadleaved ones by beech and oak (disperse porous and ring porous). From these three species, there were prepared test specimens with dimensions of 100 mm  50 mm  15 mm (Fig. 1). In all specimens, there were irradiated their radial surfaces. Before the radiation, the surfaces were ground with a sandpaper with a grain size of 240. The moisture content in specimens ranged within 8−10 %.

The specimens were engraved with a CO2 laser CM-1309, provided by the firm EAGLE and performing with a maximum power of 135 W (Fig. 2a, b). The distance between the irradiated surface and the lens focus was 17 mm. The laser head was moving over the specimen surface parallel to the grain course, with a constant speed of 350 mm·s 1 . Under constant laser power, the radiation intensity varied with varying raster density (number of paths per a one-millimetre width). The raster scanning course followed the perpendicular-to-grain in the radial direction, the numbers of paths per a width of one millimetre were 2, 5, 10, 20, and 30. Within one wood species, there were altogether five combinations, each comprising three specimens, plus three referential (control) specimens.

Under the pre-set conditions, each specimen set was irradiated uniformly along its entire length and width.

Changes to wood surface morphology corresponding to different irradiation modes with a CO2 laser were studied based on the roughness parameters and based on the morphological changes on the irradiated surface inspected with a light microscope.

Roughness was measured on specimens´ radial surfaces parallel to the grain course, and evaluated through the values of roughness parameters Ra (mean arithmetic deviation), Rz (the maximum peak height plus the maximum depression depth within the cut-off, or sampling length), Rt (the maximum peak height plus the maximum depression depth within the entire evaluation length) a RSm (mean distance between the trenches – arithmetic mean calculated from distances between the profile unevennesses within the sampling length).

Fig. 1 Test specimen: shape and dimensions

Fig. 2 Wood surface engraving

Surface roughness was measured with a mechanic profilometer Surfcom 130A (Carl Zeiss, Germany), consisting of two units: measuring one and evaluating one (Fig. 3). The 37

profilometer was tuned in such a way as to measure the profile ranging from −800 µm to +800 µm from the central line (so the total measured profile was 1600 µm). The total measured length consisted of the starting-run segment, five sampling segments and stop-way segment. The sampling length was in all the cases the same - 8 mm (together making evaluation length 40 mm). The sampling length was established from the measured values of roughness parameters Ra and Rz. During the roughness evaluation, the waviness was filtered away from the roughness profile measured with the profilometer, and the obtained roughness curve was transferred onto the base line.

Fig. 3 Mechanic profilometer Surfcom 130A..1 – measuring unit, 2 – evaluating unit, 3 – PC

The structure of engraved surface of the tested wood species was studied with the aid of a digital microscope Dino-Lite EDGE; with the aid of a light microscope Leica MZ 9,5 on micro-cuts; and with a camera Leica EC 3. Microscopic slides representing transverse cuts were prepared from specimens softened in glycerine and modified with a transparent lacquer.

Then, the micro-cuts were sealed in euparal. The carbonised cell layer on the surface had a very low stability, and as such, peeled off during the preparation of the micro-cuts. This required applying of an additional method for micro-cuts preparation. In this case, for each tested wood species, the micro-cuts were made from wood prisms modified with a synthetical resin (Technovit). Then the micro-cuts were sealed in euparal.

RESULTS AND DISCUSSION The physical evaluation of surface morphology of the studied wood species subject to laser engraving was accomplished based on the roughness profiles obtained experimentally, parallel to and perpendicular to the grain course in the radial direction. All the measured profiles were evaluated through their roughness parameters Ra, Rz, Rt and RSm. The average values of these parameters together with other statistical characteristics for all the concerned three species, the entire raster density range, and the two anatomical directions are in Table 1.

The results of three-way variance analysis confirmed that all the three parameters inspected (raster density, wood species, anatomical direction) exerted significant impacts on the roughness parameters concerned; either acting separately or in mutual interactions.

In all three wood species, the roughness parameters Ra, Rz and Rt of control specimens exhibited the lowest values both parallel to and perpendicular to the grain course. The speciesrelated differences among these specimens were mainly due to different qualitative and quantitative presence of their cell elements. This fact was most obviously demonstrated on the roughness measured perpendicular to the grain course. Table 1 shows that the values of these 38

parameters significantly increased with increasing raster density during CO2 laser engraving.

The raster-density dependent variance of all roughness parameters is illustrated in Fig. 4.

At the highest raster density (30 mm−1), the roughness parameters Ra, Rz and Rt values parallel to the grain increased up to several times. The variability of these parameters was relatively high. Thus, we cannot state unequivocally in which wood species the CO2 laserinduced changes in the surface geometry were the most obvious. Figure 4a shows that unlike the raster density, the wood species-related influence is omissible in practical context.

Tab. 1 Basic statistical characteristics of roughness parameters parallel to and perpendicular to the grain in engraved surfaces of tested wood species, dependent on different raster density values (laser power 8 %).

Wood roughness parameter Raster Basic Parallel to grain Perpendicular to grain density statistical Ra Rz Rt Sm Ra Rz Rt RSm character. mm−1 [µm] Spruce wood x̅ 7.25 50.73 76.25 1473.2 11.64 89.32 114.10 692.9 Refer. s 2.86 21.13 26.55 660.8 1.58 19.20 26.28 284.2 x̅ 11.41 73.26 101.88 2070.5 37.85 214.36 239.85 504.9 2 s 3.27 21.34 22.81 1150.2 1.34 16.32 19.01 25.4 x̅ 21.66 107.83 136.00 3107.3 44.34 229.59 257.39 214.7 5 s 8.80 38.50 42.02 2015.3 2.17 22.33 22.63 19.0 x̅ 28.34 185.31 242.47 1173.8 103.72 497.03 560.16 1809.9 10 s 8.13 44.99 44.36 558.8 18.18 76.13 67.72 509.2 x̅ 27.67 180.38 237.37 1118.3 210.73 903.06 1034.55 1950.8 20 s 9.74 63.07 55.01 842.7 34.97 103.13 84.95 624.4 x̅ 36.27 231.14 299.69 869.4 273.03 1132.51 1244.20 1968.4 30 s 9.69 60.18 61.25 327.6 17.56 70.12 43.23 691.4 Beech wood x̅ 6.96 51.44 69.36 1770.8 6.83 63.13 78.96 701.8 Refer. s 1.81 11.70 14.14 753.6 1.21 12.03 10.06 265.4 x̅ 6.79 47.23 68.60 1774.9 38.82 195.37 217.32 499.5 2 s 2.52 18.58 22.81 860.3 1.57 15.88 23.73 8.3 x̅ 13.35 79.92 113.49 2233.4 39.63 204.90 240.45 221.1 5 s 3.97 24.61 22.03 1032.6 2.06 30.92 58.31 14.2 x̅ 24.85 148.49 188.58 1765.3 38.33 242.65 279.50 630.1 10 s 7.30 38.80 44.07 681.0 3.51 25.12 27.41 164.6 x̅ 30.46 207.66 313.56 1371.1 62.21 378.18 476.48 1031.2 20 s 10.03 75.32 105.08 664.1 4.64 63.01 100.56 375.8 x̅ 42.61 248.38 330.65 1606.8 76.42 474.87 583.07 1249.7 30 s 7.96 45.15 32.82 687.7 8.63 87.82 151.82 431.5 Oak wood x̅ 5.15 38.18 88.59 1172.4 13.98 143.94 198.55 1937.7 Refer. s 5.60 35.26 42.00 889.9 6.39 43.46 23.82 1081.1 x̅ 20.20 106.08 163.25 3030.5 50.03 264.29 293.34 499.9 2 s 8.72 47.15 51.41 1703.8 4.20 24.01 12.28 16.5 x̅ 26.71 131.04 197.15 3733.2 48.97 328.43 400.99 589.6 5 s 9.07 50.23 59.36 1796.1 7.46 51.44 54.56 233.7 x̅ 25.70 150.89 260.14 2443.6 69.95 437.17 530.21 1721.0 10 s 13.06 77.78 71.78 1506.8 12.66 55.72 52.56 605.6 x̅ 35.09 202.19 303.87 2243.6 93.93 557.73 679.23 1870.2 20 s 14.46 71.89 43.89 1315.2 19.34 76.24 46.09 681.7 x̅ 35.53 194.48 287.90 2055.7 115.36 674.38 845.69 1714.0 30 s 17.16 77.85 78.80 1216.5 25.16 124.54 124.22 524.3 The number of measurements for parameters Ra, Rz, and RSm for each testing variant was 60. For parameter Rt, n = 12.

39

50

300

a) Parallel to grain

Ra [μm]

Ra [μm]

30 20

150 100 50

0

0 5

10

15

20

25

30

0

300

1200

250

1000

200

800

Rz [μm]

Rz [μm]

200

10

0

b) Perpendicular to grain

250

40

150 100

15

20

25

30

600 400

0

0 0

5

10

15

20

25

0

30

350

1400

300

1200

250

1000

Rt [μm]

Rt [μm]

10

200

50

200 150 50

200

0

0 10

15

20

25

30

RSm [μm]

4000 3500 3000 2500 2000 1500 1000 500 0 0

5

10

15

Raster density

20

25

30

10

15

20

25

30

0

5

10

15

20

25

30

0

5

10

15

20

25

30

600 400

5

5

800

100

0

RSm [μm]

5

4000 3500 3000 2500 2000 1500 1000 500 0

[mm−1]

Raster density spruce

beech

[mm−1]

oak

Fig. 4 Raster density influence on roughness parameters Ra, Rz, Rt and RSm in engraved surfaces of spruce, beech and oak wood. Laser power 8 %.

40

The parallel-to-the-grain values of roughness parameter RSm representing the mean distance between the trenches were increasing with increasing raster density up to 5 mm-1, then a moderate decrease followed. This variation was qualitatively similar in all the three wood species; on the other hand, there were present interspecific differences in quantity (Fig. 4a).

The roughness parameters Ra, Rz and Rt measured perpendicular to the grain were significantly higher over the entire raster range. This was true for all three wood species.

With increasing raster density, the values of these parameters manifested higher increase rates compared to those measured parallel to the grain (Fig. 4b). At the maximum raster density, the Ra, Rz and Rt values were by order higher compared to the referential specimens.

On the other hand, their variability was found lower. Also, in the case of roughness measured perpendicular to the grain, the parameter RSm manifested different course compared to the roughness parameters discussed above. This parameter course with the varying raster density was opposite than in the case parallel to the grain (Fig. 4a, b).

In accordance with GURAU et al. 2017, GURAU and PETRU 2018, LI et al. 2018, KÚDELA et al. 2020, ANDREJKO et al. 2020), the morphological changes on the engraved wood surface depended on the energy amount supplied onto this surface with a laser beam. In the case of roughness measured perpendicular to the grain, the roughness parameters seem also significantly influenced by the wood species. In our case, the laser power, the laser head movement, and the laser focus distance were constant; the energy amount was only dependent on the varying raster density.

The energy concentrated in the laser beam and supplied onto the specified spot on the engraved surface was transformed to heat. The experimental wood temperature measurements performed with a thermo-camera up to its upper performance limit of 1000 °C demonstrated that, at moment of the beam reaching the surface, the temperature was mostly above 650 °C, episodically even close to the limit value that the thermo-camera could record. Such a high temperature concentrated within a very small diameter of the laser beam contacting the wood surface caused immediate burning and sublimation of a thin surface wood layer.

Experimental roughness measurements and microscopic observations revealed that the sublimated layer thickness was, beside the amount and concentration of the energy supplied, to a considerable extent influenced by the wood structure and properties (density, hardness) differing not only between the species but also varying within the particular ones. This fact has also been confirmed by (ARAI and KAWASUMI 1980, BARCIKOWSKI et al. 2006, WUST et al. 2005, HALLER et al. 2014, DOLAN 2014, KÚDELA et al. 2020, ANDREJKO et al. 2020).

The wood engraving at different raster density values tracked the wood surface with trenches causing surface roughness enhancement. With the raster path width (engraved track) of 0.14 mm, the treatment of the wood surface at a raster density of 10 mm−1 and more resulted in the tracks overlapping. This means that spots with raster overlapping had been supplied with energy repeatedly, which caused more deepening the relevant trenches.

Figure 4b demonstrates that the most conspicuous changes in roughness parameters measured perpendicular to the grain under specific CO2 laser-engraving conditions were observed in spruce wood, the least ones in beech wood. As the examined surfaces were radial, the roughness variance concerned was to a considerable extent due to the density differences between the early and late wood.

In the case of spruce, the major density differences were between the early and late wood. MOLIŃSKI et al. (2009) report for spruce early wood an average density value of 300 kg·m −3, while for late wood about 750 kg·m−3. The last cited work reports that neither early nor late wood density varied significantly with the cambium age. Nevertheless, the cambium age affected the proportions of early and late wood in growth rings, and, in this way, also the width of these rings. With increasing cambium age, the width of growth rings became narrower, due to thinner early wood bands.

41

Figure 5a shows roughness profiles measured perpendicular to the grain: on a referential specimen and on a specimen engraved with the CO2 laser performing at a power of 8 % and density of 20 mm−1. Figure 5a demonstrates how the non-uniform wood substance degradation within a growth ring, occurring due to the density differences between the early and the late wood, was mainly reflected in the apparent depth increase of trenches in the early wood and, to some extent, also in the distance between the profile unevennesses. The scans of a radial surface and a transverse section (Fig. 5b, c) illustrate the related changes in the wood morphology. Figure 5c manifests the existence of distinctive boundary between early and late wood at the beginning of the growing season. At this phase, the early wood density is the lowest. This is also evident from the intensive wood substance degradation. During the growing season, early wood is transformed to late wood, the wood density gradually increases. Accordingly, the wood substance degradation rate under laser beam treatment decreased. Cell walls of early and late wood tracheids on the engraved spruce wood surface a)

b)

c)

d)

Fig. 5 Surface morphology of spruce wood engraved with a CO2 laser perpendicular to the grain course, at a laser power of 8 % and raster density of 20 mm−1. a) roughness profiles for referential specimen and specimen engraved with the laser, b) radial section, c) transverse section d) microscopic slides of transverse sections.

42

were carbonised (Fig. 5c). In the case of the wood surface layer consisting of early wood tracheids exclusively, the walls of these tracheids were often impaired due to the scorching with the laser beam, and due to the removal of the carbonised layer. This layer was very poor stable and, as such, frequently impaired during the specimen preparation.

Qualitatively similar trends were found for oak wood engraving. The differences compared to the spruce wood were on the background of the different anatomical structure.

Figure 6 shows that, analogically as in spruce, the trench after the laser beam was deeper in the early wood than in the late wood. This fact was primarily responded by higher values of a)

b)

c)

d)

Fig. 6 Surface morphology of oak wood engraved with a CO2 laser perpendicular to the grain course, at a laser power of 8 % and raster density of 20 mm−1. a) roughness profiles for referential specimen and specimen engraved with the laser, b) radial section, c) transverse section d) microscopic slides of transverse sections.

43

the Rz parameter. Oak wood is ring-porous, with large lumina, namely in thin-walled early vessels – compared to libriform fibres with small lumen diameter and thick cell wall (POŽGAJ et al. 1997, KURJATKO et al. 2010). Roughness profiles measured perpendicular to the grain course in the radial direction were namely shaped by the displacement of these elements across the width of the growth rings. The energy supplied exerted more effects on the spot containing early-wood vessels than on the spots with the surface consisting of wood fibres (Fig. 6b−c).

Similar results concerning oak wood roughness after CO2 laser treatment can be found in ANREJKO et al. (2020). The authors of the cited work used the same raster density parameters for three power values. Their results suggests that the impact on wood morphology is stronger from the raster density than from the laser power. The same follows from the results reported

by GURAU et al. (2017), KÚDELA et al. (2020). a)

b)

c)

d)

Fig. 7 Surface morphology of beech wood engraved with a CO2 laser perpendicular to the grain course, at a laser power of 8 % and raster density of 20 mm−1. a) roughness profiles for referential specimen and specimen engraved with the laser, b) radial section, c) transverse section d) microscopic slides of transverse sections.

44

In the case of beech wood, the influence of raster density on roughness parameters was found the lowest (Fig. 4b, 7a). This was caused by the fact that the beech growth rings are more homogeneous than in spruce and oak. Beech wood belongs to hard disperse-porous species. The vessels are dispersed across the entire growth ring width and the transition from the early wood to the late wood is smooth. The radial surface consists of vessels alternating with libriform fibres rather regularly. If the laser beam meets the fibres, the engraving is less conspicuous, because beech fibres have very thick cell walls and small lumina. In the case of laser beam contacting a surface with vessels, the beam penetrated more in depth (Fig. 7d), because the cell lumina were bigger and cell walls thinner compared to the fibres. Also, in the case of beech wood, there was observed more wood substance degradation in the early wood compared to the late wood (Fig. 7c). This transition, however, was not as abrupt as in spruce wood.

The study of surface morphology of spruce, beech and oak wood irradiated with a CO2 laser revealed that the interactions occurring at the contact spot wood – laser beam are very complex. Their thorough and deep understanding needs, apart from morphological studies, also investigation of other surface properties and their relations. The issue is especially relevant meeting the needs of wood gluing and surface treatment.

CONCLUSIONS The morphological changes of the spruce, beech and oak wood surface engraved with a CO2 laser was the subject of the study. These changes were assessed based on the roughness parameters of the wood surface and based on microscopical observations. The obtained results allow us to derive the following conclusions: Significant impacts on roughness parameters were confirmed for all the three tested factors (raster density, anatomical direction, and wood species) acting during the surface engraving.

Over the entire concerned raster density range, increasing raster density produced increasing roghness parameters both parallel to and perpendicular to the grain.

Significantly bigger roughness variance was recorded perpendicular to the grain than parallel to it.

Significant influence of wood species on roughness parameters was explicitly confirmed in the case of measurements performed perpendicular to the grain in radial direction. The most obvious changes were observed in spruce, the lowest ones in beech.

The microscopical analysis inferred that the roughness differences between the tested wood species were mainly due to the differences in their structural heterogeneity across the growth ring width.

REFERENCES ANDREJKO, M., KÚDELA, J., MIŠÍKOVÁ, O., KMINIAK, R. 2020: Štúdium morfológie povrchu dubového dreva po opracovaní gravírovacím CO2 laserom. (Oak wood surface morphology inspected after engraving with a CO2 laser). Chip and Chipless Woodworking Processes, 12(1): 5–13.

ARAI, I., KAWASUMI, H. 1980: Thermal analysis of laser machining in wood III. Mokuzai Gakkaishi, 26, 773−782.

BARCIKOWSKI, S., KOCH, G., ODERMATT, J. 2006: Characterisation and modification of the heat affected zone during laser material processing of wood and wood composites. Holz als Roh –u.

Werkstoff, 64, 94−103.

45

CZANADY E., MAGOSS E., 2011: Mechanics of wood machining. Sopron: University of West Hungary, 243 p.

DOLAN, J. A. 2014: Characterization of Laser Modified Surfaces for Wood Adhesion. (Thesis for the degree of Master of Science In: Macromolecular Science and Engineering). The Faculty of Virginia Polytechnic Institute, Blacksburg, VA, 100 p.

GUO, Q., WU, Z., ZHANG, C., YANG, C., MA, Y., XU, F., CAO, Z. 2021: Study on a new clean machining method instead of sanding technology for wood. Alexandria Engineering Journal, 60(2): 2369−2380.

GURAU L. 2013: Analyses of roughness of sanded oak and beech surface. ProLigno, 9(4): 741–75.

GURAU L., IRLE M. 2017. Surface Roughness Evaluation Methods for Wood Products: A Review. Current Forestry Reports, 3(2), 119−131.

GURAU, L., PETRU, A. 2018: The influence of CO2 laser beam power output and scanning speed on surface quality of Norway maple (Acer platanoides), BioResources, 13, 8168−8183.

GURAU, L., PETRU, A., VARODI, A., TIMAR, M. C. 2017: The influence of CO2 laser beam power output and scanning speed on surface roughness and colour changes of beech (Fagus sylvatica), BioResources, 12, 7395−7412.

HALLER, P., BEYER, E., WIEDEMANN, G., PANZNER, M., WUST, H. 2014: Experimental study of the effect of a laser beam on the morphology of wood surfaces. https://www.researchgate.net/ publication/237543545 KAČÍK, F., KUBOVSKÝ, I. 2011: Chemical changes of beech wood due to CO2 laser irradiation. J. Photochem. Photobiol. A, 222, 105–110.

KUBOVSKÝ, I., KAČÍK, F. 2013: Changes of the wood surface colour induced by CO2 laser and its durability after the xenon lamp exposure. Wood Research. 58, 581−590.

KUBOVSKÝ, I., KAČÍK, F. 2014: Colour and chemical changes of the lime wood surface due to CO2 laser thermal modification, Appl. Surf. Sci., c321:261−267. http://dx.doi.org/ 10.1016/j.apsusc.2014.09.124 KUBOVSKÝ, I., KRIŠT’ÁK , Ľ., SUJA, J., GAJTANSKA, M., IGAZ, R., RUŽIAK, I., RÉK, R. 2020: Optimization of Parameters for the Cutting of Wood-Based Materials by a CO2 Laser. Appl. Sci., 10, 8113.

KÚDELA, J., MRENICA, L., JAVOREK, Ľ. 2018: Influence of milling and sanding on wood surface KURJATKO, S. et al. 2010: Parametre kvality dreva určujúce jeho finálne použitie. (Wood quality parameters determining its final utilization). Zvolen: Technická univerzita vo Zvolene, p. 352. morphology. Acta Facultatis Xylologiae Zvolen, 60(1): 71−83.

KÚDELA, J., KUBOVSKÝ, I., ANDREJKO, M. 2020: Surface properties of beech wood after CO2 laser engraving. Coatings, 10(1): 77.

KÚDELA, J., REINPRECHT, L., VIDHOLDOVÁ, Z., ANDREJKO, M. 2019: Surface properties of beech wood modified by CO2 laser. Acta Facultatis Xylologiae Zvolen, 61, 5−18.

LI, R., XU, W., WANG, X.A., WANG, C. 2018: Modelling and predicting of the color changes of wood surface during CO2 laser modification, J. Clean. Prod., 183, 818−823.

LIN, C. J., WANG, Y. C., LIN, L. D., CHIOU, C. R., WANG, Y. N., TSAI, M. J. 2008: Effects of feed speed ratio and laser power on engraved depth and color difference of Moso bamboo lamina. J. Mater. Process. Technol., 198, 419−425.

MOLIŃSKI, W., ČUNDERLÍK, I., KRAUSS, A., FABISIAK, E., JUREK, P. 2009. Gradient of density and tensil strength along the grains of spruce wood (Picea abies) within individual annualrings. Ann. WULS - SGGW, For. and Wood Technol., 69, p. 87–92.

PATEL, CH., PATEL, A. J., PATEL, R. C. 2017. A review on laser marking process for different materials. IJSRD, 5(1): 147−150.

POŽGAJ, A., CHOVANEC, D., KURJATKO, S., BABIAK, M. 1997. Štruktúra a vlastnosti dreva. (Wood structure and properties). Bratislava: Príroda 1997, 488 p.

VIDHOLDOVÁ, Z., REINPRECHT, L., IGAZ. R. 2017. The impact of laser surface modification of beech wood on its color and occurrence of Molds. BioResources, 12(2): 4177−4186.

46

WUST, H. HALLER, P., WIEDEMANN, G. 2005. Experimental study of the effect of a laser beam on the morphology of wood surfaces. In Proceedings of the Second European Conference on Wood Modification, Göttingen, Germany, 6−7 October 2005.

YANG, CH., JIANG, T., YU, Y., BAI, Y., SONG, M., MIAO, Q., MA, Y., LIU, J. 2019. Water-jet assisted nanosecond laser micro cutting of northeast China ash wood: Experimental Study. BioResources, 14, 128−138.

YUNG, K. CH., CHOY, H. S., XIAO, T., CAI, Z 2021. UV laser cutting of beech plywood. J. Adv. Manuf. Technol. 112(9): 1−23.

ACKNOWLEDGMENTS This work was supported by the Slovak Research and Development Agency under the contract No. APVV-16-0177 and No. APVV-16-0326.

ADDRESSES OF AUTHOR Prof. Ing. Jozef Kúdela, CSc. Ing. Michal Andrejko Ing. Oľga Mišíková, PhD. Technical University in Zvolen Faculty of Wood Sciences and Technology Department of Wood Science T. G. Masaryka 24 960 01 Zvolen Slovak Republic kudela@tuzvo.sk misikova@tuzvo.sk michalandrejko207@gmail.com

47

48

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 49−62, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.05

COMPUTING THE 2D TEMPERATURE DISTRIBUTION IN LOGS STORED FOR A LONG TIME IN AN OPEN WAREHOUSE IN WINTER AND DURING SUBSEQUENT AUTOCLAVE STEAMING Nencho Deliiski – Ladislav Dzurenda – Peter Niemz – Dimitar Angelski – Natalia Tumbarkova ABSTRACT An approach to computing and research on the 2D non-stationary temperature distribution in horizontally situated logs in an open warehouse under the influence of periodically changing atmospheric temperature in winter and during their subsequent steaming in autoclaves is described in the paper. Mathematical descriptions of the changing atmospheric temperature and also of the temperature of the steaming medium in the autoclaves and of the conditioning air medium after steaming are introduced as boundary conditions in own mutually connected 2D non-linear mathematical models of the log freezing and defrosting processes. Numerical solutions of the coupled models in the calculation environment of Visual FORTRAN Professional are given as an application of the suggested approach. The results from a simulative investigation of the change in the 2D temperature field and average mass temperature of beech (Fagus sylvatica L.) logs with a diameter of 0.4 m, length of 0.8 m, and moisture content of 0.6 kg∙kg-1 during their 5 dayand night-long continuous alternating freezing and defrosting at sinusoidal change of the air temperature with different initial values below and equal to 0 °C and with different amplitudes, and also during steaming of such frozen logs with different initial temperature in an autoclave and their subsequent conditioning are graphically presented and analysed.

Key words: autoclave steaming, beech logs, 2D coupled models, atmospheric temperature, freezing, defrosting, model based control.

INTRODUCTION It is known that different layered items are subjected to steaming or heating in agitated water for plasticization of the logs and other kinds of wood materials in the production of veneer (CHUDINOV 1966, SHUBIN 1990, STEINHAGEN 1986, 1991, TREBULA – KLEMENT 2002, VIDELOV 2003, DELIISKI 2003, 2009, 2011, PERVAN 2009, DELIISKI – DZURENDA 2010), etc.

Steaming in autoclaves is one of the most intensive and energy effective processes for plasticization of the wood materials (RIEHL et al. 2002, DELIISKI 2003, 2004, SOKOLOVSKI et al. 2007, DELIISKI et al. 2015).

For the development and automatic achieving of energy saving and regimes with an optimal duration for autoclave steaming of logs, it is very important to know the initial temperature of the materials of each batch subjected to the thermal treatment.

49

The initial temperature of the separate batches depends on the duration of log storing in an open warehouse at periodically changing air temperature (DELISKI et al. 2020a, 2020b).

The influence of this temperature in winter on the temperature field and on the average mass temperature of the logs is of considerable scientific and practical interest.

The aim of the paper is to suggest an approach to computing the 2D temperature distribution in logs at periodically changing atmospheric temperature during many days and nights in winter and to study the influence of the computed average mass temperature of such logs on the duration of the regimes for their steaming in autoclaves.

MATERIAL AND METHODS Mathematical models of the 2D temperature distribution in logs during their alternating freezing and defrosting in the air environment In (DELIISKI and TUMBARKOVA 2019, TUMBARKOVA 2019), the following coupled models describing the 2D non-stationary temperature distribution in logs situated in the air environment subjected to freezing and subsequent defrosting horizontally were created, solved, and verified: a) During the log freezing process: ceff -fr  

  2T (r , z , ) 1 T (r , z , )   r -eff -fr T (r , z , ) =  r -eff -fr  + . +  r r T r 2  

+  p -eff -fr

2

 T (r , z , )

+

z 2

under an initial condition:

2

 T (r , z , )    + r

(1)

 p -eff -fr  T (r , z , )    + qv T z 2

T (r , z,0) = T01

(2)

and the following boundary conditions: • Along the radial coordinate r on the log frontal surface:





 p -fr (r ,0, ) T (r ,0, ) =− T (r ,0, ) − Tair -fr () , r  p -eff -fr (r ,0, )

(3)

• Along the longitudinal coordinate z on the log cylindrical surface:





 r -fr (0, z, ) T (0, z, ) =− T (0, z, ) − Tair -fr () . z  r -eff -fr (0, z, )

(4)

b) During the log defrosting process: ceff -dfr  

  2T (r , z , ) 1 T (r , z , )   r -eff -dfr T (r , z , ) =  r -eff -dfr  + . +  r r T r 2  

+  p -eff -dfr

2

 T (r , z , ) z 2

+

 p -eff -dfr  T (r , z , )    T z

under an initial condition 50

2

2

 T (r , z , )    + r

(5)

T (r , z,0) = T (r , z, frе )

(6)

and boundary conditions • Along the radial coordinate r:





(7)





(8)

 p -dfr (r ,0, ) T (r ,0, ) =− T (r ,0, ) − Tair -dfr () , r  p -eff -dfr (r ,0, )

• Along the longitudinal coordinate z:  r -dfr (0, z, ) T (0, z, ) =− T (0, z, ) − Tair -dfr () . z  r -eff -dfr (0, z, )

where сеff-fr and сеff-dfr are the effective specific heat capacities of the wood during the freezing and defrosting of logs in temperatures when the free and the bound water crystallizes or melts (DELIISKI and TUMBARKOVA 2017), J·kg-1·K-1; λr-eff-fr and λp-eff-fr – effective thermal conductivities of the wood in radial and longitudinal direction during the separate stages of the log freezing process, W·m-1·K-1; λr-eff-dfr and λp-eff-dfr – effective thermal conductivities of the wood in radial and longitudinal direction during the separate stages of the log defrosting process, W·m-1·K-1; ρw – wood density, kg·m-3; r – coordinate of the separate points along the log radius: 0 ≤ r ≤ R, m; R – radius of the log, m; z – coordinate of the separate points along the log length: 0 ≤ z ≤ L/2, m (Fig. 1); L – length of the log, m;  – time, s; fre – terminal time of the freezing process and the time when the coupled defrosting process begins, s; T – temperature, K; T01 – initial average mass temperature of the log subjected to freezing, K; T(r,z,0) – temperature of all points in the log volume at the beginning of the freezing or defrosting process, K; T(r,0,) – temperature of all points on the log frontal surface during the freezing or defrosting process, K; T(0,z,) – temperature of all points on log cylindrical surface during the freezing or defrosting process, K; Tair-fr and Tairdfr – temperature of the ambient air environment during the log freezing and defrosting processes, K; qv – internal heat source in the log volume caused by the release of the latent heat of both the free and bound water in the wood during their crystallization (DELIISKI and TUMBARKOVA 2017, 2019), W·m-3; αr-fr and αp-fr – convective heat transfer coefficients between the log surfaces and ambient air environment in radial and longitudinal direction during the freezing process respectively, W·m-2·K-1; αr-dfr and αp-dfr – convective heat transfer coefficients between the log surfaces and ambient air in radial and longitudinal direction respectively during the thawing process,W·m-2·K-1.

The models (1) to (8) can be used for the calculation of the 2D temperature distribution in logs, which length, L, is larger than their diameter, D, not more than 3 ÷ 4 times.

Modelling of the 2D temperature distribution in frozen logs during their steaming in an autoclave and subsequent cooling in an air medium The mechanism of 2D change in the temperature in the longitudinal sections of frozen logs during their steaming and subsequent conditioning is mathematically described by eq. (5) (DELIISKI 2003) under an initial condition (DELIISKI et al. 2020b) T (r , z,0) = T02

(9)

where T02 is an initial average mass temperature of the log subjected to steaming, K and the following boundary conditions: • During the steaming process – at a prescribed temperature of the steaming medium: 51

T (r,0, ) = T (0, z, ) = Tm ()

(10)

where Tm is the temperature of the steaming medium in the autoclave, K. • During the cooling the steamed logs in an air environment – under convective boundary conditions, which are presented above by eqs. (7) and (8).

Equations (1) to (10) represent a common form of the coupled mathematical models of 2D heat distribution in logs subjected to alternating freezing and defrosting in an air environment and after that to steaming and subsequent cooling in an air medium.

The positioning of the coordinate axis r and z, and also of 4 representative knots of the calculation mesh is shown in Fig. 1. Subsequently, it was used for the numerical solving the coupled models containing the eqs. (1) to (10).

Fig. 1 Positioning of the coordinate axes and the knots of the calculation mesh with representative points T1, T2, T3, and T4 on ¼ of the longitudinal section of log subjected to freezing, defrosting, and autoclave steaming.

Mathematical description of the thermo-physical properties and icing degree of the logs Mathematical descriptions of the effective specific heat capacities of the wood during its freezing and defrosting, сеff-fr and сеff-dfr, respectively, and also of the effective thermal conductivities of the wood in radial and longitudinal direction during its freezing and defrosting, λr-eff-fr, λp-eff-fr, and λr-eff-dfr, λp-eff-dfr respectively, were suggested in DELIISKI (2003, 2004, 2009, 2011, 2013a, 2013b, DELIISKI – DZURENDA 2010, DELIISKI et al. 2010, 2019, 2020b) based on the experiments mentioned in the dissertations of CHUDINOV (1966) and KANTER (1955) related to the change in c and λ as a function of t and u.

These relations are used in both the European (CHUDINOV 1968, SHUBIN 1990, POŽGAJ et al. 1997, TREBULA and KLEMENT 2002, VIDELOV 2003, PERVAN 2009, HRČKA and BABIAK 2017) and the American professional literature (STEINHAGEN 1986, 1991, STEINHAGEN and LEE 1988, KHATTABI and STEINHAGEN 1992, 1993, 1995) when calculating various processes of wood thermal treatment.

Mathematical descriptions of the effective specific heat capacities of the logs during their freezing and defrosting, ceff-fr and ceff-dfr respectively, which participate in eqs. (1) and (5) respectively, were given in (DELIISKI and TUMBARKOVA 2019, TUMBARKOVA 2019, TUMBARKOVA et al. 2018). Mathematical descriptions of the wood density above the hygroscopic range, ρ, and of the internal heat source in the log volume, qv, are also given in the literature.

For the calculation of the heat transfer coefficients of the horizontally situated beech logs during their freezing and defrosting at the free convection of the periodically changing 52

ambient air temperature, the following experimentally verified equations (3), (4), (7), and (8) (TELEGIN et al. 2002, TUMBARKOVA 2019, DELIISKI et al. 2020c) were used: • In the radial direction on the cylindrical surface of the logs:

 r -fr = 1.123T (0, z, ) − Tm-fr ()0.46 ,

(11)

 r -dfr = 1.123T (0, z, ) − Tm-dfr ()0.26 .

(12)

• In the longitudinal direction on the frontal surface of the logs:  p-fr = 2.56T (r ,0, ) − Tm-fr (τ)0.46 ,

(13)

 p-dfr = 2.56T (r ,0, ) − Tm-dfr ( τ)0.26 .

(14)

The heat transfer coefficients αr-cond = αr-dfr in eq. (8) and αp-cond = αp-dfr in eq. (7) of the beech logs subjected to the air conditioning immediately after their autoclave steaming are equal to (DELIISKI 2003, 2013b)

T (0, z,

 r -cond = 0.3801.026

reg

T (r ,0,

 p-cond = 0.6761.026

reg

) −Tair

  Т (0, z, 

, ( )  Т (r ,0, reg ) − Tair (),

) −Tair ( )

reg ) − Tair ()

(15) (16)

where Tair is the value of the surrounding air environment during the log conditioning, K; τreg – duration of the steaming regime, s.

Mathematical description of the periodically changing atmospheric temperature Mathematical description of the change in the atmospheric temperature, Tair, near the logs stored many days and nights in an open warehouse in winter is needed for the numerical solution of the coupled mathematical models (1) to (8).

The periodic change of the atmospheric temperature Tair during a long time when its maximum value Tair-max remains constant can be described using the following equation (GUZENDA and GANOWICZ 1986, DELIISKI 1988, OLEK and GUZENDA 1995): Т air = Т air0 + (Tair -max − Tair0 )  sin(  )

(17)

where Tair0 is the initial value of Tair, K; Tair-max – maximal value of Tair during its sinusoidal change, K; ω – angular frequency of Tair, s−1; τ – time, s.

The angular frequency of Tair in eq. (17) is equal to =

2 0

(18)

where τ0 is the period of change in Tair, s.

It is needed to use π = 3.14159 for the precise solution of the tasks with the participation of eqs. (17) and (18).

For a periodic change of the air temperature during one day and night, i.e. at τ0 = 1 d = 24 h = 86,400 s, according to eq. (18) it is obtained that =

2 2  3.14159 = = 7.2722  10 −5 s−1 0 86400

53

When Tair0 and Tair-max gradually increase or decrease during the time compared to their initial values, Tair0-in and Tair-max-in respectively, the temperature Tair can be calculated according to the equation Т air = Т air0 -in  (1  Kair0  ) + [(Т air -max -in − Tair0 −in )  (1  Kair -a  )]  sin(  ) (19)

where Kair0 and Kair-a (in s-1) are coefficients determining how much the change in Tair0 and in the amplitude of Tair equal to Tair-a = Tair-max-in – Tair0-in, respectively, is over a period of time. These coefficients can be calculated according to following equations (DELIISKI et al.

2020a): Tair0 -  0 K air0 =

Tair0 -in 0

(20)

Т air -max -  0 Т − Т air0 -in K air -a = air - max -in 0

(21)

where Т air0 − 0 is the change in Tair0 during the time interval equal to τ0, K; Т air -max − 0 – change in Tair-max during one period of τ0, K; Tair0-in – initial value of the periodically changing temperature Tair, K; Tair-max-in – initial value of Tair-max, K. The signs “+” and “–“ on the right side of eq. (19) are used when Tair0 and Tair-max increase or decrease respectively during the periodical change in Tm. For the purpose of the analysis of the current log temperature condition at the initial temperature of the logs before their steaming, synchronously with solving the models (1) to (8), the average mass temperature of the logs, Т avg , for each moment of their alternating freezing and defrosting is calculated according to the equation

Т avg =

1 Tin, k dS  S

(22)

S

where the area of ¼ of the log longitudinal section, S, is equal to

S = R

L 2

(23)

RESULTS AND DISCUSSION The mathematical descriptions of the thermo-physical representatives of non-frozen logs considered above, and also of the periodically changing atmospheric temperature were introduced in the mathematical models (1) to (8).

For numerical solution of the models aimed at computing the 2D temperature fields in logs, a software program was prepared. It was an input in the calculation environment of Visual FORTRAN Professional. An explicit form of the finite-difference method were used for transformation of the models in a form suitable for programming (DELIISKI 2003, 2011, 2013b, DELIISKI and TUMBARKOVA 2019). The calculation mesh was built on ¼ of the

54

longitudinal section of the logs due to the circumstance that this ¼ is mirror symmetrical towards the remaining ¾ of the same section (refer to Fig. 1).

Using the program, computations were made to determine the 2D non-stationary temperature distribution in the longitudinal sections of beech logs under different boundary conditions given below.

During the solution of the models, the mathematical descriptions of the thermophysical characteristics of non-frozen and frozen beech logs with industrial dimensions (diameter D = 0.4 m and length L = 0.8 m), basic density ρb = 560 kg·m−3 (DELIISKI and DZURENDA 2010), moisture content of 0.6 kg·kg−1, and standardized fibre saturation point 293.15 = 0.31 kg  kg −1 , were used. at 293.15 K (i.e. at 20 oC), ufsp

The change in the following parameters during the alternating atmospheric freezing and defrosting or autoclave steaming of the logs were studied in this work: temperature of the processing mediums tair during storing the logs in an open warehouse and tm during their steaming in an autoclave, log surface and average mass temperatures, ts and tavg respectively, and also t of 4 representative points in the logs.

The coordinates of the four representative points in the longitudinal section of the logs were equal to: Point 1 with temperature t1: r = R/4 = 50 mm and z = L/4 = 200 mm; Point 2 with t2: r = R/2 = 100 mm and z = L/4 = 200 mm; Point 3 with t3: r = 3R/4 = 150 mm and z = 3L/8 = 300 mm; Point 4 with t4: r = R = 200 mm and z = L/2 = 400 mm. These coordinates of the points allow the determination and analysis of the 2D temperature distribution in logs during their storing for a long time in an open warehouse and during their subsequent autoclave steaming and conditioning.

Computation of 2D temperature field in logs at changing atmospheric temperature Three options of 120 h (i.e. 5 d) continuous periodic freezing and defrosting of beech logs with an initial temperature t01 = 0 °C (refer to eq. (2)) were studied as follows: • for Log 1: at a constant values of the initial air temperature Tair0 = 268.15 K (i.e. tair0 = –5 °C) and of the maximum value of the sinusoidal changing air temperature Tair-max = 288.15 K (i.e. at the amplitude value of the air temperature tair-a = Tair-max – Tair0 = 20 °C); • for Log 2: at a constant value of tair0 = –5 oC and gradual decrease in the value of tairo o a-in = 20 C by 2 C/d; • for Log 3: at a gradual decrease in the initial values of tair0-in = –5 oC and tair-a-in = 20 o C by 2 oC/d.

To provide a decrease in tair0-in = 20 oC and tair-a-in by 2 °C/d, the following values of the coefficients Kair0 and Kair-a according to eqs. (20) and (21) were used: Kair0 = –8.6325·10-8 s-1 and Kair-a = –1.15741·10-6 s-1.

Fig. 2, Fig. 3, and Fig. 4 present the calculated change in tair, ts, tavg, and t of 4 representative points in Log 1, Log 2, and Log 3 during their continuous 5 day- and nightlong (i.e. 120 h) periodic freezing and defrosting under the described conditions of the atmospheric temperature influence.

It can be seen that during the second period of the change in tair (i.e. between 24th and 48th hour) the temperature in all points of the studied logs drops below 0 oC. It means that there are conditions for water crystalizing in the entire volume of the logs. The temperature in the central point, t4, changes the least. It remains very long in the range from 0 oC and –1 o C, in which the free water in the wood freezes (DELIISKI and TUMBARKOVA 2017). Only in the 81st h for Log 1, in the 91st h for Log 2, and in the 71st h for Log 3 t4 drops to –1 oC and the freezing process of the whole amount of the free water in the log centre ends and then the freezing of the bound water begins (TUMBARKOVA 2019).

55

30 Log 1: t air0 = –5 oC = const, t air-a = 20 oC = const, D = 0.4 m

Temperature t , ° C

tair

t 01 = 0 oC, ρb = 560 kg.m-3, u = 0.6 kg.kg-1, L = 0.8 m

20

ts

10

t1

0

t2 t3

-10

t4

-20 tavg

-30 0

12

24

36

48

60

72

84

96

108 120

Time τ, h

Fig. 2 Change in tair, ts, tavg, and t of 4 representative points of the Log 1 during its 120 h periodical freezing and defrosting at constant values of tair0 and tair-a.

30 Log 2: t air0 = –5 oC = const, t air-a = 20 oC - 2 oC/d, D = 0.4 m

Temperature t , ° C

20

tair

t 01 = 0 oC, ρb = 560 kg.m-3, u = 0.6 kg.kg-1, L = 0.8 m ts

10

t1

0

t2 t3

-10

t4

-20 tavg

-30 0

12

24

36

48

60

72

84

96

108 120

Time τ, h

Fig. 3 Change in tair, ts, tavg, and t of 4 representative points of the Log 2 during its 120 h periodical freezing and defrosting at constant value of tair0 and decreasing value of tair-a.

In Fig. 2 it is seen that at the constant values of tair0 = –5 oC and tair-a = 20 oC after 96th h, i.e. after the 4th period of tair, a periodical change in the log temperature with practically constant amplitudes for the separate points comes. The further is distance of the point from the log surfaces, the smaller is the amplitude of the periodic change in the temperature in that point. The amplitudes of tair, ts, tavg, and t in the separate representative points after the 4th period are equal to: tair-a = 20.0 °C, ts-a = 8.4 °C, t1a = 6.4 °C, t2a = 4.3 °C, t3a = 2.0 °C, t4a = 1.6 °C, and tavg-a = 4.1 °C. When tair0 = –5 oC remains constant and the initial value of tair-a = 20 oC decreases by o 2 C/d (Fig. 3) or when both mentioned initial values of tair0 and tair-a decrease by 2 oC/d (Fig. 4) during the time, the amplitudes of the temperature in the separate points gradually decrease and during the second half of the last 5th period of tair they are equal to: tair-a = 10.0 o C, tsa = 5.0 oC, t1a = 3.6 oC, t2a = 2.4 oC, t3a = 1.2 oC, and t4a = 1.0 oC. 56

30 Log 3: t air0 = –5 oC - 2 oC/d, t air-a = 20 oC - 2 oC/d, D = 0.4 m

Temperature t , ° C

20

o

-3

tair

-1

t 01 = 0 C, ρb = 560 kg.m , u = 0.6 kg.kg , L = 0.8 m

10

tsL t1

0

t2 t3

-10

t4

-20 tavg

-30 0

12

24

36

48

60

72

84

96

108 120

Time τ, h

Fig. 4 Change in tair, ts, tavg, and t of 4 representative points of the Log 3 during its 120 h periodical freezing and defrosting at decreasing values of tair0 and tair-a.

The average mass temperature of the logs, tavg strongly affecting the duration and energy consumption of the regimes for autoclave steaming of frozen logs (DELIISKI 2013b) is equal to the following values: ▪ at 24th h: –2.8 oC for Log1, –2.5 oC for Log2, and –3.2 oC for Log3; ▪ at 48th h: –6.0 oC for Log1, –5.0 oC for Log2, and –7.8 oC for Log3; ▪ at 72nd h: –8.2 oC for Log1, –6.4 oC for Log2, and –11.9 oC for Log3; ▪ at 96th h: –9.8 oC for Log1, –7.2 oC for Log2, and –15.1 oC for Log3; ▪ at 120th h: –10.4 oC for Log1, –7.4 oC for Log2, and –16.7 oC for Log3.

Computation of 2D temperature field in logs during their steaming in an autoclave Two options of autoclave steaming and subsequent conditioning of the frozen beech logs (named as Log 4 and Log 5) were studied: • Log 4 was with an initial temperature t02 = –1 °C (refer to eq. (9)); • Log 5 was with an initial temperature t02 = –20 °C. At the beginning of the steaming process, the Log 4 contains only frozen free water in it, but the Log 5 apart from this contains significant amount of frozen bound water.

During solving the equation (5) under an initial condition (9) and boundary conditions (10), 3-stage regimes for autoclave steaming of the logs were used. The typical temperature time profile of the processing medium temperature tm in a steaming autoclave and of the air medium for the consequent conditioning of the heated wood materials is shown in (DELIISKI 2003, DELIISKI and DZURENDA 2010, DELIISKI et al. 2020b).

In Fig. 5 and Fig. 6, the calculated change in tm, ts, and t of 4 representative points of the Log 4 and Log 5 during their autoclave steaming and subsequent conditioning at tm = 20 °C is presented.

One of the aims of this study was to determine how the average mass temperature of the logs at the beginning of the steaming process influences the duration of the steaming regime. That is why the initial temperature of Log 4 and Log 5, t02, which in the practice is equal to tavg of the logs after their storing in an open warehouse, were assumed to be too different from each other and equal to t02 = –1 °C and t02 = –20 °C respectively, during the simulations with the model (5) ÷ (8), using eqs. (15) and (16).

57

160 Log 4: D = 0.4 m, L = 0.8 m, ρ b = 560 kg·m-3, u = 0.6 kg·kg-1, t 02 = -1 oC

140 tm

Temperature t , oC

120

ts

100 t1

80

t2

60

t3

40

t4

20

tmax tmin

0 -20 0

2

4

6

8 10 Time τ, h

12

14

16

18

Fig. 5 Change in tm, ts, and t in 4 representative points of the Log 4 with t02 = –1 °C during its steaming in an autoclave and its subsequent conditioning at tair = 20 °C. 160 Log 5: D = 0.4 m, L = 0.8 m, ρ b = 560 kg·m-3, u = 0.6 kg·kg-1, t 02 = -20 oC

140 tm

o

Temperature t , C

120 ts

100

t1

80

t2

60

t3

40

t4

20

tmax

0

tmin

-20 0

2

4

6

8

10 12 Time τ, h

14

16

18

20

Fig. 6 Change in tm, ts, and t in 4 representative points of the Log 5 with t02 = –20 °C during its steaming in an autoclave and its subsequent conditioning at tair = 20 °C.

In Fig. 5 and Fig. 6, the minimum and maximum values of the temperature, tmin = 62 °C and tmax = 90 °C, are also shown. It is well known that the temperature of all representative points of the logs during the veneer cutting process between these optimum values of tmin and tmax is needed to obtain the quality veneer from plasticized beech wood (DELIISKI 2003, DELIISKI and DZURENDA 2010).

The temperature fields in the logs was computed for the processes of their steaming in an autoclave and their subsequent conditioning in an air environment.

This means that the calculation of the non-stationary 2D change in the temperatures in the longitudinal sections of the logs during the time of their conditioning begins from the already reached temperature during the time of calculations distribution of temperature at the end of the steaming regime. Based on the calculations, the time of reaching the 58

temperature in the entire volume of the heated logs needed for cutting the veneer can be determined (between tmin and tmax in Fig. 5 and Fig. 6).

Fig. 5 and Fig. 6 show that temperatures of all representative points enter between tmin = 62 °C and tmax = 90 °C after conditioning of the heated logs in an air environment equal to approximately 60 min for Log 4 and 90 min for Log 5.

The analysis of Fig. 5 and Fig. 6 shows that the duration of the steaming regimes of the studied logs is equal to τreg = 10.5 h for Log 4 and to τreg = 14.0 h for Log 5. It means that a decrease in the initial log temperature t02 from –1 oC to –20 oC (i.e. by 19 °C causes a decrease in τreg by 3.5 h, i.e. each decrease in t02 by 1 °C in our case causes an increase in τreg by approximately 11 min.

The comparison of the obtained results with the results in (DELIISKI et al. 2020b) show that a decrease in the initial temperature of the logs from 0 oC to –1 oC when the whole amount of the free water is fully crystallized and there is still no frozen bound water in the wood causes elongation of the regime for the autoclave steaming by 30 min.

CONCLUSIONS An approach to mathematical modelling and research on the 2D non-stationary temperature distribution in logs stored for a long time in an open warehouse in winter under the influence of periodically changing atmospheric temperature and during their subsequent steaming in autoclaves was described in the paper.

Mathematical descriptions of the periodically changing atmospheric temperature, of the temperature of the autoclave steaming regimes of the logs, and also of their subsequent conditioning in an air environment were presented. These descriptions were introduced as boundary conditions in our own 2D non-linear coupled mathematical models of the 2D temperature distribution in logs during their freezing and defrosting.

A software program for solving the models and computing the 2D temperature field of logs during the processes were prepared in the calculation environment of Visual FORTRAN Professional. The paper showed and analysed, e.g. the application of the suggested approach, diagrams of the change in tm, ts, tavg, and 2D temperature distribution in beech logs with industrial dimensions (diameter of 0.4 m and length of 0.8 m), basic density of 560 kg·m-3, and moisture content of 0.6 kg·kg-1 in the following two cases: • during 5 day-long (i.e. 120 h) continuous alternating freezing and defrosting of logs with an initial temperature of 0 oC under the influence of periodically changing atmospheric temperature at its constant initial value tair0 = –5 °C and constant amplitude tair-a = 20 °C (Log 1); at a constant value of tair0 = –5 oC and gradual decrease in tair-a-in = 20 oC by 2 oC/d (Log 2) and at gradual decrease in both initial values of tair0-in = –5 oC and tair-a-in = 20 oC by 2 oC/d.

It was computed that at the end of 120 h periodically freezing and defrosting of the studied logs their average mass temperature was equal to –10.4 oC for Log 1, to –7.4 oC for Log 2, and to –16.7 oC for Log 3; • during 3-stage regimes for autoclave steaming of two logs with an initial temperature of –1 °C (Log 4) and –20 °C (Log 5) and during 2.5 h of their subsequent conditioning (cooling) at the air temperature of 20 °C. It was determined that the duration of the steaming regimes of the studied logs was equal to τreg = 10.5 h for Log 4 and to τreg = 14.0 h for Log 5. It means that each decrease in the initial log temperature by 1 °C in the considered temperature range causes an increase in τreg by approximately 11 min.

59

Comparing these data with previous results of the authors it was determined that a decrease in the initial temperature of the logs from 0 oC to –1 oC when the whole amount of the free water is fully crystallized and there is still no frozen bound water in the wood causes elongation of the regime for autoclave steaming by 30 min.

The presented approach to the computing the 2D temperature field in logs and their average mass temperature at periodically changing atmospheric temperature can help to determine accurately the initial temperature of the logs before steaming, depending on the duration of the log storing in an open warehouse during all seasons. This creates a basis for the development of energy saving steaming regimes with an optimal duration depending on the initial temperature of the frozen and non-frozen logs of each batch subjected to thermal treatment.

The suggested approach and the results from the solutions of the coupled two models can be used for development and implementation of advanced systems for model-based automatic control (DELIISKI 2004, 2011, HADJISKI and DELIISKI 2016, HADJISKI et al. 2019) of different thermal treatment processes of logs and other wood materials.

REFERENCES CHUDINOV, B. S. 1966. Theoretical Research of Thermo Physical Properties and Thermal Treatment of Wood. Dissertation for DSc., Krasnojarsk, USSR : SibLTI. CHUDINOV, B. S. 1968. Theory of the Thermal Treatment of Wood. Moscow : Nauka, 255 pp.

DELIISKI, N. 1988. Thermische Frequenzkennlinien von wetterbeanspruchten Holzbalken. In Holz als Roh- und Werkstoff, 46(2): 59−65.

DELIISKI, N. 2003. Modeling and Technologies for Steaming of Wood Materials in Autoclaves. DSc. thesis, Sofia : University of Forestry, 358 pp.

DELIISKI, N. 2004. Modelling and Automatic Control of Heat Energy Consumption Required for Thermal Treatment of Logs. In Drvna Industrija, 55(4): 181−199.

DELIISKI, N. 2009. Computation of the 2-dimensional Transient Temperature Distribution and Heat Energy Consumption of Frozen and Non-Frozen Logs. In Wood Research, 54(3): 67−78.

DELIISKI, N. 2011. Transient Heat Conduction in Capillary Porous Bodies. In Ahsan A. editor.

Convection and Conduction Heat Transfer. Rieka : InTech Publishing House, 149−176.

DELIISKI, N. 2013a. Computation of the Wood Thermal Conductivity during Defrosting of the Wood.

In Wood Research, 58(4): 637−650.

DELIISKI, N. 2013b. Modelling of the Energy Needed for Heating of Capillary Porous Bodies in Frozen and Non-Frozen States. Saarbrücken : Lambert Academic Publishing, Scholars’ Press, Germany, 106 pp.

DELIISKI, N., DZURENDA, L. 2010. Modelling of the Thermal Processes in the Technologies for Wood Thermal Treatment. Zvolen : TU vo Zvolene, Slovakia, 224 pp.

DELIISKI, N., DZURENDA, L., MILTCHEV, R. 2010. Computation and 3D Visualization of the transient Temperature Distribution in Logs during Steaming. In Acta Facultatis Xylologiae Zvolen, 52(2): 23−31.

DELIISKI, N., DZURENDA, L., TUMBARKOVA, N., ANGELSKI, D. 2015. Computation of the Temperature Conductivity of Frozen Wood during its Defrosting. In Drvna Industrija, 66(2): 87−96.

DELIISKI, N., DZURENDA, L., ANGELSKI, D., TUMBARKOVA, N. 2019. Computing the Energy for Warming up of Prisms for Veneer Production during Autoclave Steaming with a Limited Power of the Heat Generator. In Acta Facultatis Xilologiae Zvolen, 61(1): 63−74.

DELIISKI, N., TUMBARKOVA, N. 2017. An Approach and an Algorithm for Computation of the Unsteady Icing Degrees of Logs Subjected to Freezing. In Acta Facultatis Xilologiae Zvolen, 59(2): 91−104.

60

DELIISKI, N., TUMBARKOVA, N. 2019. Numerical Solution to Two-Dimensional Freezing and Subsequent Defrosting of Logs. In Iranzo A. editor. Heat and Mass Transfer - Advances in Science and Technology Applications, IntechOpen, 20 pp.

DELIISKI, N., NIEMZ, P., TUMBARKOVA, N. 2020a. Computing the 2D Temperature Distribution and Icing Degree in Logs at Changing Atmospheric Temperature in Winter. PRO LIGNO, 16(3): 8−18.

DELIISKI, N., DZURENDA, L., TRICHKOV, N., TUMBARKOVA, N. 2020b. Computing the 2D Temperature Field in Non-frozen Logs at Changing Atmospheric Temperature and during their Subsequent Autoclave Steaming. In Acta Facultatis Xilologiae Zvolen, 62(2): 47−59.

DELIISKI, N., DZURENDA, L., TUMBARKOVA, N. 2020c. Modelling of the Two-Dimensional Thawing of Logs in an Air Environment. In Valdman J. editor. Modeling and Simulation in Engineering – Selected Problems, IntechOpen, 19 pp.

GUZENDA, R., GANOWICZ, R. 1986. Temperaturänderungen in brettschichtverleimten Holzbalken bei periodisch wechselnden Umgebungtemperaturen. Holz als Roh– und Werkstoff, 44(1): 61−67.

HADJISKI, M., DELIISKI, N. 2016. Advanced Control of the Wood Thermal Treatment Processing. In Cybernetics and Information Technologies, Bulgarian Academy of Sciences, 16(2): 179−197.

HADJISKI, M., DELIISKI, N., GRANCHAROVA, A. 2019. Spatiotemporal Parameter Estimation of Thermal Treatment Process via Initial Condition Reconstruction using Neural Networks. In Hadjiski M, Atanasov KT, editors. Intuitionistic Fuzziness and Other Intelligent Theories and Their Applications. Springer International Publishing AG, Cham, Switzerland: 51−80.

HRČKA, R., BABIAK, M. 2017. Wood Thermal Properties. In Wood in Civil Engineering, Giovanna Consu, InTechOpen: 25−43.

KANTER, K. R. 1955. Investigation of the Thermal Properties of Wood. PhD Thesis, Moscow, USSR : MLTI. KHATTABI, A., STEINHAGEN, H. P. 1992. Numerical Solution to Two-Dimensional Heating of Logs. In Holz als Roh- und Werkstoff, 50 (7−8): 308−312.

KHATTABI, A., STEINHAGEN, H. P. 1993. Analysis of Transient Non-linear Heat Conduction in Wood Using Finite-difference Solutions. In Holz als Roh- und Werkstoff, 51(4): 272−278.

KHATTABI A., STEINHAGEN H. P. 1995. Update of “Numerical Solution to Two-dimensional Heating of Logs”. Holz als Roh- und Werkstoff, 53(1): 93−94.

OLEK, W., GUZENDA, R.1995. Prediction of Temperature Changes in Glued Laminated Beams. Holz als Roh– und Werkstoff, 53(4): 249−252.

PERVAN, S. 2009. Technology for Treatment of Wood with Water Steam. Zagreb : University in Zagreb.

POŽGAJ, A., CHOVANEC, D., KURJATKO, S., BABIAK, M. 1997. Structure and Properties of Wood. 2nd edition, Bratislava : Príroda a.s., 485 pp.

RIEHL, T., WELLING, J., FRÜHWALD, A. 2002. Druckdämpfen von Schnittholz, Arbeitsbericht 2002/01: Institut für Holzphysik, Hamburg: Bundesforschungsanstalt für Forst- und Holzwirtschaft.

SHUBIN, G. S. 1990. Drying and Thermal Treatment of Wood. Moscow : Lesnaya Promyshlennost, 337 pp.

SOKOLOVSKI, S., DELIISKI, N., DZURENDA, L. 2007. Constructive Dimensioning of Autoclaves for Treatment of Wood Materials under Pressure. In Woodworking Techniques, Zalesina, Croatia, 117−126.

STEINHAGEN, H. P. 1986. Computerized Finite-difference Method to Calculate Transient Heat Conduction with Thawing. In Wood and Fiber Science 18(3), p. 460−467.

STEINHAGEN, H. P. 1991. Heat Transfer Computation for a Long, Frozen Log Heated in Agitated Water or Steam – A Practical Recipe. In Holz Roh- Werkstoff, 49(7–8): 287−290.

STEINHAGEN, H. P., LEE, H. W. 1988. Enthalpy Method to Compute Radial Heating and Thawing of Logs. In Wood and Fiber Science, 20(4): 415−421.

TELEGIN, A. S., SHVIDKIY, B. S., YAROSHENKO, U. G. 2002. Heat- and Mass Transfer. Moscow : Akademkniga, 456 pp.

TREBULA, P., KLEMENT, I. 2002. Drying and Hydrothermal Treatment of Wood. Zvolen : TU vo Zvolene, 449 pp.

61

TUMBARKOVA, N. 2019. Modeling of the Logs’ Freezing and Defrosting Processes and their Energy Consumption. PhD Thesis, Sofia : University of Forestry, 198 pp.

TUMBARKOVA, N, DELIISKI, N., PENKOVA, N., MIHAILOV E. 2018. Computation of the Temperature Field at Cooling of Logs below the Freezing Point of the Moisture. Information Technologies and Control, 16(1): 8−15.

VIDELOV, H. 2003. Drying and Thermal Treatment of Wood. Sofia: University of Forestry, 335 pp.

ACKNOWLEDGEMENTS This document was supported by the APVV Grant Agency as part of the project: APVV-17-0456 as a result of work of authors and the considerable assistance of the APVV agency.

AUTHORS’ ADDRESSES Prof. Dr. Nencho Deliiski, DSc. University of Forestry Faculty of Forest Industry St. Kliment Ohridski Blvd. 10 1797 Sofia Bulgaria deliiski@netbg.com Prof. Ing. Ladislav Dzurenda, PhD. Technical University in Zvolen Faculty of Wood Science and Technology T. G. Masaryka 24 960 01 Zvolen Slovakia dzurenda@tuzvo.sk Prof. Dr. Eng. habil. Peter Niemz Institut für Baustoffe Stefano-Franscini-Platz 3 CH 8093 Zürich niemzp@retired.etzh.ch

Assoc. Prof. Dimitar Angelski, PhD. Eng. Mag. Natalia Tumbarkova, PhD. University of Forestry Faculty of Forest Industry St. Kliment Ohridski Blvd. 10 1797 Sofia, Bulgaria d.angelski@gmail.com ntumbarkova@abv.bg

62

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 63−73, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.06

BONDING OF THE THERMALLY MODIFIED NORWAY SPRUCE WOOD WITH THE PUR AND PVAc ADHESIVES Zuzana Vidholdová – Dávid Ciglian – Ladislav Reinprecht ABSTRACT The quality of the bonded joints in wood structures and furniture products is influenced by the surface characteristics of wood and the physical-chemical characteristics of adhesive.

In this study the polyurethane (PUR) and polyvinyl acetate (PVAc) adhesives were used for bonding the Norway spruce (Picea abies /L./ Karst.) wood which was firstly dried at 100 °C and subsequently thermally modified for four hours at the temperatures of 160 °C, 180 °C, 200 °C, and 220 °C. The shear strength of the lap joints in bonded wood specimens, determined in accordance with the standard EN 205:2016, was analysed in relation to (a) the type of adhesive, (b) the temperature at thermal modification of wood, and (c) the composition of the bonded wood specimen formed from two slabs which were prepared from unmodified, unmodified and thermally modified, and thermally modified timbers, respectively. The shear strength of the lap joints was lower (a) at using the PVAc adhesive compared to PUR adhesive, (b) at using the wood slabs prepared from timber thermally modified at higher temperatures from 160 °C to 220 °C, and (c) for the specimens formed from slabs both thermally modified. In comparison to the reference specimens, there the highest decrease in the shear strength – by 42.3%, from 9.3 MPa at using PVAc adhesive, or by 56.1%, from 11.2 MPa at using PUR adhesive – was found for specimens formed from two slabs thermally modified both at 220 °C. Only at the highest modification temperatures, a cohesive type of failure occurred directly in the wood adherent.

Key words: polyvinyl acetate, polyurethane, shear strength, spruce, thermal modification.

INTRODUCTION The use of wood in the construction industry is currently experiencing a renaissance.

New joining methods and construction principles, as well as the discovery of the classic and modified wood-based material for modern architectural solutions, have opened up new possibilities for building with wood (ŠTEFKO and REINPRECHT 2004). The advent of different engineered wood products from sawn timber and veneer such as Cross-Laminated Timber, Glued-Laminated Timber, Laminated Veneer Lumber, etc. as a versatile material that can be fully combined with other building materials has led to its increased use for building detached houses, especially for multi-storey apartment buildings and in high-tech architecture. On the other hand, the world of the engineered wood products is rapidly changing and dynamic innovation process in this field is continuously expanding (SANDBERG et al. 2018).

63

The bonding quality of engineered wood products from sawn timber and veneers is affected by various factors to which mainly belong (1) the structural characteristics and physical properties of the wood used in a function of adherent such as its: - chemical composition which can be changed at its chemical and thermal modifications, as well as at its physical surface modifications with electrical discharge-corona, cold plasma treatments, laser treatment and mercerization (KAMKE and LEE 2007, PETRIČ 2013, NOVÁK et al 2015 KONNERTH et al. 2016, AICHER et al. 2018, BEKHTA et al. 2018, JAKES et al. 2018, REINPRECHT et al. 2020), - porosity, - surface free energy, - moisture content, - roughness which is influenced by wood morphology and its surface machining by sawing, sanding or planing (HASS et al. 2014, KNORZ et al. 2015), (2) the properties of the adhesive such as its: - chemical composition, - modification with additives, - solid content, - viscosity and surface free energy, - pH value, - buffering capacity, - hardening time, - rate of its curing or solidification (AYDIN 2004, HUNT et al. 2018, TRAN et al. 2020), and (3) the processing parameters of bonding such as: - spread rate, - open and closed assembly time, - temperature, - pressure (FOLLRICH et al. 2007, ŠMIDRIAKOVÁ and KOLLÁR 2010, BEKHTA et al. 2014).

Bonding of thermally modified wood can pose some issues. Changes in the chemical composition, anatomy, physical and mechanical properties of wood after thermal modification can affect the ability of adhesives at jointing the wood surfaces (SERNEK et al.

2008). The improved dimensional stability of thermally modified wood commonly improves the bonding performance, because the stresses due to shrinking or swelling on the cured adhesive bond of wood are reduced (REINPRECHT and VIDHOLDOVÁ 2008). However, heat treatment of wood can be expected to cause significant changes related to its adhesion with adhesives, which makes it necessary to adapt the bonding process (KRYSTOFIAK et al. 2013).

Strong adhesion between the adhesive and the wood is achieved by appropriate liquid flow of the adhesive, its penetration into wood and following curing.

Thermally modified wood is less hygroscopic (HILL 2006, REINPRECHT and VIDHOLDOVÁ 2008, VIDHOLDOVÁ et al. 2019, KUČEROVÁ et al. 2019), which can alter the distribution of the adhesive on the wood surface and the penetration of the adhesive into porous of wood (FOLLRICH et al. 2006). The intensity of water absorption from the waterborne adhesive could affect its hardening process and subsequently the quality of the adhesive bond. Several studies have shown that the wettability of wood with water decreases after heat treatment (WANG et al. 2015, HUANG et al. 2012, BUDHE et al. 2020, KÚDELA et al. 2020), mainly because the surface of the heat-treated wood is more hydrophobic, less polar and significantly repellent to water (REINPRECHT and REPÁK 2019, BAAR et al. 2020).

BASTANI et al. (2016a) found that the processing time needed for the adhesive to be absorbed into the thermally modified wood is higher due to slower penetration rate. Changes of the pH value of the thermally modified wood surface might retard or accelerate the curing of adhesives, depending on their type (CAI et al. 2018). Adhesives penetrates relatively easily into the voids and porous structure of wood tissue (KAMKE and LEE 2007, HUNT et al. 2018), also after its initial thermal modification (BASTANI et al. 2016b). Due to thermal modification of wood at higher temperatures, there in its anatomical structure are created other free spaces like cracks in the cell walls (TIRALOVÁ and MAMOŇOVÁ 2005, BOONSTRA et al. 2006). Additionally, there are created smaller substances due to depolymerisation reactions in its lignin-polysaccharide components, and changes occur also in the chemical reactivity of some chemical components of wood cell walls (INARI et al. 2007).

Both glue-laminated wood and thermally modified wood offers the interesting opportunities in area of the engineered wood products. In some studies (SERNEK et al. 2008, KRYSTOFIAK et al. 2013, MIRZAEI et al. 2017, and PULNGERN et al. 2020) the bonding performance of the thermally modified wood and also the glulam made only from thermally

64

treated timbers were evaluated. However, there is no information about quality of the joints in bonded timbers when thermally modified and unmodified wood is bonded together.

The aim of this experiment was to determine the bonding performance of the thermally modified wood prepared at various modification temperatures – through the adhesive bond strength and the type of delamination.

MATERIALS AND METHODS Wood materials The sound Norway spruce (Picea abies /L./ Karst.) timber with the moisture content of 10% ± 2% were machine-milled to a thickness of 10 mm and dried at a temperature of 100 °C for four hours. The dry timbers were then exposed to thermal modification processes at the temperatures of 160 °C, 180 °C, 200 °C, and 220 °C, lasting four hours under atmospheric pressure in the laboratory heating oven Memmert UFE 500 (Schwabach, Germany). Finally, the timbers were cooled down and 14 day-long conditioned at a temperature of 23 ± 2 °C and a relative air humidity of 50 ± 5 %. Equilibrium moisture content of the unmodified timber was w = 9.1% ± 0.1%, while of the thermally modified timber was lower: 160 °C: w = 6.73% ± 0.10%, 180 °C: w = 6.23% ± 0.03%, 200 °C: w = 5.34% ± 0.15%, and 220 °C: w = 4.52% ± 0.18%.

Adhesives Two different adhesives were used in the experiment: (a) one-component polyurethane (PUR) Kestopur 1030 (Kiilto Oy, Tampere, Finland), and (b) one-component polyvinyl acetate (PVAc) Rakoll® 4330 (H.B. Fuller Europe, Zürich, Switzerland). The specific characteristics of adhesives as well as the recommended processing conditions of these adhesive systems are summarized in Table 1.

Tab 1. Adhesive systems and processing conditions.

Adhesive Type Viscosity at 20 °C [mPa·s] Density [kg/m3] Colour pH value Recommended spread rate [g/m2] Open time [min] Pressing time [min]

Kestopur 1030 Polyurethane (PUR) 7000 1200 Transparent, light after drying 160-200 30 90-120

Rakoll® 4330 Polyvinyl acetate (PVAc) 13000 1100 White, transparent after drying 3 160-180 8-12 10-15

Lap joint shear “laminated” specimens The spruce timbers dried at the temperature of 100 °C, as well as those following thermally modified at the temperatures from 160 °C to 220 °C, were machine-planed to a thickness of 5 mm and subsequently machine-grinded with sandpaper number of 120. The lap joint shear specimens were prepared according to the standard EN 205:2016 with these requirements: (a) only straight cut wood - parallel with the fiber orientation, and (b) the growth ring angle of the wood adherents only between 30° and 90°. The single-lap joints with the overlap of 10 mm were prepared from two wood slabs with dimensions of 80 mm × 20 mm × 5 mm.

The lap joint shear specimens were made from unmodified and thermally modified wood slabs as follows: I. Slabs from unmodified timber (timber dried at 100 °C) – variant I (reference);

65

II. The combination of slabs of unmodified timber and thermally modified timber – variant II; III. Slabs from thermally modified timber – variant III. The schematic of single-lap joints for shear strength test – laminated spruce specimens – are shown in Figure 1. Due to the facts, that the curing reaction of PUR adhesives requires water, the thermally modified slabs were additionally moisturized by spring with water to rise the moisture content of their surface on 12 ± 2%. The spread rate of adhesives slabs surfaces was 180 g/m2.

F F

a)

b)

c)

d)

Fig. 1 The single-lap joint shear – laminated spruce specimens used in the experiment. a) Schematic configuration of specimen according to the standard EN 205:2016. b−d) Single lap-joint of three tested variants: b - slabs from timber dried at 100 °C, c - the combination of slabs of timber dried at 100 °C and thermally modified timber, and d - slabs from thermally modified timber.

Shear strength test The shear strength of the laminated single-lap joint shear specimens was tested in the machine LabTech 4.050 (LaborTech s.r.o., Opava, Czech Republic) with 5 kN head.

Specimens were placed into the testing machine directly after being removed from the standard climate (20 °C, 65% RH, 7 days after bonding), and loaded with a speed of 50 mm∙min-1 until breakage occurred according to EN 205:2016.

The shear strength was computed as the ratio between the maximal force and the bonded area (10 × 20 = 200 mm2). Delaminating failures in the specimens were estimated visually.

RESULTS AND DISCUSSION The highest shear strength of the bonded wood specimens was recorded for the reference ones – composed of two slabs prepared from unmodified timbers dried at 100 °C (variant I.)

66

– bonded with the PUR adhesive. At using the PVAc adhesive, the shear strength capacity of the reference specimens achieved only 83% comparing to using the PUR adhesive (Table 2).

The shear strength of the bonded specimens from unmodified and thermally modified timber (variant II) and only from thermally modified timber (variant III.) continuously decreased with increasing the temperatures 160-220 °C used at thermal modification. A higher reduction in the shear strength was determined with application of the PUR adhesive than with the PVAc adhesive (Table 2 - see average values, Duncan test, and p-level of significance, Figures 2 a 3). This result can be explained by the more significant effect of the increased hydrophobicity of thermally modified wood surfaces on the deteriorating bond quality of the PUR adhesive.

Tab. 2. The shear strength of the bonded specimens formed from the spruce slabs and the PVAc or PUR adhesives.

Shear strength PVAc Duncan test

Slab’s combination Average SD Wood Average SD in bonded specimen failure [MPa] (p-level) [%] [MPa] I. Slabs from unmodified timber (Reference) 100°C/100°C 9.3 0.8 < 10 11.2 1.6 II. Slabs from unmodified timber and thermally modified timber 100°C/160°C 9.2 1.1 d (0.811) 10-20 9.5 1.5 100°C/180°C 8.3 1.0 d (0.080) 20-30 9.4 2.1 100°C/200°C 7.9 1.2 c (0.011) 40-50 7.9 1.7 100°C/220°C 7.0 0.8 c (0.011) 90-100 6.2 1.3 III. Slabs from thermally modified timber 160°C/160°C 8.0 1.2 c (0.022) 10-20 8.2 1.3 180°C/180°C 7.8 1.2 c (0.016) 40-50 8.0 1.4 200°C/200°C 6.6 1.3 a (0.000) 90-100 7.5 1.3 220°C/220°C 5.4 0.6 a (0.000) 90-100 4.9 1.1

PUR Duncan test (p-level)

Wood failure [%]

-

< 10

d d a a

(0.056) (0.056) (0.000) (0.000)

10-20 10-30 90-100 90-100

b b a a

(0.001) (0.001) (0.000) (0.000)

10-20 90-100 90-100 90-100

Notes: Average - mean values from 10 replicates of tested single-lap joint of laminated specimens; SD - standard deviations; a, b, c, d - indexes of the Duncan test characterizing the significance level of shear strength in relation to the reference laminated specimens 100/100 (a very significant decrease > 99.9%, b - significant decrease > 99%, c - less significant decrease > 95%, d -insignificant decrease < 95%).

The lowest shear strength of bonded specimens was determined in the case of using spruce slabs thermally modified with the highest temperature of 220 °C – i.e., drop in comparison to the reference specimens by 42.3%, from 9.30 MPa at using PVAc adhesive, or by 56.1%, from 11.20 MPa when using PUR adhesive. Results of the shear strength valued by the Duncan test (Table 2) show statistically significant differences in relation to the shear strength of reference specimens – from “c” (p-level lower than 0.05, at 95% level of confidence) to “a” (p-level lower than 0.001, at 99.9% level of confidence). The statistically lower shear strength was determined mainly for bonded specimens formed from slabs prepared only from the thermally modified timber (variant III. – significant and continuous decrease of strength from 160°C/160°C to 220°C/220°C). On the contrary, for the bonded specimens formed from the reference slabs “i.e., from timber dried at 100 °C” and the thermally modified slabs “i.e., from thermally modified timber”, there at using slabs modified at lower temperatures (variant II. - combinations 100°C/160°C and 100°C/180°C) were not determined statistically significant decreases (p-level higher than 0.05) By the linear correlations was analysed the decrease in the shear strength of the bonded specimens in dependence of the increased temperature during the thermal modification of spruce timbers (Figure 2). A significantly negative effect of the increased modification temperature (t) was confirmed by the coefficient of determination r2 and the p-level = 0.000

67

of the linear correlation “ = a + b · t”. The r2 was 0.390 for PVAc adhesive and 0.469 for PUR adhesive at the combination of slabs from timber dried at 100 °C and from thermally modified timber (variant II.), respectively, 0.648 and 0.654 at the combination of slabs only from thermally modified timber (variant III.).

14



PVAc

= 11.47 - 0.018 · t r2 = 0.390 p = 0.000

14

10 8 6 4 2 0

100/160

100/180

6 4

100/100

 = 12.68 - 0.031 · t

PVAc

r2 = 0.648

14

p = 0.000

100/160

100/180

100/200 100/220 [°C]



PUR

= 16.05 - 0.047 · t r2 = 0.654 p = 0.000

12

Shear strength -  [MPa]

Shear strength -  [MPa]

8

0

100/200 100/220 [°C]

12 10 8 6 4 2 0

10

2

100/100

14

= 15.57 - 0.039 · t r2 = 0.469 p = 0.000

12

Shear strength -  [MPa]

Shear strength -  [MPa]

12



PUR

10 8 6 4 2

100/100

160/160

180/180

200/200 220/220 [°C]

0

100/100

160/160

180/180

200/200 220/220 [°C]

Fig. 2 Linear correlations between the shear strength () of the bonded spruce specimens and the modification temperatures – including the drying temperature of 100 °C (t).

Generally, the shear strength of the bonded spruce specimens decreased apparently less for those ones made by combination of spruce timber dried at 100 °C and thermally modified timber, in comparison to the bonded spruce specimens made only from thermally modified timber (Figure 3).

The lower shear strength values of the bonded thermally modified timber determined some other researchers, as well. For example, UZUN et al. (2016) found out that the reduction of density and changes in surface properties of heat-treated wood, as well as the physicalchemical characteristics of adhesive, could potentially affect the bonding performance of thermally modified wood. ANDROMACHI and EKATERINI (2018) also mentioned that the shear strength reduction of bonded wood can be due its degradation during its previous thermal treatment in connection with its density reduction and not due to the reduction of the adhesive bond capacity.

The summarised view for the reduction of the shear strength of the bonded thermally modified timber were offered in study of TAGHIYARI et al. (2020). The shear strength reduction of bonds created from thermally modified timber can be attributed to: (a) a reduction of polar groups in the cell walls of wood due to the degradation of amorphous polysaccharides by the heat treatment, resulting in less sites available for bonding, (b) an increased stiffness of the cell 68

walls after heat treatment, which results in a reduction of internal surfaces for chemical bonding or mechanical interlocking of adhesives, and (c) a reduction in wettability that may retard the proper penetration and curing of water-based adhesives such as PVAc adhesive. The formation of micro-cracks and checks due to the heat treatment at temperatures above 180 °C might also contribute to a declined shear strength of heat-treated wood.

Fig. 3 The shear strength of the bonded spruce specimens formed from spruce slabs exposed firstly to temperatures from 100 °C to 220 °C.

The failures in the bonded specimens created during the shear tests were located mainly in the wood adherent thermally modified at the highest temperatures of 200 °C and 220 °C – cohesive type of failures (Table 2, Figure 4). It means that the adhesion of the used adhesive to the thermally weakened wood was higher than the internal cohesion strength of the thermally damaged wood.

69

PVAC

100/100

100/160

100/180

100/200

100/220

100/100

160/160

180/180

200/200

220/220

PUR

100/100

100/160

100/180

100/200

100/220

100/100

160/160

180/180

200/200

220/220

Fig. 4. Failure modes of bonded spruce specimens at the shear test by EN 205:2016

CONCLUSIONS •

The shear strength of the single-lap joints, valued in the dry state of the reference bonded specimens prepared from the unmodified spruce slabs, was 11.2 MPa at using the polyurethane (PUR) adhesive or 9.3 MPa at using the polyvinyl acetate (PVAc) adhesive.

•

Applying the thermally modified timber, the shear strength decreased more apparently if the bonded specimens were formed only from the thermally modified

70

timber and less apparently if they were formed both from the reference and the thermally modified timber – at which the shear strength continuously reduced with increased modification temperature of timber from 160 °C to 220 °C, comparable at using both adhesives.

•

At the shear strength test the cohesive type of failure in the wood adherent occurred mainly if the spruce slabs were prepared from timber thermally modified at the highest temperatures of 200 °C and 220 °C.

REFERENCES AYDIN, I. 2004. Activation of wood surfaces for glue bonds by mechanical pre-treatment and its effects on some properties of veneer surfaces and plywood panels. In Applied Surface Science, 233(1-4), 268-274. DOI: 10.1016/j.matdes.2008.07.001.

AICHER, S., AHMAD, Z., HIRSCH, M. 2018. Bondline shear strength and wood failure of European and tropical hardwood glulams. In European Journal of Wood and Wood Products, 76(4): 1205−1222. DOI: 10.1007/s00107-018-1305-0.

ANDROMACHI, M., EKATERINI, R. 2018. Adhesive bond performance of heat-treated fir wood (Abies Borrissiregis). In Wood Research, 63(5): 909−16.

BAAR, J., BRABEC, M., SLÁVIK, R., ČERMÁK, P. 2020. Effect of hemp oil impregnation and thermal modification on European beech wood properties. In European Journal of Wood and Wood Products, 15 pp. DOI: 10.1007/s00107-020-01615-9.

BASTANI, A., ADAMOPOULOS, S., MILITZ, H. 2016a. Effect of open assembly time and equilibrium moisture content on the penetration of polyurethane adhesive into thermally modified wood. In The Journal of Adhesion, 93(7): 575−583. DOI: 10.1080/00218464.2015.1118621.

BASTANI, A., ADAMOPOULOS, S., KODDENBERG, T., MILITZ, H. 2016b. Study of adhesive bondlines in modified wood with fluorescence microscopy and X-ray micro-computed tomography. In International Journal of Adhesion and Adhesives, 68: 351−358. DOI: 10.1016/j.ijadhadh.2016.04.006.

BEKHTA, P., ORTYNSKA, G., SEDLIAČIK, J. 2014. Properties of modified phenol-formaldehyde adhesive for plywood panels manufactured from high moisture content veneer. In Drvna industrija, 65(4): 293−301. DOI: 10.5552/drind.2014.1350.

BEKHTA, P., SEDLIAČIK, J., JONES, D. 2018. Effect of short-term thermomechanical densification of wood veneers on the properties of birch plywood. In European Journal of Wood and Wood Products, 76(2): 549−562. DOI: 10.1007/s00107-017-1233-4.

BOONSTRA, M. J., RIJSDIJK, J. F., SANDER, C., KEGEL, E., TJEERDSMA, B., MILITZ, H., VAN ACKER, J., STEVENS, M. 2006. Microstructural and physical aspects of heat treated wood. Part 1. Softwoods.

In Maderas. Ciencia y Tecnología, 8(3): 193−208. DOI: 10.4067/S0718-221X2006000300006.

BUDHE, S., BANEA, M. D., GHUGAL, S., DE BARROS, S. 2020. Effects of heat treatment on the behavior of teak wood adherends bonded joints. In Applied Adhesion Science, 8(1): 1. DOI: 10.1186/s40563-020-00124-5.

CAI, Ch., ANTIKAINEN, J., LUOSTARINEN, K., MONONEN, K., HERÄJÄRVI, H. 2018. Wetting-induced changes on the surface of thermally modified Scots pine and Norway spruce wood. In Wood Science and Technology, 52: 1181–1193. DOI: 10.1007/s00226-018-1030-1.

EN 205:2016. Adhesives. Wood adhesives for non-structural applications. Determination of tensile shear strength of lap joints. European Committee for Standardization: Brussels, Belgium, 2016.

FOLLRICH, J., MÜLLER, U., GINDL, W. 2006. Effects of thermal modification on the adhesion between spruce wood (Picea abies Karst.) and a thermoplastic polymer. In Holz als Roh-und Werkstoff, 64(5): 373−376. DOI: 10.1007/s00107-006-0107-y.

FOLLRICH, J., TEISCHINGER, A., GINDL, W., MÜLLER, U. 2007. Tensile strength of softwood butt end joints. Part 1: Effect of grain angle on adhesive bond strength. In Wood Material Science & Engineering, 2(2): 83−89. DOI: 10.1080/17480270701841043.

71

HASS, P., KLÄUSLER, O., SCHLEGEL, S., NIEMZ, P. 2014. Effects of mechanical and chemical surface preparation on adhesively bonded wooden joints. In International Journal of Adhesion and Adhesives, 51: 95−102. DOI: 10.1016/j.ijadhadh.2014.02.014.

HILL, C. A. 2006. Wood modification: chemical, thermal and other processes. John Wiley & Sons., Chichester, UK. HUANG, X., KOCAEFE, D., CAO, J., BOLUK, Y., KOCAEFE, Y., PICHETTE, A. 2012. Effect of surface preparation on the wettability of heat-treated jack pine wood surface by different liquids. In Wood Products, 70: 711–717. DOI: 10.1007/s00107-012-0605-z.

HUNT, C. G., FRIHART, C. R., DUNKY, M., ROHUMAA, A. 2018. Understanding wood bonds–going beyond what meets the eye: a critical review. In Reviews of Adhesion and Adhesives, 6(4): 369−440.

DOI: 10.7569/RAA.2018.097312.

INARI, G. N., PETRISSANS, M., GERARDIN, P. 2007. Chemical reactivity of heat-treated wood. In Wood Science and Technology, 41(2): 157. DOI: 10.1007/s00226-006-0092-7.

JAKES, J. E., FRIHART, C. R., HUNT, C. G., YELLE, D. J., PLAZA, N. Z., LORENZ, L. F., CHING, D. J. 2018. Integrating multiscale studies of adhesive penetration into wood. In Forest Products Journal, 68(4): 340−348. DOI: 10.13073/FPJ-D-17-00067.

KAMKE, F. A., LEE, J. N. 2007. Adhesive penetration in wood—a review. In Wood and Fiber Science, 39(2): 205−220.

KNORZ, M., NEUHAEUSER, E., TORNO, S., VAN DE KUILEN, J. W. 2015. Influence of surface preparation methods on moisture-related performance of structural hardwood–adhesive bonds. In International Journal of Adhesion and Adhesives, 57: 40−48. DOI: 10.1016/j.ijadhadh.2014.10.003.

KONNERTH, J., KLUGE, M., SCHWEIZER, G., MILJKOVIĆ, M., GINDL-ALTMUTTER, W. 2016. Survey of selected adhesive bonding properties of nine European softwood and hardwood species. In European Journal of Wood and Wood Products, 74(6): 809−819. DOI: 10.1007/s00107-016-1087-1 KRYSTOFIAK, T., LIS, B., MUSZYNSKA, M., SOBOTA, K. 2013. Gluability of thermally modified ash wood with EPI adhesives. In Proceedings of the COST FP0904 and FP1006 International Workshop on Characterization of modified wood in relation to wood bonding and coating performance, Rogla, Slovenia.

KUČEROVÁ, V., LAGAŇA, R., HÝROŠOVÁ, T. 2019. Changes in chemical and optical properties of silver fir (Abies alba L.) wood due to thermal treatment. In Journal of Wood Science, 65(1): 21. DOI: 10.1186/s10086-019-1800-x.

KÚDELA, J., LAGAŇA, R., ANDOR, T., CSIHA, C. 2020. Variations in beech wood surface performance associated with prolonged heat treatment at 200 °C. In Acta Facultatis Xylologiae Zvolen, 62(1): 5−17. DOI: 10.17423/afx.2020.62.1.01.

MIRZAEI, G., MOHEBBY, B., EBRAHIMI, G. 2017. Glulam beam made from hydrothermally treated poplar wood with reduced moisture induced stresses. In Construction and Building Materials, 135: 386-393. DOI: 10.1016/j.conbuildmat.2016.12.178.

NOVÁK, I., POPELKA, A., ŠPITALSKÝ, Z., MIČUŠÍK, M., OMASTOVÁ, M., VALENTIN, M., SEDLIAČIK, J., JANIGOVÁ, I., KLEINOVÁ, A., ŠLOUF, M. 2015. Investigation of beech wood modified by radiofrequency discharge plasma. In Vacuum, 119: 88−94. DOI: 10.1016/j.vacuum.2015.04.038.

PETRIČ, M. 2013. Surface modification of wood: A Critical Review. In Reviews of Adhesion and Adhesives, 1(2): 216−247. DOI: 10.7569/RAA.2013.097308.

PULNGERN, T., UDTARANAKRON, T., CHANTO, K. 2020. Physical and mechanical behaviors of thermally modified rubberwood glulam beam under sustained and cyclic loading. In Wood and Fiber Science, 52(3): 298-312. DOI: 22382/wfs-2020-028.

REINPRECHT, L., VIDHOLDOVÁ, Z. 2008. Termodrevo – príprava, vlastnosti a aplikácie. Monograph, Zvolen: Technická univerzita vo Zvolene, 89 p. ISBN 978-80-228-1920-6.

REINPRECHT, L., REPÁK, M. 2019. The impact of paraffin-thermal modification of beech wood on its biological, physical and mechanical properties. In Forests, 10: 14 pp. DOI: 10.3390/f10121102.

REINPRECHT, L., TIŇO, R., ŠOMŠÁK, M. 2020. The impact of fungicides, plasma, UV-additives and weathering on the adhesion strength of acrylic and alkyd coatings to the Norway spruce wood. In Coatings, 10(11/1111), 15 pp. DOI: 10.3390/coatings10111111.

72

SANDBERG, D., KUZMAN, K. M., GAFF, M. 2018. Kompozitní výrobky na bázi dřeva - Dřevo jako kompozitní a konstrukční material. Prague: Czech University of Life Sciences in Prague (CULS), 185 p. ISBN 978-80-213-2869-3.

SERNEK, M., BOONSTRA, M., PIZZI, A., DESPRES, A., GÉRARDIN, P. 2008. Bonding performance of heat treated wood with structural adhesives. In Holz als Roh-und Werkstoff, 66(3): 173−180. DOI: 10.1007/s00107-007-0218-0. ŠMIDRIAKOVÁ, M., KOLLÁR, M. 2010. Modifikácia polyuretánových lepidiel biopolymérmi na lepenie dreva s vyšším obsahom vlhkosti. In Acta Facultatis Xylologiae Zvolen, 52(1): 75−83. ŠTEFKO, J., REINPRECHT, L. 2004. Dřevěné stavby – konstrukce, ochrana a údržba. Bratislava: Jaga group, spol. s.r.o., 207 p. ISBN 80-88905-95-8.

TIRALOVÁ, Z., MAMOŇOVÁ, M. 2005. Aktivita celulózovornej huby Gloeophyllum trabeum na termicky upravenom dreve - mikroskopická analýza. In Drevoznehodnocujúce huby 2005, Zvolen: Technická univerzita vo Zvolene, p. 65−68. ISBN 80-228-1535-7.

TRAN, A., MAYR, M., KONNERTH, J., GINDL-ALTMUTTER, W. 2020. Adhesive strength and micromechanics of wood bonded at low temperature. In International Journal of Adhesion and Adhesives, 103: 102697. DOI: 10.1016/j.ijadhadh.2020.102697.

TAGHIYARI, H. R., ESMAILPOUR, A., ADAMOPOULOS, S., ZERESHKI, K., HOSSEINPOURPIA, R. 2020.

Shear strength of heat-treated solid wood bonded with polyvinyl-acetate reinforced by nanowollastonite. In Wood Research, 65(2): 183−194.

UZUN, O., PERCIN, O., ALTINOK, M., KURELI, I. 2016. Bonding strength of some adhesives in heattreated hornbeam (Carpinus betulus L.) wood used for interior and exterior decoration. In BioResources, 11(3): 7686-7696. DOI: 10.15376/biores.11.3.7686-7696.

VIDHOLDOVÁ, Z., SANDAK, A., SANDAK, J. 2019. Assessment of the chemical change in heat treated pine wood by near infrared spectroscopy. In Acta Facultatis Xylologiae Zvolen, 61(1): 31−42. DOI: 10.17423/afx.2019.61.1.03.

WANG, W., ZHU, Y., CAO, J., SUN, W. 2015. Correlation between dynamic wetting behavior and chemical components of thermally modified wood. In Applied Surface Science, 324: 332–338. DOI: 10.1016/j.apsusc.2014.10.139.

ACKNOWLEDGMENTS This work was supported by the Scientific Grant Agency of the Ministry of Education of Slovak Republic Grant No. VEGA 1/0729/18 and by the Slovak Research and Development Agency under the contract no. APVV-17-0583.

ADRESSES OF AUTHORS Zuzana Vidholdová Dávid Ciglian Ladislav Reinprecht Technical University in Zvolen Faculty of Wood Sciences and Technology Department of Wood Technologies T. G. Masaryka 24 960 01 Zvolen Slovak Republic zuzana.vidholdova@tuzvo.sk davidciglian123@gmail.com reinprecht@tuzvo.sk

73

74

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 75−83, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.07

COMPARISON OF THE FRACTURE TOUGHNESS OF PINE WOOD DETERMINED ON THE BASIS OF ORTHOGONAL LINEAR CUTTING AND FRAME SAWING Daniel Chuchała – Kazimierz A. Orłowski – Gerhard Sinn – Aleksandra Konopka ABSTRACT In this paper, the values of the fracture toughness of Scots pine determined by cutting tests are presented. The cutting tests were carried out using the samples of Scotch pine (Pinus sylvestris L.) from Pomeranian Region, Poland. These experiments were carried out on two research stands: orthogonal linear cutting tests were conducted using the microtome instrument and the frame saw PRW-15M was used for sawing tests. The values of the fracture toughness were determined following the recorded values of cutting power during the cutting tests (PRW-15M) and cutting forces (microtome instrument) with the use of models based on the elements of fracture mechanics. It was observed that the fracture toughness values determined following the orthogonal linear cutting tests were significantly lower, what could be caused by bending the wood fibers under the pressure of the cutting force.

Key words: fracture toughness, orthogonal cutting, cutting force, cutting power, frame sawing, pine wood

INTRODUCTION The Atkins model (2003, 2005, 2009) of cutting forces (cutting power) takes into account the geometry of cutting tool, the friction conditions in the cutting zone and mechanical properties (fracture toughness R and the shear yield stress in the cutting zone τ) of the raw material of the workpiece. This model was applied to determine the mechanical properties of wood on the basis of experimental cutting tests (ORLOWSKI and ATKINS 2007, ORLOWSKI and PAŁUBICKI 2009, CHUCHALA and ORLOWSKI 2016, HLASKOVA et al. 2018, 2019, 2020). The methodology compiled by ORLOWSKI and ATKINS 2007, and further developed by Orlowski et al. (2017), ORLOWSKI and OCHRYMIUK (2017), SINN et al. (2020), was used successfully to forecast the cutting forces (cutting power) for different sawing processes of wood (CHUCHALA et al. 2020, OTTO and PARMIGIANI 2015, ORLOWSKI et al.

2020, SINN et al. 2020). The fracture toughness R is a material property representing the fracture mechanics in discussed models and defines an internal specific work required for the formation of a new surface. BEER et al. (2005) also adopted Atkins model (2003, 2005, 2009) to determine the value of fracture toughness from orthogonal cutting processes with the use of the microtome instrument. This investigation showed that microtome instrument is suitable for determination of the fracture toughness for wood composite materials like 75

particle-boards. Moreover, the study conducted by KOPECKY et al. (2014) and HLASKOVA et al. (2020) also showed that the fracture toughness values for wood-based-materials can be determined following the circular saw machining tests. On the other hand, ATANASOV and KOVATCHEV (2019) proposed a model of the cutting power for the particleboard milling process based only on the feed speed and cross-section of the cutting layer.

The fracture toughness is one of the mechanical properties of the material on which the Atkins model of cutting forces is based. Therefore, it indirectly enables optimization of the cutting process by forecasting the power and/or cutting forces based on this model. NASIR and COOL (2019, 2020 and 2021) showed that optimization of the cutting process has a large impact on the machining quality.

The goal of this study was to investigate differences of the fracture toughness values which were determined using the methods based on element of fracture mechanics in two separate cutting tests (orthogonal and semi-orthogonal) for pine wood (Pinus sylvestris L.).

The mentioned material property could be useful for the proper estimation of the cutting forces (cutting power) demand.

MATERIALS AND METHODS Materials Scots pine (Pinus sylvestris L.) species were used to prepare the samples. One log was randomly selected among others at the yard in the sawmill. The middle part (2 m length) of 4 m long log was cut into rectangular samples with dimensions W = 60 mm × H = 60 mm × L = 600 mm (width × height ×length, respectively). The prepared ten samples were dried and conditioned under laboratory conditions assuring constant air temperature of 20°C and relative humidity of 65 % by three months. The final moisture content MC was obtained at the level around 12 %. The density of the tested wood was 536.34 ±7.1 kg·m-3 for final MC 12 % (416.4 kg·m-3 under oven dry conditions). The structure of the examined pine wood was characterized by an average width of annual rings 2.12 ±0.4 mm and an average width of the late wood in annual rings 0.44 ±0.05 mm. These rectangular samples were used in sawing tests conducted using a frame saw. Lamellas with the thickness of around 5 mm resulted from the sawing process using the frame saw. Lamellae were used to prepare small samples with dimensions Ws = 5 mm × Hs = 30 mm × Ls = 50 mm. Sample dimensions were adapted to the material holder of the microtome instrument.

Frame sawing tests The dried ten samples were sawn using the frame saw PRW-15M equipped with a hybrid dynamically balanced driving system and elliptical teeth trajectory movement (WASIELEWSKI and ORLOWSKI 2002). The tests were carried out at the laboratory of the Gdańsk University of Technology (GUT). Every investigated sample was cut with feed speed set up at two levels, around 0.3 and 1.1 m/min. The exact values of the feed speed and corresponding feeds per tooth were determined on the basis of actual recorded courses of cutting power, following the works by CHUCHALA and ORLOWSKI (2016) and SINN et al. (2020). The use of electric power (active and passive) during idling and cutting cycles was continuously monitored and recorded with the power converter PP54 (LUMEL S.A., Zielona Góra, Poland). The average cutting power Pc while frame sawing was calculated as the difference between the mean total power PT and the mean idle power Pi following to CHUCHALA and ORLOWSKI (2016) and SINN et al. (2020), and expressed in equation (1):

Pc = PT − Pi

(1)

76

The average idle power Pi of the frame sawing process was determined each time before starting the regular cutting cycle. The average cutting power in a working stroke Pcw was calculated as in equation (2), following the works (ORLOWSKI and PALUBICKI 2009, CHUCHALA and ORLOWSKI 2016; SINN et al. 2020):

Pcw = 2  Pc

(2)

A specific list of frame saw settings and parameters of the applied tool is shown in Table 1.

Tab. 1 Machine tool and tools settings for frame sawing and linear cutting processes.

Parameter

Symbol

Value

Unit

nF HF m vc vf1 vf2 fz1 = h1 fz2 = h2

685 162 5 3.69 0.92 1.35 0.11 0.16

spm mm – m·s-1 m·min-1 m·min-1 mm mm

– St s L0 b tp γf αf σN

– 2 0.9 318 30 13 9 14 300

– mm mm mm mm mm ° ° MPa

vc h1 h2 h3

0.05 0.10 0.15 0.20

m·s-1 mm mm mm

– γf βf

– 15 55

– ° °

frame saw setting number of strokes of saw frame per min saw frame stroke number of saws in the gang average cutting speed

slow fast slow feed per tooth fast frame blade parameters the sharp saw blades with stellate tipped teeth overall set (kerf width) saw blade thickness free length of the saw blade blade width tooth pitch tool side rake angle tool side clearance tension stresses of saws in the gang linear cutting settings average cutting speed small uncut chip thickness mid large linear cutting blade parameters sharp knife blade made of High-Speed Steel (HSS) tool side rake angle (tool-in machine system) tool wedge angle feed speed

Orthogonal cutting tests The analyzed linear cutting process was conducted using the microtome instrument (Figure 1b) which is located at the University of Natural Resources and Life Sciences, Vienna (BOKU) laboratory. The investigated linear cutting process was performed in perpendicular direction to wood fiber (direction 90°−90° according to KIVIMAA (1952)) (Figure 1a). The same direction in relation to wood fibers was performed while sawing using the frame saw. The uncut chip thickness h was set at three levels 0.1, 0.15 and 0.2 mm. For each level of tested h, five repetitions were conducted. The linear cutting tests were conducted with sharp knife blades made of High-Speed Steel (HSS) by Leitz GmbH & Co. KG., Germany. The other detailed parameters of cutting process are shown in Table 1. The cutting forces in cutting speed vc direction were recorded during this process.

Methodology for determination of the fracture toughness from machinability tests The both recorded values, cutting power and cutting forces, correspondingly while frame 77

sawing and linear cutting processes were conducted can be described by ATKINS (2003, 2005, 2009) model proposed. This model was adopted for determined fracture toughness of wood based on sawing process by ORLOWSKI and ATKINS (2007) and based on linear cutting

by BEER et al. (2005). Equation (3) shows model described cutting power for frame sawing process: Pcw =

m  H p      St Q tp

h  vc +

m  H p  R  St Q tp

 vc

(3)

where: R – is the fracture toughness in J·m-2, τγ – is the shear yield stress (in the cutting zone) in MPa, m – is the number of saw blades in the gang, HP – is workpiece height (cutting depth) in mm, h – is the uncut chip thickness (corresponding to the feed per teeth fz for frame sawing) in mm, tp – is the tooth pitch in mm, St – is the overall set (kerf width) in mm, vc – cutting speed in m·s-1, γ – is the shear strain along the shear plane and can be calculated according to equation (4), assuming that Φc corresponds to the shear angle: =

cos f

(4)

cos( c −  f )  sin  c

where: γf – tool side rake angle (tool-in machine system).

The coefficient of friction correction Q represents an effect of friction between tool rake face and separated material, it can be calculated using equation (5): Q = 1−

sin    sin  c

(5)

cos(  −  f )  cos( c −  f )

where: βμ = tan-1μ is a friction angle (rad) directly related to the coefficient of friction μ.

Equation (6) shows the model described cutting force Fc for analysed linear cutting process: Fc =

     St Q

h+

R  Ws Q

(6)

where: Ws – is the width of sample (analogue to the kerf width in frame sawing) in mm. a)

b)

Fig. 1 Orthogonal linear cutting process: a) schema of cutting process, b) microtome instrument while cutting.

Both equations (3) and (6) can be expressed as a linear regression functions: Pcw (h) = c1  h + c0

(7) 78

Fc (h) = c1  h + c0

(8)

In those cases, c1 and c0 correspond to the slope and intercept, respectively.

An independent variable of the regression is the uncut chip thickness h. This makes it possible to determine the values of the fracture toughness R⊥ by matching the regression equation (8) with the experimental data from the cutting tests. The equations (9) and (10) express mathematical procedure for calculation values of fracture toughness R⊥ based on frame sawing tests and linear cutting tests, respectively:

R⊥ = R⊥ =

c0  t p  Q

(9)

m  H p  St  vc c0  Q

(10)

Ws

RESULTS AND DISCUSSION Obtained experimental results from the series of sawing and linear cutting performed on the pine wood samples in 90° – 90° direction to the wood grain (KIVIMMA 1952) are summarized in Figures 2 and 3. Figure 2 presents two test point groups that correspond to the mean value and standard deviations of measured cutting powers at two levels of feed speed vf (Table 1). Applied values of feeding are represented by the basic geometrical parameter of the cutting process, i.e. uncut chip thickness h. The exact values of the uncut chip thickness determined individually for each processed sample based on recorded experimental data were clustered around values of h1 = 0.11 mm and h2 = 0.16 mm. The data fitting curve (linear regression), as well as regression equation with coefficient c1 and intercept c0, is provided.

The linear regression for three test point groups is presented in Figure 3. These point groups correspond to the mean value and standard deviations of measured cutting forces at three levels of uncut chip thickness: h1 = 0.1 mm, h2 = 0.15 mm and h3 = 0.2 mm. Figure 3, similarly like in Figure 2, includes regression equation with coefficient c1 and intercept c0, which is the basis for determination of fracture toughness value according to equations (9) and (10). The both presented linear regressions are characterized by high values of determination coefficient, about r2 = 0.95. The determined average values of fracture toughness R⊥ for Scots pine are shown in Table 2, together with their standard deviations.

Fig. 2 Cutting power per one saw blade versus uncut chip thickness when sawing on frame saw pine wood.

79

Fig. 3 Cutting force versus uncut chip thickness when orthogonal linear cutting pine wood.

Tab. 2 Fracture toughness R⊥ of Scots pine (mean value and its standard deviations) determined from two different cutting tests. name of cutting process frame sawing orthogonal linear cutting

mean value 747.93 273.13

fracture toughness, R⊥ standard deviations ±327.2 ±75.24

unit J·m-2 J·m-2

Differences in the determined average fracture toughness R⊥ values from two different machining tests are noticeable. Despite large standard deviations for both values, these differences are significant. This case is very puzzling because the material properties of the same wood sample in the same direction in relation to the fibres should be the same or at least very similar. The reason for this phenomenon might be the flexible bending of wood fibres during linear cutting. The cutting process was carried out in the 90°– 90° direction, which meant that the cutting force was also the bending force of the fibres. Applied a small values of the uncut chip thickness, the fibres under the pressure of the cutting edge were partially tilt (bend) before the shearing process occured (Figure 4).

a)

b) Fig. 4 Bending of the pine wood fibers while orthogonal linear cutting in the 90° – 90° direction in relation to the fibers: a) scheme of cutting process with bending of wood fibers, b) example of a sample with bent and destroyed fibers.

80

Mentioned tilt of wood fibers might change the cross-section of the cutting layer and consequently reduce the cutting forces. As a result, the reduced values of the cutting forces cause the experimentally determined linear regression (Figure 3) to cross the axis of ordinates lower than it should and the determined values of fracture toughness might be underestimated. The early wood has lower mechanical properties (GONÇALEZ et al. 2018, BENDTSEN and SENFT 1986) and is therefore more susceptible to such deflections and its percentage share in annual growth may have a significant effect on this occurrence.

The phenomenon of the fiber tilt does not occur during frame sawing. Fibers loaded by cutting force, which also gives bending torque, do not tilt because they are supported by unloaded material on both sides of the cut kerf. The frame sawing process is called quasiorthogonal cutting process because during this process the main cutting edge works mostly.

However, minor cutting edges also take part in cutting process, but to a smaller share (ORLOWSKI 2007). The minor cutting forces do not significantly affect cutting process, but material located on both sides of kerf stabilizes process, which is most notifiable while sawing in 90° – 90° direction in related to the wood fibers.

The analysed cutting processes also differ very significantly in their cutting speeds.

The cutting speed for the frame sawing process (vc = 3.69 m·s-1) is 74 times higher than for the linear cutting process (vc = 0.05 m·s-1) (Table 1). Nevertheless, both KIVIMAA (1950, 1952) and MCKENZIE (1961) have shown in their works that cutting speed does not significantly affect the values of cutting forces (cutting power).

CONCLUSIONS The conducted research and obtained results allow the following conclusions to be drawn: 1. The value of the fracture toughness for pine wood based resulting from the orthogonal linear cutting tests in 90° – 90° direction related to fibers, are more than 2.5 times lower than values based on frame sawing tests.

2. While the orthogonal linear cutting of soft wood (e.g. pine wood) in 90° – 90° direction to fibers can cause fiber bending and result in disruption of the crosssectional dimension of the cutting layer, what directly affects values of cutting forces.

3. In order to analyze the phenomenon of wood fiber bending in more detail during the orthogonal linear cutting of softwood, it would be necessary to conduct research using a high frame rate camera and Digital Image Correlation system (DIC).

REFERENCES ATANASOV, V., KOVATCHEV, G. 2019. Determination of the cutting power during milling of woodbased materials. In Acta Facultatis Xylologiae Zvolen, 61(1): 93−101. https://doi.org/10.17423 /afx.2019.61.1.09 ATKINS, A.G. 2003. Modelling metal cutting using modern ductile fracture mechanics: quantitative explanations for some longstanding problems. In International Journal of Mechanical Sciences, 45: 373–396.

ATKINS, A.G. 2005. Toughness and cutting: a new way of simultaneously determining ductile fracture toughness and strength. In Engineering Fracture Mechanics, 72: 849–860.

81

ATKINS, A.G. 2009. The science and engineering of cutting. The mechanics and process of separating, scratching and puncturing biomaterials, metals and non-metals. Oxford: ButterworthHeinemann is an imprint of Elsevier, 2009, 413 p.

BEER, P., SINN, G., GINDL, M., TSCHEGG, S. 2005. Work of fracture and of chips formation during linear cutting of particle-board. In Journal of Materials Processing Technology, 159: 224–228.

BENDTSEN B.A., SENFT J. 1986. Mechanical and anatomical properties in individual growth rings of plantation-grown eastern cottonwood and loblolly pine. In Wood and Fiber Science. 18(1): 23−38.

CHUCHALA, D., OCHRYMIUK, T., ORLOWSKI, K.A., LACKOWSKI, M., TAUBE, P. 2020. Predicting cutting power for band sawing process of pine and beech wood dried with the use of four different methods. In BioResources, 15(1): 1844−1860. https://doi.org/10.15376/biores.15.1.1844-1860.

CHUCHAŁA, D., ORŁOWSKI, K. 2016. Shear yield stresses and fracture toughness of Scots pine (Pinus sylvestris L.) according to the raw material provenance. In Trieskove a beztrieskove obrabanie dreva, X, 49−55.

GONÇALEZ, J.C., SANTOS, N., DA SILVA, F.G.JR., SOUZA, R.S., DE PAULA, M.H. 2018. Growth ring width of Pinus Caribaea and its relationship with wood properties. In Scientia Forestalis/Forest Sciences, 46(120): 670−678. https:/doi.org/10.18671/scifor.v46n120.15.

HLÁSKOVÁ, L., ORLOWSKI, K.A., KOPECKÝ, Z., SVITÁK, M., OCHRYMIUK, T. 2018. Fracture toughness and shear yield strength determination for two selected species of central European Provenance. In BioResources, 13(3): 6171−6186. https://doi.org/10.15376/biores.13.3.6171-6186.

HLÁSKOVÁ, L., KOPECKÝ, Z., SOLAŘ, A., POTOČKA, Z. 2019. Cutting test as a source of fracture toughness and shear yield strength for axial-perpendicular model of wood cutting. In Wood and Fiber Science, 51(1): 1−11. https://doi.org/10.22382/wfs-2019-006.

HLÁSKOVÁ, L., KOPECKÝ, Z., NOVÁK, V. 2020. Influence of wood modification on cutting force, specific cutting resistance and fracture parameters during the sawing process using circular sawing machine. In European Journal of Wood and Products, (e-print). https://doi.org/10.1007/s00107-02001581-2.

KIVIMAA, E. 1950. Cutting force in woodworking. State Institute for Technical Research, Helsinki.

KIVIMAA, E. 1952. Die Schnittkraft in der Holzbearbeitung. (The cutting force in wood processing).

In Holz Roh- Werkst, 10(3): 94–108.

KOPECKÝ, Z., HLÁSKOVÁ, L., ORLOWSKI, K.A. 2014. An innovative approach to prediction energetic effects of wood cutting process with circular-saw blades. In Wood Research, 59(5): 827−834.

MCKENZIE, W.M. 1961. Fundamental analysis of the wood-cutting process. PhD thesis, University of Michigan, Department of Wood Technology, MI, USA. NASIR, V., COOL, J. 2019. Optimal power consumption and surface quality in the circular sawing process of Douglas‑fir wood. In European Journal of Wood and Wood Products, 77: 609–617. https://doi.org/10.1007/s00107-019-01412-z.

NASIR, V., COOL, J. 2020. A review on wood machining: characterization, optimization, and monitoring of the sawing process. In Wood Material Science & Engineering, 15(1): 1−16. https://doi.org/10.1080/17480272.2018.1465465.

NASIR, V., COOL, J. 2021. Cutting power and surface quality in sawing kiln-dried, green, and frozen hem-fir wood. Wood Science and Technology. https://doi.org/10.1007/s00226-020-01259-1 ORLOWSKI, K.A., ATKINS, A. 2007. Determination of the cutting power of the sawing process using both preliminary sawing data and modern fracture mechanics. In Proceedings of the Third International Symposium on Wood Machining. Fracture Mechanics and Micromechanics of Wood and Wood Composites with regard to Wood Machining, 21–23 May, Lausanne, Switzerland. Eds. Navi, P., Guidoum, A. Presses Polytechniques et Universitaires Romandes, Lausanne, 2007, 171– 174.

ORŁOWSKI, K. 2007. Experimental studies on specific cutting resistance while cutting with narrowkerf saws. In Advances in Manufacturing Science and Technology, 31(1): 49–63.

ORŁOWSKI, K.A., PAŁUBICKI B. 2009. Recent progress in research on the cutting process of wood.

A review COST Action E35 2004–2008: Wood machining – micromechanics and fracture. In Holzforschung, 63:181–185.

82

ORLOWSKI, K.A., OCHRYMIUK, T., SANDAK, J., SANDAK, A. 2017. Estimation of fracture toughness and shear yield stress of orthotropic materials in cutting with rotating tools. In Engineering Fracture Mechanics, 178: 433−444. https://doi.org/10.1016/j.engfracmech.2017.02.023.

ORLOWSKI. K.A., OCHRYMIUK, T. 2017. A newly-developed model for predicting cutting power during wood sawing with circular saw blades. In Maderas Cienc. Tecnol. 19(2): 149–162. https://doi.org/10.4067/S0718 -221X2 01700 50000 13.

ORLOWSKI, K.A., OCHRYMIUK, T., HLASKOVA, L., CHUCHALA, D.; KOPECKY, Z. 2020. Revisiting the estimation of cutting power with different energetic methods while sawing soft and hard woods on the circular sawing machine: a Central European case. In Wood Science and Technology, 54(2): 457−477. https://doi.org/10.1007/s00226-020-01162-9.

OTTO, A., PARMIGIANI, J. 2015. Velocity, depth-of-cut, and physical effects on saw chain cutting. In BioResources, 10(4): 7273−7291. https://doi.org/10.15376/biores.10.4.7273-7291.

SINN, G., CHUCHAŁA, D., ORLOWSKI, K.A., TAUBE, P. 2020. Cutting model parameters from frame sawing of natural and impregnated Scots pine (Pinus sylvestris L.). In European Journal of Wood and Wood Products, 78(4): 777−784. https://doi.org/10.1007/s00107-020-01562-5.

WASIELEWSKI, R., ORLOWSKI, K. 2002. Hybrid dynamically balanced saw frame drive. In Holz RohWerkst 60(3): 202–206.

ACKNOWLEDGEMENTS Financial support of NAWA (Polish National Agency for Academic Exchange) within the project PPN/BIL/2018/1/00100/U/00001 and BMBWF trough OEAD within the project PL 06/2019 is gratefully acknowledged.

The authors gratefully acknowledge the Ministry of Science and Higher Education, Poland, for support the maintenance of scientific and research equipment – PRW-15M frame saw, grant number 21/E-359/SPUB/SP/2019.

AUTHORS ADDRESSES Kazimierz Orlowski (ORCID id: 0000-0003-1998-521X) Daniel Chuchala (ORCID id: 0000-0001-6368-6810) Aleksandra Konopka (ORCID id: 0000-0003-0733-249X) Gdansk University of Technology Faculty of Mechanical Engineering Narutowicza 11/12 80-233 Gdansk Poland Gerhard Sinn (ORCID id: 0000-0002-8365-3650) University of Natural Resources and Life Sciences Institute of Physics and Materials Science Peter‑Jordan‑Str. 82 1190 Vienna Austria

83

84

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 85−92, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.08

DETERMINATION OF VIBRATION DURING LONGITUDINAL MILLING OF WOOD-BASED MATERIALS Georgi Kovatchev − Valentin Atanasov ABSTRACT The study of vibrations during the operation of the cutting mechanism in a woodworking shaper is presented in the paper. Wood-based details used in the furniture manufacturing like plywood and medium-density fibreboard (MDF) were milled. The experiments were performed using the universal milling machine with lower spindle position FD-3 located in a laboratory at the Department of Woodworking Machines, University of Forestry – Sofia.Vibration velocity was measured using a specialized device model Bruel & Kjaer Vibrotest 60 at 6000 min-1 rotation frequency. The exact vibration state measurements were made in two mutually perpendicular directions (AxandAy). On the basis of the experiment, regression equations were developed. Determination of the influence of feed speed U (from 2m·min-1to 10m·min-1) and milling area А (from 48 mm2 to 144 mm2) at the vibration velocity V mm·s-1 (r·m·s) was performed by conducting planned two-factor regression analysis. The vibration velocity values at the point Ax were from 1.85 mm·s-1 to 2.76 mm·s-1, and at the point Ay were from 1.75 mm·s-1 to 2.97 mm·s-1. The measured roughness Rz of the test specimens was from 20 μm to 157 μm. The investigation results can be used as a base for making some recommendations concerning an increase in reliability of the wood shapers as well as the accuracy and quality of their production.

Key words: MDF, plywood, milling, cutting mechanism, vibrations.

INTRODUCTION Wood shapers are widespread in practice. Their universality allows them to be used in diverse industries in the woodworking and furniture manufacturing industry. Their main use is in the manufacture of furniture, windows, doors, construction products and many other items used in everyday life. Wood shapers allow different devices to be attached to them, Their technological capabilities are increased this way. Contemporary wood shapers should be able to work at different cutting speed. Most often the speed ranges between 30–60 m/s (GOCHEV 2005, OBRESHKOV 1997). This inevitably is associated with the machinery resources to work at different rotational speed. They are a precondition for the emergence of different cutting forces that create conditions for loads in the mechanisms which lead to errors during operation (VUKOV et al. 2012). Dynamic effects are constantly changing, which is a premise for permanent shifting loads in the bearings.

Not just the milling machines, but the rest of the woodworking machines operate at high cutting speeds. This is certainly a precondition for an increase in overall vibration throughout the mechanical system. Important attention should be paid to the preparation of 85

the cutting tools. The use of cutting tools exceeding the required imbalance is also a reason to increase the vibration intensity of the particular mechanical system (GOCHEV et al. 2017, ORLOWSKI et al. 2007, VITCHEVet al. 2020). Also, increased levels of mechanical oscillations have an influence on the production noise and the quality of finished products (BREZIN et al. 2015, ROUSEK et al. 2010, VITCHEV et al. 2019). Clearing the wood cutting tool after work and correct installation are factors greatly affecting the ability of each cutting tool to process the wood quality. The choice of the technological operation mode depends on the processed tree species and wood-based materials, type and preparation of the cutting tool, the complexity of machining, the quality of the milling, etc. This set of factors during cutting creates forced oscillations throughout the mechanical system.These vibrations directly affect the quality of each milling wood detail.

The aim of this work is the measurement and analysis of vibration velocity at working tread by universal wood shaper with bottom location of the working shaft. The load on the bearings is researched while milling different wood-based materials and changing some technological factors. The study is aimed at improving the reliability and efficiency of a wood shaper machine to ensure the accuracy and quality of products.

MATERIAL AND METHODS To conduct the experiment, a universal wood shaper with bottom location of the working shaft was selected Fig.1. The cutting mechanism of the selected machine is of relatively simple construction, which helps the more accurate execution of the experiment.

The cutting mechanism is driven by an asynchronous electric motor at 3 kW of power and rotation frequency of 2880 min-1.The rotation frequency used for the experiments was 6000 min-1. This is one of the most commonly used frequency in milling machines.The selected rotational speed wass performed by pulleys mounted on the electric motor shaft and the machine shaft.

Fig. 1 Wood shaper general view.

Fig. 2 Groove cutter D=140 mm.

A cutter with a diameter D = 140 mm was used Fig 2. The technical data of the cutting tool are shown in Table 1. The inscriptions in the table are: D - diameter of the milling cutter,d – diameter of the bore, B – milling width, α – back angle of cutting,  – angle of sharpening,  – front angle of cutting, z – number of teeth.

Tab. 1 Technical data of the cutting tool.

Type of instrument Groovecutter

D mm 140

d mm 30

α  16

B mm 12

86

  55

  19

z бр 6

Material of the teeth HM

The cutting speed was calculated by the formula 1 (VLASEV 2007). At a rotation frequency of 6000 min-1/(100 s-1), the calculated cutting speed was v = 44 m·s−1.

V = π·D·n, m·s-1,

(1)

D – diameter of the cutting tool, m; n –rotation frequency of the cutting tool, s-1.

The research in this paper was conducted at a calculated speed.The experimental part included milling of MDF and plywood samples. Some of the samples can be seen in Fig. 3.

The paper examines the influence of some important factors in the cutting process on the impact on the vibration velocity measured in the bearing housings is examined in the paper.

Table 2 shows the studied factors and their levels in open and coded form.

Fig. 3 MDF and Plywood test samples.

Tab. 2 Survey factors.

Factors

Open 2 48

Feed speed U, [m.min-1] Milling area А, [mm2]

Cod. −1 −1

Factor levels Open Cod. 6 0 96 0

Open 10 144

Cod. 1 1

Determination of the influence of feed speed (U, m.min-1) and milling area (А, mm2) on vibration velocity V mm.s-1 (r.m.s) was performed by conducting a planned two-factor regression analysis. Table 3 shows the experimental matrix (VUCHKOV et al.1986). The feed speed is indicated by X1 and milling area X2. The results were calculated by the software products QstatLab5 and Microsoft Excel.

Tab. 3 Experimental matrix. № 1 2 3 4 5 6 7 8 9

U, m.min-1 10 10 2 2 6 6 10 6 2

X1 1 1 -1 -1 0 0 1 0 -1

87

A, mm2 144 48 144 48 96 144 96 48 96

X2 1 -1 1 -1 0 1 0 -1 0

The intensity of the vibrations was assessed on the basis of the root mean square value of the vibration velocity V mm∙s-1 (r∙m∙s) measured at the different working modes of the machine. The measurements were performed at two measuring points located on the upper bearing housings of the main shaft of the machine. In the present work, vibration measurement in the lower bearing housing was not done. In the previous studies conducted by the authors, it was found that an increase in oscillations in the lower bearing housing was very small during work (KOVATCHEV 2018). The measurement points on each bearing housing are located mutually perpendicular, radial to the main shaft of the machine Fig. 4. (БДСISO 10816 – 1:2002).

Fig. 4 Measurement points.

Vibration speed was measured using a specialized device model Bruel & Kjaer Vibrotest 60 shown in Fig. 5. The measurement points are located on the bearing housing of the machine. It significantly responds to the dynamic state. The exact vibration state measurements needed to be made in two mutually perpendicular directions Fig. 6.

Fig. 6 Measuring sensor.

Fig. 5 Bruel & Kjaer Vibrotest 60.

The roughness parameter (Rz, μm) was measured on each detail after milling.The Rz parameter was determined using the mean average value from the five measurements in accordance to БДС EN ISO 4287:2006.The roughness was measured using a specialized surface roughness tester Mitutoyo Surftest SJ-210shown in Fig.7.

88

Fig. 7 Mitutoyo Surftest SJ-210.

RESULTS AND DISCUSSION The work trials in milling of MDF and plywood were included in the experiment. The intensity of the vibrations is assessed on the basis of the root mean square value of the vibration velocity V mm·s−1 (r·m·s) measured at the different feed speed (U, m.min-1) and milling area (А, mm2). The measurement points are indicated by A: Ax– in direction parallel to the feed direction, Ay – in direction perpendicular to the feed direction.

The regression equations 2 and 3 show the influence of factors at the milling of MDF and plywood in the point Ax. АхMDF = 1.916+0.026x1+0.057x2−0.005x1x1+0.009x2x2+0.008x1x2

(2)

АхPlywood = 2.109+0.067x1+0.156x2−0.053x1x1−0.069x2x2+ 0.003x1x2

(3)

x1 – feed speed, coded; x2 – milling area, coded; As it can be observed in the regression equations obtained, the strongest influence on the vibration velocity V mm.s-1 (r.m.s) in the longitudinal milling of MDF and plywood is the first and the most important factor in the milling area (А, mm2). The feed speed of the processed material (U, m.min-1) is the second most important factor. Graphical, the results at point Ax can be seen in Fig.8 and Fig. 9.

Fig. 8 Vibration speed measured at point Ax in the milling of MDF components.

Fig. 9 Vibration speed measured at point Ax in the milling of Plywood components.

Figure 8 shows that by increasing the feed rate of the processed material, the vibration velocity increases. This tendency was observed in all three milling areas. Similar results were observed by the authors (VITCHEVet al.2020). Figure 9 shows the same trend as in Fig.8.

89

The regression equations 4 and 5 show the influence of factors in the milling of MDF and plywood in point Ay. АyMDF = 2.695+0.049x1+0.053x2−0.016x1x1−0.017x2x2+0.004x1x2

(4)

АyPlywood = 2.663+0.033x1+0.093x2+0.095x1x1+0.101x2x2−0.011x1x2 (5) Here, as in point Ax, the strongest influence on the vibration velocity V mm·s−1 (r.m.s) in the longitudinal milling of MDF and plywood is the factor of the milling area (А, mm2).

The feed speed of the processed material (U, m·min-1) is the second most important factor.

Graphically, the results at point Ay can be seen in Fig. 10 and Fig.11.

Fig. 10 Vibration speed measured at the point Ay in the milling of MDF components.

Fig. 11 Vibration speed measured at the point Ay in the milling of Plywood components.

Figure 9 shows us again that by increasing the feed rate of the processed material, the vibration velocity increases. This tendency was observed in all three milling areas. Figure 10 shows that as the feed speed of the treated material increases, the vibration velocity initially decreases and then rises again. The local minimum vibration velocity at the point Ay can be explained by the diffraction of the cutting force in a different direction.At that time the magnitude of the cutting force is not high enough to increase the vibration velocity.

After the local minimum, the vibration velocity is rising again, which is a sign that the cutting force is concentrating in the Ay direction. The same trend was observed by the authors in some of their previous studies as well(KOVATCHEV et al. 2018).

At the end of the experiment, the roughness of a part of the test samples was measured.

The samples of MDF and plywood were selected. Graphically, the results can be seen in Fig. 12 and Fig. 13.

Fig. 12 Surface roughness (Rz) depending on the feed speed in the milling of MDF components.

90

Fig. 13 Surface roughness (Rz) depending on the feed speed in the milling of Plywood components.

CONCLUSION On the basis of the conducted experimental studies, the following more important conclusions and recommendations can be drawn: Тhe strongest influence on the vibration velocity V mm·s-1 (r·m·s), at the measured points Ax and Ay was the first and the most important factor in the milling area (А, mm2).

The feed speed of the material (U, m·min-1) was the second most important factor.

When the feed rate increases, the surface roughness increases. This trend was maintained for all milling areas and for all MDF and plywood samples. A similar trend was also observed by the authors in their publications: Higher feeding speeds of processed material lead to rougher surface (KMINIAK et al. 2016, KMINIAK et al. 2017, SEDLECKY et al.

2018, SIKLIENKA et al. 2016, VITCHEV et al. 2018, VITCHEV 2019).

REFERENCES BREZIN,V., ANTOV, P.2015. Engineering ecology, Sofia, ISBN 978-954-332-135-3, 259p.(in Bulgarian).

GOCHEV, Z. 2005. Manual for Wood Cutting and Woodworking Tools. Sofia, 232 p.(in Bulgarian).

GOCHEV, Z., VUKOV, G., VITCHEV, P., ATANASOV, V., KOVATCHEV, G. 2017. Study On The Vibration Severity Generated By Woodworking Spindle Moulder Oachine. In International Scientific and Technical Conference Wood Technology and Product Design. Skopje: Ss. Cyril and Methodius University of Skopje, pp. 55–60.

KMINIAK, R., SIKLIENKA, M., SUSTEK,J. 2016. Impact of tool wear on the quality on the quality of the surface in routing of MDF boards by milling with reversible blades. In Acta Facultatis Xylologiae Zvolen, 58(2): 89−100, DOI: 10.17423/afx.2016.58.2.10.

KMINIAK, R., BANSKI, A., CHAKHOV,DK. 2017. Influence of the thickness of removed layer on the quality of created surface during milling the MDF on CNC machining centers. In Acta Facultatis Xylologiae Zvolen, 59(2): 137−146,DOI: 10.17423/afx.2017.59.2.13.

KOVATCHEV, G.,ATANASOV, V.2018.Determination of vibration during milling process of some deciduous wood species. In8thInternational Science Conference Hardwood Conference, Sopron, ISBN 978-963-359-095-9, ISSN 2631-004X, pp. 112–113.

KOVATCHEV, G. 2018. Influence Of The Belt Type Over Vibration Of The cutting Mechanism In Woodworking Shaper.In 11thInternationalScience Conference Chip and Chipless Woodworking Processes.Zvolen: Technical University in Zvolen,pp. 105–110.

OBRESHKOV P. 1997. Woodworking Machines, Sofia. 182 p.(in Bulgarian) ORLOWSKI K., SANDAK, J., TANAKA, C. 2007.The critical rotational speed of circular saw: Simple measurement method and its practical implementations. In Journal of Wood Science 53(5): 388−393.

ROUSEK, M., KOPECKY, Z., SVATOS,M. 2010. Problems of the quality of wood machining by milling stressing the effect of parameters of machining on the kind of wood. In Annals of Warsaw University of Life Sciences – SGGW №72: 233-242, ISSN: 1898-5912.

SEDLECKY, M., KVIETKOVA, M., KMINIAK,R., KAPLAN,L. 2018. Medium-density fibreboard and edge-glued panel after milling-surface waviness after machining with different parameters measured by contact and contactless method. In Wood research 63(4), 683-697.

SIKLIENKA, M., JANDA, P., JANKECH,A. 2016. The influence of milling heads on the quality of created surface. In Acta Facultatis Xylologiae Zvolen, 58(2): 81−88, DOI: 10.17423/afx.2016.58.2.10.

VITCHEV, P.2019.Evaluation of the surface quality of the processed wood material depending on the construction of the wood milling tool.In Acta Facultatis Xylologiae Zvolen, 61(2): 81−90, 2019, DOI: 10.17423/afx.2019.61.2.08.

91

VITCHEV, P., ANGELSKI, D.,MIHAILOV, V. 2019.Influence of The Processed Material on the Sound Pressure Level Generated by Sliding Table Circular Saw.In Acta Facultatis Xylologiae Zvolen, 61(2): 73−80, 2019, DOI: 10.17423/afx.2019.61.2.07.

VITCHEV, P., GOCHEV, Z., VUKOV, G. 2020. The influence of some factors on the vibrations generated by woodworking spindle moulder machine when processingspecimens from beech wood.In Acta Facultatis Xylologiae Zvolen, 62(2): 97−107, 2020, DOI: 10.17423/afx.2020.62.2.09.

VITCHEV, P., GOCHEV, Z.2018. Study on quality of milling surfaces depending on the parameters of technological process. In 29th ICWST International Conference on Wood Science and Technology, Zagreb, pp 193−199.

VLASEV, V. 2007. Exercise And Drilling Manual On Woodworking Machines. Sofia, 78 p. (in Bulgarian) VUCHKOV, I., STOYANOV, S. 1986. Mathematical Modeling And Optimization Of Technological Objects. Sofia 341 p. (in Bulgarian) VUKOV, G., GOCHEV, Z., SLAVOV, V. 2012, Torsional Vibrations In The Saw Unit of A Kind Of Circular Saws. NumericalInvestigationsof the NaturalFrequencies and Mode Shapes. In Proceedings Of Papers, 8thInternationalScience Conference Chip and Chipless Woodworking Processes.Zvolen: Technical University in Zvolen,ISBN 978-80-228-2385-2, pp. 371–378. БДС ISO10816-1:2002, Evaluation Of Machine Vibration By Measurement On Non – Rotating Parts – Part 1: General Guidelines, 25 p. БДС EN ISO 4287:2006, Surface texture: Profile method - Terms, definitions and surface texture parameters, 35p.

AUTHORS’ ADDRESSES Chief Assist. Prof. Georgi Kovatchev, PhD. Chief Assist. Prof. Valentin Atanasov, PhD. University of Forestry Faculty of Forest Industry Kliment Ohridski Blvd. №10, 1797 Sofia Bulgaria g.kovatchev@gmail.com vatanasov_2000@ltu.bg

92

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 93−101, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.09

PRODUCTION OF WOODEN HOLLOW SPIRAL BALUSTERS ON TURN-MILLING CNC MACHINES Stanislav Korchagin – Mikhail Chernykh – Vladimir Štollmann ABSTRACT The esthetic perception of a hollow baluster with spiral grooves, the influence of the spiral line elevation angle value and number of passes on perception are presented in the paper. The variants of balusters implementation depending on the direction of the mill and workpiece feed during machining are demonstrated. The formulas characterizing the crosssection geometry with odd and even amounts of grooves are given.

It is shown that the elevation angle of, approximately, 45˚ provides the most favorable esthetic perception. The production of balusters from wood and medium density fiberboard on a turn-milling CNC machine is examined. The recommendations on the groove milling sequence are proposed, the values of cutting depth and mill feed for certain typical sizes of balusters from birch and beech are recommended.

Key words: wood, baluster, milling, groove, esthetic perception.

INTRODUCTION Barrier railings with balusters are popular in interiors and exteriors of buildings.

Balusters are distinguished by the material type – from wood, metal, stone, concrete, etc.; by shape – rotation bodies, flat, sculptural; by implementation style prevailing in a certain time period − gothic, baroque, renaissance, modern, art-deco, etc. Depending on the material type, several techniques are used to produce balusters – turning, milling, forging, casting, manual engraving.

The production of balusters from wood, esthetic and environmentally friendly material, has certain peculiarities. Universal lathes allow forming decorative elements in the form of rotation surfaces on balusters (PILIKINA 2005), (SIKLIENKA et al. 2016). The use of turn-milling machines gives additional opportunities for decorating balusters due to the formation of longitudinal and spiral grooves (channelures) on them. And the availability of computerized numerical control (CNC) in turn-milling machines (Figure 1) provides the identity of baluster sizes, formation of relief patterns on rotation surfaces and improved productivity (CHERNYKH et al. 2018, VITCHEV 2019).

93

Fig. 1 Turn-milling machine “Robor-D” (Semil, Russia) with computerized numerical control.

When processing on turn-milling machines, the mill rotation is the main motion, its movement in coordinates X, Z and workpiece rotation are the feed motion (Figure 2, а). The look of channelures is defined by the availability or absence of one or another feed movement (Figure 2, b).

a)

b)

Fig. 2 Scheme of processing workpieces on turn-milling machines (a) and implementation of balusters produced with different directions of the mill and workpiece continuous feed (b).

The paper investigates the production possibilities of the most decorative and, at the same time, the most complicated by design and production technology, hollow balusters with spiral channelures on turn-milling CNC machines, their geometry and esthetic perception.

94

MATERIALS AND METHODS Geometry of hollow baluster cross-section If the milling depth h is not less than R (half of the external diameter D of the baluster), the opposite channelures link up inside the baluster (Figure 3) forming the longitudinal cavity, which produces the effect of decorative grid and improves the esthetic expressiveness of the baluster and barrier railings in general. The obtained cavity can be also used for placing lighting elements providing a new decorative effect for the barrier railings.

а)

b)

Fig. 3 Cross-section of the hollow baluster with the odd number of channelures with the different milling depth h: equaled to (а) and more than (b) the half of external diameter D of the baluster.

The diameter of circumference dв inserted into the cavity is defined by the diameter of mill dф and angle α between the axes of adjoining channelures (Figure 3, а). With the odd number of channelures the cavity can be extended due to cutting off the apexes of the opposite load-carrying (supporting) elements of the baluster (Figure 3, b). There is no such possibility with the even number of channelures.

The main ratios of sizes of cross-section of hollow balusters are given in Table 1.

Tab. 1 Main ratios of sizes of the cross-section of hollow balusters.

Size

Angle between the axes of the adjoining channelures α, degrees Diameter of the circumference inserted into the cavity 𝑑в Cross-section area of one loadcarrying element S

With the even number of channelures, as well as with their odd number and uncut apexes of opposite load-carrying elements 𝛼= 𝑑в =

𝑑ф sin

𝜋𝐷 2

𝛼 2

360° 𝑁

, where N – number of passes

, where 𝑑ф – mill diameter 𝐷𝑑ф

2 𝑑ф

𝛼

1

2

sin

𝑑в =𝑑ф (𝑐𝑡𝑔 +

𝛼

𝑆 = 4𝑁 − 2 + 4 c𝑡𝑔 2 , where D – baluster external diameter

95

With the odd number of channelures and cut off apexes of opposite load-carrying elements

𝜋𝐷 2

S= 4𝑁 −

𝐷𝑑ф 2

𝛼 2

)

Assessment of esthetic perception The sizes of baluster elements should be selected due to their estheticism and functionality, as well as the milling mode. When describing the parameters of spiral balusters, the terms characterizing the spiral line are used for the terminology unification – elevation angle φ and number of passes N, which equals the number of both the channelures and load-carrying elements of the baluster.

The influence of elevation angle φ and number of passes N onto the esthetic perception of balusters was investigated on 3D models of balusters by the method of expert assessment (GUTSYKOVA 2020) to decrease the investigation costs. The method of expert assessment is applied when it is difficult or impossible to measure the studied parameter instrumentally. This is appropriate for aesthetic perception of design objects, balusters, in particular. The method involves information processing by a complex expert group. When forming the group, the collective assessment of each expert’s competence was used. The collective assessment is used for experts in aesthetic areas. The experts had backgrounds in design, experience in artistic works and information of each other as specialists. The three-point competence assessment system, with 1 – as the highest level, 0.5 – average and 0 – low, was applied. The coefficient of each expert’s competence level exceeded 0.5 being appropriate for the method requirements. The assessment was carried out by 20 experts on a scale from 1 to 5. Thus, the maximum possible values of the total points for each value of the elevation angle, as well as the number of passes, were 100.

The elevation angle changed from 15 to 75 degrees in increments of 15 degrees, and the number of passes – from 1 to 9. The models had the similar length to diameter ratios of 5:1. 45 models were studied.

For the convenience of the experts’ work, the background and lighting for all models were the same, and the view plane was placed vertically in accordance with the actual location of the baluster in the barrier railings.

Samples, equipment and tools Based on the experts’ assessment results, the balusters were worked out and produced (Figure 4) from wood of birch, beech and pine with 12% humidity and from the medium density fiberboard (MDF). The ultimate strength of the materials was found by GOST 16483.5-73 on three samples for each material. The baluster diameter was 70 mm, length – 690 mm, elevation angle – 45 degrees, width of channelures – 12 mm, width of load-carrying elements (size “a” in Figure 3, а) – 20 mm, thickness of load-carrying elements with the cutoff apex (size “t” in Figure 3, b) – 50 mm. The number of passes was taken as 5.

a)

b)

c)

Fig. 4 – Balusters produced from wood of different species: а – birch, b – beech, c – pine.

96

The workpieces for balusters were produced from edged boards. The boards were sawed by disc mills into lumps on sawing machines. The lumps were planed from four sides on the thickness planer and glued with wood glue by GOST 18992-80 on hydraulic press “Elbrus” 2G 3000 (Russia). The wood fibers were placed along the baluster axis. The balusters were processed on turn-milling CNC machine “Robor-D” (Figure 1) produced by Semil (Russia). The machine characteristics: power – 3.2 kW, overall dimensions – 2,500 × 1,100 × 2,100 mm, workpiece continuous rotation – 0…3,200 rpm, workpiece programmable rotation – 0…10 rpm, mill rotation – up to 20,000 rpm, range of operating feeds – up to 3.5 m/min, mill movement accuracy – ±0.15 mm, discreteness of mill movement set up – 0.1 mm, maximum diameter of the workpiece machined – 250 mm, maximum length of the workpiece machined – 1,400 mm.

The machining was performed by new sharp mills. The cylindrical surface was machined by the straight double-helical mill 12 mm in diameter at 16,000 rpm of the mill and 2,000 rpm of the workpiece, and mill feed of 4 mm/sec. The profiled surfaces of rotation were machined by the spiral mill with a spherical head 8 mm in diameter in the range of 9,000…20,000 rpm of the mill and 650…1,300 rpm of the workpiece. The channelures were machined by the quadruple-helical mill from the tool steel 12 mm in diameter with chip splitter – toothed-like cutting edges on the side surface – in the range of 12,000…20,000 rpm of the mill and its feed – from 0.5 up to 3 mm/sec. The workpiece turning speed during the grooves machining was defined by the mill feed and elevation angle φ taken for one pass, the allowance was from 0.5 up to 2 mm.

RESULTS AND DISCUSSION The ultimate strength of the materials investigated is given in Table 2. The data obtained comply with the known results (UGOLEV, 2004).

Tab. 2 Average values of the ultimate strength based on the results of shear tests along the fibers, MPa. Material Birch 5.95

Beech 7.18

Pine 2.79

MDF 1.33

The total of expert assessments of esthetic perception of the models (total scores) is given in Table 3 and Figure 5.

Tab. 3 Total expert assessments of the balusters esthetics, scores.

Number of the model group

φ [degrees]

1 2 3 4 5

15 30 45 60 75

Number of passes, N 1

2

3

4

5

6

8

9

27 47 80 62 54

31 74 87 62 46

43 71 70 51 35

55 78 55 40 35

61 64 54 39 25

51 74 51 44 34

46 61 66 53 50

38 49 57 59 53

From the graphs it is seen that regardless of the number of passes the overall tendency of changing the esthetic perception with the change in the elevation angle φ is observed.

With the increased angle value, the esthetic perception, at first, increases reaching the

97

Total points

maximum, then it decreases. From the point of esthetic perception, the elevation angle φ equaled to, approximately, 45° is the most favorable.

The number of passes does not significantly influence the esthetic value of the baluster (Figure 6). It can be only pointed out that with the increase of N the range of assessment spread is narrowed, since the difference in the number of passes becomes less visible.

ϕ [degrees]

Total points

Fig. 5 Dependence of esthetic perception of the spiral baluster from elevation angle φ for models with different number of passes (from 1 to 9).

Number of passes N Fig. 6 – Influence of the number of passes onto esthetic perception of the spiral baluster with elevation angles changing in the range of 15º…75º With the decreased size of the cross-section of the load-carrying elements of the balusters, its decorativeness increases. However, the minimum values of sizes are limited, moreover, rather by the load emerging in the process of grooves milling than the functional load onto the baluster in the barrier railings, where the load is distributed between several balusters and in the baluster – uniformly between all load-carrying elements. The cutting 98

forces act mainly on those two load-carrying elements of the baluster, which contact the mill.

The elements deform under the action of the cutting forces and the baluster bends producing vibrations. Due to the fibrous composition of wood, the high values of the cutting mode and machining productivity are limited, first of all, by the material ultimate shearing strength along the fibers. With the excessive cutting depth and feed values the load-carrying elements can chop off along the fibers (Figure 4, b). The theoretical calculation of the cutting mode elements – cutting depth, cutting speed and feed – due to the baluster design complexity and variation of the material strength properties is a complicated task and is not considered in this paper. The experimental investigation allowed finding out the following.

When selecting the material from the investigated ones, the birch and beech should be preferred. The balusters from MDF were destructed when milling the grooves even at the small cutting depth (0.3 mm) due to the low material ultimate strength. The balusters from pine were also destructed at the cutting depth of 0.4 mm due to the pine inclination to chopping and insufficient ultimate strength. Positive results were achieved on the workpieces from birch and beech.

To decrease the possibility of destructing the baluster load-carrying elements by more uniform distribution of cutting forces along the baluster cross-section, the groove should not be milled to the whole depth at once. First, the groove should be deepened to one value turning the workpiece step-by-step to angle α, then to another one, then to the third one, etc. until we get a through hole.

Due to the baluster decreased rigidity when the depth of the grooves increases while milling, the work rest should be used, and the cutting depth, i.e. the material layer thickness removed during one pass, should be decreased with the groove depth increase. At the same time, the cutting is performed both with the feed in straight direction, i.e. from left to right, and in reverse direction, i.e. from right to left (Figure 1).

For the investigated typical size of the balusters from birch and beech the following values of the cutting modes are rational. The cutting depth, i.e. the wood layer thickens removed by the mill at the feed in one direction (both in straight and reverse) equals 1.0 mm at the beginning of the groove formation and 0.5 mm – in the end; the feed equals 1 mm/seс.

The values less than the indicated ones result in decreased productivity, the greater ones – to decreased machining quality, the baluster vibration increase and increased destruction possibility. The number of mill revolutions was 19,000 rpm and was limited by the machine characteristics. Operation at the mode ultimate for this machine (20,000 rpm) will result in accelerated wear. To improve the milling quality, as indicated (DZURENDA, 2008), it is practicable to use high-speed machines.

CONCLUSION The sizes of the baluster elements should be selected due to its estheticism and functionality, as well as the milling mode.

The cross-section geometry analysis allowed finding the dependence of the cavity diameter of hollow balusters with odd and even amounts of channelures on their number and mill diameter. It is demonstrated that in balusters with odd number of channelures the cavity can be extended when cutting off the apexes of the load-carrying elements of the baluster opposite to the channelures. The formula for finding the cross-section area of the baluster load-carrying element is obtained.

With the increased elevation angle of the baluster spiral groove its esthetic value first increases reaching the maximum at the angle of, approximately, 45˚, then – decreases. The number of passes does not have significant influence on the esthetic value. It is pointed out 99

that with the increased number of passes the range of the experts’ assessment spread is narrowed, which is explained by the less visible difference of large number of passes for human eyes.

The minimum cross-section area of the load-carrying element is limited by the cutting forces emerging when milling the grooves. The destruction of load-carrying elements at the excessive cutting depth and feed values occurs with their chopping off along the fibers.

From the investigated materials (MDF, wood of pine, birch and beech) the positive results are achieved with birch and beech.

To decrease the destruction possibility due to more uniform distribution of cutting forces between the load-carrying elements, each groove should be milled in several steps, first, to one depth, then, after milling all the grooves, to another depth, then – to the third one, etc. gradually decreasing the cut off wood layer with the increased groove depth.

For the investigated typical size of the balusters from birch and beech the rational cutting depth at the first milling stage is 1.0 mm, at the final stage – 0.5 mm, and the feed is 1 mm/seс. The values less than the indicated ones result in decreased productivity, the greater ones – in decreased machining quality, the baluster vibration increase and increased destruction possibility.

REFERENCES PILIKINA, N.N. Wooden architecture of the North. Arkhangelsk: Pravda Severa Publishers, 2005.

160 p.

SIKLIENKA, M., JANDA, P., JANKECH, A. 2016. The influence of milling heads on the quality of created surface. In Acta Facultatis Xylologiae Zvolen, 2016, 58(2): 81−88. ISSN 1336-3824.

CHERNYKH, M., OSTANINA, P., STOLLMANN, V. 2018. Aesthetic properties and quality of relief surfaces of wooden products in automated manufacturing. In Acta Facultatis Xylologiae Zvolen, 2018, 60(2): 153−160. ISSN 1336-3824.

VITCHEV, P. 2019. Evaluation of the surface quality of the processed wood material depending on the construction of the wood milling tool. In Acta Facultatis Xylologiae Zvolen, 2019, 61(2): 81−90.

ISSN 1336-3824.

GUTSYKOVA, S.V. 2019. Method of expert assessments: theory and practice. 2nd edition. Мoscow: Institute of Psychology of RAS Publishers, 2019. 144 p. ISBN 978-5-9270-0209-2. Text: Electronic E-library system IPR BOOKS: [webpage] – URL: http://www.iprbookshop.ru (BB 355.html (refernce date: 02.02.2020).

UGOLEV, B.N. 2002. Wood and Forestry Commodity. Moscow: Publishiny Center The Academy, p. 272.

DZURENDA, L. 2008. English Impact of technology on the quality of split and machined wood.

Zvolen: Technical university in Zvolen, p.140. ISBN 978-80-228-1923-7.

GOST 16483.5-73. Wood. Methods for determination of ultimate shearing strength parallel to grain. М.: IPK Standard Publishers, 1999. p. 6.

GOST 18992-80. Polyvinylacetate homopolymeric coarse breakup dispersion. Specifications. М.: IPK Standard Publishers, 2001. p. 19.

Robor model “D” [E-resource]. Accessible at: https://semil.ru/machine_tools_wood (refernce date: 27.02.2020).

Hydraulic press Elbrus 2G 3000 [E-resource]. Accessible at: https://tiu.ru/P97609184-pressgidravlicheskij-elbrus.html (refernce date: 27.02.2020).

ACKNOWLEDGMENT The paper has been prepared under the financial support of the Ministry of Education, Science and Sports of the Slovak Republic in the frameworks of the project KEGA 007TU Z-4/2019

100

AUTHORS‘ ADDRESSES Stanislav Korchagin, master student Kalashnikov Izhevsk State Technical University V.A. Shumilov Institute of Construction and Architecture Department “Technology of Industrial and Artistic Processing of Materials” Studencheskaya St. 7 Izhevsk Russian Federation azgard45@yandex.ru Prof. Dr. Michael Chernykh Kalashnikov Izhevsk State Technical University V.A. Shumilov Institute of Construction and Architecture Department “Technology of Industrial and Artistic Processing of Materials” Studencheskaya St. 7 Izhevsk Russian Federation rid@istu.ru Associate professor Vladimír Štollmann Technical University in Zvolen Faculty of Forestry Department of Forest Harvesting, Logistics and Amelioration T. G. Masaryka 24 960 01 Zvolen Slovak Republic stollmannv@tuzvo.sk

101

102

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 103−118, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.10

CHIP SUCTION SYSTEM IN CIRCULAR SAWING MACHINE: EMPIRICAL RESEARCH AND COMPUTATIONAL FLUID DYNAMICS NUMERICAL SIMULATIONS Jacek Baranski – Przemyslaw Dudek ABSTRACT The experimental analysis of the wood chip removing system during its redesigning in the existing sliding table circular saw and computational fluid dynamic (CFD) numerical simulations of the air flow process is presented in the paper. The attention was focused on the extraction hood and the bottom shelter of the actual existing system. The main aim was to perform experimental research on the pressure distribution inside the hood and at the exit of the bottom shelter and the air flow distribution during operation of wood chip removal system. In the work a systematic experimental study of pressure and numerical modelling of the air flow distribution in the upper cover and bottom shelter during operation for the selected rotational speed of saw blade of 3 500 and 6 000 min-1 with a diameter of  300 mm and  450 mm were carried out. The analyses of results obtained from the experimental measurements and numerical simulations allowed the estimation of the areas with improper air flow hindering the controlled transport of wood chips and to optimize the shape of extraction hood and the bottom shelter. As the result, a new design of the chip suction system was obtained, noticeably improving the chips extraction from the tool operation space.

Key words: sliding table saw circular sawing machine, chip removing system, experimental study, numerical simulations.

INTRODUCTION Nowadays, the recent development of woodworking machines design, introduction of new technologies and also the material machining and feed speeds result in the need to provide more efficient systems of wood particles (e.g. chips, dust) removing. The manufacturers of woodworking machines quite freely approach to the problem of chip removing from the workspaces of their machines. Ideas used in some devices for shapes and dimensions of the suction system elements, linking these suctions to common, interior collectors, separators, moving suctions, etc. do not always work well in reality (BARAŃSKI et al. 2016). It is possible to give examples of machines manufactured by well-known and less-known companies, whose suction arrangements are ineffective during standard operation. In operation manuals there are often presented the essential parameters of the extraction installation for given woodworking machine, e.g. the necessary air velocity, its amount and the value of vacuum pressure. However in the area around the operating tool even higher vacuum pressure is required. A modern sliding table saw operating with

103

improper designed extraction installation, loses its performance and service life immediately (BARAŃSKI and PIKAŁA 2017, BARAŃSKI et al. 2018, ORŁOWSKI et al. 2020).

In the all technological processes of machine wood chips and dust, besides the main product, a by-product is also generated. These by-product are called chips Their shape, size and amount mainly depend on the form, physical and mechanical properties of sawed wood as well as on the shape, dimensions, type of machine, sharpness of the cutting tool, and technical and technological conditions of sawing process (BELJO LUČIĆ et al. 2005, OČKAJOVÁ et al. 2006, KOPECKÝ and ROUSEK 2007, KLEMENT and Detvaj 2007, OČKAJOVÁ et al. 2016, KMINIAK et al. 2020, ROGOZINSKI et al. 2020).

On the other hand, the space around the rotating tool is usually surrounded by different shape and dimension upper hood, connected with an exhaust fan using partially flexible pipes. Thus, the shape and dimensions of chip removing system are especially designed and made for specific woodworking machine requirements.

The machine operation parameters together with material properties with its drying process type strongly determine the particle size distribution of chipped wood (BARAŃSKI et al. 2016, ORŁOWSKI et al. 2019).

Particles of wood substance formed in individual chipping and machining processes are called “bulk wood substance” (DZURENDA 2007). In this reason, wood industry workers are exposed to airborne wood dust particles in the surrounding air of the workplace and may cause different occupational health problems (KOHLER 1995). Currently, the nature of the present production and chips features require their continuous removal from the place where they are generated. As far as sanding dust is concerned, it is removed by means of an airtechnical device – suction installation. To develop such suitable suction system, it is important to know the size and shape of bulk particles, which are the basic data for characterizing the bulk substance. The above characteristics affect physical and mechanical properties of bulk substance (bulk density and angle, tilt angle, aerodynamic properties of particles in the pipes of the suction system) and properties of separation or filtration in the separating device (DZURENDA 2007, KOHLER 1995). Likewise, those characteristics affect the service life of equipment in the workplace, where chips are generated and transported and filtering elements and last but not least the safety of the working environment.

Removal of dust is very difficult when working space is large and when the tool moves during processing at relatively high velocity. Depending on the woodworking machine type and the shape and size of its dust area, serious problems may occur concerning the effective dust remove through a suction system during certain machining processes. The chips’ dispersion in different directions in the space of the treatment zone is very unfavorable in this respect. When movement direction of the created chips during machining does not coincide with that of the air flow forced by an extraction system, many chips are still not removed and can become dispersed in the environment around the machine and workers.

This takes place during sawing when the whole tool goes into the material piece. For this reason there are many problems with the direct chips discharge from the working space and operating tools. Also the chips dispersion in all directions occurs due to rotational speed of those tools. Many researchers investigated the possibility of reducing the particle (airborne) by the control of machining parameters and by varying the cutting and feed speed respectively, tool type and its size, cutting angles, number of blades and processed material respectively (FUJIMOTO and TAKANO 2003, HEMMILÄ et al. 2003, FUJIMOTO et al. 2011, PAŁUBICKI and ROGOZIŃSKI 2016, HLÁSKOVÁ et al. 2015). According to those results, the average chip thickness is one of the most important parameters of wood machining.

Furthermore, sanding process as the major source of airborne dust generation was investigated by OČKAJOVÁ and BELJAKOVÁ (2004), KUČERKA and OČKAJOVÁ (2018).

Wood processing companies, as sources of air pollution, emit into the air wood dust 104

classified as solid pollutants. Clean Air for Europe is the initiative taken by the EU Commission, by which one of its main aims set in 2002 was: “… To reach such a quality of the environment where the level of pollutants coming from human activities doesn’t cause any significant impacts and risks for human health …” (EU Decision 2002).

In the paper, the results of the systematic empirical research together with numerical simulations of air flow distribution process of the chip removing system in existing sliding table circular sawing machine which were carried out are presented. The aim of the full research was to optimize and redesign the chip extracting system of the device. The special attention was focused on upper hood and bottom shelter conditions respectively.

MATERIALS AND METHODS Experiments During the experimental research, pressure distribution inside the upper hood and velocity measurement at its exit were accurately recorded with the selected rotational speeds of 3 500 min-1 and 6 000 min-1 of saw blade with the diameter of  300 mm and  450 mm respectively (BARAŃSKI et al. 2016, BARAŃSKI et al. 2018). Simultaneously to the experimental research, the numerical simulations for selected boundary conditions and rotational speeds with chosen saw diameter blades were performed. The analysis of the results of these experiments and numerical modelling helped to identify in the upper cover areas with insufficient vacuum value (and even with small overpressure), which caused chip ejection from the upper hood and decrease operating volume of bottom shelter. On the basis of the obtained results, a new design of the hood and bottom shelter shapes has been proposed.

The pressure and velocity measurements were performed using the wide and narrow upper hood and the bottom shelter of the suction system with different diameter and rotational speed of the saw blades respectively. For the saw blade No. 1 two of its extreme rotational speeds 3 500 min−1 and 6 000 min−1, and for saw blade No. 2 its nominal rotational speed 3 500 min−1 were considered respectively. Table 1 shows the parameters of the saw blades used during experiments.

Tab. 1 The main parameters of the sliding table saw blades during experiments.

Tool diameter, D (mm) Number of teeth, z High of tooth high, h (mm) Overall set (kerf width), St (mm) Thickness of saw blade, a (mm) Rotational speed, n (min-1)

Circular saw 1  300 96 9.82 3.2 2.2 3 500 6 000

Circular saw 2  450 72 19.63 4.4 3.2 3 500

The relative pressure was measured using the digital multi-function measuring instrument TESTO 480 (TESTO SE & Co., Germany). The results were averaged for 10 seconds at each point. Measuring range was from -10 to +10 kPa, resolution 0.01 kPa and accuracy ±0.3 Pa +1% of measured value in the lower range. Velocities at the exit of the hood as well as from bottom shelter were measured using the same device equipped with a hot wire probe. Measuring range was 0−20 m/s with resolution 0.01 m/s and accuracy was ±0.03 m/s + 5% of measured value.

The base case of the wide upper hood inner dimensions are length 710 mm, high 240 mm and width 144 mm respectively and the base case of the narrow hood inner dimensions are length 710 mm, high 157 mm and width 50 mm respectively, Fig. 1. The

105

distance between the saw blade and the inside the cover is 245 mm and 165 mm for saw blade  300 mm and  450 mm respectively. a)

b)

Fig. 1 The view of the base case upper hood shape with all dimensions: a) side view, b) fron view.

The empirical research consisted of measuring the relative pressure distribution zeroreferenced against ambient air pressure at several points on the wide and narrow upper hood of suction-chip removing system (points “A”-“I” on the upper part and “a”-“i” on the side part of the wide hood and points “b*”-“h*” on the side of narrow hood). Locations of all measuring points are shown in Fig. 2 and Fig. 3.

D

E

F

G

C

H B A

d

b

h

f

e

i a

Fig. 2 The view of measurement points on a wide cover location.

b*

d*

e*

f*

h*

Fig. 3 The view of measurement points on a narrow cover location.

106

I

The obtained results from measurements were analyzed and verified experimentally using different shape design of the upper hood, Fig. 4.



Fig. 4 The chosen examples of the upper hood shape with  angle at its outlet: a) base case, b−e) modification cases.

The bottom shelter shape was modified to minimize its operating volume and leakages and the outlet channel shape in the upper part was changed.

Numerical simulations To create 2D numerical models of the sliding table sawing machine complex geometry 3D model of the device was created in CFD ANSYS Software. CFD Fluent, which is a steady state/transient, finite volume program that can solve three-dimensional fields of pressure, velocity, temperature, kinetic energy of turbulence and dissipation rate of turbulence belongs to this group. The code operates by solving the governing differential equations of the flow physics by numerical means on a computational mesh and is able to predict air velocity.

In the second step the 2D models were created taking into account simplifications of 3D model with necessary boundary conditions and circular saw blade and scoring blade properties respectively (Fig. 5). The 2D numerical model consists of an unstructured mesh.

Thus, it was possible to divide the flow region into finite elements of small size. This approach was very important to achieve convergence and accuracy of the solution. For grid generation, the unstructured finite volume grid was used to divide the very complicated 107

geometry of the flow domain into discrete segments with high grid quality. This approach is very important to get a convergent and accurate solution. The number of unstructured elements (cells) was about 1 892 600. Combining the body-fitted meshing capabilities with unstructured non-orthogonal grids and arbitrary coupling between mesh blocks gives great flexibility in the representing highly complex geometries.

The finite volume method and first-order upwind scheme were used to convert elliptical partial differential equations into algebraic equations, which are solved using the iteration method.

The standard Simple scheme was used for pressure-velocity coupling, while the underrelaxation method was used to control the update of computed variables during the iteration process.

For the turbulent flow modelling the standard model of turbulence k- (k- standard) and a mesh of the entire dust extraction system with machining material and with or without a separating knife were used.

The numerical modeling simulations for each case were performed until the values of the solution results of the equations’ behavior (the so called residuals) were stabilized in the narrow range of the variation at the level of 10-4.

Fig. 5 The view of the sliding table saw 3D numerical model.

In the selected cases, the air flow in the dust extraction system was analyzed without taking into account the occurrence of the wood chips movement. In Fig. 6 the examples of the circular sawing machine numerical models with the diameter  300 mm and  450 mm circular saw blades respectively and separation knife are presented.

The vacuum pressure values at the outlets from the upper hood and bottom shelter were assumed on the basis of the results from the experimental measurements and set 200 Pa at the upper hood exit and 400 Pa at the lower shelter exit respectively. The maximum value of the vacuum pressure at both outlets, e.g. 1 500 Pa, was adopted in accordance with the information provided by the panel saw Manufacturer (REMA S.A., Reszel, Poland) and the data available in literature (DZURENDA 2007). Simulations of air flow process for vacuum pressure at both exits with values of 200 Pa and 400 Pa respectively were also performed.

108

a) upper hood exit

upper hood

air inlet to the wood chip

air inlet to the wood chip machining material

removing system

removing system sliding table

sliding table

scoring blade

separating knife

circular saw

bottom shelter

bottom shelter exit

b) upper hood exit

upper hood

air inlet to the wood chip

air inlet to the wood chip machining material

removing system

removing system sliding table

sliding table

scoring blade

separating knife bottom shelter circular saw

bottom shelter exit

Fig. 6 The cross section view of the sliding table circular sawing machine numerical model with boundary conditions: a) saw diameter blade  300 mm, b) saw diameter blade  450 mm.

109

The numerical simulations were performed in the case of different boundary conditions and circular saw blade diameter, as shown in Table 2.

Tab. 2 The boundary conditions assumed for numerical simulations for the sliding table circular sawing machine with the chip removing system.

Ambient pressure [Pa] 1.

2.

3.

4.

5.

6.

101 130 101 130 101 130 101 130 101 130 101 130

7.

8.

9.

101 130 101 130 101 130

Vacuum pressure Vacuum pressure at bottom shelter exit at upper hood exit [Pa] [Pa] circular saw  300 mm - 400 - 200 - 400 - 400 - 1 500 - 1 500 - 400 - 200 - 400 - 400 - 1 500 - 1 500 circular saw  450 mm - 400 - 200 - 400 - 400 - 1 500 - 1 500

Circular saw rotational speed [min-1]

Scoring blade Rotational speed [min-1]

3 500 3 500 3 500 6 000 6 000 6 000

8 500 8 500 8 500 8 500 8 500 8 500

3 500 3 500 3 500

8 500 8 500 8 500

For each modeling case, which was carried out, the chosen simulation results in the ANSYS Fluent software will be presented as velocity vector field in cross section of sliding table circular saw and chip removing system, the upper hood and bottom shelter.

RESULTS AND DISCUSSION Numerical simulations A series of computer simulations were performed. Each simulation was performed in the case of various boundary conditions. The differences concerned the rotational speed of the saw, the vacuum pressure it the both outlet and the diameter of the circular saw.

Figure 7 presents the results of numerical simulations of the air flow process through the sliding table circular saw without separating knife and with chip removing system and material machining for parameters, which were assumed in the numerical methodology. We can see the behavior of the air flow process, which after entering the dust removing system is directed to the outlets of lower shelter and upper hood.

On the other hand, Figure 8 shows the resutls of numerical modeling of the air flow process of the same device equipped with separation knife. It can be noticed that the occurrence of the knife affects the less favorable distribution of the air flow in the rear part of the dust extraction system. Additionally, in the front part of the upper hood, it can be seen that the air is pushed into the surrounding environment with greater intensity compared to the previous case (marked with the red circle). The velocity value is higher in Fig. 7. There is no air turbulence zone in the upper part of the upper hood, near to the handle (marked with the blue circle). This is the effect of the rotating saw marginal impact on the flowing medium.

The separating knife is an obstacle in the air flow process at the exit of the lower shelter. It reduces the saw impact on the air flow. In the rear part of the upper hood, the inflowing air is separated into two streams and flows to the outlets of the lower shelter and the upper cover.

In the upper hood the air is directed into the environment, toward its upper wall. In addition, there is an area where the air flow velocity values are small.

After many numerical simulations performed it can be concluded that there was the rotating circular saw large impact on the medium flow and intensive mixing process of the air with the air stream incoming from surrounding environment to the dust removing system.

110

This was due to the high value of the linear velocity of the circular saw and the shape of the upper hood, which had got the so called the dead space (recirculation zone) from which the air was unable to get out from the upper hood. In addition, it could clearly noticed the effect of the circular saw on pushing the air out the upper hood in its front part (Fig. 6b and 7b).

This was the total effect of the linear velocity of the circular saw and the small vacuum pressure value occurrence (200 Pa) at the upper hood exit. a)

b)

Fig. 7 Results of numerical modeling of: a) the air flow velocity field through the circular saw and chip removing system with saw diameter  450 mm without separating knife, b) the view of the air flow in the front part of the upper hood.

111

a)

b)

Fig. 8 Results of numerical modeling of: a) the air flow velocity field through the circular saw and chip removing system with saw diameter  450 mm with separating knife, b) the view of the air flow in the front of the upper hood.

In Fig. 9 the chosen examples of the upper hood of the chip removing system shapes simulation results are presented. The each modification of the upper hood of chip removing system was experimentally examined during measuring the velocity in selected points and relative pressure values along the upper hood in different locations.

After numerical simulations and empirical research the final shape of the upper hood of dust removing system was obtained which characterized high efficiency in chip collection and removing from working area of rotating tools.

112

Fig. 9 The selected examples of the results of numerical modeling of the air flow through the circular saw and chip removing system of the different upper hood shapes.

The results of relative pressure measurements for the wide and the narrow hood of existing system, referred further as "base case", showed that at most of the measurement points there was a vacuum pressure existence, as it was expected. However, there were observed some places, with a very low vacuum pressure value or even gauge pressure. These are points “A”-“B” and “a”-“b” for wide and “b*”-“d*” for narrow hoods respectively. They are located in the zone where the highest influence of the air stream created by the teeth width and rotational speed of the saw blade exists. That influence at three of the critical points can be noticed in Fig. 10 and Fig. 11. Saw blade rotational speed (Fig. 10) adversely 113

affects the pressure distribution field, causing disappearance of desired vacuum pressure.

The same result can be seen if saw blade diameter increased (Fig. 11). Contrary effect occurs in all critical points. location B

location b

location b*

2

Relative pressure [Pa]

0 -2 -4 -6 -8 -10 -12

3 500

6 000

Saw blade rotational speed [min -1] Fig. 10 Relative pressure [Pa] as a function of saw blade rotational speed at different location for wide and narrow cover; saw blade No. 1 ( 300 mm), fan motor frequency 50 Hz. location B

location b

location b*

6

Relative pressure [Pa]

4 2 0 -2 -4

-6 -8 -10 -12

300

450

Saw blade diameter [mm] Fig. 11 Relative pressure [Pa] as a function of saw blade diameter at different location for wide and narrow cover; saw blade rotational speed 3500 min−1, fan motor frequency 50 Hz.

In order to improve the performance of chip removing system without interfering with the fan model structure several modifications were carried out as follows. Firstly, the shape change of lower shelter, minimizing its leakages as well as modification of the shape of the outlet channel in upper part were executed. The results of pressure distribution after that modifications are shown for both wide and narrow hood in Fig. 12−13 as “modified case”.

114

base

modified

final

60 40

Relative pressure [Pa]

20 0

A

B

C

D

E

F

G

H

I

-20 -40 -60 -80

-100

Fig. 12 Chosen results of the relative pressure distribution along the upper part of the wide cover. base

modified

final

60 40

Relative pressure [Pa]

20 0 a

b

d

e

f

h

i

-20

-40 -60 -80

-100

Fig. 13 Chosen results of the relative pressure distribution along the side part of the wide cover.

It can be seen that higher vacuum pressure was achieved in comparison with "base case" except for points “A” and “a”.

115

A new design of the upper hood was proposed. Its final shape is already patented. It can also be seen in Fig. 12−13 (“final case”) that it resulted in satisfactory pressure distribution without modification of the fan construction, Fig. 14−15.

Fig. 14The final shape of the upper hood. a)

b)

c)

Fig. 15 The view of: a) final design, b) cross section of upper hood, c) cross section of bottom shelter.

116

CONCLUSIONS The existing dust removing system in the analyzed woodworking machine did not provide efficient and satisfactory chip extraction from the working tool area. The empirical research and numerical modeling simulations proved that: • in the area around the tool, insufficient vacuum pressure value could hinder the organized pneumatic transport of wood wastes; • after several changes in the upper hood and lower shelter shapes of the suction system, all parts of the extraction system to obtain better conditions and provide efficient performance of chip removing system were redesigned and modified except the fan construction; • the final design of the upper hood shape, which belongs to chip extraction system is shown in Fig. 15; • the dimensions of the new wide upper hood inner dimensions are length 690 mm, high 175 mm and width 137 mm and the new narrow hood inner dimensions are length 690 mm, high 175 mm and width 42 mm respectively; • from the main dimensions of the new wide and narrow upper covers the distance between the saw blade and the inside the new cover shape is 226 mm and 141 mm for saw blade  300 mm and  450 mm respectively and; • the distance above the saw blade and the inside the new cover shape is 175 mm and 25 mm for saw blade  300 mm and  450 mm respectively.

REFERENCES ANSYS, Inc. 2016. ANSYS Fluent User’s Guide, Release 17.2.

BARAŃSKI, J., JEWARTOWSKI, M., WAJS, J., PIKAŁA T. 2016. Experimental analysis of chip removing system in circular sawing machine. In Trieskove a Beztrieskove Obrabanie Dreva, Technical University in Zvolen, 10: 25−29.

BARAŃSKI J., PIKAŁA T. 2017. Application in circular sawing machines of the experimental results of investigations of the chip removing system operation. In Annals of WULS, Forestry and Wood Technology. no 100, pp. 199−205.

BARAŃSKI J., JEWARTOWSKI M., WAJS J., ORŁOWSKI K., PIKAŁA T. 2018. Experimental examination and modification of chip suction system in circular sawing machine. In Drvna Industrija, 69(3): 223−230.

BELJO LUČIĆ, R., KOS, A., ANTONOVIĆ, A., VUJASINOVIĆ, E., ŠIMIČIĆ, I. 2005. Svojstva usitnjenog materijala nastaloga pri mehaničkoj obradi drva. In Drvna industrija, 56 (1): 11−19.

DZURENDA, L. 2007. Sypká drevná hmota, vzduchotechnická doprava a odlučovanie. 1st ed., Zvolen: Technical University in Zvolen. 182 p.

FUJIMOTO, K., TAKANO, T. 2003. Mass concentration and particle size distribution of suspended during circular sawing. In The Proceedings of the 16th International Wood Machining Seminar.

Matsue, Japan, pp. 724−731.

FUJIMOTO, K., TAKANO, T., OKUMURA, S. 2011. Difference in mass concentration of airborne dust during circular sawing of five wood-based materials. In J Wood Sci 57:. 149−154. https://doi.org/10.1007/s10086-010-1145-y HEMILLÄ, P., GOTTLÖBER, C., WEELLING, I. 2003. Effect of cutting parameters to dust and noise in wood cutting, laboratory and industrial tests. In The Proceedings of the 16th International Wood Machining Seminar. Matsue, Japan, pp. 375−384.

HLÁSKOVÁ, L’., ROGOZINSKI, T., DOLNY, S., KOPECKỲ, Z., JEDINÁK, M. 2015. Content of respirable and inhalable fractions in dust created while sawing beech wood and its modifications. In Drewno 58 (194): 135−146.

117

KMINIAK, R., ORLOWSKI, K.A., DZURENDA, L., CHUCHALA, D., BANSKI, A. 2020. Effect of Thermal Treatment of Birch Wood by Saturated Water Vapor on Granulometric Composition of Chips from Sawing and Milling Processes from the Point of View of Its Processing to Composites. In Appl. Sci., 10, 7545.

KOHLER, B. 1995. Wood dust and cancer. National Report - Health, Safety and Environment, IARC, France.

KOPECKÝ, Z., ROUSEK, M. 2007. Dustiness in high-speed milling. In Wood research, 52(2): 65−76.

KOS, A., BELJO LUČIĆ, R. 2004. Factors influencing particle size distribution of oak sawdust developed during circular sawing. In Trieskove a Beztrieskove Obrabanie Dreva. Zvolen: Technical University in Zvolen, Vol. 4, pp. 131−137.

KUČERKA, M., OČKAJOVÁ, A. 2018. Thermowood and granularity of abrasive wood dust. In Acta Facultatis Xylologiae Zvolen, 60(2): 43−51. DOI: 10.17423/afx.2018.60.2.04.

OČKAJOVÁ, A., BELJO LUČIĆ, R., ČAVLOVIĆ, A., TERAŇOVÁ, J. 2006. Reduction of dustiness in sawing wood by universal circular saw. In Drvna industrija, 57(3): 119−126.

OČKAJOVÁ, A., KUČERKA, M., BANSKI, A., ROGOZIŃSKI, T. 2016. Factors affecting the granularity of wood dust particles. In Chipand Chipless Wood working Processes, 10(1): 137−144.

ORŁOWSKI K., CHUCHAŁA D., MUZIŃSKI T., BARAŃSKI J., BANSKI, A., ROGOZIŃSKI T. 2019. The effect of wood drying method on the granularity of sawdust obtained during the sawing process using the frame sawing machine. In Acta Facultatis Xylologiae, Zvolen, 61(1): 83−92.

ORLOWSKI, K. A., DUDEK, P., CHUCHALA, D., BLACHARSKI, W., PRZYBYLINSKI, T. 2020. The Design Development of the Sliding Table Saw Towards Improving Its Dynamic Properties. In Appl. Sci., 10, 7386.

PAŁUBICKI, B., ROGOZIŃSKI, T. 2016. Efficiency of chips removal during CNC machining of particleboard. In Wood Research, 61(5): pp. 811−818.

ROGOZINSKI, T., CHUCHALA, D., PĘDZIK, M., ORLOWSKI, K., DZURENDA L., MUZINSKI, T. 2020.

Influence of drying mode and feed per tooth rate on the fine dust creation in pine and beech sawing on a mini sash gang saw. In Eur. J. Wood Prod. https://doi.org/10.1007/s00107-020-01608-8.

Decision No 1600/2002/EC of the European Parliament and of the Council of 22 July 2002 laying down the Sixth Community Environment Action Programme. Official Journal of the European Communities, 10.9.2002: L 242/1 - L 242/15 http://eur-lex.europa.eu/legal-content/EN/TXT/PDF /?uri=CELEX:32002D1600&from=EN (accessed on October 12, 2020) Operation manuals of panel saws.

Testo Technical Data (https://www.testo.com/resources/51/43/180850f3d452/testo-480).

ACKNOWLEDGMENTS It is kindly acknowledged that this work has been carried out within the framework of the project POIR.01.01.01-00-05888/15, which has been financially supported by the European Regional Development Fund. The authors would also like to acknowledge the company REMA S. A. in Reszel (Poland), which is the beneficiary of the project, for helping out in the experimental part of this work. The data included in this article was partially presented during 23 rd International Wood Machining Seminar, Warsaw, Poland. The sliding table saw FX550, in which the presented suction system is applied, was awarded with the Gold Medal at the International Fair DREMA in Poznan, Poland in 2017.

AUTHORS ADDRESSES Jacek Baranski (ORCID id: 0000-0001-9040-9181) Gdansk University of Technology Faculty of Mechanical Engineering and Ship Technology Narutowicza 11/12 80-233 Gdansk, Poland

118

Przemyslaw Dudek REMA S.A. B. Chrobrego 5 11-440 Reszel, Poland

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 119−130, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.11

DURABILITY OF KITCHEN FURNITURE MADE FROM MEDIUMDENSITY FIBREBOARD (MDF) Olena Pinchevska – Ján Sedliačik – Olexandra Zavorotnuk – Andriy Spirochkin – Ivan Hrabar – Rostislav Oliynyk ABSTRACT Furniture products designed of medium-density fibreboard require preliminary assessment of their durability in real conditions. The use of kinetic theory of solid strength was suggested, which had been previously confirmed for particleboard. Developed mathematical model of durability enables to determine the time of destruction of different thickness MDF with different types of surface finishing under the influence of thermal and power loads on panels with moisture content in the range from 6% to 20%. It comprises calculation of MDF thermoactivation parameters based on experimentally determined bending strength and destruction time under temperature ranging from 20 °С to 80 °С.

Results of calculations correspond to weighted average service life of furniture intended for kitchen applications.

Key words: medium density fibreboard (MDF), kitchen furniture, bending strength, predicting durability.

INTRODUCTION European market reflects the increasing tendency to use MDF in a cabinet furniture production instead of particleboard, especially for nursery furniture, kitchen or living room furniture (ENGINEERED WOOD PRODUCTS MARKET SHARE 2021). Kitchen furniture is generally used in adverse temperature and humid environment conditions (HU et al. 2020).

Given this and the action of mechanical load on the countertop causes bending, probability of cracks and their destruction. High temperature can be as an activator of the destruction and this is a factor reducing durability (GERHARDS 1982; YOUNG and CLANCY 2001; MORAES et al. 2005; GREEN and EVANS 2008a and 2008b; AYRILMIS et al. 2009). Especially this can be observed in composite materials, which includes resin. The resin has features to change their characteristics over time. The rate of these changes is mainly dependent on temperature. Temperature range of these changes – ageing – is related to glass transition temperature and temperature of other transitions specific to this polymer. Configuration and conformation of macro chain and their supramolecular organization is changing in a result of physical processes. Time factor has an important role in this case (MANIN and HROMOV 1980).

The moisture is a factor that has also an impact on a product durability. As the result of this reaction, macromolecules hydrolytically decompose and properties of the product can be changed. A plasticizer is contributing to polymer secondary structure changes and 119

relaxation of internal stresses (MIRSKI et al. 2020). Significant changes of polymeric materials take place under the moisture and heat impact (ZUBAREV et al. 2004; MARTENSSON and THELANDERSSON 1990; MOHEBBY et al. 2008).

Mechanical load also can be an external activator of loss of durability. It effects both physical and chemical transformations in the material. Mechanical load (static and dynamic) activates physical processes and leads to acceleration of relaxation processes at different temperature and moisture content in composite materials (YU and OSTMAN 1983; SHI and GARDNER 2006). It was found that MDF’s modulus of rupture (MOR) and modulus of elasticity (MOE) is decreasing when temperature is increasing (ZHOU et al. 2012).

Research on durability of construction materials was carried out during long period.

Understanding of solid materials destruction passed through several steps. Current thermoactivation step considers destruction as a process that depends on parameters of defective structure and takes into account thermal motion of atoms. Study of the timetemperature dependence of solid material strength began by ZHURKOV (1965). It has led to formulation of basic principles of kinetic theory of strength of solid materials. Investigations that were conducted in the conditions of single axis tensile for solid materials (singlecrystals, poly-crystals, polymers, composite materials) during testing at different temperatures provided an opportunity to identify dependence of durability τ on load σ and temperature Т: 𝑈 − 𝛾𝜎 𝜏 = 𝜏0 𝑒𝑥𝑝 [ 0𝑅𝑇 ] (1) where: R – universal gas constant (kJ/mol·K); U0, τ0, γ – constants of tested material; Т – temperature (°K); σ – load (МPа).

Equation (1) is the main for kinetic theory of strength. It has real physical meaning and shows that thermal motion of atoms is the reason of the destruction of solid material. The kinetic theory of strength was used for metals (SLUTSKER et al. 2002), HRABAR (2002), for composite materials (REGEL 1980), for polymer fibres and cellulose derivatives (RATNER and YARTSEV 1992). This provided fundamental reasons for using this theory for woodbased composite material – particleboard (HRABAR et al. 2008).

Later, a parameter Тm was introduced into the equation (1). Тm is a temperature of destruction, when all bonds in material are breaking during one heat vibration and when substance cease to exist (RATNER and YARTSEV 1992): 𝜏 = 𝜏𝑚 ∙ 𝑒𝑥𝑝 [

𝑈0 − 𝛾𝜎 𝑅

(Т−1 − Т−1 𝑚 )]

(2)

where: τm, U0, γ and Tm – physical (thermoactivation) material parameters (TAP); τm – minimum durability (the period of oscillation of kinetic units – atoms, atom groups, segments) (s); U0 – maximum destruction activation energy (kJ/mol); γ – structural and mechanical parameters (kJ/mol·MPa); Tm – temperature limit existence of solids (temperature degradation) (°K); R – universal gas constant (kJ/mol·K), τ – time to fracture (durability), (s), σ – stress (MPa); T – temperature (°K).

It is rational to determine the strength and rigidity of MDF board under the influence of external factors analytically. This will enable to predict the durability of future product during the design process (BOIKO et al. 2013).

The aim of this study is to develop the new method of prediction of MDF furniture durability if there is an influence of high temperature and humidity. This method takes into account a long-term action of mechanical load.

120

MATERIAL AND METHODS Furniture MDF boards with the thickness of 10, 16 and 19 mm manufactured by JSC “Korostensky MDF Plant” were selected for the study. These panels were without surface finishing, painted and finished by veneer “fine-line”. Boards were randomly taken from one consignment of samples according to EN 326-1. It was chosen 40 panels of each thickness at all. Test pieces for physical and mechanical tests (determination of moisture content (EN 322), density (EN 323) and MOR and MOE (EN 310)) were cut from these panels. Before cutting according to EN 325, panels were conditioned at 20 °C and 65% relative humidity.

Average value of panel’s density at the moisture content of 5-7 % with the thickness of 10 mm was 824 kg·m-3, for the thickness 16 mm was 811 kg·m-3 and for the thickness 19 mm was 802 kg·m-3. Samples of MDF with cross section 10×50 mm, 16×50 mm and 19×50 mm and length 250 mm, 370 mm and 430 mm were used for determination of static bending strength according to EN 310. Tests were carried out at four temperatures 20, 40, 60 and 80 °С by static 3-point bending in the universal test machine, 20 replicates were done.

Activation energy was determined using derivatograph according to (STB 1333.0-2002 «POLYMER PRODUCTS FOR BUILDING») with the mass loss of material test portion at temperature during heating with given speed using a range of temperatures. Pine wood (Pinus) with moisture content 8 %, MDF panels and urea-formaldehyde resin KF-MT were tested. Experimental plans that included such factors: moisture content within 6-20 %, and temperature action from 5 min to 5 hours (PIZHURIN 2004) were used to test moisture content influence and influence of urea-formaldehyde resin aging on MDF durability.

Equation 2 is used to calculate theoretical durability because it does not take into account influence of temperature changes, humidity, material moisture content and mechanical loads. That’s why amendment taking into account exploitation factors are used.

Moisture content has the main influence on physical and mechanical properties of wood and wood-based panels, it increases internal stresses and as a result – probability of destruction (KULMAN et al. 2017). A lot of external factors influence MDF, which has polymeric binder in its structure. During exploitation, difficult changes take place inside the system of polymer matrix. At first heat treatment leads to material strengthening because the resin is curing, and later to change of physical and mechanical properties. This is taking into account in suggested panel strength model including binder aging (ZAVOROTNUK et al. 2020): 𝜏 = 𝜏𝑚 𝑒𝑥𝑝 [

𝑈0 −𝛾𝜎 𝑅

(𝑇 −1 − 𝑇𝑚−1 )] 𝑒𝑥𝑝( 𝛼𝛽

𝑊 𝑊𝑚

)

(3)

where: α – coefficient, which takes into account the effect of material moisture content on durability; β – coefficient, which takes into account the degradation of the polymeric binder; Wm – maximum valid value of material’s moisture content within which it has sufficient durability properties, (%); W – material’s current moisture content during using, (%).

Thermoactivation parameters τm, U0, γ and Тm for MDF are calculated using the equations: U0 RT1…4

U

σ

σ

1…4 - RT0 -γ T1…4 +γ RT +lnτm =lnt1…4 m

1…4

m

(4)

where: T1…4, – temperature of four testing series, (°K); σ1…4, – maximum breaking stress at the appropriate temperature, (МPа); t1…4 – time to destruction of the sample at the appropriate temperature, (s).

Coefficients α and β were determined experimentally for each type and thickness of MDF depending on the moisture content, the effect of temperature and the time of these factors.

121

RESULTS AND DISCUSSION TAP of all testing MDF were calculated by equation system (4) using the Mathcad software, Table 1. Modulus of elasticity in bending and bending strength were tested at different temperatures with time fixation before the destruction of the sample, Table 2.

Tab. 1 Thermoactivation parameters of medium-density fibreboard.

Type of MDF

Activation energy [U₀], (kJ/mol)

Destruction temperature [Tm], (K)

Structural-mechanical parameter [γ], (kJ/ mol·MPa)

Minimal durability [τm], (s)

206

409

5.15

0.819

162 183

484 425

5.72 2.65

0.733 0.779

179

440

5.07

0.861

174 198

437 432

4.73 4.98

0.811 0.705

207

411

6.83

0.463

202 213

423 396

6.24 5.67

0.905 0.419

10 mm without finishing 10 mm painted 10 mm veneered 16 mm without finishing 16 mm painted 16 mm veneered 19 mm without finishing 19 mm painted 19 mm veneered

MOR and MOE testing showed deterioration of wood-based composite materials properties within the temperature 60 – 80 °С. MOR and MOE of MDF without finishing with thickness 10 and 16 mm decreased on average at 37.6 % and 33.2 % during temperature rising from 20 °С to 80 °С, Figure 1.

Tab. 2 Test results required to calculate the MDF slab thermoactivation parameters.

No.

MOR, [σ], MPa

1 2 3 4

σ1 σ2 σ3 σ4

38.61 33.47 31.07 27.42

1 2 3 4

σ1 σ2 σ3 σ4

39.53 31.80 31.67 30.58

1 2 3 4

σ1 σ2 σ3 σ4

43.13 42.60 36.45 33.77

1 2 3 4

σ1 σ2 σ3 σ4

32.34 26.21 26.82 26.88

1 2 3 4

σ1 σ2 σ3 σ4

23.48 20.87 17.96 22.78

Temperature, [Т], K Time until the sample is destroyed, lnτ, s MDF 10 mm without finishing T1 293 ln τ1 4.304 T2 308 ln τ2 4.174 T3 323 ln τ3 3.932 T4 353 ln τ4 3.584 MDF 10 mm painted T1 293 ln τ1 4.143 T2 308 ln τ2 3.932 T3 323 ln τ3 3.689 T4 353 ln τ4 3.219 MDF 10 mm veneered “fine line” T1 293 ln τ1 4.394 T2 308 ln τ2 4.174 T3 323 ln τ3 3.989 T4 353 ln τ4 3.829 MDF 16 mm without finishing T1 293 ln τ1 4.394 T2 308 ln τ2 4.174 T3 323 ln τ3 4.111 T4 353 ln τ4 3.611 MDF 16 mm painted T1 293 ln τ1 4.522 T2 308 ln τ2 4.263 T3 323 ln τ3 4.190 T4 353 ln τ4 3.912

122

1 2 3 4

σ1 σ2 σ3 σ4

27.13 25.89 24.99 21.27

1 2 3 4

σ1 σ2 σ3 σ4

40.54 32.92 27.82 28.39

1 2 3 4

σ1 σ2 σ3 σ4

28.79 25.12 20.52 20.12

1 2 3 4

σ1 σ2 σ3 σ4

34.94 31.99 28.52 26.83

MDF 16 mm veneered “fine line” T1 293 T2 308 T3 323 T4 353 MDF 19 mm without finishing T1 293 T2 308 T3 323 T4 353 MDF 19 mm painted T1 293 T2 308 T3 323 T4 353 MDF 19 mm veneered “fine line” T1 293 T2 308 T3 323 T4 353

ln τ1 ln τ2 ln τ3 ln τ4

4.369 4.263 4.234 4.159

ln τ1 ln τ2 ln τ3 ln τ4

4.654 4.454 4.317 4.220

ln τ1 ln τ2 ln τ3 ln τ4

4.700 4.511 4.394 4.304

ln τ1 ln τ2 ln τ3 ln τ4

4.771 4.736 4.443 4.357

* The table shows the average values of a tests series.

5500

40

5000

MOE, MPa

MOR, MPa

45

4500

35

4000

30

3500

25

3000

20 20

40

60

80

20

100

40

60

80

100

Temperature, [t] , °C

Temperature, [t] , °C

MDF without finishing with thickness 10 mm

MDF without finishing with thickness 10 mm

Fig. 1 Graphs of temperature dependence: a) MOR (MPa); b) MOE (MPa).

It was determined that the destruction temperature of MDF is less than the limit destruction temperature of urea-formaldehyde resin, due to the low content of synthetic cross-linked composition in the material (RATNER and YARTSEV 1992). When the composite material destroys, appeared cracks are the result of atomic bonds destruction. The value τm characterizes the time during which the process of composite materials destruction occurs at the temperature of destruction. The results of MDF components activation energy determination are shown in Table 3.

Tab. 3 Tested materials activation energy values.

Tested material MDF 10 mm without finishing Pine wood UF resin KF-MT

Activation energy, [E], kJ/mol 186 148 110

123

It can be seen that the activation energy of the binder has the lowest value, so the process of destruction begins with the resin. This confirmed the hypothesis of the resin aging effect on the durability of MDF, and justifies the addition into the equation (2) of the factor , which takes into account the effect of UF resin aging. Comparison of the activation energy values determined by the derivative and calculated from the accelerated test showed that they differ on average by no more than 10% for MDF 10 mm thick without finishing (U0 = 206 kJ/mol).

Adequate experimental data were obtained for calculating the coefficient α of MDF with 10, 16 and 19 mm thick without finishing: For MDF 10 mm without finishing: α = 0.02 - 0.0008∙W + 0.0028∙τ - 0.00015 W∙τ

(5)

For MDF 16 mm without finishing: α = 0.02 - 0.0011∙W + 0.0055∙τ - 0.0003 W∙τ

(6)

For MDF 19 mm without finishing: α = 0.04 - 0.0017∙W + 0.0042∙τ - 0.0002 W∙τ

(7)

where: W – material moisture content, (%); τ – the duration of the increased value of relative humidity, (h).

Checking the obtained equations for a significance level of 5% confirmed the adequacy of the models (for MDF 10 mm without finishing Fcalc = 1.376< Ftab = 3.03; for MDF 16 mm without finishing Fcalc = 0.265< Ftab = 3.03; for MDF 19 mm without finishing Fcalc = 0.813< Ftab = 3.03).

Based on the results of the studies, a regression dependence was obtained to determine the coefficient of destruction of the resin β from the temperature and time of its action: β = 8.14 - 0.0082∙T - 0.0049∙τ - 0.005∙t∙τ

(8)

140 120 100 80 60 40 20 0

135,54 119,43131,44 104,15 70,56 93,39 102,51 63,23 55,05

without painted finishing

Thikness 19 mm Thikness 16 mm Thikness 10 mm

MDF thickness

Durability, year

where: T – temperature, (°C), τ – duration of the temperature action, (h) Both factors and their interaction have a strong negative effect on the destruction of the resin. Verification of equations confirmed the adequacy of the model at the level of 5% significance (Fcalc = 2.22< Ftab = 6.0), all the coefficients of the equation are significant.

According to the results of theoretical and experimental studies, the durability of MDF of different thickness and type of finishing intended for use in the room conditions with parameters t = 20 °C and φ = 60% was determined, Figure 2.

finished by veneer

Type of finishing Fig. 2 Durability of MDF with different thickness and finishing type.

124

Panels surface finishing contributes to the increase in panel’s durability of the same thickness, as well as there is a tendency to increase the durability in the case of increasing the thickness of the panels.

The influence of temperature and humidity on the durability of furniture is the most notable when using the kitchen countertops. To evaluate the durability of the countertop, a kitchen set consisted of a cabinet with swing doors, a cabinet for the oven and a hob and a section with drawers for storing kitchen utensils, were considered. The countertop was made of MDF 16 mm thick and finished with natural oak veneer. The case of a curbstone was made from the laminated chipboard 16 mm thick of production "Krono-Ukraine".

To determine the quantitative values of the parameters of equation (3), namely the value of the internal stresses, that occur during the kitchen countertop exploitation, was calculated using the finite element method. It was ensured that the kitchen countertop is rigidly screwed to the lower cabinets. Bottom cabinets were mounted on the floor of the room on supports, which are closed by a plinth. The total weight of household items on the countertop was 20 kg and was evenly distributed. The construction strength criterion is the value of internal stresses that occur during loading and the deflection of the material from which the product is made.

Based on the calculations, the values of internal stresses that occur in the material under the action of uniformly distributed load of 0.16 MPa were determined, Figure 3. Given that the MOR of the MDF is 20 MPa, it can be stated that the material has a considerable margin of safety.

Fig. 3 Stress distributions of the fragment when applying a uniformly distributed load.

The amount of countertop deflection was determined by its deformation by computer simulation in «SolidWorksSimulations», Figure 4.

Fig. 4 Movement distribution of the fragment when applying a uniformly distributed load.

125

When the value of internal stresses is 0.16 MPa, the deformation value is insignificant − 0.165 mm and does not exceed the value of 5 mm, which leaves the required gap between the countertop and the facade.

“SolidWorks Simulations” software was also used to determine the temperature in the area of the hob and oven. The “SolidWorks Simulations” application allowed us to simulate the process of temperature action on a material by setting the initial temperature of the product T = 20 °C and the time during which it would affect the construction τ = 3600 s = 1 hour, Figure 5.

Fig. 5 Simulation of thermal impact in the application «SolidWorks Simulations».

Based on the calculations, it was determined that the temperature has the greatest influence on the side walls of the oven section, where they are adjacent to the oven door and in the contact area of the countertop with the hob. When using the oven and the hob simultaneously for one hour, the maximum temperature was 42.4 ° C. To check the temperature level obtained from the simulation, experimental studies were carried out with the use of the FLIRi3 thermal imager to determine the temperature field at the junction of the lower table with the countertop to the heating appliances. During the study oven and cooking hobs were switched on, Figure 6.

а)

b)

Fig. 6 Sample of kitchen countertop during determining its temperature: a) the sample temperature at the limit of "cooking surface - countertop" at the beginning of heating; b) sample temperature at the limit of "cooking surface - countertop" after 1 hour of oven and cooking hobs working.

The temperature values obtained experimentally (Texp = 44.4 °C) have sufficient similarity to the temperature values obtained by simulation (Tmodel = 42.4 °C). Therefore, the 126

Durability, year

following parameters were accepted for the calculations: load on the hob – 20 kg, causing internal stresses in the material σ = 0.16 MPa, temperature T = 44 °C or T = 317 K, total time of temperature and humidity influence in the kitchen furniture, given that the average daily cooking is 2 hours, then during the year this period of operation will be 30 days.

Observations on the work in the kitchens of residential buildings allowed us to establish that the average duration of the effect of temperature on the kitchen countertop is about 2 hours a day. The result of the calculation with equation (2) and (3) for MDF with different types of finishing are shown in Figures 7−9.

100 50 0

55,05 63,23 70,56 28,9 33,2 37,7 calculation with equation 2 calculation with equation 3

Thickness, mm

Durability, year

Fig. 7 Calculated durability of the kitchen countertop without finishing.

150

93,39 104,15 119,43

100 50

33,9

37,4

39,4 calculation with equation 2 calculation with equation 3

0

Thickness, mm

Durability, year

Fig. 8 Calculated durability of the painted kitchen countertop.

150 100 50 0

102,51 131,44 135,54 36,8

38,7

41,2 calculation with equation 2 calculation with equation 3

Thickness, mm Fig. 9 Calculated durability of the kitchen countertop finished by veneer.

127

The results of the calculated kitchen countertops durability meet the recommended service life of kitchen furniture, which average 15−20 years.

CONCLUSION The established technique for predicting the durability of MDF furniture products allows us to find the TAP of the material within the change of heat and humidity load.

Application of this methodology can optimize the thickness of MDF and type of surface finishing during the product design. This example of the application of the proposed method for calculating the countertop durability proves the possibility of using thin MDF for kitchen countertops, which meets current trends.

REFERENCES AYRILMIS N., LAUFENBERG L.T., WINANDY J.E. 2009. Dimensional stability and creep behavior of heat-treated exterior medium density fiberboard. In European Journal of Wood and Wood Products 67: 287–295.

BOIKO L.M., HRABAR I.G., KULMAN S.M. 2013. Durability particleboards in furniture. Osvita Ukrainy, Ukraine, 210 p.

EN 310: 1993. Wood-based panels. Determination of modulus of elasticity in bending and of bending strength.

EN 322: 1993. Wood-based panels. Determination of moisture content.

EN 323: 1993. Wood-based panels. Determination of density.

EN 325: 2012. Wood-based panels. Determination of dimensions of test pieces.

EN 326-1: 1994. Wood-based panels. Sampling, cutting and inspection. Part 1: Sampling and cutting of test pieces and expression of test results.

GERHARDS C.C. 1982. Effect of moisture content and temperature on mechanical properties of wood: An analysis of immediate effects. In Wood and Fiber Science 14(1): 4–36.

GREEN D.W., EVANS J.W. 2008a. Effect of cyclic long-term temperature exposure on bending strength of lumber. In Wood and Fiber Science 40(2): 288–300.

GREEN D.W., EVANS J.W. 2008b. The immediate effect of temperature on the modulus of elasticity of green and dry lumber. Wood and Fiber Science 40(3): 374–383.

HRABAR I.H. 2002. Thermoactivation analysis and synergetics of destruction. ZhTTI, Ukraine. 312 p.

HRABAR I.H, BOIKO L. М, KULMAN S. М. 2008. Design and method of forecasting the resource and ultimate strength of cabinet furniture. In Scientific journal. Lviv. 18.10. p. 81–89.

HU, W.G., LIU, N., XU, L., GUAN, H.Y. 2020. Study on cold/warm sensation of materials used in desktop of furniture. In Wood Research 65(3): 497–506.

KULMAN S., BOIKO L. PINCHEVSKA O., SEDLIAČIK J. 2017. Durability of wood-based panels predicted using bending strength results from accelerated treatments. In Acta Facultatis Xylologiae Zvolen, 59(2): 41−52.

MANIN V.N., HROMOV А.N. 1980. Physico-chemical resistance of polymeric materials under operating conditions. Khimiya, Russia. 248 p.

MARTENSSON A., THELANDERSSON S. 1990. Effect of moisture and mechanical loading on wooden materials. In Wood Science and Technology 24(3): 247–261.

MIRSKI, R., DERKOWSKI, A., DZIURKA, D. 2020. Construction board resistance to accelerated aging.

In BioResources 15(2): 2680–2690.

MOHEBBY B., ILBEIGHI F., KAZEMI-NAJAFI S. 2008. Influence of hydrothermal modification of fibers on some physical and mechanical properties of medium density fiberboard (MDF). In Holz als Roh- und Werkstoff 66(3): 213–218.

MORAES P.D., ROGAUME Y., BOCQUET J.F., TRIBOULOT P. 2005. Influence of the temperature on the embedding strength. In Holz als Roh- und Werkstoff 63: 297–302.

128

PIZHURIN A.A. 2004. Bases of Scientific Researches in the Woodworking. Moscow State Forest University, Moscow, Russia, 166 p.

SHI S.Q., GARDNER D. J. 2006. Hygroscopic thickness swelling rate of compression molded wood fiberboard and wood fiber/polymer composites. In Composites Part A: Applied Science and Manufacturing 37(9): 1276–1285.

RATNER S.B., YARTSEV B.P. 1992. Physical mechanics of plastics. How to predict the performance?

Moscow: Khimiya, Russia, 320 p.

Engineered Wood Products Market Share 2021 – Global Trends, Market Demand, Industry Analysis, Growth, Opportunities and Forecast 2025. https://www.marketwatch.com/press-release/ REGEL V. Р. 1980. Research on the physics of strength of composites. In Mechanics of Composite Materials 15: 684–701.

SALA, C.M., ROBLES, E., KOWALUK, G. 2020. Influence of adding offcuts and trims with a recycling approach on the properties of high-density fibrous composites. In Polymers 12(6): 1327.

SLUTSKER A.I., POLIKARPOV Y.I., VASILEVA K.V. 2002. Determination of the activation energy for complicated relaxation processes. In Physics of the Solid State 44(8): 1604–1610.

STB 1333.0-2002 «POLYMER PRODUCTS FOR BUILDING. The method of determination of longevity on an activation energy of thermal-oxidizing destruction of polymer materials». 13 p.

YU D.X., OSTMAN B.A.L. 1983. Tensile strength properties of particleboards at different temperatures and moisture contents. In Holz als Roh- und Werkstoff 41: 281–286.

YOUNG S.A., CLANCY P. 2001. Compression mechanical properties of wood at temperatures simulating fire conditions. In Fire and Materials 25(3): 83–93.

ZHOU J., HU C., HU S., YUN H., JIANG G., ZHANG S. 2012. Effects of temperature on the bending performance of wood-based panels. In BioResources 7(3): 3597–3606.

ZHURKOV S.N. 1965. Kinetic concept of the strength of solids. In International Journal of Fracture Mechanics 1(4): 311–322.

ZAVOROTNUK O.V., HOLOVACH V.M., PINCHEVSKA O.O., SIRKO Z.S. 2020. The method of predicting the durability of wood products and wood composite materials. Patent 139534–UA. ZUBAREV H.N., BOYTEMIROV F.A., HOLOVINA V.M. 2004. Wood and plastic structures. Academia, Russia. 304 p.

ACKNOWLEDGEMENTS This work was supported by the Ukrainian Ministry of Education and Science under Program No. 2201040: “The research, scientific and technological development, works for the state target programs for public order, training of scientific personnel, financial support scientific infrastructure, scientific press, scientific objects, which are national treasures, support of the State Fund for Fundamental Research”. The authors are grateful to Ministry of Education and Science of Ukrainian for financial support of this study.

This work was supported by the Slovak Research and Development Agency under the contracts No. APVV-17-0583, APVV-18-0378 and APVV-19-0269 and VEGA project No. 1/0556/19.

AUTHOR’S ADDRESS Prof. Ing. Olena Pinchevska, DrSc. Olexandra Zavorotnuk, PhD. Assos. Prof. Andriy Spirochkin, PhD. National University of Life and Environmental Sciences of Ukraine Department of Technology and Design of Wood Products Geroiv Oborony str. 15 03041 Kyiv Ukraine olenapinchevska@nubip.udu.ua

129

Prof. Ing. Ján Sedliačik, PhD. Technical University in Zvolen Department of Furniture and Wood Products T. G. Masaryka 24 960 01 Zvolen Slovakia sedliacik@tuzvo.sk Prof. Ing. Ivan Hrabar, DrSc. Polisskiy National University Department of Processes, Machines and Equipment Staryy Bulvar 7 1002 Zhytomyr Ukraine ivan-grabar@ukr.net Assoc. Prof. Rostislav Oliynyk, PhD. Kyiv National Taras Shevchenko University Geography Faculty Meteorology and Climatology Department Akademika Glushkova 2a 02000 Kyiv Ukraine rv_oliynyk@ukr.net

130

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 131−142, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.12

MISMATCH BETWEEN THE ANTHROPOMETRIC PARAMETERS AND CLASSROOM FURNITURE IN THE SLOVAK PRIMARY SCHOOLS Nadežda Langová – Sylvia Blašková – Jozef Gáborík – Denisa Lizoňová – Andrej Jurek ABSTRACT The aim of the study was to compare classroom furniture sizes in four primary schools with the anthropometric measurements of the Slovak school children in the region of Central Slovakia in order to evaluate the potential mismatch between them. Following the measure in their body dimensions, functional dimensions relating to the appropriate size, shape and ergonomic design of classroom furniture were evaluated. The measurements of 295 school children of four primary schools in Central Slovakia were carried out. The basic anthropometric measures such as stature, popliteal height, shoulder height, sitting, elbow height, sitting, thigh thickness, hip width were measured. Following the measured dimensions, the appropriateness of functional dimensions of the classroom furniture used in individual grades was evaluated. The seat height, seat to desk clearance, the seat width was evaluated. The seat to desk clearance resulted in proposing a formula for calculating the height of the storage shelf. Results indicated that seat height, which should be considered the starting point for the design of classroom furniture, the appropriate students' popliteal height was only in the case of 23% of the 6 year old pupils, the seat height was appropriate for 87.89% of 14 year old pupils. In the classrooms of the first and the second grade, according to the age of the school children, there is usually only one dimension of furniture used. Therefore, it does not meet the needs of all children.

Key words: classroom furniture, school children mismatch, anthropometry.

INTRODUCTION Classroom furniture is a very special group of furniture, which must meet various dimensional and safety requirements. It is especially due to the target users, children, attending primary schools whose body is still developing, and the body measurements change rather quickly during this period. Therefore, this period is very important, nothing must be ignored and preventing the health problems at the older age must be ensured. When there is ignored anything at the time of body development it can result in various health problems and chronic pain by study BREZIN and ANTOV (2015).

Specific workplace for a student is a sitting position at the desk, designed for writing, reading and various other activities related to the educational process. In general, a chair and a desk are considered to be an interacting set. Due to the length of sitting time during classes, a quality chair is the basis for providing optimal performance in the workplace. Slovak 131

primary school children spend about a quarter of the day in a sitting position in the workplace. It is 5 hours on average and children must be fully concentrated on work for 45 minutes and the only opportunity to stretch their body is a 10-minute break. Considering the amount of time spent in the sitting position, furniture manufacturing industry must provide the classroom furniture supporting correct sitting posture. Comfortable sitting affects our feeling of comfort and ability to concentrate. The mismatch between school children and classroom furniture is likely to result in several negative effects, such as uncomfortable body posture, pain, and ultimately (ILIEV et al. 2019, DOMLJAN 2019, BRANOWSKI et al. 2020).

Moreover, the teaching-learning process can be affected in negative way as well. Due to the situation, the most intensive interest in classroom furniture can be observed, particularly regarding the study and design of school furniture. Determining the standards dealing with classroom furniture can be considered a milestone. The furniture sizes are defined with the aim at accommodating school children with different anthropometric measurements (CASTELLUCCI et al. 2014, ALIBEGOVIĆ et al. 2020).

Quality and ergonomically balanced seating is especially important for students of the first grade at primary schools. According to a study of the State Institute of Public Health of the Slovak Republic from 2003, the biggest increase in the postural problems among children in the Slovak Republic occurs between the ages of 7 to 11. In the classrooms at the first grade in any country, the height children of the same age differ reaching varieties of 200 mm (CASTELLUCCI et al. 2010). This significant height difference must be taken into account when designing fixed or adjustable school desks and chairs.

Several scientific works are focused on the mismatch between classroom furniture sizes and anthropometric measurements of students in terms of national levels (DIANAT et al.

2013, FASULO et al. 2019, CASTELLUCCI et al. 2014, MÁCHOVÁ et al. 2019, LANGOVÁ et al.

2019). The results showed that there is a considerable mismatch between body measurements of the school children and the existing classroom furniture. The seat height, seat width and desktop height are the furniture sizes with a level of mismatch more than 52%. The levels of mismatch varied between the grades and between genders indicating the special requirements and possible difficulties. The assumption that children could use the most appropriate available size significantly improved the match indicating that the limited provision of one size per cluster of grades does not accommodate the variability of anthropometry even among children of the same age (PARCELLS et al. 1999). Fewer than 20% of students can find acceptable chair/desk combinations. Most students sit on chairs with seats that are too high or too deep and at desks that are too high. Even after controlling the body stature, girls are less likely to find fitting chairs ( LEE et al. 2018). MOLENBROEK and RAMAEKERS (1996) stated that based on the anthropometric data, every country can design fitting furniture for school children. This would require to update measures from the relevant population (age 4–20 years) including at least 40 subjects from each age group and gender.

Furthermore, the system was applied on Dutch, English, and German children (MOLENBROEK et al. 2003, PARCELLS et al. 1999) The study (CARNEIRO et al. 2017) evaluates the relationship between student anthropometry and a height-adjustable school set. The results showed that the systems of present desks used to modify the height is compatible with the height of only half of the children. A drawer attached underneath the board of the can be the reason. To increase the degree of matching, new systems were developed for the desks and chairs using an algorithmic approach. The anthropometric data of small children are different than the elder ones. Hence, classroom furniture should be designed separately for them following ergonomic criteria and concentrating on users’ comfort, adjustability, or possibility to choose set dimensions (YANTO et al. 2017, GOUVALI et al. 2006).

132

A considerable increase in stature (body height) of the population, both in adults and children, is the reason to carry out the research (HITKA et al. 2018a, HITKA et al. 2018b, RÉH et al. 2019, HITKA et al. 2020, BONENBERG et al. 2019). The aim of the study was to compare classroom furniture sizes in four primary schools with the anthropometric measurements of the Slovak school children in the region of Central Slovakia in order to evaluate the potential mismatch between them. A chair and a desk with no possibility of height adjustment is considered a school set. The fact that the Slovak Technical Standard STN EN 1729-1 (2017) defining the dimensions of classroom chairs and desks is available, but it does not deal with the set as a whole is important. Therefore, we would like to focus on designing the correct combination of the set as a whole. The obtained results seem to be relevant as they provide a scientific basis for the design and they are compatible with the anthropometric measurements of the studied population of users.

EXPERIMENTAL PART Based on available data from the years 1976 to 2011, a database of existing anthropometric measurements of school children was created. The gathered data were from the years 1976, 1980, 1985, 2001, 2009 and 2011. The data from 1976 were based on the measurements performed in Slovakia in the 70s. Relevant data used in the research were associated with children aged 7 and 8 years. The data from 1980 and 1985 followed the measurements made during the Czechoslovak Spartakiads. The data useful in the research were about children aged 6, 7, 8, 13, 14 and 15. The data from 2001 and 2011 were provided by the Regional Public Health Authority of the Slovak Republic. It was a source of data on children aged 7, 8, 13, 14 and 15. The data from 2009 were based on the measurements of children in the Bratislava region and from there we used data on children aged 7 and 8 years.

Subsequently, in 2019, our own pilot measurements were performed, which show an increase in the stature of pupils compared to years 1976 to 2011. The pilot measurements were performed considering the ethical principles with the consent of students and their parents.

Anthropometric measurements were performed in four different primary schools in the region of Central Slovakia. In total, 295 children were measured, out of which 158 were primary school children aged 6, 7, 8 and 137 were school children of the second grade 13, 14 and 15. The focus was given on these aged groups because of their most significant increase in height. Moreover, at the age of 6.7, children undergo an intensive learning process for the first time. At the age of 14, 15, their body dimensions are similar to those of adults. All data were gathered by the method of direct detection, measuring individual anthropometric characteristics of children. A detailed overview of the number of boys and girls and their age is given in Table 1.

Tab. 1 Sample distribution based on age, grade and gender.

Primary school

I grade

II grade

Age 6 7 8 13 14 15

Number of respondents in pilot measurements in 2019 Girls Boys Total 20 18 38 25 43 68 26 26 52 12 23 35 39 37 76 5 21 26

The total number of respondents in the database in the years 1976−2019 Girls Boys Total 80 78 158 205 223 428 206 206 412 132 143 275 159 157 316 125 141 266

The measurement of anthropometric characteristics was conducted according to the methodology specified in the standard STN EN ISO 7250-1 (2017). The following 133

anthropometric data (Fig.1) needs to be considered to estimate the most important furniture sizes: a) Stature (S): vertical distance between the floor and the top of the head and measured with the subject erect and looking straight ahead (Frankfort plane), b) Shoulder Height Sitting (ShH): vertical distance from subject seated surface to the acromion, c) Elbow Height Sitting (EH): taken with a 90° angle elbow flexion, as the vertical distance from the bottom of the tip of the elbow (olecranon) to the subject seated surface, d) Popliteal Height (PH): measured with 90° knee flexion, as the vertical distance from the floor or footrest and the posterior surface of the knee (popliteal surface), e) Thigh Thickness (TT): the vertical distance from the highest uncompressed point of the thigh to the subject’s seated surface, f) Hip Width (HW): the horizontal distance measured at the widest point of the hip in the sitting position, g) Knee Height (KH) – vertical distance from floor to suprapatellar in anthropometric sitting posture.

Fig. 1 Anthropometric characteristics used in this study.

In Fig. 2, the sizes of the classroom furniture of the four different schools measured are described.

Fig. 2 Illustration of present school furniture and dimensions considered in this study. Desk height (DH): vertical distance from the floor to the tip of the front edge of the board of the desk; underneath desk height (UDH): vertical distance from floor to lowest point below the drawer; seat height (SH): vertical distance from floor to middle point of the front edge of the sitting surface; seat width (SW) horizontal distance between the right and left edges of the seat.

The seat height evaluated following the measured popliteal height and the desk height, or the distance between the seat and the upper surface of the desk top evaluated following the elbow height in seating were primary evaluated parameters. To evaluate the mismatch, the methodology according to CASTELLUCCI et al. (2010) and FIDELIS et al. (2020) was used. In the case of the two-way equations, three categories were defined: high mismatch is described as a lower limit of the criterion inequality higher than furniture size and a low mismatch is described as a higher limit of the inequality lower than the furniture size. A match is when 134

furniture size is between the lower and higher limits of the criterion inequality. For the one-way equations only 2 categories or levels were defined match and mismatch (FIDELIS et al. 2020).

In the case of a high mismatch, school children will not be able to rest their feet on the floor, their legs will hang from the chair. This way the pressure on the inside of the thigh will increase and blood flow to the legs will be reduced and the legs start going numb.

Popliteal Height (PH) against Seat Height (SH) The seat height (SH) is required to be balanced in respect to the popliteal height (PH) and enabling the knee to be flexed so that the lower legs shape a greatest of 30° edge with respect to the vertical. PH ought to be higher than the SH (PARCELLS et al. 1999). The lower leg constitutes a 5–30° point with respect to the vertical and furthermore the shin-thigh edge is in the vicinity of 95 and 120°. Typically, PH does not have an esteem higher than 4 cm or 88% of the PH. PH and SH are characterized when the seat stature is either >95% or <88% of the popliteal height and it is conceivable to build up a model for SH. Correction for shoe stature (SC) may naturally vary according to culture, fashion, and country. For this work, 2 cm correction for shoe stature is incorporated to the popliteal height (CASTELLUCCI et al.

2010). This way, a match model is built up as follows (1): (𝑃𝐻 + 𝑆𝐶) ∙ cos 30° ≤ 𝑆𝐻 ≤ (𝑃𝐻 + 𝑆𝐶) ∙ cos 5°

(1)

Sitting Elbow Height (SEH) against Desk Height (DH) Various reviews (GARCIA-ACOSTA, LANGE-MORALES 2007, MOLENBROEK et al. 2003, CASTELLUCCI et al. 2010) demonstrated that the elbow height is measured as the central point for the work area stature. As the load on the spine decreases, the arms are upheld on the desk and the desk height is liable to the shoulder flexion and shoulder snatching edge which is obtained by the fifth percentile. Thus, the work area stature ought to be 3–5 cm higher than the SEH. Subsequently, a match measure is set up with a changed condition (2) that acknowledges the SEH as the most minimal stature of DH and considering that the extraordinary height of DH ought not to be higher than 5 cm over the SEH. 𝑆𝐸𝐻 ≤ 𝐷𝐻 ≤ 𝑆𝐸𝐻 + 5

(2)

Hip width against seat width To avoid discomfort and mobility restrictions, the SW should be higher than hip width (HW) (HELANDER 1997, CASTELLUCCI et al. 2010). In this case the match criterion was one-way, when the: 𝑆𝑊 > 𝐻𝑊

(3)

Thigh thickness against seat to desk clearance Seat to Desk Clearance (SDC): SDC is considered appropriate when it is higher than thigh thickness (TT) (MOLENBROEK et al. 2003). Also, CASTELLUCCI et al. (2010), GARCIAACOSTA and LANGE-MORALES (2007) proposes that the SDC should be 2 cm higher than TT. The equation for this furniture dimension is: 𝐷𝐶 > 2 + 𝑇𝑇

(4)

Furniture Sizes in the Selected Schools In the Slovak primary schools, there are seven different size groups of school desks and chairs for the first and the second grade. Chairs with yellow markings are used in the first grade of primary schools for students aged 6, 7 and 8. According to STN EN 1729-1 (2017), the height of the chair is 35 cm, the width of the chair is 32 cm and the height of the table is 59 cm with corresponding colour. These sizes are for children with a height in the range between 119–142 cm. At the second grade of primary schools, for the seventh, eighth and ninth year of study, i.e. for school children aged 13, 14 and 15, chairs with green markings are used. According to STN EN 1729-1, this colour designation corresponds to a seat height of 43 cm, a width chair of 36 cm and a desk height of 71 cm. These sizes are for school children with a height range from 146 to 176 cm.

135

Descriptive statistics was used to evaluate the data and the overview of the anthropometric characteristics of school children was provided. According to the principle of the restrictive measures, a properly designed piece of furniture should take into account the sizes associated with the dimensions of the users, at least 90 % of the population, it means those, with dimensions falling between the values of the 5th and 95th centile. Thus, the percentage of people, whom a usable space or piece of furniture will not be adjusted, will amount to 5 and 10 %, respectively (SMARDZEWSKY 2015).

RESULTS AND DISCUSSION Confirmation of the hypothesis of an increase in the body stature of school children The trend of increasing body stature in the case of boys and girls is shown in Fig. 3. The aim of the comparison was to determine the changes in the anthropometric characteristics of children during the observed period.

Fig. 3 Trend of increasing the stature of children according to the age and gender.

The average body stature of girls at the age of 6 increased by 4.29 cm, at the age of 7 years by 4.13 cm, at the age of 8 years by 0.06 cm, at the age of 13 years by 2.15 cm, at the age of 14 years by 0.43 cm and at the age of 15 years by 3.09 cm. The average body stature of boys aged 6 years increased by 5.92 cm, at age 7 years by 2.23 cm, at age 8 years by 6.23 cm, at age 13 years by 8.62 cm, at age 14 years by 5.8 cm and at the age of 15 years by 0.93 cm.

Comparing the increments of stature values, it was found out that the biggest difference in boys was at the age of 13 with an increase by up to 8.62 cm on average. Comparing the increments of body stature values in the case of girls, we found out that the biggest increase compared to the past occurred at the age of 6 years, specifically by 4.29 cm. Except for the age group of 7 years, there was an increases in body stature greater in boys than in girls. At some ages, the differences are smaller comparing to others, but the results show that there has been a clear increase in stature over the last 40 years.

Anthropometric Measurements of School children in the Selected Schools Children in lower grades are usually smaller than school children in the second grade.

Therefore, there is not recommended to design the same furniture set for both grades of the primary school. The size of classroom furniture is divided into 7 groups in STN EN 1729-1.

The eighth group marked 0 is intended for children in kindergarten. It is important to describe both groups of students in terms of anthropometry. These descriptions are given in Table 2.

136

Tab. 2 Anthropometric data of students (cm).

Age/gender/n 6 year old girls n=20

6 year old boys n=18

7 year old girls n=20

7 year old boys n=18

8 year old girls n=20

8 year old boys n=18

13 year old girls n=20

13 year old boys n=18

14 year old girls n=20

14 year old boys n=18

15 year old girls n=20

15 year old boys n=18

Descriptive statistics 5th 50th 95th mean SD 5th 50th 95th mean SD 5th 50th 95th mean SD 5th 50th 95th mean SD 5th 50th 95th mean SD 5th 50th 95th mean SD 5th 50th 95th mean SD 5th 50th 95th mean SD 5th 50th 95th mean SD 5th 50th 95th mean SD 5th 50th 95th mean SD 5th 50th 95th mean SD

Stature (cm) 116.03 124.55 132.03 124.31 5.38 120.28 126.20 131.44 125.81 3.796 121.50 128.50 135.00 128.69 5.46 118.81 127.80 138.26 128.38 5.54 123.1 132.3 144.02 132.52 6.147 126.3 133.9 147.78 134.93 6.635 152.275 165.00 167.45 162.71 5.402 160.55 169.5 180.35 170.14 7.668 151.75 165.00 172.7 164.09 7.022 160.4 172.00 185.00 171.74 7.509 160.8 162.5 173.00 165.30 5.085 163.00 175.00 187.5 176.48 8.581

Shoulder height, sitting (cm) 41.04 43.55 47.82 43.94 2.38 41.53 43.75 47.175 44.08 2.136 41.80 45.10 50.50 45.21 2.78 40.85 44.30 48.88 44.35 2.34 41.4 45.8 49.965 45.38 2.572 43.125 46.5 51.535 46.28 2.582 53.24 58.9 62.32 58.33 3.097 52.31 57.2 62.63 57.38 3.254 54.7 58.5 62.66 58.32 2.979 52.48 59.3 67.3 58.75 4.555 55.18 60.9 75.14 63.18 8.074 52.6 61.1 64.1 56.47 20.731

137

Elbow height, sitting (cm) 15.31 18.80 22.53 18.76 3.09 15.71 18.85 21.84 18.99 2.092 15.58 18.90 23.68 19.26 2.44 16.11 19.00 21.99 18.82 1.95 17.225 20.10 24.095 19.95 2.596 16.55 19,6 25.185 20.25 3.318 20.685 24.85 27.53 24.58 2.490 20.4 22.3 28.31 22.90 3.073 20.91 24.4 29.66 24.75 2.832 18.76 22.7 28.56 23.04 2.943 21.76 27.5 27.8 26.14 2.879 19.1 24.5 29.7 24.40 3.569

Knee height, sitting (cm) 36.36 40.45 43.00 40.05 2.16 38.41 40.9 43.46 40.79 1.694 38.42 42.10 45.26 42.09 2.10 38.18 41.90 45.46 41,70 3.20 39.45 43.9 48.425 43.87 3.035 40.025 44.1 47.895 44.10 2.700 49.655 53.25 56.925 53.05 2.495 52.78 56.5 61.83 56.91 3.173 48.34 53.9 57.69 53.42 3.162 52.46 57.6 60.72 57.30 2.526 52.66 54.5 57.5 54.76 1.899 51.4 58.9 63.6 58.16 4.706

Popliteal height, sitting (cm) 30.39 33.00 35.35 33.02 1.79 28.31 33.8 37.2 33.54 3.352 32.36 35.20 36.66 34,74 1.61 29.65 34.20 37.88 34.16 2.55 31.75 35.3 40.82 35.76 3.056 32.8 35.8 38.07 35.65 1.745 39.495 43.8 46.105 43.05 2.516 41.66 45.4 50.83 45.85 4.056 39.44 43.8 47.83 43.49 2.496 43.08 46.3 49.98 45.90 2.907 40.34 42.7 47.04 43.24 2.658 41.7 46.8 52.3 46.80 3.933

Mismatch between anthropometric data and classroom furniture The percentage of school children necessary for evaluating the mismatch level is given in Fig. 4. The evaluation of the seat height is in Fig. 4a, and seat to desk height according to the age of the school children is in Fig. 4b.

a) b) Fig. 4 Percentages of school children by match/mismatch level for evaluating the seat height and seat to desk height.

Seat Height: The seat height is the basic size affecting the size of the desk, so it is evaluated as the first one. Following the measured value of Popliteal Height (PH) and the equation (1) for determining the height limit of the chair, we found out that when studying 158 primary school children, the seat height of 35 cm has a match of only 48.73% children, 2.53 % used a higher seat (High mismatch) and 48.73 % used a seat that was too low (Low mismatch).

In the first grade, the seat height was appropriate for 52.11% of girls, 2.81% of girls used a higher seat and 45.07% of girls used a seat that was too low. The seat height was appropriate for 45.97% of boys, 2.29% of boys used a higher seat and 51.72% of boys used a seat that was too low.

A total of 137 (56 girls and 81 boys) second grade children said the seat height of 43 cm was appropriate for 76.64% children, 5.10 % used a higher seat (High mismatch) and 18.24 % used a seat that was too low (Low mismatch). In the second grade, the seat height was appropriate for 67.85% of girls, and 32.14% of girls used a seat that was too low. The seat height was appropriate for 82.71% of boys, 8.64% of boys used a higher seat and 8.64% of boys used a seat that was too low.

Seat to desk height: In terms of evaluating the seat to desk height, we observed that in the first grade, the seat to desk height is optimal for 45.56% of students, for 4.43% of students the height was too high (High MM) and for 47.46% of students is too low (Low MM). In the first degree, this assessment was almost identical in the case of girls as well as boys. The girls achieved a seat to desk height match rating for 47.88% of them, it was too low for 49.29% and too high for 2.81% of girls. In the seat to desk height evaluation, boys achieved a match for 43.67% of them, it was too low for 50.57% and too high for 5.76% of boys. In terms of seat to desk height evaluation, it was observed that in the second grade, the seat to desk height was optimal for 48.90% of students, for 10.22% of students the height was too high (High MM) and for 40.87% of students was too low (Low MM). Girls achieved a seat to desk height match rating for 62.50% of them, it was too high for 10.71% and too low for 26.78% of girls.

138

Boys achieved a seat to desk height match rating for 39.50% of them, it was too high for 9.87% and too low for 51.61% of boys.

Mentioned results were compared to the personal evaluation of children at the second grade of the primary school aged 13-15 years, children spend 6 hours sitting at an average.

Following the questionnaire, it was found out that, in most cases, boys aged were not satisfied with the height of the chair or desk. The height of the backrest was sufficient for them, they had enough leg room and a sufficiently large desk top. For boys aged 14 and 15 the seat height and desk did not suit them. The backrest height, the leg room and the height desk top fit them. In most cases, girls aged 13 sat on the chair for 6 hours a day and the height of chairs and desk met tier requirements. The backrest height, the leg room and the height desk top fit them. For girls aged 14 and 15 the seat height and desk did not fit them.

Fig. 6. shows a personal evaluation of the seat height and desk height of all students in the second grade.

Fig. 6 Personal evaluation of seat and desk height by children aged 13−15.

Seat width and seat to desk clearence The evaluation of the seat width and the seat to desk clearance is shown in Fig. 7.

Fig. 7 Percentage of children by match/mismatch level for evaluation seat width (a) and seat to desk clearance.

The seat width fit 96.15% of children in the first grade, while at the age of 6 and 7 it was appropriate for 100% of children. At the age of 8, in the first grade, the width of the chair fit 100% of girls and 92.30% of boys. In the second grade, the width of the chair was appropriate for 70.07% of students, of which 66.07% are girls and 72.83% are boys.

As the construction of the desk consists only of the supporting part and the worktop of the desk, the seat to desk clearance will fit all children in the first and the second grade.

School children have enough leg room and space for the movement on a chair. Following the dimensions of the desk, chair, and thigh thickness (TT) and the requirement (Eqv. 3), the equation 4 to calculate the position of the storage shelf was derived. The position of the storage shelf is not specified in the standard STN EN 1729-1 or in regulation of the Slovak Republic No. 527/2017. Therefore, based on the anthropometric characteristics of the school 139

children and the construction dimensions of the furniture set, the calculation of the height of the storage shelf was proposed using the equation 5. 𝑝1 = ℎ1 − (𝑇𝑇 + 2 + ℎ8 )

(5)

where: p1 – the height of the storage shelf, i.e. the distance between the upper surface of the desk top and the lower surface of the storage shelf (cm), h1 – desk height (cm), TT – thigh thickness (cm), h8 – seat height (cm).

CONCLUSION The aim of this paper was to evaluate the relationship between the classroom furniture and school children anthropometric characteristics studying the sample set consisting of 295 school children from primary schools in Central Slovakia. Based on our pilot measurement, we found an increase in the stature of pupils compared to the period 1976 to 2011. This increase was specific in the case of each age and gender. The major increase in stature was recorded in the case of boys aged 13, by 8.62 cm. The biggest increase in height in girls was recorded at the age of 6, specifically by 4.29 cm.

Measurements and calculations showed that the greatest mismatch at seat height was found in the first grade children, aged 6 to 8 years. In this case, their seat was too low. Also, evaluating a seat desk height showed that the furniture set used was small for children in the first and the second grade of primary school. This mismatch causes pain in the shoulders and neck as well as discomfort in the knees and inner thighs. Any pain or discomfort reduces the quality of sitting and concentration. The seat width is satisfactory. There was an increase in the percentage of children in the second grade, aged 15, whose seat width is inconvenient.

When evaluating the distances between the desktop and the seat height, which is evaluated based on the thigh thickness, there can be seen positive reaction in all students.

This is justified by the fact that school desks do not have storage space. Therefore, the equation to calculate the height of the storage shelf was proposed.

While according to the STN EN, up to 7 different sets can be used. It is necessary to follow the recommendation that there should be at least three different furniture size in the classrooms. Of course, the ideal situation could be when the children stature was marked on the desk and chairs, so that the teacher would be able to assign a table and a chair of the desired height to the student according to this basic anthropometric characteristic.

REFERENCES ALIBEGOVIĆ, A., MAČAK HADŽIOMEROVIĆ, A., PAŠALIĆ, A., DOMLJAN, D. 2020. School Furniture Ergonomics in Prevention of Pupils’ Poor Sitting Posture. In Drvna industrija, 71(1): 88−99. https://doi.org/10.5552/drvind.2020.1920.

BONENBERG, A., BRANOWSKI, B., KURCZEWSKI, P., LEWANDOWSKA, A., SYDOR, M., TORZYŃSKI, D., ZABŁOCKI, M. 2019. Designing for human use: Examples of kitchen interiors for persons with disability and elderly people. In Human Factors and Ergonomics in Manufacturing & Service Industries, 29(2): 177–186. https://doi.org/10.1002/hfm.20772.

BRANOWSKI, B., STARCZEWSKI, K., ZABŁOCKI, M., & SYDOR, M. 2020. Space for innovation in the design of furniture fasteners. In BioResources, 15(4): 8472–8495. https://doi.org/10.15376/ biores.15.4.8472-8495.

140

BRANOWSKI, B., ZABŁOCKI, M., & SYDOR, M. 2019. The Material Indices Method in the Sustainable Engineering Design Process: A Review. In Sustainability, 11(19), https://doi.org /10.3390/su11195465.

BREZIN, V., ANTOV, P. 2015. Engineering Ecology, 1st ed.; Sofia : Publishing House of the UF, 259 p.

CARNEIRO, V., GOMES, Â., RANGEL, B. 2017. Proposal for a universal measurement system for school chairs and desks for children from 6 to 10 years old. In Applied Ergonomics, 58: 372–385. https://doi.org/10.1016/j.apergo.2016.06.020.

CASTELLUCCI, H. I., AREZES, P. M., MOLENBROEK, J. F. M. 2014. Applying different equations to evaluate the level of mismatch between students and school furniture. In Applied Ergonomics, 45(4): 1123–1132. https://doi.org/10.1016/j.apergo.2014.01.012.

CASTELLUCCI, H. I., AREZES, P. M., VIVIANI, C. A. 2010. Mismatch between classroom furniture and anthropometric measures in Chilean schools. In Applied Ergonomics, 41(4): 563–568. doi:10.1016/j. apergo.2009.12.001.

CASTELLUCCI, I., AREZES, P., MOLENBROEK, J. 2014. Applied Anthropometrics in School Furniture Design: Which Criteria Should be Used for Standardization? In Proceedings of the 5th International Conference on Applied Human Factors and Ergonomics AHFE 2014, Kraków, Poland 19−23 July 2014.

DIANAT, I., KARIMI, M. A., ASL HASHEMI, A., BAHRAMPOUR, S. 2013. Classroom furniture and anthropometric characteristics of Iranian high school students: Proposed dimensions based on anthropometric data. In Applied Ergonomics, 44(1): 101–108. https://doi.org/10.1016 /j.apergo.2012.05.004 DOMLJAN, D. 2019. Enhancing school environment design by observing user's behaviour.

In Proceedins of the implementation of Wood Science in Woodworking Sector, Conference on Wood Science and Technology - ICWST 2019, Zagreb, pp. 1−9.

FASULO, L., NADDEO, A., CAPPETTI, N. 2019. A study of classroom seat (dis)comfort: Relationships between body movements, center of pressure on the seat, and lower limbs’ sensations. In Applied Ergonomics, 74: 233–240. https://doi.org/10.1016/j.apergo.2018.08.021.

FIDELIS., O. P., OGUNLADE, B., ADELAKUN S. A. 2020. Incidence of School Furniture Mismatch and Health Implications in Primary School Children in Akure, South-West Nigeria. In Journal of Occupational Therapy, Schools, & Early Intervention, 1–13. https://doi.org/10.1080 /19411243.2020.1787292.

GARCIA-ACOSTA, G., LANGE-MORALES, K. 2007. Definition of sizes for the design of school furniture for Bogota´ schools based on anthropometric criteria. In Ergonomics, 50: 1626–1642.

GOUVALI, M. K., BOUDOLOS, K. 2006. Match between school furniture dimensions and children’s anthropometry. In Applied Ergonomics, 37(6): 765–773. https://doi.org/10.1016/j.apergo.2005.11.009.

HELANDER, M. 1997. Anthropometry in workstation design. In: Helander, M. (Ed.), A Guide to the Ergonomics of Manufacturing. Taylor & Francis, London, pp. 17–28.

HITKA, M., JOŠČÁK, P., LANGOVÁ, N., KRIŠŤÁK, Ľ., BLAŠKOVÁ, S. 2018. Load-carrying capacity and the size of chair joints determined for users with a higher body weight. In Bioresources, 13(3): 6428−6443.

HITKA, M., JOŠČÁK, P., LANGOVÁ, N., KRIŠŤÁK, Ľ., BLAŠKOVÁ, S. 2018a. Load-carrying capacity and the size of chair joints determined for users with a higher body weight. In Bioresources, 13(3), DOI: 10.15376/biores.13.3.6428-6443.

HITKA, M., SEDMÁK, R., JOŠČÁK, P., LIŽBETINOVÁ, L. 2018b. Positive Secular Trend in Slovak Population Urges on Updates of Functional Dimensions of Furniture. In Sustainability, 10(10): 3474.

DOI: 10.3390/su10103474.

ILIEV, B., DOMLJAN, D., VLAOVIĆ, Z. 2019. Compliance of Preschool Chair Dimensions. In Drvna industrija, 70(2): 175−182. https://doi.org/10.5552/drvind.2019.1850.

LANGOVÁ, N., RÉH, R., IGAZ, R., KRIŠŤÁK, Ľ., HITKA, M. JOŠČÁK, P. 2019. Construction of Wood-Based Lamella for Increased Load on Seating Furniture. In Forests, 10: 525. DOI: 10.3390/f10060525.

LEE, Y., KIM, Y. M., LEE, J. H., YUN, M. H. 2018. Anthropometric mismatch between furniture height and anthropometric measurement: A case study of Korean primary schools. In International Journal of Industrial Ergonomics, 68: 260–269. https://doi.org/10.1016/j.ergon.2018.08.010.

MÁCHOVÁ, E., LANGOVÁ, N., RÉH, R., JOŠČÁK, P., KRIŠŤÁK, Ľ., HOLOUŠ, Z., IGAZ, R., HITKA, M. 2019. Effect of Moisture Content on the Load Carrying Capacity and Stiffness of Corner Woodbased and Plastic Joints. In Bioresources, 14(4): 8640−8655. DOI: 10.15376/biores.14.4.8640-8655.

141

MOLENBROEK, J., RAMAEKERS, Y. 1996. Anthropometric design of a size system for school furniture. In: Robertson, S.A. (Ed.), Proceedings of the Annual Conference of the Ergonomics Society: Contemporary Ergonomics. Taylor & Francis, London, pp. 130–135.

MOLENBROEK, J.F.M., KROON-RAMAEKERS, Y.M.T., SNIJDERS, C.J. 2003. Revision of the design of a standard for the dimensions of school furniture. In Ergonomics, 46: 681–694.

ORBORNE, D.J., 1996. Ergonomics at Work: Human Factors in Design and Development,third ed.

John Wiley & Sons, Chihester. Oxford, H.W., 1969. Anthropometric data for educational PARCELLS, C., STOMMEL, M., & HUBBARD, R. P. 1999. Mismatch of classroom furniture and student body dimensions. In Journal of Adolescent Health, 24(4): 265–273. https://doi.org/10.1016/s1054139x(98)00113-x.

RÉH, R., KRIŠŤÁK, Ľ., HITKA, M., LANGOVÁ, N., JOŠČÁK, P., ČAMBÁL, M. 2019. Analysis to Improve the Strength of Beds Due to the Excess Weight of Users in Slovakia. In Sustainability, 11(3): 624. https://doi.org/10.3390/su11030624.

SMARDZEWSKI, J. 2015.

Furniture Design.

Springer International Publishing. https://doi.org/10.1007/978-3-319-19533-9.

STN EN 1729-1 (2017). Furniture. Chairs and tables for educational institutions. Part 1: Functional dimensions.

STN EN ISO 7250-1 Basic human body measurements for technological design - Part 1: Body measurement definitions and landmarks (ISO 7250-1:2017) VYHLÁŠKA Č. 527/2007 Z. z. Vyhláška Ministerstva zdravotníctva Slovenskej republiky o podrobnostiach o požiadavkách na zariadenia pre deti a mládež.

YANTO, LU, C.W., LU, J.M. 2017. Evaluation of the Indonesian National Standard for elementary school furniture based on children’s anthropometry. In Applied Ergonomics, 62: 168–181. https://doi.org/10.1016/j.apergo.2017.03.004.

ACKNOWLEDGMENTS This work was supported by the grant agency VEGA under the project No. 1/0556/19 Lightweight wood veneers-based materials and their application in products.

AUTHORS‘ADDRESS Ing. Nadežda Langová, PhD. Ing. Silvia Blašková, PhD. Assoc. prof. Ing. Jozef Gáborík, CSc. Ing. Denisa Lizoňová, ArtD. Ing. Andrej Jurek Technical University in Zvolen Faculty of Wood Sciences and Technology Department of Furniture and Wood Products T.G. Masaryka 24 960 01 Zvolen Slovakia langova@tuzvo.sk gaborik@tuzvo.sk

142

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 143−150, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.13

UNDERSTANDINGS OF DESIGN IN CIRCUMSTANCES OF HUMANITY Elena Farkašová – René Baďura ABSTRACT We are living in the world where we should accept all sides of human life. That is reflected in daily work of designer. What are basic for such point of view? That is necessary to describe, analyze and convert to a design solution and practice. It is necessary to line-up all known information and giving these weights for understanding role or position in design research and development. The article gives brief orientation in that problem to give designers or similar other view on daily practice. The main goal of the paper is thus to point out the importance of human-centered design. The partial goals are to analyze current state of affairs of humanity in design, analysis of the concept human-centered design and presentation of good practices of human-centered design. The main method used is analysis of literature and design works along with conceptual analysis. Results of given analyses are as follows: there are several terms used for description of concepts related to human-centered design, only minority of contemporary design can be understood as human-centered, there are several good examples of inspiring ideas for the wood using in everyday human needs, despite the predominance of many other types and forms of design.

Key words: design, human-centered, accessibility, diversity, age-friendly.

INTRODUCTION “We can't work for a company that pays us; we must work for society and create important values for it. In the twenty-first century, we can no longer use humanity to serve technology – we must use technology to serve humanity" Marcel Wanders (FAIRS 2007).

The word-term human-centered began to be used in the 1980s in connection with the development variety of interactive systems. Today, human-centered design (HCD) is already a traditional approach, which is applied in various areas such as management, design or in solving general social problems. Its specificity lies in the fact that it consistently brings a human perspective into all steps in the entire problem-solving process.

The renaissance of humanity can be associated with the ambition to direct social development towards a vision of a sustainable (BAĎUROVÁ 2015), ethically and culturally advanced democratic society, based on the human rights model of functioning and on the ideas of social inclusion. In practice, this means, among other things, that the need to ensure equal opportunities for minority groups in relation to the majority, to accept the diversity and individuality of people, to strengthen the rights and active participation of all members of society in economic, social, political and cultural life. In accordance with the application of human rights principles, everyone should be able to exercise their human potential, 143

regardless of age, gender or permanent or temporary functional limitations (EU2020).

Several international, European and national documents have been adopted in support of these principles since the turn of the millennium, which are not only declare rights and freedoms but also guarantee opportunities and oversee the rights of people with different limits. Their implementation into legal regulations, norms and international standards is now somewhat more urgent in modifying the existing frameworks that determine design seriously. In period of creating the environment, products, services, information and communication systems, methods and approaches are used and take into such as account of the diversity of users, as accept the heterogeneity of human needs, demands and abilities.

In this frame context, human-centric creation is a framework (not only) of a design approach to the development of products, services, environments or systems, the aim of which is to make them universally accessible, usable and useful regarding to the requirements and needs of all users. That focus is on man, respecting its individual unique value and its personal human values. The deepening and application of knowledge about human factors - considering ergonomics, information about whole users, their limitations but also possibilities, knowledge, user techniques, etc. – that should deeply contribute to achieving this goal. Important is that basis of all steps on the way to the solution is the consistent application of the human perspective. The benefit is to improve the efficiency and effectiveness of solutions, increase user satisfaction, well-being, accessibility and sustainability, as well as prevent their possible adverse effects on individuals and society as a whole – human health, safety, performance, social incoherence or exclusion (ISO 9241210 2019).

The aim of this work is focused and presents examples of wood application used in furniture, products & interior same as wood used in combination with other materials in design creation which is based on human needs.

MATERIALS AND METHODS Human-centric work is often appreciated precisely for its ability to actively solve wide range problems of social inclusion. Today, the term Human-Centered Design (HCD) covers a range of design methods that contribute to the creation of an inclusive / universal environment and products.

After analysis of representative resources oriented on design it is possible to see that in this context, there are more related terms used. For example, like by followings CARLA B. ZOLTOWSKI, WILLIAM C. OAKES, MONICA E. CARDELLA (2013) understand human-centered design like an idea that designers keep in mind the people they are designing for.

Other similar concepts are as follows: ▪ Universal Design (e.g. STEINFELD, MAISEL 2012) ▪ Design for all, (e.g. FEO, HURTADO 2008) ▪ User-friendly design, People-Friendly Design, User Centered Design (e.g.

KUANG, FABRICANT 2019; VREDENBURG, MAO, SMITH, CAREY 2002) ▪ Inclusive Design, (e.g. KEATES, CLARKSON, HARRISON, ROBINSON 2000) ▪ Design for all ages, Life-span Design, Trans-generation Design, (e.g. STORY, MUELLER, MACE 1998; PIRKL, 1994; PULOS 1994) ▪ Barrier-free design, Accessible Design, (e.g. STORY, MUELLER, MACE 1998; HOLMES-SIEDLE 1996) The concepts mentioned above are in listed from the broadest (focusing on everyone) to those focusing on more specific criteria (age, disability). However, all of them carry similar focus on humanity in design also from ethical point of view.

144

Different terminology and content nuances are related to the cultural-geographical and historical background of their origin in different countries (ČEREŠŇOVÁ, ROLLOVÁ 2015), and the beginning of the evolution towards inclusive design can be traced back to the 1950s (HUMANCENTEREDDESIGN 2020).

In today's understanding of the outlined issues, all these strategies have a common denominator. They accept human diversity and perceive otherness as part of human nature.

Through design, they create opportunities for people to function equally, equally, freely and autonomously, thus emphasizing and accepting human dignity in the wider social context – acknowledging the value of the human being and helping to fulfill the human dimension.

The accessibility achieved by applying these design approaches can not only maximize the range of potential users of environments, products, goods, and services. It also has enormous social value, as it increases the proportion of the population that can participate fully and independently in society daily life, which is also linked to the goals of social sustainability as an integral part of sustainable development.

In that context, the importance of the designer's application not only in traditional design processes but also in team interdisciplinary dialogues and participatory design is intensifying and strongly growing, which can lead to more significant processes of social transformation.

On the other hand, it should be noted that, if ideas of social inclusion and social sustainability are a broader framework of humane-centric creation, it should also be borne in mind that in certain cases artificially induced connections of heterogeneous social forms may no longer represent development but may lead to undesirable averaging, loss of originality and individuality too.

And this also applies to design too. Even the design practice itself involves much more than just the aspect of universal applicability or accessibility. The designer must include other important aspects in his creative processes, which requires a deep understanding of the design activity. In any case, the application of any method should not lead to a reduced view of the design. A key factor in using the method is the degree of creativity which is applied in that. After all, the most creative solutions in the history of design often arose from breaking the rules or standards or by ignoring generally accepted practices or customs.

RESULTS AND APPLICATIONS 1/Project ThisAbles, 2019, IKEA Israel – Part of IKEA's long-term strategic effort is its commitment to sustainability as well as the immediate support of the communities in which it operates. The company's global vision is to improve the quality of everyday life for as many people as possible. This also applies to people with various types of functional disability or limit, which represent up to 10% of the world's population. As part of this vision, the Israeli IKEA has teamed up with domestic non-profit organizations that focus on creating solutions for people with special needs or disabilities. The ThisAbles project was aimed at enabling these people to enjoy the IKEA product range and improve their quality of life through them. The cooperation resulted in a vision of a new product line that can bridge the gap between existing products and the specific needs of minority groups. As part of the codesign, representatives of the target group - people with a specific form of health restriction – were also involved in the development. "Accessibility" was maximized in all aspects of each project stage – from determining an accessible project location, built and adapted to the needs of disabled people who participated in the project, to the products themselves, made available to the general public in the form of shared documents for 3D printing, including instructional instructions and videos for production, assembly and use. The project website 145

also became a crowdsourcing platform for the issue. The public can initiate any special problem, but also share their own ideas, which can be integrated into IKEA products and make them available to all those who need it most.

The individual products, designed primarily to help minority groups, ultimately universally expand the range of usability and accessibility for all of us – for children, the elderly or for temporarily functionally indisposed "healthy" people. For example, mirror connectors that can be integrated into storage furniture, which make the contents visible on high shelves, serve people in wheelchairs as well as children or people of short stature (THISABLES 2019).

Watch videos about the ThisAbles project by using QR above.

2/The Alternative Limb Project, design: Sophie de Oliveira Barata – The author of the project uses a unique medium of prosthetics to create wearable works of art. Her work combines the latest technologies with traditional crafts. She is examining by her design the theme of the appearance of the human body, the possibilities of its modification, evolution, and transhumanism. This project is a support platform for a positive perception of disability at the same time. It celebrates the diversity of the human body, breaking not only physical but also mental, social and cultural barriers. In her original, tailor-made works, she implements current material and technological innovations. In the creation of each product, it participates with experts in the fields of 3D modeling, electronics, and other cutting-edge technologies.

One of solutions, which has been created in frame of Alternative Limb Project mentioned before, was named as MATERIALISE (2017). That design dedicated for Kelly Knox has two parts. The underside is a realistic representation of the hand, which represents Kelly's physical body and is made of silicone skin indistinguishable from human. The upper half is designed with interchangeable parts, each of which represents different elements reflecting its emotional and spiritual selves. The limb was created by a combination of 3D printing, CNC milling, hand carving and sculpture, using steel, rock, earth, wood, moss, oil, cork, wool, bronze, rhodium and gold. The two halves of one forearm form a strong visual connection between the physical and the metaphysical (THEALTERNATIVELIMBPROJECT 2020).

3/Innovation from Matell – a step towards equality. In recent years, the manufacturer of the iconic doll has brought a multidimensional view of beauty and fashion and in its own way has reflected on the current social issues of social inclusion. Through atypical toys, it turned its attention to the adoption of various forms of human "otherness" and thus implemented issues of social diversity into the education of the youngest generation. The 2016 collection first changed the ideal of the stylized white beauty Barbie presented for many years - her new slender, fluffy, and racial mutations were introduced. In 2019, the company went even further, responding to customer requests to bring toys closer to real life. It launched several genders-neutral dolls and other atypical ones appeared, with aesthetic flaws or disabilities (ITV 2020).

4/Artro chair, 2018, design: Ondrej Bukovec, Technical University in Zvolen – Design creativity applied in marginal user groups.

That wooden chair is 146

typologically designed for interiors where a larger number of people with limited lower limb function is expected to stay. The solution reduces physical exertion – it helps to get up and promotes self-sufficiency. Seating in the sense of "senior-friendly design" integrates requirements aimed at maintaining independence and mobility. Awareness of the value of the humanitarian message is characteristic of the solution. A detail with a hint of subtle nostalgia or retro memories can be as good a support for the user's soul as a physical barrel for its body (BUKOVEC 2018).

5/Transgenerational furniture, 2014, design: Ondrej Bukovec, Technical University in Zvolen – Furniture system is supporting the development of interpersonal relationships from birth to adulthood. With a simple modification, the cot becomes a dynamic or static seat and then the application of a part that raises the legs of the furniture can also be used as a table.

The design accepts the life changes that the individual undergoes during the life of human life as well as the individual developmental stages of family life (ARCHIVE 2014).

6/Wait, 2018, design: Ivan Jedinák, Technical University in Zvolen – The concept of a sticks stand for the Elderly House in Sliač solves the problem of storing orthopedic crutches (ARCHIVE 2018).

147

7/Interior of the Ľ. Štúr library in Zvolen – department for children and youth, 2019, design: Karolína Štefániková, Technical University in Zvolen – The project respects the essence of the children's world, while the individual functional zones of the children's department of the library are sensitively conceived regarding the peculiarities of the growth periods of young people (ARCHIVE 2019).

DISCUSSION AND CONCLUSIONS HCD is part of design thinking and taking problem. That sequence is very important to accept, it should be included naturally in designer practice. HCD must be strongly connected with designer education, technical education as well. That should make interface to a daily research, development, creation, production. That human way of combining technical, environmental, social, artistic, economical and many other views related to a production of a new design is necessary for correct and socially responsible future and development trends in 21st century.

For some design products, the ambiguity of identifying the target group for which the products are intended is also felt. In many cases, designers address them in general, but as the results of real-life mapping suggest, users do not perceive similar products as necessary or useful. Based on this, it then appears that the target group of products are the designers themselves, or the design community or sometimes the media. Another observed phenomenon may be related to the aesthetic trend and publicity – although the formal features of similar products as the materials used and the modest to poor aesthetics suggest that it should be a low-cost product, in many cases it is a cost-effective product, which it can also affect the resulting interest or lack of interest of the target group, especially for example in seniors.

REFERENCES ARCHIVE. Archive of Department of Furniture and Interior design. Technical University in Zvolen, 2014–2019.

BAĎURA, R., BAĎUROVÁ, B. 2016. Dizajn a etika. 1. ed., Zvolen: TU in Zvolen, 191 p., ISBN 97880-228-2909-0.

BAĎUROVÁ. B. 2018. The potential of virtue ethics in ethical education in Slovakia. In Metodički ogledi : časopis za filozofiju odgoja. - Zagreb : Hrvatsko filozofsko društvo, 2018. ISSN 0353-765X. Vol. 25, no. 2, 2018, pp. 67–84.

BHAMRA, T., HERNANDEZ, R. 2021. Thirty years of design for sustainability: an evolution of research, policy and practice. In Design Science, Volume 7, 2021, e2, ISSN: 2053-4701

148

BUKOVEC, O. 2018. Dizajnérska kreativita aplikovaná v okrajových užívateľských skupinách.

Dissertation thesis, 2018, Zvolen: Technical University in Zvolen.

BURGSTAHLER, S. 2007. (Project, DO-IT & Publications), Washington. Universal design in education: Principles and applications. Available on: https://www.washington.edu/doit/universaldesign-education-principles-and-applications CAMBURN. B. et al. 2017. Design prototyping methods: state of the art in strategies, techniques, and guidelines. In: Design Science, Volume 3, 2017, e13, ISSN: 2053-4701 ČEREŠŇOVÁ, Z., ROLLOVÁ, L. 2015. Tvorba inkluzívneho vysokoškolského prostredia. Bratislava: STU Publishing, 176 p., 2015, ISBN: 978-80-227-4452-2, p. 46.

FARKAŠOVÁ, E. 2020. Human-centered design – stale aktuálny rámec v dizajne. In Designum, year XXVI., No. 2, p. 72–79, SCD publishing, 2020, ISSN 1335-034x.

FEO, R.; HURTADO, R. 2008. Optimastudio Diseños para Todos/Designs for AllMadrid ISBN 97884-691-3870-0 GARDNER, H. 1999. Dimenze myšlení. Teorie rozmanitých inteligencí. Prague: Portál, 1999. ISBN 80-7178-279-3.

HEDDEN, B. 2020. Consequentialism and Collective Action. In: Ethics, Volume 130, Number 4 july 2020, The University of Chicago Press Books Publ.

HOLMES-SIEDLE J. 1996. Barrier-free Design: A Manual for Building Designers and Managers, Butterworth architecture. Routledge Publ., 190p. ISBN 978-07-506-1636-2 HUMANCENTEREDDESIGN 2020. Available on: https://www.humancentereddesign.org/inclusivedesign/history ISO 9241-210:2019 Ergonomics of human-system interaction, Part 210: Human-centered design for interactive systems ITV. 2020. Available on: https://www.itv.com/news/2020-01-28/barbie-dolls-to-feature-no-hairand-vitiligo/ KANDIYALI, J. 2020. The Importance of Others: Marx on Unalienated Production. In: Ethics, Volume 130, Number 4, July 2020 The University of Chicago Press Books Publ.

KEATES, S., CLARKSON, P., J., HARRISON, L., A., ROBINSON, P. 2000. Towards a practical inclusive design approach. In: The 1st ACM Conference on Universal Usability (CUU 2000), 2000-11- to -pp. 45–52. Available on: http://web.mit.edu/16.459/Keates.pdf KUANG, F. 2019. User Friendly: How the Hidden Rules of Design are Changing the Way We Live, Work & Play. Publisher Random House, 416 p. ISBN 978-03-742-7975-2 NUSSBAUM, M. 2000. Aristotle, Politics, and Human Capabilities: A Response to Antony, Arneson, Charlesworth, and Mulgan. In Ethics, Volume 111, Number 1, The University of Chicago Press Books Publ.

PESSOA, M.V.P. 2020. Smart design engineering: leveraging product design and development to exploit the benefits from the 4th industrial revolution. In Design Science , Volume 6 , 2020 , e25 PLATZ, J. 2020. Democratic Equality and the Justification of Welfare-State Capitalism In: Ethics, Volume 131, Number 1, October 2020, The University of Chicago Press Books Publ.

POZNIC, M. et al. 2020. Designing as playing games of make-believe. In: Design Science , Volume 6 , 2020 , e10, ISSN: 2053-4701 PULOS, A. J. 1994. Forward to: Transgenerational Design: Products for an Aging Population, by James J. Pirkl. New York: Van Nostrand Nostrand. pp. viii. ISBN 0-442-01065-6 STEINFELD, E., MAISEL, J. eds. 2012. Universal Design: Creating Inclusive Environments Wiley. pp.

408 pages. ISBN 978-04-703-9913-2 STORY, M., MUELLER, J., MACE, R.L. 1998. The Universal Design File: Designing for People of All Ages and Abilities. Revised Edition. https://files.eric.ed.gov/fulltext/ED460554.pdf THEALTERNATIVELIMBPROJECT 2020. Available on: http://www.thealternativelimbproject.com/ project/materialise/ THISABLES 2019. Available on: https://thisables.com/en/about/ VERMEERSCH. P.W., HEYLIGHEN, A. 2018. Involving Blind User/Experts In Architectural Design: Conception And Use Of More-Than-Visual Design Artefacts. In CoDesign, International Journal of CoCreation in Design and the Arts, Volume 17, 2021 Issue 1, ISSN: 1571-0882

149

VREDENBURG, K., MAO, J., SMITH, P., CAREY, T. 2002. A Survey of User-Centered Design Practice.

CHI '02: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems April 2002 Pages 471–478 Available on: https://doi.org/10.1145/503376.503460 ZOLTOWSKI, C., B., OAKES, W., C., CARDELLA, M., E. 2013. Students' Ways of Experiencing Human‐Centered Design Available on: https://onlinelibrary.wiley.com/doi/abs/10.1002/j.21689830.2012.tb00040.x

AUTHORS‘ADDRESS Elena Farkašová René Baďura Technical University in Zvolen T.G. Masaryka 24 960 01 Zvolen Slovakia

150

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 151−164, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.14 pal

A COMPARISON OF THE IMPORTANCE OF THE FURNITURE MANUFACTURING IN EU COUNTRIES USING CLUSTER ANALYSIS AND HELLWIG'S METHOD Emilia Grzegorzewska – Mariana Sedliačiková – Ján Kalafús ABSTRACT The goal of the paper was to compare the importance of the furniture manufacturing industry in the European Union and indicate the groups of countries showing similarities in terms of the selected economic and production features. For this purpose, cluster analysis and multivariate linear ordering analysis were conducted. Italy, Poland and Germany were at the forefront of the list taking into account selected economic and production indicators of the furniture industry. The last places in the ranking were taken by small countries (Luxembourg, Cyprus, Malta, Greece) characterised by a low level of the furniture production and a relatively small number of enterprises. The conducted research allowed to determine the position of the furniture manufacturing industry of the EU countries as well as to identify the most relevant competitors within certain strategic groups. This may influence the national strategy of building a competitive advantage on the international market.

Keywords: furniture manufacturing, EU countries, cluster analysis, Hellwig's method.

INTRODUCTION Furniture manufacturing with the wood and pulp and paper industries belong to the woodworking sector. These are based on processing wood, i.e. domestic ecological resource (POTKÁNY et al. 2018). Wood raw material obtained in forestry is the starting link in the chain of forestry and wood products. Value-added wood products are primary wood products that have been further processed into secondary products (FOREST PRODUCTS… 2017-2018).

A significant part of them comprises furniture.

The furniture industry is an important component of the EU economy, and the EU plays a particularly significant role in the global furniture market. Approximately onequarter of the world’s furniture is manufactured within the European Union (BARBARITANO et al. 2019). The dominance of micro-firms and small medium-sized enterprises in the furniture industry is driving the furniture market growth in the region (ZION MARKET...

2018). The dynamic development of this industry has been observed over the years. In 2017, the value of global furniture production reached nearly USD 450 billion. For comparison, in 2003, it amounted to USD 223 billion, and in 2008, USD 278 billion. In just 15 years, furniture production doubled, and in the last 10 years, it has grown by over 60% (JIVKOV 2019). It is estimated that the size of the furniture market in 2019 was valued at approximately USD 609 billion (GLOBAL MARKET INSIGHT 2020). European and North American furniture manufacturing has lost considerable market share since the 1990s, 151

mostly to Asian countries. In Europe furniture manufacturing has grown most rapidly in the Czech Republic, Poland, Portugal and Romania (FOREST PRODUCTS…2017-2018).

The largest furniture manufacturers include China, USA, Germany, India, Italy and Poland (HAN et al. 2009, GRZEGORZEWSKA and STASIAK-BETLEJEWSKA 2014, JIVKOV 2019).

Additionally, apart from China, Germany, Italy and Poland are the largest furniture exporters (GRZEGORZEWSKA et al. 2020). In the face of increasing competition, the international competitive ability of wood-based industries (including furniture production) is of growing importance, which has been the subject of research in relation to various countries (HAN et al. 2009, HAJDÚCHOVÁ and HLAVÁČKOVÁ 2014, HITKA et al. 2017, MILIĆEVIĆ et al. 2017, HITKA et al. 2019).

Many factors are used to determine the development of the furniture industry. Due to the close links with forestry, geographic and natural factors are taken into account, especially resources and availability of wood raw material. The development and competitiveness of the furniture industry is influenced by economic and market conditions, e.g. demand, economic situation, fluctuations in exchange rates, competition, changes in prices of wood raw materials and other materials. Market trends and demand are closely related to drivers.

These include economic growth, urbanization, family incomes, and trends in housing and construction (FOREST PRODUCTS…2017-2018, TRACOGNA 2013, KAUFINGER 2014).

Additionally, market tendencies in the furniture industry are determined by design, demographics, and changes in consumer preferences (PAKARINEN and ASIKAINEN 2001, KOZAK 2004).

In studies concerning the development of the wood-based industry and the level of its competitiveness, the necessity to improve the production capacity, introduce innovations (CAO and HANSEN 2006, GRZEGORZEWSKA and WIĘCKOWSKA 2016, SEDLIACǏ KOVÁ et al.

2019, RATNASINGAM et al. 2018), enhance the quality of products (TOIVONEN 2012, LOUCA ̌ NOVÁ et al. 2015) and take environmental issues into account (MALÁ et al. 2018, WIĘCKOWSKA, GRZEGORZEWSKA 2019) tends to be emphasised. CHOBANOVA and POPOVA (2015) identified the challenges before the furniture industry and characterised the state of the art and trends among others in the manufacture of the furniture, consumption, import/export, availability of raw materials and components. STASIAK-BETLEJEWSKA et al.

(2020) analysed global macroeconomic trends affecting the furniture market and gave recommendations to furniture companies. OBLAK et al. (2020) analysed the European furniture market in terms of challenges, effective strategies, and key factors.

Considering variable circumstances, processes of economic globalisation and internationalisation of enterprises, particularly in export-oriented sectors, which in some European countries include the furniture industry, there is a necessity to constantly monitor the position of a given country in the European and global furniture market. This is of increasing importance because the group of countries eagerly aiming to develop this industry, and seek fresh sources of competitive advantage, are constantly growing. The goal of this paper is to compare the importance of the furniture industry in the European Union member states, indicate the groups of countries showing similarities in terms of the selected economic and production features associated with the production and operation of business entities as well as to specify the positions of particular countries in the EU market based on these variables.

METHODOLOGY The main source of statistical data was Eurostat (2020). The research concerned the furniture industry, which in the NACE Rev. 2 classification is placed in section C31 152

Production of furniture. The analysis covers years 2010–2016, because complete and reliable statistical data were obtained during this period. However, Ireland was excluded from the analysis due to the significant volume of missing data. During the first stage of the study, kmeans clustering was conducted to identify which of the analysed objects (EU member states) were the most related to each other in terms of the highlighted criteria. Indicators concerning the furniture production and economic entities functioning in the European furniture market were specified. These included furniture production value [million euro], number of enterprises [pcs.], furniture production value per 1 enterprise [thousand euro] which is the ratio of the furniture production value to the number of enterprises and the share of the production value of the furniture manufacturing in the GDP [%]. Similar economic and production indicators were also included in previous analyzes regarding the importance and position of the furniture industry (e.g. SMARDZEWSKI 2009, RATAJCZAK 2014, CHOBANOVA and POPOVA 2015, STASIAK-BETLEJEWSKA et al. 2020).

Cluster analysis is an approach used for studying similarities of observations with respect to the analysed phenomenon (FRĄTCZAK et al. 2009). It consists of finding homogeneous subsets in a heterogeneous set of objects, i.e. determining objects, which are more similar to objects forming a particular group (cluster) than to objects from outside of this group (BALICKI 2009, KISIELIŃSKA 2009). Hence, the aim of the cluster analysis was to identify the groups of EU member states, which were similar to each other in terms of particular aspects of activity in the furniture industry. Due to the diversified measuring scales of the variables, the procedure of their standardisation was implemented in order to normalise the measurement units as well as to eliminate the diversification of variables in terms of the location and variability of the studied population (BALICKI 2009). The factorial k-means analysis started with determining the number K of classes (clusters), into which the set of observations should be divided. Subsequently, the average vectors, also known as the centres of gravity, were calculated in each iteration. These were the points determining the value, within which the points included in the cluster were located. The determination of these median values enabled the classification of the objects into clusters. Every object Oi (i = 1, ..., n) was assigned into a group with the closest centre of gravity, i.e. Oi ∈ Sj where d(Oi, Mj) = d(Oi, Mj), where d refers to the Euclidean distance. A given object (country) was assigned to the cluster whose centre was the closest. In subsequent iterations, attempts were made to enhance the assignment of the objects. This was possible because the hierarchy of clusters was not previously specified, and the objects could move between groups. This means that for Sj (j = 1, ..., k), new gravity centres were calculated as arithmetic averages of all objects belonging in a particular group (FRĄTCZAK et al. 2009).

The cluster analysis was complemented by multivariate data analysis utilising linear classification methods. Considering its wide range of applications and universality in studying economic and social phenomena, the Hellwig's method was also introduced in the analysis. Initially, the degree of differentiation of the studied features was specified using the coefficient of variation. Features for which the coefficient of variation exceeded the value of 0.1 were adopted for subsequent analyses. All analysed variables met the variability criterion. Additionally, uncorrelated features were taken into account, i.e. those below the value of 0.7 (KISIELIŃSKA 2012). In the conducted analyses, the calculated coefficient of variation confirmed the required level of differentiation in the case of all features constituting the base for the synthetic measure.

Hellwig's method requires finding a potential influence of variables on the examined phenomenon.

The calculation scheme of Hellwig's method is as follows: 1.

Determining the stimulant and destimulant among the studied variables.

2.

Performing zero unitarisation for stimulants and destimulants (eq. 1 and 2).

153

3.

Finding an abstract development pattern - for stimulants, the pattern is the maximum value of a feature from a given set of objects/countries (eq. 3), and for destimulants, the pattern of development is the minimum value of a feature from a given set of objects/countries (eq. 4).

4.

Calculating the Euclidean distance of the feature value of each object/country from the adopted development pattern (eq. 5).

5.

Determining a development measure that takes into account all the examined features (eq. 6 and 7).

6.

Assigning individual objects/countries an appropriate natural number, following the criterion of the decreasing value of the development measure each year.

7.

Calculating of the arithmetic mean of Hellwig's development measure in the whole analysed period.

8.

Calculating of the final position of objects/countries in the analysed period on the basis of the calculated arithmetic mean.

Stimulants are specific features, the high values of which are desired from the viewpoint of this phenomenon, whereas low values are considered undesirable. In the case of destimulants, the opposite is true. In the conducted research, all features were stimulants.

Subsequently, the specified features were normalised, which involved assigning appropriately processed (transformed) variables to the original variables. To this end, reset unitisation was used, as it fulfils all the requirements for data normalisation. The transformations of the variables were performed according to the following formulas (KUKUŁA 2014): − for stimulants: 𝑥𝑖𝑗 −min 𝑥𝑖𝑗

𝑧𝑖𝑗 = max 𝑥 𝑖

𝑖

(1)

𝑖𝑗 −min 𝑥𝑖𝑗 𝑖

− for destimulants: max 𝑥𝑖𝑗 −𝑥𝑖𝑗

𝑧𝑖𝑗 = max𝑖𝑥 𝑖

(2)

𝑖𝑗 −min 𝑥𝑖𝑗 𝑖

Subsequently, an abstract object, i.e. a so-called development model z0j with the most optimal values for each variable as well as an anti-model z_0j with the worst values for each variable were specified. These were determined according to the following relationships (STEC 2011): 𝑧0𝑗 = max 𝑧𝑖𝑗 , when xj is a stimulant 𝑙𝑢𝑏 𝑗 = 1,2, 𝐿, 𝑚 { 𝑧0𝑗 = min 𝑧𝑖𝑗 , when xj is a destimulant

(3)

𝑧_0𝑗 = min 𝑧𝑖𝑗 , when xj is a stimulant 𝑙𝑢𝑏 𝑗 = 1,2, 𝐿, 𝑚 { 𝑧_0𝑗 = max 𝑧𝑖𝑗 , when xj is a destimulant

(4)

In the subsequent step, the similarities of the objects with respect to the best abstract object were analysed by calculating the Euclidean distance of each object from the development model (BALICKI 2009):

154

𝑝

𝑑𝑖0 = √∑𝑗=1(𝑧𝑖𝑗 − 𝑧𝑜𝑗 )2

(5)

In the next step, the so-called measures of development were determined for each object (country) according to the following formula: 𝑑 𝑚𝑖 = 1 − 𝑑𝑖0 (6) 0

where: 𝑝

𝑑0 = √∑𝑗=1(𝑧0𝑗 − 𝑧_0𝑗 )2

(7)

𝑚𝑖 – development measure for the ith object, 𝑑0 – distance between the model of development and the anti-model.

The values of the Hellwig's development measure were in the range from 0 to 1, whereby the measure of development calculated for the model was equal to one, and for the anti-model – zero. A higher value of the development measure indicated a higher level of the studied complex phenomenon. Based on the obtained values of the development measure, rankings were built separately for each year (2010–2016). Then, using the arithmetic mean, the average value of Hellwig's development measure over the entire period was calculated. All analyses were performed using the MS Excel Office 2016 and SPSS Statistics 24.0 package. In the case of studying the significance of the diagnostic variables in the cluster analysis, the inference was made at the significance level of  = 0.05.

RESULTS AND DISCUSSION The preliminary k-means clustering conducted for the EU countries showed that with three cluster centres, the convergence in the cluster centres was achieved after three iterations. The addition of one more cluster centre allowed for the convergence to be achieved after two iterations. Finally, the data were analysed and grouped into four cluster centres. There was one country in the first centre (i.e. Lithuania), 19 countries in the second centre (Austria, Belgium, Finland, France, Greece, Spain, Netherlands, Luxembourg, Portugal, Sweden, United Kingdom, Bulgaria, Croatia, Cyprus, Czech Republic, Malta, Slovakia, Slovenia, Hungary), 2 countries (Denmark, Germany) in the third and 5 countries (Italy, Estonia, Latvia, Poland, Romania) in the fourth. The summary of the affiliation of particular countries to the centres is presented in Table 1.

Tab. 1 Affiliation of EU countries to various cluster centres due to the importance of the furniture industry Country Lithuania Austria Belgium Finland France Greece Spain Netherlands Luxembourg

Cluster Distance 1 2 2 2 2 2 2 2 2

0.000 0.767 0.792 0.725 0.694 0.745 0.476 0.302 1.345

Country

Cluster Distance Country

Portugal Sweden United Kingdom Bulgaria Croatia Cyprus Czech Republic Malta Slovakia

155

2 2 2 2 2 2 2 2 2

0.604 0.812 1.133 0.822 0.910 0.862 0.661 0.718 0.698

Slovenia Hungary Denmark Germany Italy Estonia Latvia Poland Romania

Cluster

Distance

2 2 3 3 4 4 4 4 4

0.804 0.701 1.368 1.368 1.220 1.590 1.108 0.745 0.710

Due to the diversified measuring scales of the variables, the procedure of their standardisation was implemented in order to normalise the measurement units as well as to eliminate the diversification of variables in terms of the location and variability of the studied population. The conducted analysis of variance revealed that the strongest diagnostic variable was the share of the furniture production value in the GDP [%]. The second strongest variable was the furniture production value per 1 enterprise [thousand euro] and the third variable was the furniture production value production [million euro]. The number of enterprises [pcs] was determined as an unsuitable diagnostic variable differentiating particular clusters. The summary of the analysis of variance is presented in Table 2.

Tab. 2 The analysis of variance of diagnostic variables related to the furniture industry.

Cluster

Itemisation Furniture production value [million euro] Number of furniture enterprises [pcs.]

Furniture production value per enterprises [thousand euro] Share of the furniture production value in GPD [%]

Error df

F test

Significance

0.056

23

3.501

0.009

3

0.037

23

0.909

0.452

6.819

3

0.361

23

18.894

0.000

7.228

3

0.282

23

25.597

0.000

Mean square

df

Mean square

0.140

3

0.034

Table 3 shows the intra-class mean values (average values per country) of selected diagnostic variables indicating the importance of the furniture industry in those countries.

The first cluster included only Lithuania. This cluster was characterised by the lowest number of enterprises and a relatively low total production value. However, attention should be paid to the value of this production per enterprise, which was on average twice as high as the average in the EU13 countries. Moreover, Lithuania stood out as the country with the highest importance of the furniture industry for the national economy, which was confirmed by the large share of the furniture production in the GDP value. It is a relatively small country, which in terms of the production value was in the 15th place at the end of the studied period. However, the country showed the highest dynamics of the production growth in terms of value.

Tab. 3 Intra-class means of selected variables characterising the furniture industry in the EU countries Itemisation Furniture production value [million euro] Number of furniture enterprises [pcs.]

Furniture production value per enterprises [thousand euro] Share of the furniture production value in GPD [%]

Cluster 1 1065.3

2 2013.4

3 10651.1

4 5974.7

1468.0

4000.1

4825.5

9057.2

725.4

538.2

3103.1

585.5

3.21

0.51

0.72

1.50

The second cluster comprised as many as 19 EU countries, including those classified as new and old member states. These countries showed on average the lowest share of the furniture industry in creating GDP – lower than the EU average in the studied period.

Furthermore, the cluster was characterised by a relatively low production value per enterprise. It should be noted that this cluster included countries with a low as well as a relatively high level of economic development.

156

The third cluster included two countries – Denmark and Germany, which demonstrated the highest total value of the furniture industry production as well as per enterprise. In addition, these countries are characterised by high labour productivity. In contrast, the furniture industry was less important for the national economies of these countries than in the case of Lithuania and the countries belonging to the last cluster. Moreover, Germany is one of the world's largest producers of furniture and wood-based panels.

Five countries were included in the fourth cluster, which was characterised by a relatively high share of the value of the furniture production in generating GDP. This indicated that the furniture industry in these countries was very important to their national economies. This cluster was also distinguished by the significant furniture production value.

However, due to the large number of economic entities operating in the industry, the furniture production value per enterprise placed the countries below the EU average.

Thanks to the cluster analysis, countries were identified that are similar to each other in terms of selected economic and production factors. This means that it is possible to identify countries that may be the biggest competitors in the European furniture market.

Countries with similar production capacities and a similar number of enterprises should first look for sources of advantage within their own strategic group. Small countries, where the furniture industry is not strongly developed, have limited opportunities to compete with the largest furniture manufacturers.

Cluster analysis showed which countries are in clusters and therefore are similar to each other. However, this method did not provide information on the position of individual countries. Hence, it was important to develop a ranking of countries taking into account the importance of their furniture sector on the EU market.

Tables 4 and 5 show that Italy is the leader in the ranking of the EU member states, which was developed using the Hellwig's method based on the indicators concerning the production activity of enterprises. The position of this country did not change in the analysed period. However, the measure of Hellwig's development pattern slightly decreased from 0.519 to 0.459. The Italian economy recorded the highest values of the furniture production and the largest number of economic entities. Considering the level of production per enterprise, Italy was also at the top of the list, and the importance of this industry for the national economy was twice as high as the EU average.

In the list of key production indicators, Poland ranked a high, second place. The measure of Hellwig's development pattern was at a similar level in years 2010-2016 (coefficient of variation V = 4.23%) and at the end of the analysed period, it was equal to 0.426. Poland moved up from the 6th to the 4th place in terms of the furniture production value. At the same time, similarly to the case of Italy, a relatively large number of economic entities actively operating in the market caused a decrease in production per enterprise. In turn, the importance of the furniture production for the national economy was confirmed by the relatively large share of the value of this production in generating the GDP. This indicator was three times higher than the average in the EU countries.

Somewhat different trends were noted in the case of Germany in 2016, where a slight increase in the value of the Hellwig's index was observed (i.e. to a level of 0.387).

Furthermore, a relatively high value of the furniture industry production and the value of that production per enterprise were observed. In addition, the importance of the furniture industry for the national economy was not as evident as in the case of the previous countries, which was confirmed by the share of furniture production in the aggregate value of goods and services, which was at the level of approx. 0.7%.

In 2010–2016, Italy, Poland and Germany were at the top of the ranking, and the positions of these countries remained unchanged. The distinctive role of these countries in the furniture market was emphasised, among others, by HAN et al. (2009), ZHELEV (2013), 157

GRZEGORZEWSKA and STASIAK-BETLEJEWSKA (2014), VU et al. (2019) and STASIAKBETLEJEWSKA et al. (2020). These countries show the highest production value among the EU states, and for years have been in the group of the largest furniture exporters, behind the leader, China. However, it should be pointed out that Poland and Italy are relevant net exporters of the wood-based industry, while in the case of Germany, for years there has been a surplus in the value of furniture imports over exports (DIETER and ENGLERT 2007, VU et al. 2019). Additionally, the Polish furniture manufacturers and exporters exhibited a relatively large international competitiveness in the foreign trade in relation to Italy and Germany (GRZEGORZEWSKA et al. 2020). This is also due to the fact that in Poland furniture is wood products of special importance because of its high value added and huge share within foreign trade. The Polish furniture industry generated the highest value added amongst the wood industries (RATAJCZAK 2014). The most important challenges for the Polish furniture industry include maintaining the current level and dynamics of furniture sales to Western European countries and the development of the potential of Eastern markets, as well as the sale of furniture under own brands. According to STASIAK-BETLEJEWSKA et al. 2020, this requires greater control in the distribution of products to the end customer and much greater involvement in marketing activities on foreign markets.

Tab. 4 Measure of the Hellwig's development pattern for EU countries according to selected criteria in 2010–2016 Country 2010 Italy 0.519 Poland 0.434 Germany 0.386 Lithuania 0.273 United Kingdom 0.249 Estonia 0.247 Spain 0.298 Denmark 0.285 France 0.232 Austria 0.228 Romania 0.227 Sweden 0.197 Netherlands 0.179 Portugal 0.192 Czech Republic 0.181 Belgium 0.156 Latvia 0.158 Slovakia 0.164 Croatia 0.156 Slovenia 0.182 Bulgaria 0.152 Finland 0.125 Greece 0.161 Hungary 0.099 Malta 0.140 Cyprus 0.107 Luxembourg 0.015 *SD- standard deviation **V- coefficient of variation

2011 0.517 0.409 0.404 0.285 0.270 0.259 0.294 0.262 0.240 0.224 0.224 0.195 0.186 0.193 0.167 0.161 0.148 0.145 0.150 0.171 0.144 0.129 0.154 0.093 0.104 0.106 0.018

2012 0.497 0.432 0.405 0.299 0.249 0.258 0.268 0.254 0.243 0.225 0.214 0.207 0.180 0.178 0.171 0.160 0.140 0.139 0.148 0.150 0.140 0.135 0.138 0.093 0.087 0.087 0.021

2013 0.465 0.382 0.403 0.307 0.250 0.244 0.235 0.235 0.229 0.207 0.203 0.193 0.172 0.157 0.159 0.143 0.136 0.124 0.139 0.124 0.134 0.127 0.093 0.085 0.068 0.072 0.023

2014 0.472 0.414 0.392 0.306 0.256 0.245 0.218 0.232 0.234 0.206 0.209 0.190 0.174 0.161 0.152 0.150 0.145 0.126 0.133 0.113 0.132 0.124 0.067 0.084 0.078 0.067 0.025

2015 0.460 0.416 0.385 0.315 0.276 0.226 0.211 0.214 0.228 0.193 0.203 0.170 0.167 0.157 0.143 0.140 0.142 0.136 0.117 0.111 0.126 0.117 0.066 0.081 0.073 0.055 0.028

2016 Mean 0.459 0.484 0.426 0.416 0.387 0.394 0.313 0.300 0.295 0.263 0.238 0.245 0.208 0.248 0.216 0.243 0.221 0.232 0.187 0.210 0.203 0.212 0.167 0.186 0.173 0.176 0.163 0.172 0.140 0.157 0.148 0.151 0.140 0.144 0.137 0.143 0.124 0.138 0.105 0.136 0.124 0.136 0.114 0.124 0.074 0.108 0.084 0.089 0.073 0.089 0.049 0.078 0.027 0.023

SD* 0.026 0.018 0.009 0.016 0.017 0.012 0.039 0.026 0.008 0.016 0.010 0.014 0.006 0.016 0.012 0.008 0.007 0.025 0.014 0.031 0.010 0.007 0.042 0.007 0.026 0.023 0.005

V** 5.46 4.23 2.28 5.18 6.62 4.70 15.63 10.54 3.23 7.73 4.79 7.59 3.60 9.37 7.56 5.48 5.02 17.24 10.32 22.76 7.37 5.61 39.20 7.67 28.98 29.70 21.21

In the analysed period, Lithuania was also ranked high in the ranking of the Hellwig's development pattern. The measure of the development pattern in the considered period increased from 0.273 to 0.313. It should be emphasised that Lithuania advanced by two 158

places compared to 2010. It is the only country among the EU member states that was alone in a cluster. Although Lithuania was characterised by a relatively low furniture production value, the level of this production per enterprise placed the country high in the ranking. In addition, it should be emphasised that the Lithuanian furniture industry is of great importance for the domestic economy. During the studied period, this indicator increased from 2.4% to 3.8%. Moreover, the share of the furniture industry in the total industrial production value increased from 5.8% to 7.9%. These were the highest values among the group of EU member states. As mentioned by KALĖDIENĖ et al. (2010), in Lithuania, the wood manufacturing and furniture industry are very old and traditional industries. This is because the processing of wood in Lithuania has favourable conditions, nearly a third of Lithuania‘s territory is covered with forests. Quite cheap labour force and high qualifications are big competitive advantages for Lithuania in foreign markets.

Tab. 5 Positions of EU countries in the Hellwig's ranking in 2010–2016.

Country Italy Poland Germany Lithuania United Kingdom Estonia Spain Denmark France Austria Romania Sweden Netherlands Portugal Czech Republic Belgium Latvia Slovakia Croatia Slovenia Bulgaria Finland Greece Hungary Malta Cyprus Luxembourg

2010 1 2 3 6 7 8 4 5 9 10 11 12 16 13 15 20 19 17 21 14 22 24 18 26 23 25 27

2011 1 2 3 5 6 8 4 7 9 10 11 12 14 13 16 17 20 21 19 15 22 23 18 26 25 24 27

2012 1 2 3 4 8 6 5 7 9 10 11 12 13 14 15 16 20 21 18 17 19 23 22 24 26 25 27

2013 1 3 2 4 5 6 7 8 9 10 11 12 13 15 14 16 18 21 17 22 19 20 23 24 26 25 27

2014 1 2 3 4 5 6 9 8 7 11 10 12 13 14 15 16 17 20 18 22 19 21 26 23 24 25 27

2015 1 2 3 4 5 7 9 8 6 11 10 12 13 14 15 17 16 18 20 22 19 21 25 23 24 26 27

2016 1 2 3 4 5 6 9 8 7 11 10 13 12 14 16 15 17 18 19 22 20 21 24 23 25 26 27

Mean 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

Change 0 0 0 2 2 2 -5 -3 2 -1 1 0 4 -1 0 5 2 -1 2 -8 2 3 -6 3 -2 -1 0

The top ten countries in the ranking included both old and new member states, i.e.

United Kingdom, Estonia, Spain, Denmark, France, Austria and Romania. In this group, it is possible to identify countries belonging to EU-15 (United Kingdom and France), which are characterised by a significant production of furniture and a relatively high number of enterprises; however, their furniture industries are two times less important for the national economies than the EU average. Nevertheless, the top ten also included smaller EU-13 countries, such as Estonia or Romania, where the share of the furniture industry in generating GDP was twice as high as the EU average. In recent years, the furniture production and export in Romania have increased significantly, primarily due to industry restructuring and large investments in new technologies (BURJA and MĂRGINEAN 2013). The furniture industry also plays a relevant role in the Romanian economy, due to its great impact on the 159

commercial balance through the generated financial flows. However, producers from Romania have to still invest to become more competitive, so that they can increase their profits and raise the exports (PUIU et al. 2012). Research, innovation, and development are neglected in a Romanian furniture sector that is activity-dependent on creation, development, and innovation, as these elements may improve performance (MARINESCU 2008).

During the considered period, as many as 12 countries improved their position in the ranking, which was prepared based on the Hellwig's method. The largest improvements were observed in the case of Belgium (5 places higher) and the Netherlands (4 places higher); however, these countries were still placed in the middle of the ranking, taking the 13th and 16th place, respectively. On the other hand, a decrease in the importance of selected features concerning the production activity of furniture enterprises was recorded for Slovenia (by 8 places) and Greece (by 6 places). The last places in the ranking were occupied by small countries (i.e. Luxembourg, Cyprus, Malta), which showed a low level in the case of all examined diagnostic variables. These are the countries characterised by a low level of production in the furniture industry, which is accompanied by a relatively small number of enterprises. The furniture production does not play a significant role in the economies of these countries, which was confirmed by the low share of the furniture industry in creating the value of industrial production and GDP. This group also showed the highest variability of the Hellwig index in the studied period (V>20%).

The methods used in the research have some limitations. K-means cluster analysis requires the number of groups to be defined, although it is usually not known how many groups there are in the set being processed. Moreover, the starting centroids are chosen at random while their selection has a decisive influence on the quality of the resulting clustering. Despite many disadvantages, it is still one of the most used iterative methods since it is easy to implement. Likewise, among the methods of linear ordering, Hellwig's method is one of the most used. However, this method may turn out to be not very objective due to the relative nature of many variables and the need to select variables for analysis. In futures studies, an attempt can be made to use non-hierarchical iterative methods, as well as the method of standardised sums and sum (or mean) ranks. However, in the latter, the use of continuous variables in the research means that the natural numbers assigned to the objects do not reflect the real differences between the objects.

CONCLUSION The furniture industry is an important component of the EU economy, and the EU plays a special role in the global furniture market. The EU largest furniture manufacturers include Germany, Italy and Poland. However, the importance of the furniture industry for the national economies of individual countries as well as their role in the EU and global markets is diversified. Competition in the furniture market also comes from smaller EU 13 countries, where the furniture industry is of great importance to the national economy.

The conducted research allowed to determine the position and economic and production importance of the furniture manufacturing industry in the EU countries, as well as to identify the most relevant competitors within certain strategic groups. This is important information for representatives of the furniture manufacturing industry and government that may influence the national strategy of building a competitive advantage on the international market.

Knowledge of the most important competitors in the EU and global market helps in the search for sources of competitive advantage. One of the major challenges of most EU countries is to increase their role in the global furniture market. It is not only about 160

maintaining the current level and dynamics of furniture sales. In the case of the EU13 countries, it is also important to develop the sale of furniture under producers of private labels, which requires a greater involvement in marketing activities. It is also important to emphasise strong traditions, use of modern design, and emphasise ecology and sustainable development. In this regard, it is worth following good practices applied, inter alia, by Italian furniture manufacturers. Hence, it is important to build a development path for the furniture industry and prepare a development strategy at the national level.

European countries, especially EU13, should strengthen their business model based on added value for the consumer. It is important to assess the market opportunities of individual countries and anticipate potential threats, and consequently, reduce the risk of wrong decisions in order to cope better in highly competitive markets. However, the effective implementation of the strategy at the national economic level requires a thorough knowledge of internal and external factors that determine the development opportunities of the industry.

The conducted research makes it possible to determine the patterns of competition in the sector and to further analyse the environment, including actual competitors.

In subsequent research stages, it is also essential to take into account the labour productivity, labour costs as well as foreign trade as important sources of building a competitive advantage in the global and European furniture manufacturing market.

REFERENCES BALICKI, A. 2009. Statistical dimensional analysis and its socio-economic applications (Statystyczna analiza wymiarowa i jej zastosowania społeczno-ekonomiczne), Wydawnictwo Uniwersytetu Gdańskiego, 2009, Gdańsk. ISBN: 978-83-7326-654-4 BARBARITANO, M., BRAVI, L., SAVELLI, E. 2019. Sustainability and quality management in the Italian luxury furniture sector: A circular economy perspective. In Sustainability, 2019, 11(11): 3089. DOI: 10.3390/su11113089 BURJA, V., MAR ̆ GINEAN, R. 2013. The furniture industry in Romania and the European Union - A comparative approach. In Revista Economica, 2013, 65(4): 107−120.

CAO, X., HANSEN, E.N. 2006. Innovation in China’s Furniture Industry. In Forest Products Journal, 2006, 46(11/12): 33−42.

CHOBANOVA, R., POPOVA, R. 2015. Furniture manufacturing challenges on the world market: the Bulgaria's case. In 8-th International Scientific Conference Wood Processing and Furniture Manufacturing Challenges on the World Market, WoodEMA International Association for Economics, Management, Marketing, Quality and Human Resources in Forestry and Forest Based Industry: pp.47−56.

DIETER, M., ENGLERT, H. 2007. Competitiveness in the global forest industry sector: An empirical study with special emphasis on Germany. In European Journal of Forest Research, 2007, 126: 401−412. DOI: 10.1007/S10342-006-0159-X. EUROSTAT 2020. Annual detailed enterprise statistics for industry (NACE Rev. 2) (sbs_na_ind_r2) [online]. [cit. 2020-01-20] Available online: <http://appsso.eurostat.ec.europa.eu/nui/show.do? dataset=sbs_na_ind_r2&lang=en>.

FOREST PRODUCTS ANNUAL MARKET REVIEW 2017−2018, UNECE/FAO, https://unece.org/DAM/timber/publications/FPAMR2018.pdf FRĄTCZAK. E., GOLATA, E., KLIMANEK, T., PTAK-CHMIELEWSKA, A., PĘCZKOWSKI M. 2009.

Multivariate statistical analysis. Theory and examples of applications with the SAS system (Wielowymiarowa analiza statystyczna. Teoria i przykłady zastosowań z systemem SAS).

Warszawa: Oficyna Wydawnicza Szkoły Głównej Handlowej w Warszawie, 2009.

GLOBAL MARKET INSIGHT 2020. Industry treds – Furniture market[online]. [cit. 2020-09-29] Available online: <https://www.gminsights.com/industry-analysis/furniture-market> GRZEGORZEWSKA, E., SEDLIAČIKOVÁ, M., DRÁBEK, J. BEHÚN, M. 2020. Evaluating the international competitiveness of Polish furniture manufacturing industry in comparison to the

161

selected EU countries. In Acta Facultatis Xylologiae Zvolen, 2020, 62(2): 149−164. DOI: 10.17423/AFX.2020.62.2.14.

GRZEGORZEWSKA, E., STASIAK-BETLEJEWSKA, R. 2014. The influence of global crisis on financial liquidity and changes in corporate debt of the furniture sector in Poland. In Drvna Industrija, 2014, 65(4): 315−322. (DOI: 10.5552/drind.2014.1342).

GRZEGORZEWSKA, E., WIEC ̨ KOWSKA, M. 2016. Selected aspects of innovation in the furniture industry – Empirical research findings. In Drewno, 2016, 59(198): 147−161. DOI: 10.12841/wood.1644-3985.175.12.

HAJDÚCHOVÁ, I., HLAVÁČKOVÁ, P. 2014. Impact of global economy of forest-based industry in the Czech and Slovak Republic. In Acta Facultatis Xylologiae Zvolen, 2014, 56(2): 135−146.

HAN, X., WEN, Y., KANT, S. 2009. The global competitiveness of the Chinese wooden furniture industry. In Forest Policy and Economics, 11(8): 561−569. DOI: 10.1016/J.FORPOL.2009.07.006.

HITKA, M., LORINCOVÁ, S., LIŽBETINOVÁ, L., BARTÁKOVÁ, G. P., MERKOVÁ, M. 2017. Cluster analysis used as the strategic advantage of human resource management in small and medium-sized enterprises in the wood-processing industry. In BioResources, 2017, 12(4): 7884−7897. DOI: 10.15376/BIORES.12.4.7884-7897.

HITKA, M., LORINCOVÁ, S., GEJDOŠ, M., KLARIĆ K., WEBEROVÁ, D. 2019. Management approach to motivation of white-collar employees in forest enterprises. In BioResources, 2019, 14(3): 5488−5505. DOI: 10.15376/biores.14.3.5488-5505.

JIVKOV, V. 2019. State and Trends in the development of the world, European and Bulgarian Furniture Industries. In IX International Scientific and Technical Conference Innovations In Forest industry and engineering design – INNO 2018, Innovation in Woodworking Industry and Engineering Design, 1/2019 (15): 7−16.

KAUFINGER, G.G. 2014. Macroeconomic factors affecting US retail furniture/home furnishings industry sales. In Research in Business and Economics Journal, 2014, 10: 1−10.

KALĖDIENĖ, L., LUBYTĖ, E., KARLAVIČIENĖ, V. 2010. Wood manufacturing and furniture industry in Lithuania. In Linnaeus ECO-TECH ́10, November 22−24, Kalmar, Sweden, 2010, pp. 234−246.

KISIELIŃSKA, J. 2012. Basics of econometrics in Excel (Podstawy ekonometrii w Excelu).

Warszawa: Wydawnictwo SGGW w Warszawie, 2012. ISBN 978-8375833669.

KOZAK, R.A., COHEN, D.H., LERNER, J., BULL, G.Q. 2004. Western Canadian consumer attitudes towards certified value-added wood products: an exploratory assessment. In Forest Products Journal, 2004, 54(9): 21−24.

KUKUŁA, K. 2014. Zero unitarisation metod as a tool in ranking research. In Economic Science for Rural Development, 2014, 36: 95−100.

LOUCA ̌ NOVÁ, E., KALAMÁROVÁ, M., PAROBEK, J. 2015. Konkurencieschopnosť produktov dreva z pohľadu použitého materiálu. In Acta Facultatis Xylologiae Zvolen, 2015, 57(2): 155−163.

MALÁ, D., SEDLIAČIKOVÁ, M., BENČIKOVÁ, D. 2018. How customer of small and medium woodprocessing Slovak enterprises perceive a green product. In BioResources, 2018, 13(1):1930−1950.

DOI: 10.15376/biores.13.1.1930-1950.

MARINESCU, N. 2008. Romanian furniture exports in the context of European integration. In Pro Ligno, 2008, 4(2): 51−56.

MILIĆEVIĆ, S., NIKOLIĆ, M., CVETANOVIĆ, S. 2017. The competitiveness of wood processing industry in the Republic of Serbia during the period 1995-2015. In Industrija, 2017, 45(3): 131−150.

OBLAK, L., AYRILMIS, N., KITEK KUZMAN, M. 2020. The European furniture industry: market, design and trends. In Sustainability of forest-based industries in the global economy. Proceedings of scientific papers, Vinkovici, Croatia 2020, pp. 113−116.

PAKARINEN, T.J., ASIKAINEN, A.T. 2001. Consumer segments for wooden household furniture. In Holz als Roh- und Werkstoff, 2001, 59(3): 217−227.

PUIU, S., BUDICĂ, B.A., OCHETAN, D. 2012. The export of furniture from Romania at present. In Management & Marketing, 2012, 10(2): 393−400.

POTKÁNY, M., GEJDOŠ, M., DEBNÁR, M. 2018. Sustainable Innovation Approach for Wood Quality Evaluation in Green Business. In Sustainability, 2018, 10(9): 2984.

162

RATAJCZAK, E. 2014. Polish furniture industry in the light of the concept of smart specialisation. In Annals of Warsaw University of Life Sciences – SGGW. Forestry and Wood Technology, 2014, 86:193−203.

RATNASINGAM, J., CHIN, K.A., LATIB, H., SUBRAMANIAM, H., KHOO, A. 2018. Innovation in the Malaysian Furniture Industry: Drivers and Challenges. In BioResouces, 2018, 13(3): 5254−5270.

SEDLIAČIKOVÁ, M., STROKOVÁ, Z., DRÁBEK, J., MALÁ, D. 2019. Controlling implementation: What are the benefits and barriers for employees of wood-processing enterprises? In Acta Facultatis Xylologiae Zvolen, 2019, 61(2): 163−173. DOI: 10.17423/afx.2019.61.2.15.

SMARDZEWSKI, J. 2009. The Polish furniture industry – a vision of the future. In. Drewno, 2009, 52(182): 103−113.

STASIAK - BETLEJEWSKA, R., GRZEGORZEWSKA, E. 2020. Directions of development of the furniture industry in Poland based on trend analysis and market trends. In Sustainability of forestbased industries in the global economy. Proceedings of scientific papers, Vinkovici, Croatia 2020, pp. 123−127.

STEC, M. 2011. Development conditions of voivodships in Poland - statistical and econometric analysis (Uwarunkowania rozwojowe województw w Polsce – analiza statystycznoekonometryczna). In Social Inequalities and Economic Growth (Nierówności Społeczne a Wzrost Gospodarczy), 2011, 20: 235−238.

TOIVONEN, R.M. 2012. Product quality and value from consumer perspective: An application to wooden products. In Journal of Forest Economics, 18: 157–173.

TRACOGNA, A. 2008. The furniture industry in the United States and Canada, Centre for Industrial Studies.

VU, T.T.H., TIAN, G., KHAN, N., ZADA, M., ZHANG, B., NGUYEN, T.V. 2019. Evaluating the international competitiveness of Vietnam wood processing industry by combining the variation coefficient and the entropy method. In Forests, 2019, 10, 901.

WIEC ̨ KOWSKA, M., GRZEGORZEWSKA, E. 2019. The industrial significance of new technology in the process of asymmetrical veneering of wood-based composites, In Drewno, 62(204): 157–169.

DOI: 10.12841/wood.1644-3985.297.03.

ZHELEV, P. 2013. Analysis of the international competitiveness of the Bulgarian furniture industry.

In Trakia Journal of Sciences, 2013, 11: 227−236.

ZION MARKET RESEARCH REPORT 2018. Furniture market by type, by material and by application (residential and commercial): global industry perspective, comprehensive analysis and forecast, 2017–2024.

Available online: https://www.globenewswire.com/newsrelease/2018/09/07/1567975/0/en/Global-Furniture- Market-Will-Reach-USD-472-30-Billion-By2024-Zion-Market-Research.html (accessed on 20 January 2021).

ACKNOWLEDGEMENT The authors are grateful for the support of the National Science Centre Poland, Grant No. 2019/03/X/HS4/01342, Slovak project agency APVV – projects APVV-18-0520, APVV-18-0378, APVV-17-0456 and APVV-17-0583, agency KEGA – project KEGA 005TU Z-4/2020 and project LignoPro - ITMS: 313011T720.

ADDRESSES OF THE AUTHORS Emilia Grzegorzewska, D.Sc. Warsaw Univerisity of Life Sciences – SGGW Department of Technology and Enterpreneurship in Wood Industry Nowoursynowska 159 02-776 Warsaw Poland emilia_grzegorzewska@sggw.edu.pl ORCID ID 0000-0002-7532-9287

163

doc. Ing. Mariana Sedliačiková, PhD. Technical University in Zvolen Department of Economics, Management and Business T. G. Masaryka 24 960 01 Zvolen Slovakia sedliacikova@tuzvo.sk ORCID ID 0000-0002-4460-2818 Ing. Ján Kalafús Technical University in Zvolen Department of Economics, Management and Business T. G. Masaryka 24 960 01 Zvolen Slovakia xkalafus@is.tuzvo.sk

164

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(1): 165−178, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.1.15

INFLUENCE OF ASPECTS OF CHANGE MANAGEMENT ON THE PERFORMANCE OF ENTERPRISES IN THE WOOD PROCESSING INDUSTRY Andrea Sujová – Ľubica Simanová ABSTRACT The ability to change and the agility of the enterprise are becoming competitive advantages, which are currently replacing the classic ones. The aim of the article is to present a summary of scientific knowledge in the field of change management and partial results of research in enterprises in the wood processing industry. The research was focused on specific areas and types of implemented changes in enterprises in the wood processing industry (WPI), at the levels of process optimization, as well as on the reasons or impulses that led enterprises to implement change. These aspects of change management were examined in relation to the level of performance presented by the return on equity ROE and their impact was evaluated using statistical methods such as contingency tables, Chi-square test and Cramer's contingency coefficient. The results showed that WPI enterprises, which implemented several types of changes and especially a complex transformational change, which included a change in the system and management methods, achieved higher levels of performance. The impulse to implement changes in these enterprises was based on changes in customer requirements and increased pressure from competitors.

Key words: corporate performance, change management, process, wood processing industry.

INTRODUCTION Change is an integral part of humanity. Effective change management is therefore an important and necessary precondition for the survival of enterprises in a predatory competitive environment. Active change management should thus contribute to the optimization of the processes performed, to the correct use of resources and to the correct management of the enterprise. The factor of success or failure of change is not only its management, but also the ability to involve people in the process of change, because change should become a daily part of the work of employees of the enterprise. Change is an integral part of the existence and functioning of enterprises. Capacity building and the creation of preconditions for the implementation of change are currently considered a strategic necessity. The future of an organization depends on the ability of its managers to manage change. Successful managers must understand the need to implement change and take their implementation as an integral part of their responsibilities (MIKUŠ 2010). According to the authors KUBÍČKOVÁ and RAIS (2012), the changes can be divided into three groups. The authors divide changes into incremental changes, transformational changes, and changes 165

based on a combination of the previous two methods. Incremental changes (incremental, gradual) are especially suitable for a stable economic environment, where only some of the specified parameters of the enterprise are "fine-tuned". Radical change (transformation, jumping), which involve a substantial interference in the enterprise. These changes are suitable for a turbulent economic environment. Factors promoting the need for change in an organisation include globalisation, uncertain economic conditions and diversity in the workforce among others (MULLINS 2013). It has been reported that 38% would leave their comfort zone and go for change (MURPHY 2016). The way of interpreting the idea of the change creates positive emotional reactions and consequently leads to desired outcomes (MURPHY 2016). However, the majority of people prefer to maintain the status quo. Firstly, employees for example, get complacent with the status quo at work when everything is good.

Change would interfere with their autonomy as people feel they lose control over their territory (KANTER 2012). Change is a process that must be consistent with the organisation's goals. The basis of effective change processes is a systemic view of the organization. Change is not an event but a process; it is an ever-present feature of organizational life, both at an operational and strategic level (RICK 2013). Change requires a systematic approach from both the perspective of an organization and at individual level (PITAGORSKY 2011). Change takes place when organizations introduce new projects and initiatives. These changes impact on organizational structure, systems, processes and job roles. The success of this change is the result of individuals doing their work differently (GALANTI, NDIAYE and ST-HILAIRE 2012). The ever increasing pace of change requires organizations to develop dynamic, competitive change management strategies on an ongoing basis.

Change management can be defined as a management line consisting in ready reactions to external and internal environment and it is focused on choice of change object, its flexible preparation, realization and use. It is based on prediction of next enterprising challenge and it enables being ready to changes on time. (KOTTER 2000). Among authors dealing with this topic belong: PALÁN (Substance of change management), KOTTER (Psychological aspects of change management), HAMMER and CHAMPY (Reengineering changes). The general philosophy for process of change has been specified and published by DRDLA and RAIS (2001), RAIS and KUBÍČKOVÁ (2012), BOROVSKÝ (2005), SLÁVIK, (2005).

The authors of this paper have been dealing with change management focused on processes in several publications during years 2015-2018. The change and change management are closely related to process and process management. The authors SVOZILOVÁ (2011), ZÁVADSKÝ and KOVAĽOVÁ (2011), PAPULOVÁ et al. (2014), RAJNOHA et al. (2013), and SUJOVÁ and ČIERNA (2018) agree that process-driven organizations are customer-centred, create higher value for the customer, focus on process management through analyses and metrics, use concepts, methods and approaches to improve processes as well as optimize and model them for to make them more radical changes and improving their performance.

Corporate performance is an object of interest not only of owners (shareholders) but also of other interested subjects such as managers, employees, creditors, customers, suppliers, municipality and state. In our opinion the corporate performance can be defined as an ability to reach required effects and outputs in measurable units, to evaluate spent resources and to create a profit. Traditional way of business performance comes out from evaluation of achieved financial indicators: profit, turnover, market share. Authors VARCHOLOVÁ et al.

(2007), BREALEY (2000), DUBOVICKÁ (2007), MAŘÍK (2003), RUČKOVÁ (2010) agree that financial indicators allow for a quick and inexpensive picture of the corporate performance.

Profitability ratios is a form of expression of the resource efficiency that serves as the main criterion for capital allocation in a market economy. The Return of equity ROE is one of the most used and expresses the return on equity of an enterprise.

166

The authors of this article conducted several research studies in the previous period, which dealt with the influence of various factors on business performance with a focus on the impact of aspects of business process management on business performance (SUJOVÁ 2013, SUJOVÁ and MARCINEKOVÁ 2015, 2017, GEJDOŠ and SIMANOVÁ 2017, SUJOVÁ and REMEŇ 2018). The aim of this paper is to point out other factors influencing the performance of WPI enterprises, namely factors in the field of change management.

MATERIALS AND METHODS Relevant data from the field of change management and information from enterprises of the wood processing industry in the Slovak Republic were obtained through an online research questionnaire and a direct controlled interview with managers of randomly selected wood processing enterprises.

Research sample determination: The core sample size for the research was a database of 300 wood processing enterprises, out of which 83 respondents represent the research sample. According to the calculation of the minimum statistical research sample the research sample size of 83 respondents is a representative statistical sample with 92% confidence and 8% standard deviation. Before the compilation of the questionnaire, from the basic set created by the WPI enterprises of the Slovak Republic (according to the Statistical Office of the Slovak Republic, SLOVSTAT database 13,983 enterprises), a selection of respondents was created using a typical non-random selection. After creating a list of the most important WPI enterprises of the Slovak Republic using the information provided by the report SARIO Forests and Wood Processing Industry, enterprises were selected according to the average number of employees, the branch of the wood processing industry and the type of production. The databases of member companies of the Association of Wood Processors of the Slovak Republic, the Association of the Pulp and Paper Industry of the Slovak Republic and the database of WPI enterprises of the Slovak Republic listed on the seznam.sk website were also used in the selection of enterprises. This way, a database of 300 enterprises was created.

If the random selection procedure of respondents were followed and, under the specified conditions, the results were generalized to the whole basic set, it would be necessary to obtain information from at least 375 enterprises. Assuming the usual feedback when responding to enterprises (20%), it would be necessary to contact at least 1875 enterprises. The calculation of the random sample size is as follows: 𝑧 2 𝑝𝑞 1.962 𝑥 0.5 𝑥 0.5 𝑛0 = 2 = = 385 𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑒𝑛𝑡𝑠 𝑒 0.052

(1)

Where: z - confidence interval (for 95% = 1.96) p, q - percentage of respondents who know the issues or are inclined to the variant of one (p) and ignorant issues or who are inclined to the variant of the other (q). Since these numbers have not been known, the maximum product p x q = 50% x 50% was generated. e - specified maximum permissible error - 5% n0 - minimum sample size

167

n=

n0 385 = = 375 𝑟𝑒𝑠𝑝𝑜𝑛𝑑𝑒𝑛𝑡𝑠 (n0 -1) 385-1 1+ 13 983 1+ N

(2)

Given that the base set is smaller than 20 000 enterprises, it was necessary to use another formula to calculate the sample size.

Where: n - recalculated sample size n0 - minimum sample size calculated according to the previous formula N - base file size It was for this reason that the survey was chosen and a non-random typical selection was made so that the created database, which contained 300 companies, included companies meeting the criteria of the study. Since the selection of respondents was made using a typical non-random selection, where the obtained results could not be generalized to the whole basic set, a calculation was given with what error the results from the distributed questionnaires could be applied to the whole basic set provided the procedure of random selection of respondents was followed. The already mentioned relation was used to calculate the sample size of the respondents, while the unknown e in this case was unknown. 𝑧 2 𝑝𝑞 𝑒2 2 1.96 𝑥 0.5 .0.5 85 = 𝑒2 𝑒 = 0.106 = 10.6% 𝑛0 =

(3)

It is clear from the above calculation that if the random selection procedure were followed, it would be possible to apply the results to the whole basic set with an error of 10.6%.

The questionnaire consists of 5 general, classification questions and 30 business-area management issues. General questions concerned the size of an enterprise, the main subject of the business, the ownership, the branch of business and the reached value of ROE. Next questions concerned the research core. They were divided into following parts: - Change management (13 questions): perception of changes in a company, types, reasons, areas, goals of performed changes, the procedures by leading changes in a company.

- Quality, production and process management (5 questions): conceptions, models and methods used by improvement of quality, production and process performance.

- Financial aspects (7 questions): cost monitoring and costing, evaluation of effects after changes.

- Investment management (5 questions): evaluation of investment effectiveness, indicators for evaluation of investment effects.

The questionnaire was published online and the data collection was in the first half of year 2017. Enterprises were also interviewed directly through employees and indirectly via e-mail communication. This paper analyses partial results of the questionnaire survey of selected industry, namely the wood processing industry (WPI), which is represented by wood, furniture and cellulose industry. Mathematical - statistical methods were used to examine and evaluate the interrelationships and the effects of individual factors in the implementation of changes on the performance of enterprises. The evaluation of the questionnaire was carried out using the software program STATISTICA 12 CZ - Stat Soft. Inc. 168

(2013), where the imported database was created in MS Excel. In the analysis of research results, selected methods of descriptive statistics for one variable were used, such as absolute, relative and cumulative frequencies, pie and bar graphs. Subsequently, PivotTables in the statistical analysis were applied. A PivotTable is a method of organizing and analysing data by groups, categories, or classes that allows them to be compared. It combines the frequency distribution of two variables and represents an extension of a simple frequency table (RIMANČÍK, 2007). The results of the analysis of PivotTables consist of selected statistical indicators, namely Pearson's chi-square and the level of statistical significance "p".

Pearson's chi-square test enabled testing the significance of the relationship between qualitative variables. The chi-square test of independence was used to test the significance of contingency coefficients. The level of significance was chosen at the level of 5%. The comparison of actually found and theoretical frequencies was the basic idea of the chi-square test of good agreement. Approaching the value of the contingency coefficient to the value of 1 was a signal of an increasing intensity of the dependence between the characters A and B. (PACÁKOVÁ et al. 2018).

The paper presents the partial research results in the areas of change management. The findings in the first part of the questionnaire are focused on the change management issues in relation to business performance represented by the ROE indicator. The 10-year period of carrying out the changes was determined by two reasons: the effects of the changes are visible within a longer period, and the previous research has focused on the change management more than 10 years ago. The research hypotheses were presented in order to supplement the findings of previous research studies conducted by the authors and verify the importance of the changes by improving the competitiveness and performance of companies, as claimed by authors dealing with the changes and their management in companies (RAIS and KUBÍČKOVÁ 2012, BOROVSKÝ 2005).

The following hypotheses established in the research were tested by the mentioned methods: • Ha: There is a statistically significant relationship between the areas of changes implemented over the last 10 years and the level of return on equity of ROEs in WPI enterprises.

• Hb: There is a statistically significant relationship between the types of changes implemented over the last 10 years and the level of return on equity in ROPs. • Hc: There is a statistically significant relationship between the level of process optimization and the level of return on equity of ROEs in WPI enterprises.

• Hd: There is a statistically significant dependence between the impulses or reasons of the WPI enterprises that led to the change and the level of return on equity of the WPI enterprises.

RESULTS AND DISCUSSION The article presents partial research results that show which aspects of change management affect the performance of WPI enterprises. Research in the wood processing industry was carried out in 83 WPI enterprises. Their percentage distribution into the woodworking, furniture and pulp and paper sector can be seen in Figure 1.

169

Research sectors of the woodprocessing industry in % sector of furniture 22%

sector of pulp and papaer 5%

sector of woodworking 73%

Fig. 1 Researched sectors of the wood processing industry.

The examined sample of WPI enterprises showed that 45.12% of respondents employ from 0 to 10 employees, 25.61% of respondents stated that they employ from 11 to 20 employees and from 51 to 250 employees only employ 7.3% of respondents. It follows that most respondents from WPI enterprises are concentrated in the category of small and medium-sized enterprises. Of the 83 surveyed WPI enterprises, 86.59% are mainly focused on production activities and 79.27% have net own domestic capital. The comparison of the achieved ROE values in surveyed woodworking, furniture and pulp and paper enterprises according to the individual categories of ROE values is shown in the graph in Figure 2.

Relative multiplicity %

Return on equity ROE 80,00 70,00 60,00 50,00 40,00 30,00 20,00 10,00 0,00

Woodworking enterprises Furniture enterprises Pulp and paper enterprises ROE ˂ 0% 0%-2%

2%-4%

4%-7%

7%-10% nad 10%

ROE values Fig. 2 Return on equity of the surveyed companies.

In determining the amount of return on equity ROE of surveyed WPI enterprises, we can state that 7.23% of respondents of WPI enterprises were included in the group with ROE below 0%, in the group with ROE values from 0% to 2% were included 19.28% of respondents from woodworking enterprises and 6 .02% of furniture enterprises. The ROE value from 2% - 4% was achieved by 8.43% of respondents of woodworking enterprises, 3.61% of furniture enterprises and 3.61% of pulp and paper enterprises. A positive ROE value from 4% to 7% was recorded in 27.71% of respondents of woodworking enterprises and in 3.61% of furniture enterprises. ROE from 7% to 10% was achieved by 12.05% of respondents of woodworking enterprises and 3.61% furniture enterprises. ROEs above 10% were reported by 2.41% of respondents of woodworking and 1.20% of pulp and paper enterprises. From the above results we can conclude that most surveyed enterprises of the wood processing industry reached the ROE value in the range from 4% to 7%, namely 31.33% of the surveyed enterprises, which we can evaluate positively.

170

In the category of enterprises in the wood processing industry, dependencies were examined as areas of changes in the last 10 years, types of changes in the last 10 years, level of process optimization and impulses that led to change in relation to business performance defined by the level of return on equity ROE. Analysis by the Chi-square test and the calculated p-value made it necessary to establish two opposing hypotheses for each area, which would confirm or refute the existence of dependencies.

To test the Ha hypothesis, which examined the dependence of the areas of changes in the last 10 years and the level of return on equity ROE in WPI enterprises, the following hypotheses were tested: • H0: There is no statistically significant relationship between the areas of changes implemented over the last 10 years and the level of return on equity ROE in WPI enterprises.

• H1: There is a statistically significant relationship between the areas of changes implemented over the last 10 years and the level of return on equity ROE in WPI enterprises.

Table 1 analyses and evaluates the dependence of the areas of implemented changes in relation to the return on equity (ROE) in surveyed WPI enterprises in the wood processing industry. Regardless of the level of ROE achieved, most respondents of WPI enterprises implemented changes in three areas: the production program, production technologies and organizational structure. The least of these were changes in business processes.

Tab. 1 Dependence of areas of implemented changes and ROE. Expected frequencies Areas of changes

< 0% 3.60 5.30 4.80 1.70

0% 2% 6.82 10.04 9.09 3.22

Organizational structure Production program Production technologies System and methods management Business processes 1.40 2.65 Information system 2.20 4.17 SUM 19 36 Relative frequencies* 10% 19% Expected value is lower than actual

p = 0.5 Chi-square test: 0.93235 ROE values 2% 4% 7% > SUM RF* 4% 7% 10% 10% 5.31 13.45 5.31 1.52 36 18.95% 7.81 19.81 7.81 2.23 53 27.89% 7.07 17.94 7.07 2.02 48 25.26% 2.51 6.35 2.51 0.72 17 8.95% 2.06 3.24 28 15%

5.23 2.06 0.59 14 7.37% 8.22 3.24 0.93 22 11.58% 71 28 8 190 37% 15% 4% 100% Expected value is higher than actual

Based on the data in Table 1, the p = 0.93235 value was calculated using the Chisquare test. This value is higher than 0.05, so in this case we can accept hypothesis H0 based on the results of the analysis: There is no statistically significant relationship between the achieved profitability of ROE and areas of change for the last 10 years and the results can be marked as statistically insignificant. In the next step, more detailed tests, the test for each area of change, were made. The results of more detailed tests show that for the dependencies between the issues of return on equity ROE and the area where changes have been made over the last 10 years, statistically significant dependence (α = 5%) was confirmed in only one of six areas in the system and methods management.

As it can be seen from Table 2, the critical value of the tested criterion p = 0.01988 is less than 0.05. For this reason, we can state that the observed dependence was statistically confirmed. Among the observed features, we state a moderately strong dependence based on the value of the Cramer V parameter, which ranges between 0.3 - 0.8.

171

Tab. 2 Results Chi-square test for the relationship between the achieved ROE and the area of implementation of change in the system and methods management.

Pearson Chi-square Contingency coefficient Cramer V

Chi-square 13.40368 0.3748257 0.4043012

Degree of freedom df = 5

p value 0.01988

The second research hypothesis Hb focused on the dependence of the types of implemented changes and the level of return on equity ROE in WPI enterprises was tested by the following hypotheses: • H0: There is no statistically significant relationship between the types of changes made and the level of return on equity ROE in WPI enterprises.

• H1: There is a statistically significant relationship between the types of changes implemented and the level of return on equity ROE in WPI enterprises.

The Table 3 analyses the types of changes implemented over the last 10 years in WPI enterprises in relation to the level of return on equity ROE. Most enterprises with higher performance, with ROE levels above 4% implement gradual optimization changes and unplanned changes incrementally.

Tab. 3 Dependence between types of implemented changes and ROE. Expected frequencies Types of implemented changes

< 0% 1.26

Financial restructuring Transformational change 0.59 restructuring Radical reengineering change 0.34 Gradual optimization changes 3.95 Unplanned but necessary changes 1.43 - incremental No changes were made 1.43 SUM 9 Relative frequencies* 8% Expected value is lower than actual

p = 0.5 Chi-square test: 0.18034 ROE values 0% - 2% - 4% - 7% > SUM 2% 4% 7% 10% 10% 3.08 2.52 5.47 2.10 0.56 15

14.02%

1.44

1.18

2.55

0.98

0.26

7

6.54%

0.82 9.66

0.67 7.91

1.46 17.13

0.56 6.59

0.15 1.76

4 47

3.74% 43.93%

3.50

2.86

6.20

2.38

0.64

17

15.89%

3.50 22 21%

2.86 18 17%

RF*

6.20 2.38 0.64 17 15.89% 39 15 4 107 36% 14% 4% 100% Expected value is higher than actual

Based on the data obtained, the p = 0.18034 value was calculated using the Chisquare test. This value is greater than 0.05. Based on the analysis, we can accept hypothesis H0 from the above results: There is no statistically significant dependence between the type of change in the enterprise over the last 10 years and the achieved profitability of ROE and the results can be marked as statistically insignificant. The tests further confirmed, as shown in Table 4, that when examining the dependencies between the types of changes made over the last 10 years and the return on equity ROE, a statistically significant dependence (α = 5%) on the transformational restructuring change was confirmed. Based on the calculations in the table, we can conclude that the individual categories of enterprises differ significantly in this type of change.

172

Tab. 4 Results of the Chi-square test between the transformational restructuring change and the achieved ROE in the last 10 years.

Pearson Chi-square Contingency coefficient Cramer V

Chi-square 12.21490 0.3600685 0.3859561

Degree of freedom df = 5

p value 0.03196

The critical value of the tested criterion p = 0.03196 is less than 0.05 and therefore we can confirm that the dependence was confirmed. Based on Cramer's V 0.3859561, the observed statistical features showed a strong dependence ranging from 0.3 to 0.8.

The dependence of the level of process optimization (improvement) and the level of ROE in WPI companies was investigated by testing the third hypothesis Hc with the following hypotheses: • H0: There is no statistically significant relationship between the level of process optimization and the level of return on equity ROE in WPI enterprises.

• H1: There is a statistically significant relationship between the level of process optimization and the level of return on equity ROE in WPI enterprises.

The Table 5 analyses the relationship between the level of process optimization (improvement) and the achieved level of ROE in WPI enterprises. The aim of this analysis was to determine whether the level of optimization and improvement of processes in enterprises of wood processing industry affects the achieved ROE value of the surveyed enterprises.

Tab. 5 Dependence between types of the level of process optimization and ROE. Expected frequencies Level of process optimization Identification of optimization options Established mathematical model for quantification of total costs Employees have modern technology at their disposal for efficient processes Business standards and processes are linked to business success factors and customer requirements Created a change management program, ensuring employee loyalty None of the above applies SUM

p = 0.5 Chi-square test: 0.32751 ROE values 2% - 4% - 7% > SUM 4% 7% 10% 10%

< 0%

0% 2%

3.29

11.52

7.13

14.27

7.13

1.65

45

54.88%

0.07

0.26

0.16

0.32

0.16

0.04

1

1.22%

0.59

2.05

1.27

2.54

1.27

0.29

8

9.76%

0.66

2.30

1.43

2.85

1.43

0.33

9

10.98%

0.66

2.30

1.43

2.85

1.43

0.33

9

10.98%

0.73

2.56

1.59

3.17

1.59

0.37

10

12.20%

6

21

13

26

13

3

82

26%

16%

Relative frequencies* 7% Expected value is lower than actual

RF*

32% 16% 4% 100% Expected value is higher than actual

The results in Table 5 show that most enterprises do not optimize real processes and do not pay attention to process improvement, they only have identified optimization options.

This may be the reason why the chi-square test result did not confirm the dependence of the ROE height on the process optimization level.

173

When examining the dependence between the impulses of WPI enterprises that led to the implementation of the change and the level of ROE, the hypothesis Hd was determined and the hypotheses tested: • H0: There is no statistically significant dependence between pulses resp. the reasons for the WPI enterprises that led to the change and the level of return on equity of the ROE. • H1: There is a statistically significant relationship between the reasons or impulses of the WPI enterprises that led to the change and the level of return on equity ROE. The Table 6 presents the reasons and impulses of the WPI enterprises that led to the change and the level of return on equity. The results in Table 6 show that enterprises with a ROE level above 4% implemented changes mainly due to changes in customer requirements and increased competitive pressure, but also due to financial problems.

Tab. 6 Dependence between reasons or impulses to implement change and ROE. Expected frequencies

p=0.05

Chi-square test: 0.20347 ROE values

Reasons or impulses to implement change Financial problems

< 0% 1.87

0% 2% 3.64

2% 4% 2.75

4% 7% 5.90

7% 10% 3.05

> 10% 0.79

SUM

RF

18

9.84%

Low efficiency and quality of production Defective processes

1.97

3.84

2.91

6.23

3.22

0.83

19

10.38%

0.83

1.62

1.22

2.62

1.36

0.35

8

4.37%

Customer dissatisfaction

1.25

2.43

1.84

3.93

2.03

0.52

12

6.56%

Legislative changes

1.04

2.02

1.53

3.28

1.69

0.44

10

5.46%

Competitive pressure

4.46

8.69

6.58

14.10

7.28

1.88

43

23.50%

Changing needs and requirements of customers Existence of market opportunities SUM Relative frequencies*

4.98

9.70

7.34

15.74

8.13

2.10

48

26.23%

2.60

5.05

3.83

8.20

4.23

1.09

25

13.66%

19 10%

37 20%

28 15%

60 33%

31 17%

8 4%

183

Expected value is lower than actual

100%

Expected value is higher than actual

From the data obtained, the value of p = 0.2034 calculated by the Chi-square test was higher than 0.5. Based on the results of the above analysis, we can accept hypothesis H0: There is no statistically significant relationship between the reasons or impulses of WPI enterprises that led to the change and the level of return on equity ROE WPI enterprises, so the results can be described as statistically insignificant.

The results of other more detailed tests than those in Table 7 show that for the dependencies between the reasons or impulses that led to the change and return on equity (ROE), a statistically significant dependence (α = 5%) was confirmed for only one of the nine reasons. That was the legislative changes. The critical value of the tested criterion p = 0.00059, was less than 0.05, and therefore we can state that the observed dependence was statistically confirmed in this particular case. Among the observed statistical features, we state a moderately strong dependence on the value of the parameter Cramer V 0.5149228, which ranged from 0.3 to 0.8.

174

Tab. 7 The results of Chi-square test, the reason for legislative changes and achieved ROE. Pearson Chi-square Contingency coefficient Cramer V

Chi-square 21.74193 0.4577959

Degree of freedom df = 5

p value 0.00059

0.5149228

The first part of the research was focused on finding out the number of WPI enterprises in the examined sample, and on finding the basic data and characteristics about these enterprises. We found that woodworking enterprises had a share of 73% in the sample examined. These were mainly enterprises with a small number of employees, which are considered to be small and medium-sized enterprises. 86.59% of the surveyed WPI enterprises mentioned the production as the main subject of their activity. The ownership of the surveyed WPI enterprises consisted mainly of net domestic capital, which was stated by up to 79.27% of respondents. A positive finding was that most surveyed enterprises achieved a positive ROE value, which is a key criterion for business owners to evaluate the success of their investment. In comparison with the results of the conducted research from 2011 and 2013, it can be stated that the performance of WPI enterprises has significantly increased from the level of 0−2% to the level of 4−7%. The increase was mainly due to changes in customer requirements, increased pressure from competition, the use of new approaches to managing the improvement of business process performance, the willingness of enterprises to cope with change management and efforts to implement new methods. These performance factors correspond to the theoretical results of STANĚK (2003), WAGNER (2009), KADÁR and KADÁROVÁ (2010), FERENCOVÁ and BAŠISTOVÁ (2011), PAPULOVÁ et al. (2014), and with the research results of TUČEK and ZÁMEČNÍK (2007) and RAJNOHA et al. (2013).

In the general question category, there was a positive finding that 90.24% of the WPI enterprises in the sample examined stressed the high importance to change management and also considered change to be an important factor in performance improvement.

The relationship between the change management aspects and the performance of the WPI enterprises in the sample examined was only confirmed in three cases. In the specified areas of changes, the management systems and methods used had an impact on the amount of ROE, in the types of change it was a restructuring change, and in the case of impulses or reasons for the implementation of the change, it was legislative changes. The dependence between the level of process optimization and the level of return on equity ROE in the examined WPI enterprises was not confirmed due to the fact that surveyed WPI enterprises do not really optimize processes. We consider this finding for surveyed WPI enterprises to be a significant shortcoming in the management of business processes, which should be constantly improved and optimized.

The fact that WPI enterprises in the Slovak Republic have the maturity of processes and the level of their management at the lowest level was also stated in the results of previous research carried out by the authors of this paper. Since 2010, the results of research have shown that the level of business process management in WPI enterprises and failure to pay attention to process optimization significantly affects their performance. Greater efforts to manage and continually improve business processes would bring WPI enterprises a significant increase in their performance.

175

CONCLUSION Change is an integral and fundamental part of every enterprises and corporate governance and can be activated by many factors, whether from the external or internal environment. It is necessary to realize the importance of change, eliminate and reduce resistance to change and initiate, manage and successfully implement changes. Therefore, the ability of enterprises to adapt is an important role, because without a change in the current environment, companies will not survive. Change management is thus one of the basic elements of successful business management and effective change management becomes a tool for increasing competitiveness.

The benefits of the presented results of the research of change management in enterprises in the wood processing industry lie in finding out the current state of change management in WPI enterprises and in identifying the factors that affect their performance.

The results of the primary quantitative research carried out in the form of questionnaires were analysed and, based on statistical methods, the existing dependencies in relation to the return on equity ROE performance indicator were verified and evaluated.

The achieved results show that the performance of WPI enterprises is directly affected by changes in management systems and methods, the implementation of extensive transformational changes.

Direct dependence was also found in the implementation of changes due to changes in legislation, which may indicate the fact that legislative changes in the Slovak Republic contribute to increasing the performance of WPI enterprises. It can also be stated that enterprises with a high level of performance have decided to implement changes based on changes in customer requirements and in the activities of competitors. This means that the reason for making changes in the enterprise was the effort to improve competitiveness.

The importance of monitoring and managing change in increasing the competitiveness of enterprises also stems from the characteristics of theoretical aspects of change management and from the identified effects of successfully implemented change. Based on the achieved results presented in the article, we can state that the management of business change based on a process approach is a way to improve competitiveness.

REFERENCES BOROVSKÝ, J. 2005. Change management - the way to growth competitiveness. Bratislava: Eurounion, p.45.

BREALEY, R.A., MYERS, S.C. 2000. Theory and practice of corporate finance. Prague, Czech Republic: Victoria publishing DRDLA, M., RAIS, K. 2001. Change management in the company – reengineering. Prague: Computer Press, p. 145.

DUBOVICKÁ, L. 2007. Model approaches to measuring strategic performance: Organizational performance. Approaches to its measurement and evaluation. Kosice: PHF, p. 191.

FERENCOVÁ, M., BAŠISTOVÁ A. 2011. Management of Slovak companies. Business Economics and Management: A Scientific Journal for Economic Theory and Practice. Košice: VUSI, pp. 1−8.

GALANTI, L., NDIAYE, I., ST-HILAIRE, C. 2012. Change management in an IT methodology context: The experience of TEDO. Montreal: Software Process Improvement Network.

GEJDOŠ, P., SIMANOVÁ, Ľ. 2017. Modern Methods of the Quality Improvement and Their Application. In: Managing of Changes in a Company. In New Trends and Issues Proceedings on Humanities and Social Sciences 4 (10), pp. 64–72.

HAMMER, M., CHAMPY, J. 2000. Reengineering – radical transformation of the company. Prague: Management Press, p. 37.

176

KADÁR, G., KADÁROVÁ, J. 2010. Evaluation of factors of business competitiveness. In the 13th International Scientific Conference Trends and Innovative Approaches in Business Processes Košice: SJF, TUKE, p. 1−6.

KANTER, R. 2020.Ten Reasons People Resist Change”. Harvard Business Review [online] available from < [20 March 2020].

KOTTER, J. P. 1995. Leading Change: Why Transformation Efforts Fail. In Harvard Business Review, vol. 73(2), pp.59−67.

KOTTER, J. P. 2000. Leading the change process. In The International Journal of Bank Marketing, vol. 18, no. 7, Prague: Management Press pp. 328–337.

KUBÍČKOVÁ, L., RAIS K. 2012. Change management in companies and other organizations Prague: Grada, p.136.

MARCINEKOVÁ, K., SUJOVÁ. A. 2015. The influence of the process control level on the enterprises' ROE. Book Series: Procedia Economics and Finance, vol. 34, pp. 290−295 s.

MAŘÍK, M., MAŘÍKOVÁ, P. 2005. Modern methods of evaluation and valuation of the company.

Prague: Ekopress, pp. 164.

MIKUŠ, P. 2010. Change management - competitiveness of the organization. Ružomberok: Verbum p.143.

MULLINS, L.J. 2013. Management and organisational behaviour, London: Pearson, p. 856.

MURPHY, M. 2016.The Big Reason Why Some People Are Terrified of Change (While Others Love It). Forbes [online] available from < [21 March 2020].

PACÁKOVÁ, V., LABUDOVÁ, V., SIPKOVÁ, Ľ., ŠOLTÉS, E., VOJTKOVÁ, M. 2018. Statistical methods for economists. Bratislava: Iura Edition, p. 411.

PAPULOVÁ, Z., PAPULA J., OBORILOVÁ, A. 2014. Process Management. A comprehensive view of the process management concept. Bratislava: Kartprint, p. 223.

PITAGORSKY, G. 2011. Project managers are change managers. Retrieved February 02, 2020, from Project Management Times: http://www.projecttimes.com/george-pitagorsky/project-managers-arechange-managers.html RAJNOHA, R. 2013. Measuring and managing business performance. Zvolen: Technical University in Zvolen, p. 313.

RICK, T. 2013. Change management is a process, not an event. Retrieved 21.02, 2020, from torbenrick.eu: http://www.torbenrick.eu/blog/change-management.

RIMANČÍK, M., 2007. Statistics for practice. Kosice: Marián Rimančík, p. 200.

RUČKOVÁ, P. 2010. Financial analysis: methods, indicators, use in practice. Bratislava: Iura Edition, p. 238.

SIMANOVÁ, Ľ. 2017. The Change Management of the Manufacturing Processes in the Woodworking Company. In Proceedings of scientific papers from global scientific conference Management and economics in manufacturing, Slovakia, p. 192−198.

SLÁVIK, Š. 2005. Strategic management. Bratislava: Sprint, p. 403.

STAT SOFT. INC. 2013. STATISTICA (data analysis software), version 12. www.statsoft.com.

STANĚK, V. 2003. Improving performance through process cost management. Prague: Grada, p.236 SUJOVA, A. 2013. Business Process Performance Management – a Modern Approach to Corporate Performance Management. In Liberec Economic Forum. TU Liberec, pp. 542−550.

SUJOVÁ, A., REMEŇ, O. 2018. Management of Changes in Wood processing Companies. In Proceedings from conference: Increasing the Use of Wood in the Global Bio-Economy. Belgrade: WOODEMA Serbia, pp. 305−316.

SUJOVA, A., MARCINEKOVÁ, K., HITTMÁR, Š. 2017. Sustainable optimization of manufacturing process effectiveness in furniture production. In Sustainability 9(6): pp. 923−938.

SUJOVÁ, E., ČIERNA, H. 2018. Optimization and improvement of test processes on a production line.

In Management Systems in Production Engineering. Vol. 26, pp. 88−92.

SVOZILOVÁ. A. 2011. Improving business processes. Prague: Grada, pp. 232.

TUČEK, D., ZÁMEČNÍK, R. 2007. Management and evaluation of business process performance in practice. Elected: Technical University of Zvolen, pp.173.

VARCHOLOVÁ, T. et al. 2007. Measuring business performance. Bratislava: Ekonom, p. 167.

177

WAGNER, J. 2009. Performance measurement: how to measure, evaluate and use information about business performance. Prague: Grada, pp. 248.

ZÁVADSKÝ, J., KOVAĽOVÁ, M. 2011. Operational and strategic performance of business processes.

Bratislava: Slovak Committee for Scientific Management, ZSVTS, p.84.

ACKNOWLEDGEMENTS Authors are grateful for the support of the VEGA agency - the project No.1/0286/16 and the KEGA agency, Slovakia - the project No. 005TU Z-4/2020.

AUTHOR´S ADDRESS Assoc. prof. Ing. Andrea Sujová, PhD. Ing. Ľubica Simanová, PhD., Ph.D. Technical University in Zvolen Faculty of Wood Science and Technology Department of Economics, Management and Business T.G. Masaryk 24 960 01 Zvolen Slovakia andrea.sujová@tuzvo.sk, lubica.simanova@tuzvo.sk

178