ACTA FACULTATIS XYLOLOGIAE ZVOLEN 63 2/2021 by www.tuzvo.sk

TECHNICKÁ UNIVERZITA VO ZVOLENE DREVÁRSKA FAKULTA

ACTA FACULTATIS XYLOLOGIAE ZVOLEN

VEDECKÝ ČASOPIS SCIENTIFIC JOURNAL

63 2/2021

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 2/2021 SCIENTIFIC JOURNAL OF THE FACULTY OF WOOD SCIENCES AND TECHNOLOGY, TECHNICAL UNIVERSITY IN ZVOLEN 63 2/2021 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) 188 strán, 16,56 AH, 16,71 VH Tlač (Printed by): Vydavateľstvo Technickej univerzity vo Zvolene Vydanie I. – Október 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.

Obsah 01 BARBORA SLOVÁČKOVÁ  OĽGA MIŠÍKOVÁ: PERMEABILITY OF THREE WOOD SPECIES DEGRADED BY TRAMETES VERSICOLOR L. LLOYD ................................................................................................................... 02 VLADIMÍR IHNÁT  MÁRIA FIŠEROVÁ  ELENA OPÁLENÁ  ALBERT RUSS  ŠTEFAN BOHÁČEK: CHEMICAL COMPOSITION AND FIBRE CHARACTERISTICS OF BRANCH WOOD OF SELECTED HARDWOOD SPECIES ........................................................... 03 JAKUB KAWALERCZYK – DOROTA DZIURKA – RADOSŁAW MIRSKI – JOANNA SIUDA – JÁN SEDLIAČIK: MICROCELLULOSE AS A MODIFIER FOR UF AND PF RESINS ALLOWING THE REDUCTION OF ADHESIVE APPLICATION IN PLYWOOD MANUFACTURING ................................................................

04 SERGIY KULMAN – LIUDMYLA BOIKO – YAROSLAV BUGAENKO – JÁN SEDLIAČIK: CREEP LIFE PREDICTION BY THE BASIC MODELS OF DEFORMATION-DESTRUCTION KINETICS OF WOOD-BASED COMPOSITES ..........................................................................

05 MIROSLAV REPÁK  LADISLAV REINPRECHT: THE COLOR OF BEECH WOOD MODIFIED IN AIR, PARAFFIN OR POLYETHYLENE GLYCOL, AND AFTER FOLLOWING WEATHERING IN XENOTEST ....

06 VALENTIN ATANASOV: EXPERIMENTAL RESEARCH OF THE CUTTING FORCE DURING LONGITUDINAL MILLING OF SOLID WOOD AND WOOD-BASED COMPOSITES .............................................

07 MARIA ZYKOVA – VALERIA KASIMIOVA – MIKHAIL CHERNYKH – VLADIMIR ŠTOLLMANN – GALINA EVSTAFIEVA: METHOD OF COMPUTER TEMPLATE ADJUSTMENT FOR LASER ENGRAVING ON WOOD ...........................

08 ROMAN NÔTA – ZUZANA DANIHELOVÁ: ANALYSIS OF THE THERMAL BRIDGE OF WOOD-ALUMINUM WINDOW INSTALLATION POSITION ........................................................................

09 IVETA ČABALOVÁ – MARTIN ZACHAR – MICHAL BÉLIK – ŽANETA BALÁŽOVÁ: RESISTANCE OF SPRUCE WOOD (Picea abies L.) TREATED WITH A FLAME RETARDANTS AFTER THE RADIANT HEAT EXPOSURE .....................................................................

103

10 ROZÁLIA VAŇOVÁ – JOZEF ŠTEFKO: ASSESSMENT OF SELECTED TYPES OF STRUCTURAL ENGINEERED WOOD PRODUCTION FROM THE ENVIRONMENTAL POINT OF VIEW ........

117

11 MIROSLAVA NEJTKOVÁ  MARTIN PODJUKL: CHRISTMAS TREE IGNITION BY SPARKLERS .............................................................

131

12 ROZÁLIA VAŇOVÁ: INFLUENCE OF CARBON ACCOUNTING ON ASSESSMENT OF WOOD-BASED PRODUCTS .......................................

143

13 A.V. CHERAKSHEV  D.E. RUMYANTSEV: DENDROCHRONOLOGICAL RESEARCH OF OLD-AGED EASTERN WHITE PINE FROM KALUGA OBLAST - A NATURAL HERITAGE MONUMENT .................................................................................................

153

14 MARIANA SEDLIAČIKOVÁ – PATRIK ALÁČ – MÁRIA MORESOVÁ – IVAN SEDLIAČIK: MAPPING WOOD COLOUR PREFERENCES AMONG POTENTIAL CUSTOMERS .............................

163

15 JUSTYNA BIERNACKA: ANALYSIS OF INTRA-INDUSTRY MUTUAL TRADE OF THE FURNITURE MANUFACTURING INDUSTRY BETWEEN V4 COUNTRIES AND EU-27 USING THE GRUBEL-LLOYD INDEX ............................................................................

175

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 5−15, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.01

PERMEABILITY OF THREE WOOD SPECIES DEGRADED BY TRAMETES VERSICOLOR L. LLOYD Barbora Slováčková  Oľga Mišíková ABSTRACT Permeability of wood is an important factor in many technical processes, foremost in treating wood with various chemicals. It depends on the number of pathways in wood and their transmissivity. The more open pores and pathways are, the deeper the chemicals can penetrate wood, and protection will last longer. The focus of this paper is permeability of spruce, beech and sessile oak heartwood intentionally degraded by Trametes versicolor L. Lloyd. Porosity of the degraded wood species increased by more than 10 %. A higher permeability was expected because of the increased porosity, but this could not be confirmed, as the values showed a great variability. Longitudinal permeability of the wood degraded with Trametes versicolor L. Lloyd was similar to permeability of healthy wood. As expected, the sessile oak heartwood was not permeable at all. To support the findings, all wood species were studied under a light microscope. There were visibly thinned cell walls in the degraded spruce wood, missing toruses in bordered pits, and bore holes in tangential walls of tracheids. Permeability in the radial direction of degraded spruce wood was lower compared to permeability in tangential direction. The degraded beech wood also showed signs of degradation – thinned and disrupted cell walls, sometimes even missing cell walls, pronounced pits and bore holes turning into ruptures. There were numerous tyloses visible in the oak heartwood, as well as the presence of hyphae. Key words: permeability, light microscopy, spruce, beech, oak heartwood, degraded wood.

INTRODUCTION Wood is a natural, porous material. Pores in wood ensure the transport of water and nutrients in living trees. This ability to transport masses through pores remains active even after the tree is logged. It is important in many manufacturing processes – drying, vacuum drying, impregnating wood with chemicals, gluing ̶ to mention a few. Pores in wood are of various sizes, positions (longitudinal, radial or tangential ̶ orientation) and geometry. These properties vary in different wood species. Some wood species pores can be clogged by tyloses or other barriers, which greatly affects the ability to transport mass in wood. Permeability is the ability of wood to transport mass through its porous structure. Permeability is known to be highly variable (SIAU 1995). Factors affecting permeability can be divided into two groups – inner (anatomical structure of wood) and outer (properties of the mass passing through wood and pressure gradient). The most frequent way to express permeability is Darcy’s law. 5

Inner factors influencing permeability are related to structure of wood; the number and size of cells and elements (POŽGAJ et al. 1997). The research by BABIAK et al. (1983) showed that 96.3 % of the water transport in beech wood is provided through vessels with a radius bigger than 15 μm. Vessels with a radius smaller than 15 μm provide 3.7 % of the permeability; despite the count of these vessels in a unit of area was 41 % (BABIAK et al. 1983). KOVÁČIK (1993) presented similar results for beech wood – 95 % of axial permeability is provided through pores with a radius bigger than 18 μm. Pores with a radius smaller than 18 μm provided only 5 % of the axial permeability; despite pores with this ratio took 28.5% of the pores ratio in the unit of area. Flow pathways in wood are in axial and lateral directions. In softwoods, axial flow takes place primarily in longitudinal tracheids by passing through the bordered pits that are implemented in their end-walls (CÔTÉ 1963, ERICKSON AND BALATINECZ 1964, COMSTOCK 1965, BAILEY AND PRESTON 1969, ISAACS et al. 1971). As far as hardwoods are concerned, the flow of fluids in the longitudinal direction is largely controlled by the size and number of vessels that are un-clogged by tyloses or other obstructions (WARDROP AND DAVIES 1961, CÔTÉ 1963, ISAACS et al. 1971). CÔTÉ (1963) mentions that wider lumened fibre tracheids and vasicentric tracheids are often heavily pitted permitting better communication with adjoining cells. Longitudinal parenchyma cells are also described as being more permeable than libriform fibres (WARDROP and DAVIES 1961, HANSMANN et al. 2002). As for lateral flow, in general, wood rays offer a significant path (WARDROP and DAVIES 1961, CÔTÉ 1963). Regarding different pathways of fluids in wood in connection with specific anatomical features and the gross structural factors of wood, it is evident that movement of fluids through wood is easiest along the grain (CÔTÉ 1963). When radial permeability was compared to the tangential one, permeability was greater radially than tangentially (ISAACS et al. 1971, PALIN and PETTY 1981). Heartwood formed in both soft- and hardwoods has a decreased permeability compared to sapwood (WARDROP and DAVIES 1961, CÔTÉ 1963, COMSTOCK 1965 and 1968, ISAACS et al. 1971). Low permeability of heartwood is also due to the presence of extractives. Extractives increase the contact angle between aqueous liquids and the cell walls and lead to decreased wettability compared to sapwood (BAILEY and PRESTON 1969, MANTANIS and YOUNG 1997). Permeability of wood was studied many times before and with various treatments of wood. Experiments conducted on sound beech wood were published in works by HUDEC and DANIHELOVÁ (1992), BABIAK and KÚDELA (1993), KURJATKO et al. (1998), POŽGAJ et al. (1997), KÚDELA (1999) and BABIAK et al. (1995, 2001). Permeability on reaction wood was reported in works by HUDEC (1993) and ČUNDERLÍK and HUDEC (2002). Experiments on permeability on wood treated with microwaves were published in works by LIN and LU (2004) and NASSWETTROVÁ et al. (2014). Permeability was also studied on wood treated with wood staining fungi in works by REINPRECHT and PÁNEK (2009), PÁNEK et al. (2013), DANIHELOVÁ et al. (2018). Treatment of wood with wood decaying fungi and its impact on permeability was presented in works by KURJATKO et al. (2002), SOLÁR et al. (2003 and 2006), EMAMINASAB et al. (2015 and 2016). Degradation processes in wood caused by wood decaying fungi are driven by different mechanisms. Trametes versicolor L. Lloyd is known to thin out cell walls, cause ruptures in cell walls through creating bore holes and eventually make even cells disappear. Trametes versicolor primarily attacks lignin, followed by cellulose and hemicelluloses (BARI et al. 2015, 2018). Hence, wood becomes more porous, and cell lumina become wider, which could increase the permeability of wood. A partial removal of lignin is believed to increase porosity and open new pathways for the transport of solutions deep inside the material, which should facilitate a subsequent 6

impregnation step for further functionalization (VITAS 2019, DONALDSON et al. 2015, JAKES et al. 2015). Permeability of three wood species (Picea abies L., Fagus sylvatica L., Quercus petraea Matt. Liebl.) intentionally degraded by white rot fungus Trametes versicolor L. Lloyd was the main focus of this work. Porosity of the degraded wood species was higher compared to porosity of undegraded wood species (SLOVÁČKOVÁ 2021b). Because of a higher porosity, permeability of the degraded wood is expected to increase in comparison to permeability of undegraded wood species. The degraded sessile oak heartwood was expected to be permeable. To support the findings, all degraded wood species were studied under a light microscope. The results and experiments presented in this paper are part of a larger, ongoing experiment on thermal properties of degraded wood and proposing a new bio-based thermal insulation material based on the degraded wood. Thermal properties of the decayed wood species were described in another article (SLOVÁČKOVÁ 2021a). The results on permeability of wood degraded by a wood decaying fungus are expected to contribute to the topic of biological treatment of wood.

MATERIAL AND METHODS The lumber used to prepare the beech (Fagus sylvatica, L.) and the sessile oak heartwood (Quercus petraea, Matt. Liebl.) samples was stored at the Department of Wood Science and Technology. Spruce wood (Picea abies, L.) lumber was obtained from a wooden windows manufacture located near the University. Permeability was measured in all anatomical directions. There were 16 samples per each anatomical direction. The smallest size of 8 mm was cut according to the anatomical direction, so the sizes of the samples were (dimensions in L × R × T direction) 8 × 50 × 50 mm3 for permeability measurement in longitudinal direction; 50 × 8 × 50 mm3 for permeability measurement in radial direction and 50 × 50 × 8 mm3 for permeability measurement in tangential direction. As it was mentioned in the introduction, thermal properties were measured on these samples as well, so the dimensions were limited by calculations for the thermal properties measurement. The samples were oven-dried, then weighed and measured. They were then intentionally degraded with Trametes versicolor L. Lloyd and the degradation was prepared according to STN EN 113 and it was performed in the laboratory of Department of Wood Technology at the Technical University at Zvolen. The degradation duration was 6 months. Degradation duration according to the STN EN 113 is shorter than in this experiment. The decision for a 6 month long degradation duration was based on results from a previous set of experiments. After the exposure time passed, the samples were taken out of Kolle flasks, cleaned off visible mycelium and they were submerged into containers with distilled water. Submerging the samples caused all cavities to fill up with distilled water so as to remove any air left in the samples. Since there was not any air left, the fungus was not able to survive in these conditions and it stopped its activity. According to RYPÁČEK (1957) wood decaying fungi need at least 5–20 % air in wood in order to be able to survive. One container per wood species was used. The containers with samples were stored in a dark room with constant temperature to prevent any photochemical reactions. Distilled water was changed periodically. The samples were kept in the water until they reached maximum moisture content. Reaching the maximum moisture content was checked by double weighing the samples in water.

A follow up permeability measurement was performed on non-degraded spruce and beech samples in the longitudinal direction. Non-degraded sessile oak heartwood was not permeable at all. Measurement and calculation of permeability Measurement of permeability was performed on a patented apparatus by REGINÁČ et al. (1977). Measurement was done on wet samples after they reached maximum moisture content. Values for permeability were obtained for all anatomical directions. Thickness of each sample was measured before every procedure with a Mitutoyo Absolute Digital Digimatic calliper. The samples were fastened into the permeability measurement apparatus, a pressure was set and the valve on the water tank was opened. Water permeating through the samples was collected in a measuring cylinder placed under the apparatus on a KERN KB 1000-2 scale with 0.01g accuracy. The scale was connected to a computer and the increasing mass of the collected water was transferred automatically in desired time intervals into an Excel working sheet. Duration of each run was 2 or 5 minutes, according to the linearity of the flow and anatomical direction. Permeability was calculated according to Darcy’s law: 𝑘=𝐶∙

𝜂𝐻2𝑂 𝑆

𝑙

∙ ∆𝑝

(1)

where C is the slope of volume of the permeated water and time, ηH2O is the dynamic viscosity of water at a certain temperature, S is the surface of the nozzle of the apparatus (diameter of the nozzle is 1.00 cm), l is the samples’ thickness and Δp is the difference between set pressure for water in the apparatus and atmospheric pressure. Light microscopy Thin microtome slices were prepared from the degraded samples, one per each species per anatomical direction. Small specimen with dimensions of 3 × 3 × 7 mm were cut from a randomly selected sample. These had to be embedded in epoxy resin because it was impossible to cut thin slices due to samples brittleness. Several slices were cut from each wood species. They were cut with a sledge microtome (Reichert, Wien, Austria). Toluidine Blue stain in liquid state was applied to the samples and rinsed out with distilled water. Microslices were left to dry and then mounted permanently on a microscope slide with Euparal (BioQuip Products Inc., Rancho Dominguez, United States). Microscope slides were covered with cover slides and weighed down for one week. The permanent mounts were examined under a transilluminating microscope at magnifications of 200× and 400×. A Canon EOS 600D camera was attached to the microscope for taking pictures.

RESULTS AND DISCUSSION Permeability is known to be a very variable property (SIAU 1995). Degraded oak heartwood was expected to be permeable, but the specimen did not let any water pass through, not even under very high pressure of Δp = 55 kPa. Analysis of the sessile oak heartwood microsections showed numerous tyloses in tracheas (Fig. 3B, C) which were the main cause for the sessile oak heartwood being impermeable. As for the results of spruce and beech wood permeability, the results showed a great variation. The medians, first and third quartiles for all anatomical directions are displayed in Table 1. Average thicknesses of the samples were: 8.2 mm for non-degraded spruce wood in longitudinal direction, 8.1 mm for non-degraded beech wood in longitudinal direction. The thicknesses for degraded spruce wood samples were 7.6 mm in longitudinal direction and 8.0 mm in both radial and tangential directions. The degraded beech wood samples 8

reached a thickness of 8.0 mm in longitudinal direction, 9.1 mm and 8.4 mm for radial and tangential directions respectively. The pressures varied according to wood species and anatomical direction. In general, the pressure was lower for longitudinal direction than for transversal directions. Tab. 1 Medians of permeability coefficients for degraded spruce and beech wood for all anatomical directions and reference permeability values in longitudinal direction. N is the number of runs. First and third quartiles are under the median value (Q1 – Q3). Longitudinal direction, decayed wood

Radial direction, decayed wood

Tangential direction, decayed wood

k [m2] (n = 13; 15)

k [m2] (n = 13; 17)

k [m2] (n = 11; 15)

Spruce

1.73·1014 1.05·1015–3.33·1014

1.31·1014 8.27·10-15–2.87·10-14

2.50·1015 1.49·1015–2.91·1015

3.34·10-15 1.67·10-15– 6.36·10-15

Beech

Reference permeability values in longitudinal direction, healthy wood k [m2] (n = 15; 15)

5.28·1012 2.55·1012–8.25·1012

5.90·1013 1.48·1013–3.96·1012

7.61·1015 4.87·1015–1.05·1014

5.93·1015 4.20·1015–7.45·1015

The reference longitudinal permeability values of non-decayed wood and longitudinal permeability values of the decayed wood are similar. In fact, the median values of longitudinal permeability of the decayed wood are a little lower than longitudinal permeability values of non-decayed wood. Permeability values of undegraded beech wood were reported by many researchers: BABIAK and KÚDELA (1993) measured a longitudinal permeability of 4.90·1012 m2, POŽGAJ et al. (1997) stated a longitudinal permeability of 8.51·1012 m2, HUDEC and DANIHELOVÁ (1992) stated a permeability of 7.56·1012 m2 and according to KÚDELA (1999), longitudinal permeability of undegraded beech wood was 10.00·1012 m2. As stated by BABIAK (1990) the question “whether all variability is caused only by the change of wood characteristics or if at least a part of the variability may be attributed to different physical conditions during the experiment” arises. Permeability in longitudinal direction is the highest compared to permeability values in transversal directions in both degraded wood species. The degraded spruce wood showed a higher permeability in tangential direction than in radial direction. The degraded beech wood showed the opposite, permeability in radial direction is higher than in tangential direction. These findings are presented in Tab. 2, where ratios of permeabilities were calculated. Degraded spruce wood showed a lower permeability than degraded beech wood in all anatomical directions. According to various gas permeability data collected by COMSTOCK (1970), the ratios of permeabilities in softwoods are as follows: longitudinal to tangential permeability ratio varies from 520 to 81600, longitudinal to radial permeability ratio varies from 15 to 547000 and the tangential to radial permeability ratio varies from 0.019 to 31.3. Ratios of longitudinal to transverse permeability as high as 106 have been observed in some softwood species (COMSTOCK 1970). According to the research by LIHRA et al. (2000), the longitudinal permeability of balsam fir was about 2000 and 9000 times higher than the tangential and radial permeability respectively. COMSTOCK (1967) has shown that permeability values measured with gases and liquids are closely related, the relationships can be considered generally applicable. The L/R, L/T and T/R permeabilities ratios of degraded spruce and beech wood are presented in Table 2.

Tab. 2 Permeabilities ratios of degraded spruce and beech wood. Spruce Beech

Longitudinal/Radial 5.24 77.53

Longitudinal/Tangential 3.92 99.49

Tangential/Radial 1.34 0.78

The L/R and L/T permeability ratios in degraded spruce wood are far below the range observed by COMSTOCK (1970) and LIHRA et al. (2000), the T/R permeability ratio is within the range observed by COMSTOCK (1970). Degraded spruce wood permeability in transversal directions increased. The L/R and L/T permeability ratios in degraded beech wood are higher than in degraded spruce wood. In comparison to the degraded spruce wood, the degraded beech wood has a higher permeability in longitudinal direction than in transversal directions. Wood staining fungi like Sydowia polyspora were proven to increase permeability of wood as proven by DANIHELOVÁ et al. (2018). The coefficient of axial permeability of biotreated and subsequently dried spruce sapwood increased approximately 5-times (DANIHELOVÁ et al. 2018). However, wood staining fungi have different degrading mechanism than wood decaying fungi. Wood staining fungi are known to feed mostly on protoplasmic remnants in cell lumina and pectins in pits’ membranes. They are not able to attack cellulose; hemicelluloses or lignin hence wood-staining fungi do not cause major damage to wood structure like wood decaying fungi (REINPRECHT 2016). The research by EMAMINASAB et al. (2015) showed that white-rot fungus and also softrot fungus had a negative impact on permeabilities on both poplar normal wood and tension wood. It is suspected that the fungal hyphae in cell lumina are blocking the pathways for the mass flow. This contrasts with findings by GREEN and CLAUSEN (1999), who concluded that both white-rot and brown-rot fungi increase wood permeability in pine wood. Various species of white- and brown-rot fungi were tested in this experiment. The research of permeability on beech wood degraded by two white-rot fungi Phanerochaete chrysosporium and Ceriporiopsis subvermispora by SOLÁR et al. (2003) showed various effects of white-rot fungi degradation. Degradation of normal and tension wood specimens by P. chrysosporium increased their coefficients of axial permeability in air-dry and saturated state markedly. Degradation of normal and tension beech wood specimens with C. subvermispora reduced their axial permeability in both air-dry and saturated states unexpectedly. The following figures show an analysis of all degraded wood species under the light microscope. All wood species showed clear signs of the degradation process and an advanced stage of cell walls decomposition. Analysis of decayed wood structure showed visible changes caused by degradation. Bordered pits in degraded spruce wood had visibly damaged or missing toruses. Bore holes were visible on tangential walls of earlywood tracheids and cell walls were visibly thinned, sometimes even disrupted (Figures 1A, B). Hyphae were present in some slices. Figure 1A shows numerous hyphae present in some cell lumina. The hyphae seem to have been an obstruction for the water flow. The fungus created bore holes in both radial and tangential walls of tracheids (Figures 1B and C). Missing toruses in bordered pits and ruptures in cell walls had a bigger impact on the tangential permeability. In the end result, permeability in tangential direction was higher than permeability in radial direction of the degraded spruce wood.

Fig. 1 Degraded spruce wood. White arrows in picture A, transversal cut, shows hyphae. Black arrows show visibly thinned cell walls. Arrows in picture B point at missing toruses in bordered pits which are also visibly degraded. The arrow pointing right shows also some bore holes near the bordered pit. Some hyphae are visible under the big rupture in the middle of the picture. Horizontal white arrow in picture C point at bore holes in tangential side of the cell walls. The vertical white arrow shows disrupted cell wall in ray parenchyma. The scale in all pictures is denoted with a white stripe in the lower right corner and it equals 50 μm.

Fig. 2 Decayed beech wood, scale is marked by white stripes in lower right corner of each image, and it equals to 50 μm. Arrows in picture A show missing or severely disrupted cell walls. Pits in ray parenchyma are more pronounced. Arrows in picture B point at rupture and bore holes in cell walls. Ray parenchyma cells are mostly emptied due to activity of fungus. In picture C, arrows show bore holes turning into big ruptures in cell walls.

Similar signs of degradation were found in beech wood. Thinned, disrupted cell walls, pronounced pits and almost completely emptied ray parenchyma cells. Pits in ray parenchyma cells became visible because the fungus degraded most of the compounds stored in ray parenchyma cells. Unlike in the degraded spruce wood, activity of the fungus did not have an impact on permeability in transverse directions of the degraded beech wood (the permeability in radial direction was higher than in tangential direction). Signs of fungal attack in radial and tangential cuts of the degraded beech wood are visible in Fig. 2C and 2B. They appear to be of a similar extent in both directions. Despite a better cell-to-cell connection through the severely damaged cell walls, vessels not clogged by tyloses, permeability of the degraded beech wood in longitudinal direction was lower compared to permeability of healthy beech wood. The Fig. 2A – 2C show, that the degraded beech wood was intact even after a 6 months long degradation. As it was already mentioned in the introduction, according to the research by BABIAK ET AL. (1983) 96.3 % of the water transport in beech wood is provided in vessels with a radius bigger than 15 μm. Because the fungal activity in the degraded beech wood did not develop more pathways with such radius, the longitudinal permeability did not improve. The longitudinal permeability of the degraded beech wood could increase if the degradation duration was longer. Cell walls digestion extent could develop more pores with a radius bigger than 15 μm through a longer degradation duration.

Fig. 3 Degraded oak heartwood. The scale equals 50μm in pictures A and B and 30 μm in picture C, as it is a bigger magnification. Numerous tyloses are visible in picture A. A big tylosis overgrown with hyphae (pointed out with white arrow) is shown in picture A. The black arrow points at bore holes. Picture C shows well-visible hyphae (marked with black arrows). Cell wall disruption is visible as well.

There were numerous tyloses in the degraded sessile oak heartwood which made permeability impossible. Numerous hyphae were observed in the sessile oak heartwood specimen. The hyphae spread also over tyloses as shown in (Fig. 3B). The degradation process in cell walls was not as visible in this sample due to its lower mass loss.

Even though the hyphae visibly attacked tyloses and cell walls, the degradation process did not have an effect on the permeability of sessile oak heartwood. If the hyphae would be able to digest tyloses, the permeability of sessile oak heartwood could increase.

CONCLUSION The permeability of spruce, beech and sessile oak heartwood degraded with white rot fungus Trametes versicolor L. Lloyd was measured. The measurement was done in all anatomical directions. A higher permeability was expected because of increase in porosity, but this was not confirmed. The permeability of degraded spruce and beech showed great variations. Unexpectedly, permeability in radial direction of the degraded spruce wood was lower compared to permeability in tangential direction. Better permeability in tangential direction was the result of the degradation process. A majority of toruses was digested, hence this pathway became more open. The degraded beech wood did not have better permeability than healthy beech wood. The degradation process did not develop pathways big enough to significantly increase the longitudinal permeability of the degraded beech wood. The degraded sessile oak heartwood was not permeable at all. The fungal activity did not digest tyloses which blocked the pathways in the heartwood. REFERENCES BABIAK, M. 1990. Wood – Water System. 1st edition. Zvolen: Vysoká škola lesnícka a drevárska Zvolen, 63 p. ISBN 80-228-0105-4. BABIAK, M., ČUNDERLÍK, I., KÚDELA, J. 1983. Priepustnosť a anatomická štruktúra bukového dreva (Permeability and anatomical structure of beech wood) [Výskumná správa] Zvolen: VŠLD, 23p. BABIAK, M., KÚDELA J. 1993. Transport of free water in wood. In Vacuum drying of wood. Zvolen: Technical University in Zvolen, p. 63–75. BABIAK, M., HUDEC, J., KURJATKO, S. 1995. Nonsteady-state permeability of wood. In Vacuum drying of wood. Zvolen: Technical University in Zvolen, p. 5461. BABIAK, M., KURJATKO, S., HUDEC, J. 2001. Výskum priepustnosti dreva pre tekutiny na Katedre náuky o dreve Drevárskej fakulty vo Zvolene. In Drevo – štruktúra a vlastnosti. Zvolen: Technická univerzita vo Zvolene, p. 9–16, ISBN 80-228-1094-0. BAILEY, P. J., PRESTON, R. D. 1969. Some aspects of softwood permeability 1. Structural studies with Douglas fir sapwood and Heartwood. In Holzforschung 23(14): 113120. BARI, E., MOHEBBY, B., NAJI, H., R., OLADI, R., YILGOR, N., NAZARNEZHAD, N., OHNO, K., M., NICHOLAS, D., D. 2018. Monitoring the cell wall characteristics of degraded beech wood by whiterot fungi: Anatomical, chemical, and photochemical study. In Maderas Ciencia y Tecnologia 20(1): 35–56, ISSN 0718-221X. BARI, E., NAZARNEZHAD, N., KAZEMI, S., M., GHANBARY M., A., T., MOHEBBY, B., SCHMIDT, O., CLAUSEN, C., A. 2015. Comparison between degradation capabilities of the white rot fungi Pleurotus ostreatus and Trametes versicolor in beech wood. In International Biodeterioration & Biodegradation 104: 231–237. COMSTOCK, G., L. 1965. Longitudinal permeability of green Eastern hemlock. In Forest Products Journal, 15(10): 441–449. COMSTOCK, G., L. 1967. Longitudinal permeability of wood to gases and nonswelling liquids. In Forest Products Journal, 17(10): 4146. COMSTOCK, G., L. 1968. Relationship between permeability of green and dry eastern hemlock. In Forests Products Journal, 18(8): 20–23.

COMSTOCK, G., L. 1970. Directional permeability of softwoods. In Wood and Fiber, 1(4): 283289. CÔTÉ, W., A., JR. 1963. Structural factors affecting the permeability of wood. In Journal of Polymer Science, Part –c 2: 231242. ČUNDERLÍK, I., HUDEC, J. 2002. Axial permeability of normal and tension beech wood. In Wood Structure and Properties 2002. Zvolen : Arbora Publishers, p. 6367. DANIHELOVÁ, A., REINPRECHT, L., SPIŠIAK, D., HRČKA, R. 2018. Impact of the Norway spruce sapwood treatment with the staining fungus Sydowia polyspora on its permeability and dynamic modulus elasticity. In Acta Facultatis Xylologiae Zvolen, 60(1): 13–18. DONALDSON, L., A., KROESSE, H., W., HILL, S., J., FRANICH, R., A. 2015 Detection of wood cell wall porosity using small carbohydrate molecules and confocal fluorescence microscopy. In Journal Microsc. 2015, 259, 228236 EMAMINASAB, M., TARMIAN, A., PURTAHMASI, K. 2015. Permeability of poplar normal wood and tension wood bioincised by Physisporinus vitreus and Xylaria longpipes. In International Biodeterioration & Biodegradation, 105: 178–184. EMAMINASAB, M., TARMIAN, A., POURTAHMASI, K., AVRAMIDIS, S. 2016. Improving the permeability of Douglas-fir (Pseudotsuga menziesii) containing compression wood by Physisporinus vitreus and Xylaria longipes. In International Wood Products Journal, 7(3): 110–115. ERICKSON, H., D., BALATINECZ, J., J. 1964. Liquid flow path into wood using polymerization techniques – Douglas Fir and styrene. In Forest Products Journal, 14: 293–299. GREEN, F., CLAUSEN, C., A. 1999. Production of polygalactouranose and increase of longitudinal gas permeability in southern pine by brown-rot and whote-rot fungi. In Holzforschung, 53: 563568. HANSMANN, CH. GINDL, W., WIMMER, R., TEISCHINGER, A. 2002. Permeability of wood – A review. In Wood Research, 47(4): 1–16, 2002. HUDEC, J., DANIHELOVÁ, A. 1992. Permeability of natural and hydrothermally treated wood. In Interakcia dreva s rôznymi formami energie. Zvolen: Technická univerzita vo Zvolene, p. 69–77. HUDEC, J. 1993. Permeability of reaction and non-reaction beech wood. In Zvolen – the city of wood science and practice, Zvolen. Zvolen: Technická univerzita vo Zvolene. P. 100–107. ISAACS, C., P., CHOONG, E., T., FOGG, P., J. 1971. Permeability variation within a cottonwood tree. In Wood Science, 3(4): 231–237. JAKES, J., E., HUNT, C., G., YELLE, D., J., LORENZ, L., HIRTH, K., GLEBER, S.-C., VOGT, S., GRIGSBY, W., FRIHART, C., R. 2015. Synchroton-based X-ray Fluorescence Microscopy in Conjunction with Nanoindentation to Study Molecular-Scale Interactions of Phenol-Formaldehyde in Wood Cell Walls. In ACS Appl. Mater. Interfaces 2015, 7, 65846589. KOVÁČIK, I. 1993. Identifikácia transportu vody cez pórovitú štruktúru dreva. [Kandidátska dizertácia]. Zvolen: Technical university in Zvolen. 103 p. KÚDELA, J. 1999. Selected physical properties of tension beech wood. In XIII. International Conference of wood technology Faculty. In Wood material of comprehensive appropriation and application. Warsaw. KÚDELA, J., ČUNDERLÍK, I. 2012. Bukové drevo – štruktúra, vlastnosti, použitie. (Beech wood – structure, properties, use, in Slovak) Zvolen: Technická univerzita vo Zvolene, 2012, 152 p, ISBN 978-80-228-2318-0. KURJATKO, S., ČUNDERLÍK, I., HUDEC, J. 1998. The morphometric characteristics of Wood structure in relation to ist permeability. In Wood Structure and Properties ´98. Zvolen. Arbora Publishers. P. 71–74. KURJATKO, S., SOLÁR, R., MAMOŇOVÁ, M. HUDEC, J. 2002. Selected properties of beech wood biodegraded by white-rot fungi. In Wood Structure and Properties ´02, Zvolen: Arbora Publishers. P. 57–62. LIHRA, T., CLOUTIER, A., ZHANG, S-Y. 2000. Longitudinal and transverse permeability of balsam fir wetwood and normal heartwood. In Veda o dreve a vláknach: časopis Spoločnosti drevárskej vedy a technológie, 32(2): 164178. LIN, Z., Y., LU, J. 2004. Mechanism of several different drying methods and their effects on liquid impregnation or permeability of wood. World Forestry Research. Scietific and Technological Information Centre, 17(1): 25–30.

MANTANIS, G., I., YOUNG, R., A. 1997. Wetting of Wood. In Wood Science and Technology 31: 339–353. NASSWETTROVÁ, A., ŠMÍRA, P., ZEJDA, J., NIKL, K., SEBERA, V. 2014. Axial permeability of beech wood treated by microwave heating for distilled water. In Wood Research, 59(1): 25–38. PALIN, M., A., PETTY, J., A. 1981. Permeability to water of the cell wall material of spruce heartwood. In Wood Science and Technology, 15: 161–169. PÁNEK, M., REINPRECHT, L., MAMOŇOVÁ, M. 2013. Trichoderma viride for Improving Spruce wood Impregnability. In BioResources, 2013, 8(2): 1731–1746. POŽGAJ, A., CHOVANEC, D., KURJATKO, S., BABIAK, M. 1997. Štruktúra a vlastnosti dreva. (Structure and Properties of Wood) Bratislava, ISBN 80-07-00960-4. REGINÁČ, L., ČOP, D., ŠTEFKA, V. 1977. Zariadenie na skúšanie priepustnosti pórovitých materiálov najmä dreva pre kvapaliny a plyny. Patent ŠSR PV 5308-77, 11.8.1977. REINPRECHT, L., PÁNEK, M. 2008. Bio-treatment of spruce wood for improving of its permability of soaking Part 2: Direct treatment with the fungus Trichoderma viride. In Wood research, 53(3): 18. REINPRECHT, L. 2016. Wood Deterioration, Protection and Maintenance. Wiley-Blackwell 2016, ISBN 978-1-119-10653-1, 376 p. RYPÁČEK, V. 1957. Biologie dřevokazných hub. Praha: Nakladatelství Československé Akademie Věd, 209 p. SIAU, J., F. 1995. Wood: Influence of Moisture on Physical Properties. Department of Wood Science and Forest Products. Virginia Polytechnic Institute and State University ISBN 0-9622181-0-3, 227 p. SLOVÁČKOVÁ, B. 2021a. Thermal conductivity of spruce, beech and oak heartwood degraded with Trametes versicolor L. Lloyd. In Acta Facultatis Xylologiae Zvolen, 63 (1): 1–7. ISSN 1336-3824. SLOVÁČKOVÁ, B. 2021b. Porosity of beech and oak heartwood degraded by Trametes versicolor L. Lloyd (Fagus sylvatica, L. and Quercus petraea, Matt. Liebl). In 9th Hardwood Proceedings, Part II. Volume 9, Part II. University of Sopron press, Sopron, 2021. p 98 – 104, ISBN 978-963-334-399-9 SOLÁR, R., REINPRECHT, L., KOŠÍKOVÁ, B. 2006. Selected „structure – properties“ relationships of wood degraded by white rot fungi. In Sustainability through new technologies for enhanced wood durability, COST Action E37, Working group 1 – Understanding durability mechanisms, London – United Kingdom, 39 p. SOLÁR, R., KURJATKO, S., LIPTÁKOVÁ, E., MAMOŇOVÁ, M. 2003. Alterations of beech wood (Fagus sylvatica, L.) physical properties in the corse of degradation by selected white-rot fungi. In Wood research, 48(1/2): 1–13. VITAS, S., SEGMEHL, J., S., BURGERT, I., CABANE, E. 2019. Porosity and Pore Size Distribution of Native and Delignified Beech Wood Determined by Mercury Intrusion Porosimetry. In Materials 2019, 12: 416 WARDROP, A., B., DAVIES, G., W. 1961. Morphological factors relating to the penetration of liquids into wood. In Holzforschung, 15(5): 129141. ACKNOWLEDGEMENTS This work was supported by the Slovak Research and Development Agency under contract No. APVV 16-0177 and the Internal Project Agency under contract 17/2020.

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

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 17−30, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.02

CHEMICAL COMPOSITION AND FIBRE CHARACTERISTICS OF BRANCH WOOD OF SELECTED HARDWOOD SPECIES Vladimír Ihnát – Mária Fišerová  Elena Opálená  Albert Russ  Štefan Boháček ABSTRACT The chemical composition and fibre characteristics of branch wood of hardwood species namely European beech (Fagus sylvatica L.), common oak (Quercus Robur L.), common hornbeam (Carpinus betulus L.), sycamore maple (Acer pseudoplatanus L.), black locust (Robinia pseudoacacia L.), silver birch (Betula pendula L.), European ash (Fraxinus excelsior L.), black poplar (Populus nigra L.), black alder (Alnus glutinosa L.), white willow (Salix alba L.), small-leaved lime (Tilia cordata Mill.) and royal paulownia (Paulownia tomentosa (Thunb.) Steud.) were determined. The branch wood of the examined tree species differed in the content of ash (0.440.86%), extractives soluble in dichloromethane and hot water (2.910.63%), Klason lignin (15.323.7%), acid-soluble lignin (1.783.53%), polysaccharides glucan (40.648.9%), xylan (16.725.3%), mannan (0.81.8%), galactan (0.51.1%) and arabinan (0.31.4%). The concentrations of inorganic elements (Ca, K, Mg, Na, P, Fe, Mn, Si, Zn, Cu) in the ash of the branch wood varied depending on the tree species. Differences in the fibre coarseness (4.158.52 mg/100 m), but more differences in the arithmetic average fibre length (0.41.20 mm) and the weighted average fibre length (0.511.45 mm) were found. Branch wood of common hornbeam, European ash, European beech and silver birch had a high content of polysaccharides and a lower content of Klason lignin, which is advantageous for the production of biofuels and for the pulping process. In addition, branch wood of these species has longer fibres with lower coarseness, which is advantageous in terms of pulp and paper quality. Key words: branch wood, hardwood species, polysaccharides, lignin, extractives, ash, fibre length, fibre coarseness.

INTRODUCTION Wood is best defined as a three-dimensional biopolymer composite composed of an interconnected network of cellulose, hemicelluloses and lignin with minor amounts of extractives, and inorganics. It is a highly variable and complex material that has inherent variability among species, within species, and also within a tree (ZOBEL and VAN BUIJTENEN 1989, LINDSTRӦM 2001, MARTIN et al. 2010, KRISHNA et al. 2017, TARELKIN et al. 2019). Generally, anatomical and chemical properties are the features that are often used to identify the species of woods. Utilization of these biomass resources is critically dependant on the in-depth knowledge of their morphological and chemical characteristics. The study 17

of chemical characteristics of wood is important to exploit the potential utilization of wood such as that for pulp and paper, bioethanol, biocomposites and carbonized wood production, whereas the fibre characteristics are studied in order to discover the utilization of wood fibers, such as in pulp, paper and fibreboard production. All processes used for the conversion of biomass feedstocks are sensitive to feedstock composition and quality to various extents. Reduction in lignin content improves enzymatic hydrolysis, which along with pretreatment, is the most expensive component in the production of bioethanol. It typically results in a proportional increase in the cellulose content per unit mass (DINUS 2001). Ash content and composition, heating value, elemental ratios and proportion of lignin, cellulose, and hemicelluloses are some of the broad compositional characteristics used to screen biomass feedstocks for biofuels applications (DINUS et al. 2001). Chemical composition of wood and fibre characteristics are two important parameters which determine its suitability as raw material for the production of pulp and paper. The extractive content has a direct effect on the pulp yield,; its a high content reduces the pulp yield. On the other hand holocellulose, α-cellulose and lignin content are mainly related to pulping behaviour, (ZOBEL and VAN BUIJTENEN 1989). The fibre morphological properties are important quality parameters for pulp and paper properties. In fact, the fibre length and coarseness greatly influence the quality and properties of the final product, e.g., they are frequently correlated with the physical and mechanical properties of paper and paperboard (SETH 1995, EL-HOSSEINY and ANDERSON 1999, ANJOS et al. 2014, KEAYS et al. 2015). Proportion of stem wood of deciduous trees at harvest is 68%. The share of crown and branches ranges from 10% to 19%, the root part makes up 825% of the total weight of biomass. The collection of branch wood can substantially increase the quantity of wood fibre per area of forest harvested. According to OKAI and BOATENG (2007), approximately 35 to 50% of wood biomass is left in the forest in form of stumps, branches and crowns. The logging residues (especially branch wood) make up a significant quantity of wood volume, and its utilization can increase the yield by about 60% (SHMULSKI and JONES 2011). Forest residues (e.g. branch wood, bark, etc.) are currently highly preferred as a cover for soil amendment, industrial fuel and are possible raw materials for the production of several organic products. Particular emphasis is placed on the use of branch wood due to the decline in stem wood resources (LEICH and MILLER 2017, DADZIE et al. 2018, ZHAO et al. 2019a,b) or the low availability per unit area, for example in arid and semi-arid regions (ANDERSEN and KRZYWISKI 2007, LI et al. 2018). However, branch wood is still little used industrially, as it is associated with many disadvantages. Branch wood contains a higher amount of bark and is less uniform in comparison with stem wood (SCHMULSKI and JONES 2011). It needs an intricate treatment before its utilization, which can reduce production efficiency (NURMI 2007). When considering the use of branch wood as a potential raw material, it is time to consider what strategy should be used to manage restrictions on branch wood to be useful for commercial purposes. Branch wood as a part of a tree has been discussed broadly in recent publications (SHMULSKI and JONES 2011, DADZIE et al. 2018, ZHAO et al. 2019a, b). Investigation on chemical composition and fibre characteristics of the branch wood of the hardwood species is still rare. The results of an extensive research of branch wood of eleven species growing in India have shown that branches are not identical to stem in all the technical properties but the difference between the two is not so large as to treat the branch material separately in the manufacture of pulp, paper and boards. Furthermore, the branch diameter is an important parameter of the quality of raw material, as it affects many properties, such as the proportion of bark, the fibre length and the density of wood and bark of some species (BHAt et al. 1985). 18

Branch wood can be used for pulp papermaking or wood-based panels (ZHAO et al. 2018), low-grade paper (ZHAO et al. 2019a), or glued plates (ZHAO et al. 2019b). Branch wood of Acacia gerrardii, Tamarix aphylla, and Eucalyptus camaldulensis has several drawbacks that markedly limit its potential for commercial uses (SUANSA et al. 2020). It might not be favorable for particle board, flake board, or fibre board because of its high shrinkage. Even though all of the fibres show suitability as a raw material for pulp and paper, the quality is low due to the high density of vessels or parenchyma proportions. However, branch wood of all examined species might be used as a blending material (papermaking and glued plates) or for light construction purposes. While considering the chemical composition of branch wood, classes of green products, such as biofuel, bioenergy, and biochar might maximize the value of branch wood. The aim of this work was to determine the chemical composition and fibre characteristics of branch wood from hardwood species in order to assess the possibility of their utilization in the production of biofuels, pulp and paper.

MATERIAL AND METHODS Materials Branch woods were selected from tree species with a higher occurrence in forests of the Slovak Republic such as European beech (Fagus sylvatica L.), common oak (Quercus Robur L.), common hornbeam (Carpinus betulus L.) and sycamore maple (Acer pseudoplatanus L.) and less represented tree species black locust (Robinia pseudoacacia L.), European ash (Fraxinus excelsior L.), silver birch (Betula pendula L.), black poplar (Populus nigra L.), black alder (Alnus glutinosa L.), white willow (Salix alba L.) and small-leaved lime (Tilia cordata Mill.). Royal paulownia (Paulownia tomentosa (Thunb.) Steud.) was chosen for its high potential as a fast growing species. All samples of branch wood were collected in the region of Bratislava, Slovak Republic, from different areas and trees. The diameter of the branches ranged from 2.0 to 4.5 cm. Average samples were prepared for every tree species. Methods Preparation of samples The branch wood samples were debarked and chips were prepared on a single-knife laboratory disc chipper. Sawdust was prepared from chips in the device Brabender for determination of the chemical composition. For determining the fibre length and coarseness samples were prepared from debarked branches, from which 2 cm high discs were prepared by hand and then samples with dimensions of half-matchstick. Samples prepared in this way were macerated under reflux with a mixture of 30% hydrogen peroxide and glacial acetic acid in a ratio of 1: 1 for 1 to 2 hours until the wood samples turned white (BEREŠOVÁ and ČUNDERLÍK 1999). After cooling, the samples were filtered and washed with water, the fibres were separated after stirring the sample in distilled water. Analyses The ash content was determined according to ISO 1762. The extractives content in dichloromethane was determined according to Tappi T 204 cm-94, and the extractives content in hot water according to Tappi T 207 cm-08. The Klason lignin content was determined according to Tappi T 222 om-98 and acid-soluble lignin content according to Tappi UM 250. Standard deviations were calculated from duplicate measurements for all tree species. Polysaccharides glucan, xylan, mannan, galactan and arabinan content was calculated based on the concentrations of glucose, xylose, mannose, galactose and arabinose in the 19

hydrolysate after determination of lignin. The hydrolysate before determination of monosaccharides was treated with 4% H2SO4 at 121°C for 2 hours, to hydrolyse the oligosaccharides. Subsequently, the hydrolysate was neutralized with BaCO3. The concentration of monosaccharides was determined by using HPLC with Rezex ROA H+ column. The mobile phase was 0.005 N H2SO4 at a flow rate of 0.5 ml.min-1 at 30°C. The samples were passed through a 22 µm filter before testing. Inorganic elements in ash The concentration of K, Ca, Mg, Na, Fe, Mn, Zn, Cu elements in the branch wood was determined by atomic absorption spectrometry (AAS) using flame atomization according to the method Tappi T 266. The soluble portion of the ash in hydrochloric acid was used in their determination. The concentration of P element was determined by a spectrometric method with ammonium molybdate according to standard STN EN ISO 6878. The concentration of Si in the branch wood was calculated from the gravimetrically determined SiO2 content in the insoluble ash residue after a treatment with hydrofluoric acid. Standard deviations were calculated from duplicate measurements for all tree species. Fibre length and coarseness Fibre length was determined using ADV-3 analyser (PPRI Bratislava, Slovak Republic). The basis of the analyser is a conductivity sensor, through which a very dilute suspension of fibres passes, the surface of which is covered with an electric charge double layer and the outer part of the charges participates on conducting current in the area of the sensor electrodes. Electrical impulse arises by passing the fibre through the sensor interface at the output of the amplifier. Its length depends on the speed of the fibre passing through the capillary and on the fibre length. The fibre lengths were measured in the range of 0-5 mm with the number of 10,000 fibres in the sample. Arithmetic and weighted fibre length distributions were obtained by measuring the fibre suspension and the arithmetic average fibre lengths and weighted average fibre lengths were calculated. The arithmetic average fibre length was calculated as the sum of all individual fibre lengths divided by the total number of fibres measured. The weighted average fibre length was calculated as the sum of individual fibre lengths squared divided by the sum of the individual fibre lengths. Fibre coarseness was determined using ADV-3 analyser by a special measurement of the arithmetic average fibre length of a suspension containing 1 mg of fibres. Standard deviations were calculated from duplicate measurements of fibre length and coarseness for all tree species. RESULTS AND DISCUSSION Chemical composition The results of the chemical composition of branched woods of different tree species are shown in Tab. 1. The values of each of the chemical components varied between the examined species. The differences in the content of ash and extractives were the highest of all chemical components. The content of ash of the branch woods were found in the range from 0.44 to 0.86%. Branch wood of black locust had the highest ash content, while the smallest one was found in the branch wood of royal paulownia. In our previous work (FIŠEROVÁ et al. 1986), the ash content in branch wood of black poplar varied from 0.5 to 0.8 % depending on the diameter of the branches. According to CÁRDENAS-GUTIÉRREZ et al. (2018), the ash content in branch wood of various hardwood species ranged from 0.74 to 1.13%. The differences in ash content are caused to the fact that the content of inorganic components varies greatly depending on the environmental conditions in which the tree has grown, also on the tree species and the diameter of branch. 20

The content of extractives soluble in dichloromethane and hot water in the branch wood of the examined tree species ranged from 2.90 to 10.63% (Tab. 1). Small-lived lime branch wood had the highest content of extractives, while the lowest was determined in branch wood of European ash. The content of extractives soluble in dichloromethane ranged from 0.26 to 5.18%. The branch wood of the small-leaved lime contained the most extractives soluble in dichloromethane, while the smallest one was found in the branch wood of common hornbeam. The content of extractives soluble in hot water ranged from 2.49 to 5.50%. The branch wood of royal paulownia contained the most extractives soluble in hot water and the least one had the branch wood of European ash. The content of extractives in wood varies greatly depending on tree species, as it is controlled by genetics (ZOBEL and JACKSON 1995). The combined effect of genetics and methods and the solvents used on extraction leads to a wide range of results in the literature. The total content of extractives in black poplar branch wood ranged from 3.95 to 4.90% (FIŠEROVÁ et al. 1986) and for various hardwood species ranged from 6.9 to 15.3% (CÁRDENAS-GUTIÉRREZ et al. 2018). The content of Klason and acid-soluble lignin in the branch wood of the examined tree species ranged from 17.97 to 26.53% (Tab. 1). The highest lignin content was found in the branch wood of black poplar, while the smallest in the branch wood of small-leaved lime. The Klason lignin content ranged from 15.3 to 23.7%. The highest Klason lignin content was found in the branch wood of royal paulownia, while the lowest one in the branch wood of small-leaved lime. The content of acid-soluble lignin ranged from 1.78 to 3.53%. Branch wood of common hornbeam contained the most of acid-soluble lignin while the black alder contained the least. The total content of lignin in branch wood of various hardwood species ranged from 17.64 to 28.87% (CÁRDENAS-GUTIÉRREZ et al. 2018), while for black poplar varied from 25.35 to 28.67%, depending on the diameter of the branches (FIŠEROVÁ et al. 1986). Results obtained for the lignin content differ, which may be related to a different tree morphology. This is supported by the fact that, so far as it is known, the content of lignin and its structure differ depending on the region of the woody xylem (SJÖSTRÖM 1993). The content of polysaccharides glucan, xylan, mannan, galactan and arabinan in the branch wood of the investigated tree species ranged from 63.0 to 69.6% (Tab. 1). The highest content of polysaccharides was found in the branch wood of common hornbeam, while the lowest in the branch wood of royal paulownia. Glucan and xylan formed the highest proportion of polysaccharides. The glucan content ranged from 40.6 to 48.9%. The highest content of glucan was determined in the branch wood of black poplar, while the lowest in the branch wood of silver birch, black alder and royal paulownia. The xylan content ranged from 16.7 to 25.3%. The highest xylan content was found in the branch wood of silver birch, while the lowest in the branch wood of black poplar. The content of other polysaccharides mannan, galactan and arabinan ranged from 1.9 to 3.5%, while the highest content was determined in the branch wood of small-leaved lime and royal paulownia, the lowest in the branch wood of common hornbeam. The mannan content ranged from 0.8 to 1.8%, galactan from 0.5 to 1.1% and arabinan from 0.3 to 1.4. The polysaccharides content in the black poplar branch wood ranged from 63.43 to 67.80% depending on the diameter of the branches (FIŠEROVÁ et al. 1986). According to the results from the literature (FIŠEROVÁ et al. 1986, CÁRDENASGUTIÉRREZ et al. 2018), branch wood of various hardwood species contains less polysaccharides but more lignin, extractives and ash than stem wood. Based on the obtained results, it may be concluded that the contents of individual wood chemical components vary significantly as a function of the wood species. Therefore, it is almost impossible to find any relations between the variation in the content of one component and the contents of other examined wood chemical components. The main restrictive factor that influences these relationships lies in the fact that the contents of 21

individual components (except ash) are more or less determined by the pre-treatments that preceded their isolation (FENGEL and WEGENER 1984). Among the examined tree species the branch wood of common hornbeam, European beech, European ash and silver birch had the highest content of polysaccharides (68.569.6%), as well as glucan and xylan content (65.967.7%), from which bioethanol or biobutanol is produced by hydrolysis. A great advantage of the branch wood of these tree species is also the low content of Klason lignin (16.218.7%), as lignin plays a negative role in the production of biofuels and pulp. Lignin can reduce the strength of paper because it could be a barrier for hydrogen bonding and could be an inhibitor in the hydrolysis process in the production of bioethanol and biobutanol. The high proportion of glucan, low content of lignin and extractives should contain wood for the production of biocomposite (ŠPANIČ et al. 2018). Branch wood of black poplar meets the most of these requirements from the examined species. Tab. 1 Chemical composition of branch woods of hardwood species.

0.50 0.33 3.88 4.21 21.0 2.40 23.40 41.1

±0.01 ±0.03 ±0.14 ±0.14 ±0.20 ±0.07 ±0.16

±0.07

0.51 0.26 3.39 3,65 16.2 3.53 19.73 45.3

±0.01 ±0.03 ±0.06 ±0.10 ±0.00 ±0.02 ±0.04

±0.49

0.81 0.47 3.76 4.23 23.0 1.93 24.93 45.7

±0.02 ±0.01 ±0.01 ±0.02 ±0.03 ±0.05 ±0.02

±0.35

0.86 0.76 3.89 4.65 22.6 2.91 25.51 42.1

±0.01 ±0.03 ±0.03 ±0.01 ±0.07 ±0.03 ±0.03

±0.28

0.49 1.48 3.43 4.91 18.5 3.11 21.61 40.6

±0.01 ±0.01 ±0.04 ±0.02 ±0.07 ±0.06 ±0.13

±0.21

0.53 0.41 2.49 2.90 18.7 2.65 21.32 48.0

±0.00 ±0.01 ±0.02 ±0.03 ±0.05 ±0.08 ±0.03

±0.28

0.64 0.78 2.65 3.43 23.5 3.03 26.53 48.9

±0.02 ±0.02 ±0.01 ±0.02 ±0.18 ±0.06 ±0.12

±0.42

0.65 0.84 4.35 5.19 23.6 1.78 25.38 40.6

±0.03 ±0.06 ±0.11 ±0.11 ±0.04 ±0.01 ±0.03

±0.42

0.62 0.66 2.97 3.63 20.8 2.79 23.59 45.8

±0.01 ±0.05 ±0.02 ±0.03 ±0.18 ±0.08 ±0.08

±0.49

0.80 5.18 5.45 10.63 15.3 2.67 17.97 41.2

±0.00 ±0.05 ±0.09 ±0.15 ±0.08 ±0.01 ±0.07

±0.21

0.44 1.04 5.50 6.54 23.7 2.42 26.12 40.6

±0.01 ±0.03 ±0.01 ±0.02 ±0.06 ±0.05 ±0.04

* Klason lignin was corrected for ash content in Klason lignin.

±0.14

23.0

±0.14

22.4

±0.49

18.9

±0.14

21.6

±0.28

25.3

±0.00

18.2

±0.28

16.7

±0.35

20.4

±0.35

18.3

±0.49

20.3

±0.35

18.9

±0.21

1.10

±0.14

0.80

±0.05

1.20

±0.12

1.40

±0.08

1.50

±0.10

0.90

±0.05

1.40

±0.08

1.50

±0.11

1.70

±0.08

1.80

±0.14

1.70

±0.06

±0.03

±0.02 ±0.89

0.80 0.50

±0.03

0.50

±0.05

0.50

±0.02

0.80

±0.06

0.50

±0.01

0.50

±0.02

0.60

±0.06

0.80

±0.07

0.90

±0.04

1.00

±0.13

1.10

±0.14

Total

1.20

±0.12

Arabinan

22.1

±0.35

Galactan

±0.35

Mannan

0.54 0.31 3.40 3.71 17.5 4.40 21.90 44.3

±0.00 ±0.01 ±0.01 ±0.02 ±0.18 ±0.08 ±0.19

Xylan

Glucan

Polysaccharides [%]

Total

Acid-soluble

Lignin [%] Klason*

Total

European beech Fagus sylvatica L. Common oak Quercus robur L. Common hornbeam Carpinus betulus L. Sycamore maple Acer pseudoplatanus L. Black locust Robinia pseudoacacia L. Silver birch Betula pendula L. European ash Fraxinus excelsior L. Black poplar Populus nigra L. Black alder Alnus glutinosa L. White willow Salix alba L. Small-leaved lime Tilia cordata Mill. Royal paulownia Paulownia tomentosa (Thunb.) Steud.)

Ash [%]

Hot water

Species

Dichloromethane

Extractives [%]

0.4 68.8

1.40 67.1

±0.05 ±0.42

0.60 69.6

±0.03 ±0.64

1.30 67.6

±0.02 ±0.85

0.50 66.4

±0.02 ±0.64

0.60 68.5

±0.03 ±0.35

1.40 69.0

±0.04 ±0.67

0.30 67.9

±0.02 ±0.89

0.60 63.9

±0.02 ±0.87

0.60 67.3

±0.03 ±0.98

0.70 65.0

±0.02 ±0.71

0.70 63.0

±0.03 ±0.31

Inorganic elements in ash The inorganic elements present in the biomass form its ash content and represent the waste stream during its conversion to biofuels and are a source of biochar and slagging during thermochemical conversion. Knowledge of the ash content and composition is essential regardless of the conversion pathway or end product. Ash is composed of many major and minor elements that trees need for their growth. The major elements include Ca, K, Fe, Mg, P, Na and Mn (VESTERINEN 2003). The amount of the major elements in wood ash varies with the type of plant tissues that are part of the wood (PITMAN 2006). In Tab. 2, there are concentrations of inorganic elements (Ca, K, Mg, Na, P, Fe, Mn, Si, Zn, Cu) in the ash of the branch wood of the examined tree species. Of the major components, the elements Ca and K were the most abundant. The concentration of Ca ranged from 287 to 1776 mg·kg1, while the concentration of K was in a very wide range from 558 to 3412 mg·kg1. Most of Ca was in the branch wood of black locust, while the least was in the branch wood of royal paulownia. Therefore in the case of K, the concentration was the highest in the branch wood of small-leaved lime and the lowest in the branch wood of common hornbeam. The inorganic elements Ca and K are essential for tree metabolism and various physiological processes associated with growth. In recent years, special interest has been attributed therefore to the effect of both cations on cambial activity and xylem development (FROMM 2010). The third most common abundant chemical element in the branch wood of the examined tree species was Mg, the concentration of which ranged from 141 to 521 mg·kg1, the highest concentration was determined in the branch wood of sycamore maple and the lowest in the branch wood of common hornbeam. The concentration of P in the branch wood ranged from 170 to 408 mg·kg1, the highest concentration was determined in the branch wood of black poplar and the lowest in the branch wood of sycamore maple. The concentration of Na in the branch wood of the examined tree species ranged from 9.20 to 97.7 mg·kg1, while it was mostly found in the branch wood of black poplar and the least in the branch wood of common oak. The concentration of Mn ranged from <0.5 to 122 mg·kg1, while the highest concentration was determined in the branch wood of European beech and the lowest in the branch wood of black locust, European ash, white willow and royal paulownia. Of the major inorganic elements, Fe was the least represented in the branch wood of the examined species. The concentration of Fe ranged from 5.82 to 19.2 mg·kg1, the highest concentration was determined in the branch wood of small-leaved lime and the smallest in the branch wood of common hornbeam. The concentration of minor elements Si, Zn and Cu in ash is important from the point of view of industrial use of branch wood, as their concentration can affect some technological processes. The concentration of Si in the branch wood of the examined species ranged from 28.5 to 116 mg·kg1, while the highest concentration was determined in the wood of small-leaved lime and the lowest in the branch wood of European beech. The concentration of Zn in the branch wood ranged from 2.59 to 24.3 mg·kg1, the highest concentration was in the wood of black poplar and the lowest in the branch wood of black locust. The amount of Cu was the smallest of the above-mentioned minor elements in the branch wood, its concentration ranged from 1.47 to 4.48 mg·kg1. The branches and twigs contain more ash-forming components than stem wood. The K and Na content in the branches and twigs is significantly higher than that in stem wood (WANG and DIBDIAKOVA 2014). When wood was used as a fuel, it was found that ash slagging tendency correlated well to content of Si and K in the fuel (ÖHMAN et al. 2004).

Tab. 2 Inorganic elements concentration in branch wood of hardwood species. Species European beech Fagus sylvatica L. Common oak Quercus robur L. Common hornbeam Carpinus betulus L. Sycamore maple Acer pseudoplatanus L. Black locust Robinia pseudoacacia L. Silver birch Betula pendula L. European ash Fraxinus excelsior L. Black poplar Populus nigra L. Black alder Alnus glutinosa L. White willow Salix alba L. Small-leaved lime Tilia cordata Mill. Royal paulownia Paulownia tomentosa (Thunb.) Steud.)

Ca 1007 ±27.0

Inorganic elements [mg·kg1] K Mg P Si Na Mn 790 191 178 28.5 77.9 122

±14.1

691 1385

±16.4

1483 ±14.1

±21.9

558

±3.10

1478 1863 ±20.5

±19.6

1776 1205 ±8.50

±7.07

482 1240

±12.1

±27.0

834 1551

±24.0

±14.8

932 1517

±15.5

±17.5

1354 1832 ±21.0

±19.0

901 1733

±18.4

±24.0

1088 3412 ±17.0

±20.5

287 1325

±12.0

±10.6

±2.12

196

±2.12

141

±3.54

521

±12.1

275

±5.66

252

±6.36

244

±1.91

271

±4.95

385

±4.24

241

±1.32

440

±12.0

177

±4.24

±2.83

277

±4.95

206

±2.83

170

±5.36

220

±1.41

327

±8.48

233

±3.54

408

±2.12

368

±2.83

358

±8.19

359

±3.43

213

±4.24

±0.71

60.2

±0.92

78.6

±0.92

57.6

±0.78

103

±0.85

82.5

±2.54

107

±0.64

51.4

±1.69

74.2

±0.14

91.1

±2.26

116

±1.41

97.7

±1.98

±0.13

9.20

±0.14

11.7

±1.00

18.9

±0.78

±3.54

115

±0.71

3.11

±0.03

1.71

±0.04

60.3 < 0.5

±0.88

88.2

±0.61

±0.02

64.7

±0.78

34.4 < 0.5

±1.10

97.7

±0.60

41.0

±0.42

±0.02

1.27

±0.11

26.9

±0.42

93.7 < 0.5

±0.65

18.0

±0.21

±0.02

73.7

±0.49

11.0 < 0.5

±0.57

±0.02

Fe Zn Cu 8,62 4.70 1.71

±0.69

±0.78

14.5 3,50

±0.28 ±0.57

5.82

±0.08

9.23

±0.29

13,6

±0.35

12,1

±0.53

13.0

±0.28

9,96

±1.72

29.6

±0.57

6,98

±0.66

19.2

±0.28

19,0

±0.78

15.3

±0.42

3.94

±0.57

2.59

±0.12

23.5

±0.57

10.3

±0.46

24.3

±0.42

12.8

±0.42

21.7

±0.71

21.8

±0.42

19.1

±0.74

±0.04

2.47

±0.15

2.33

±0.14

1.47

±0.04

1.68

±0.29

2.58

±0.03

2.35

±0.11

4.21

±0.15

2.47

±0.13

2.86

±0.09

3.91

±0.11

4.48

±0.12

Wood ash has been commonly used in the past as a fertilizer applied to agriculture soil. Ca is a valuable element in ash, gives ash the properties of agricultural lime. Ash is also a good source of K, P, and Mg (SAHOTA 2007). The application of ash to the soil fulfills the idea not only sustainable soil use, but also the secondary use of waste (FAZEKAŠOVÁ 2003). Fibre characteristics Fibre length Fibre length is an important property with respect to paper and paperboard performance. Length and shape of fibres depends on the tree species (OLUWADARE and ASHIMIYU 2007). Deciduous trees produce short fibres, whereas coniferous trees produce long fibres. Typical fibres used in papermaking are short hardwood fibres (1–1.5 mm in length) and long softwood fibres (3–4 mm in length). Short fibres provide smoothness for printing papers, while long fibres provide paper strength. Long fibres create a stronger network compared to shorter fibres (DINWOODIE 1965). The arithmetic and weighted fibre length distributions of the branch wood which differed significantly for examined species are showed in Fig. 16. European beech (Fig. 1A) and common hornbeam (Fig. 2A) were characterized by a wide fibre length distribution. Silver birch (3B), white willow (5B), European ash (4A), black poplar (4B), royal paulownia (6B), black alder (5A), small-leaved lime (6A), black locust (3A) and common oak (1B) ranged in a narrower range of fibre lengths. The fibre length distributions of branch wood of sycamore maple (2B) were in a very narrow range of fibre lengths. The arithmetic average fibre length and weighted average fibre length of the branch wood of the examined tree species are given in Tab. 3. The arithmetic average fibre length and the weighted average fibre length were determined from the fibre length analysis. The arithmetic average fibre length of the branch woods ranged from 0.40 to 1.20 mm, while the weighted average fibre length ranged from 0.51 to 1.45 mm. The arithmetic average and weighted average 24

fibre length of the branch woods increased in the following order: sycamore maple < common oak < black locust < small-lived lime < royal paulownia < black alder < European ash < black poplar < white willow < silver birch < common hornbeam < European beech.

Fig. 1 Arithmetic and weighted fibre length distributions of branch wood of European beech (Fagus sylvatica L.) (A) and common oak (Quercus robur L.) (B).

Fig. 2 Arithmetic and weighted fibre length distributions of branch wood of common hornbeam (Carpinus betulus L.) (A) and sycamore maple (Acer pseudoplatanus L.) (B).

Fig. 3 Arithmetic and weighted fibre length distributions of branch wood of black locust (Robinia pseudoacacia L.) (A) and silver birch (Betula pendula L.) (B).

Fig. 4 Arithmetic and weighted fibre length distributions of branch wood of European ash (Fraxinus excelsior L.) (A) and black poplar (Populus nigra L.) (B).

Fig. 5 Arithmetic and weighted fibre length distributions of branch wood of black alder (Alnus glutinosa L.) (A) and white willow (Salix alba L.) (B).

Fig. 6 Arithmetic and weighted fibre length distributions of branch wood of small-leaved lime (Tilia cordata Mill.) (A) and royal paulownia (Paulownia tomentosa (Thunb.) Steud.) (B).

The branch wood of Betula platyphyla Roth contained significantly shorter fibre length than stem wood (ZHAO et al. 2019a). The fibre length of the branch wood of the European black alder (Alnus glutinosa Gaertn.) was approximately about 0.2 mm shorter than the stem wood (VURDU 1977). According to the results published in the literature (WANG 1998), fibres with an average length greater than 0.4 mm are suitable for papermaking. Branch wood could be suitable for the production of low-grade paper and glued plates due to its medium fibre length (ZHAO et al. 2019a). Based on the above, it can be stated that the branch wood of the examined tree species can be used for the production of paper of different quality alone or in a mixture, with the exception of branch wood of sycamore maple (Fig. 2B) and common oak (Fig. 1B), which contained which a high proportion of short fibres. Fibre coarseness Fibre coarseness is defined as the weight per unit length of fibre expressed as milligrams per 100 m. The number of fibres per unit weight is related to the weight of each individual fibre, to the fibre coarseness and to the percentage of fibre wall in the fibre volume. There is a strong correlation between the number of fibres per unit weight and the coarseness of the fibres (WATSON and BRADLEY 2009). In short, lower fibre coarseness means higher sheet tensile strength, greater bonding area, and more fibres per tonne of pulp, all of which are attributes that are highly prized by technically sophisticated papermakers. Measuring the fibre coarseness has several advantages over measuring the widths of fibres since it is not only much easier and quicker but also includes the effects of fibre thickness, the size of the central canal (or lumen), and the density of the cellulosic material composing of fibres. The coarseness of fibres has an important effect on many properties of paper. Higher fibre coarseness is associated with thicker walled fibres. These fibres produce a more open and loosened paper structure. The corresponding papers are more porous, bulkier, and more absorbent. A better consolidation of the paper web is expected with a large number of fibres per milligram in the paper structure. The

result is better interconnection and better properties dependent on bonding (tensile, folding, surface strength, surface smoothness) are achieved. The fibres coarseness of the branch wood of European beech, common oak, common hornbeam, sycamore maple, black locust, silver birch, European ash, black poplar, black alder, white willow, small-leaved lime and royal paulownia ranged from 4.15 to 8.52 mg/ 100 m (Tab. 3). The fibre coarseness of the branch wood of examined tree species decreased in the following order: sycamore maple > black locust > common oak > European ash > common hornbeam > small-leaved lime > royal paulownia > European beech > white willow > black alder > black poplar > silver birch. According to the results published in the work ZHAO (2019a), the branch wood of silver birch had significantly shorter fibres and lower coarseness than the stem wood. Branch wood of sycamore maple has the highest fibre coarseness and the shortest fibres from the examined tree species, therefore it is the least suitable for paper production. Branch wood of European beech, common hornbeam, silver birch have fibres with low coarseness and longer fibre length, so they are the most suitable of the examined tree species for pulp and paper production. In addition, they have a high content of polysaccharides and a low content of Klason lignin, which is advantageous from the point of view of pulp production. Branch wood can be used for light construction purposes or as a mixed material in paper production (SUANSA and AL-MEFARREJ 2020). The results presented in the work (SETH 2011) show that the fibres with higher coarseness have thicker walls, a smaller specific surface and a smaller bonding area as a result of which the pulp have lower strength. Tab. 3 Fibre length and coarseness of branch wood of hardwood species. Species European beech Fagus sylvatica L. Common oak Quercus robur L. Common hornbeam Carpinus betulus L. Sycamore maple Acer pseudoplatanus L. Black locust Robinia pseudoacacia L. Silver birch Betula pendula L. European ash Fraxinus excelsior L. Black poplar Populus nigra L. Black alder Alnus glutinosa L. White willow Salix alba L. Small-leaved lime Tilia cordata Mill. Royal paulownia Paulownia tomentosa (Thunb.) Steud.)

Arithmetic average fibre length ±SD [mm] 1.20 ± 0,01

0,52

± 0,00

1.09

± 0,01

0.40

± 0,00

0.60

± 0,00

0.86

± 0,01

0.74

± 0,01

0.76

± 0,00

0.72

± 0,01

0.81

± 0,01

0.61

± 0,01

0.71

± 0,01

Weighted average fibre length ±SD [mm] 1.45 ± 0,01

0.66

± 0,01

1.27

± 0,01

0.51

± 0,01

0.73

± 0,01

1.01

± 0,01

0.91

± 0,01

0.87

± 0,01

0.84

± 0,01

0.92

± 0,01

0.74

± 0,01

0.84

± 0,01

Fibre coarseness ±SD [mg/100 m] 5.36 ± 0,06

6.96

± 0,12

5.93

± 0,06

8.52

± 0,10

8.06

± 0,04

4.15

± 0,08

6.70

± 0,14

4.17

± 0,11

5.24

± 0,10

5.29

± 0,05

5.91

± 0,06

5.71

± 0,08

CONCLUSIONS Branch wood can be considered to be an important natural resource of future for production of biofuels, pulp and paper, but it has several disadvantages which limit its potential for commercial use compared to stem wood. The content of polysaccharides in the branched wood of the examined tree species ranges from 63 to 69.6%, it decreases in the following order: common hornbeam > European ash > European beech > silver birch > black poplar > sycamore maple > white willow > common oak > black locust > small-leaved lime > black alder > royal paulownia. The Klason lignin content in branch wood ranged from 15.3 to 23.7%, it increased in the following order: small-leaved lime < common hornbeam < European beech < silver birch < European ash < white willow < common oak < black locust < sycamore maple < black poplar < black alder < royal paulownia. The arithmetic average and weighted average fibre length of the branch woods increased in the following order: sycamore maple < common oak < black locust < small-lived lime < royal paulownia< black alder < European ash < black poplar < white willow < silver birch < common hornbeam < European beech. The fibre coarseness of the branch wood of examined species decreased in the following order: sycamore maple > black locust > common oak > European ash > common hornbeam > small-leaved lime > royal paulownia > European beech > white willow > black alder > black poplar > silver birch. In terms of industrial use branch wood is important Si concentration, which ranged from 28.5 to 116 mg·kg1, Zn concentration (2.5924.3 mg·kg1) and Cu concentration (1.474.48 mg·kg1). REFERENCES ANDERSEN, G.L., KRZYWINSKI, K. 2007. Mortality, recruitment and change of desert tree population in a hyper-arid environment. PloS ONE, 2(2), e208. DOI:10.1371/journal.pone.0000208. ANJOS, O., SANTOS, A., SIMÕES, R., PEREIRA, H. 2014. Morphological, mechanical and optical properties of cypress. In Holzforschung, 68:861–995. doi: 10.1515/hf-2013-0125. BEREŠOVÁ, K., ČUNDERLÍK, I. 1999. Morfológia vláknitých buniek smreka s rôznym stupňom poškodenia. In Drevo, štruktúra a vlastnosti, Zvolen: TU vo Zvolene, 1999, ISBN 80-228-0887-3, pp 1722. BHAT, K.M., BHAT, K.V., DHAMODARAN, T.K. 1985. Wood and bark properties of branches of selected tree species growing in Kerala. Research Report, Kerala Forest Research Institute, Peechi, Thrissur, India, December 1985. CÁRDENAS-GUTIÉRREZ, M.Á., PEDRAZA-BUCIO, F.E., LÓPEZ-ALBARRÁN, P., RUTIAGAQUIŃONES, J.G.R. 2018. Chemical components of the branches of six hardwood species. In Wood Research, 63(5): 795808. DADZIE, P.K., AMOAH, M., EBANYENHE, E, FRIMPONG-MENSAH, K. 2018. Characterization of density and selected anatomical features of stem wood and branch wood of E-cylindricum, Eangolense and K-ivorensis from natural forests in Ghana. In European Journal of Wood and Wood Products, 76(2): 655667. DOI: 10.1007/s00107-017-1195-6. DINUS, R.J. 2001. Genetic improvement of poplar feedstock quality for ethanol production. In Appl. Biochem. Biotechnol., 91: 2334. DINUS, R.J., PAYNE, P., SEWELL, M.M., CHIANG, V.L., TUSKAN, G.A. 2001. Genetic modifications of short rotation popular wood: Properties for ethanol fuel and fiber productions. In Crit. Rev. Plant Sci., 20: 5169. DINWOODIE, J.M. 1965. The relationship between fibre morphology and paper properties of: A Review of Literature. In Tappi J., 48: 440-447. EL-HOSSEINY, F., ANDERSON D. 1999. Effect of fibre length and coarseness on the burst strength of paper. In Tappi J., 82:202–203.

FAZEKAŠOVÁ, D. 2003. Trvalo udržateľné využívanie pôdy – vymedzenie a hodnotenie indikátorov a parametrov vývoja pôdy (Sustainable land use - definition and evaluation of indicators and parameters of soil development). Prešov: FHPV PU, 76 pp. ISBN 80-8068-228-3. FENGEL, D., WEGENER, G. 1984. Wood – Chemistry, Ultrastructure, Reactions, Berlin: De Gruyter, 1984. FIŠEROVÁ, M., OPÁLENÁ, E., KOČANOVÁ, N. 1986. Chemické zloženie stromovej hmoty topoľa (Chemical composition of poplar tree mass). In Papír a celulóza, 41, V31V 35. FROMM, J. 2010. Wood formation of trees in relation to potassium and calcium nutrition. In Tree Physiology, 30(9): 1140–1147. https://doi.org/10.1093/treephys/tpq024. KRISHNA, K.P., SHAKTI, S.C., ARUN, N.K. 2017. In Wood is good: Current trends and future prospects in wood utilization. Singapore: Springer. LEICH, M., MILLER, S. 2017. Optimizing wood utilization based on whole tree inherent property maps, in Wood is good: Current trends and future prospects in wood utilization, PANDEY, K.K., RAMAKANTHA, V., CHAUHAN, S.S., KUMAR, A.N.A. (eds.), Springer, Singapore, pp 317. KEAYS J.L., HATTON J.V., BAILEY G.R., NEILSON R.W. 2015. Present and Future Uses of Canadian Poplars in Fibre and Wood Products. Available online: http://digitalcommons.usu.edu/aspen_bib/5234/ LI, Q., JIA, Z, FENG, L., He, L., Yang, K. 2018. Dynamics of biomass and carbon sequestration across a chrono sequence of Caragana intermedia plantations on Alpine sandy land. In Scientific Reports, 8(1), 12432. DOI: 10.1038/s41598-018-30595-3. LINDSTRӦM, H. 2001. Intra-tree models of basic density in Norvay spruce as an input to simulation software. In Silva Fennica, 34(4): 411421. DOI: 10.14214/sf622. MARTIN, J.A., ESTEBAN, L.G., DE PALACIOS, P., FERNANDEZ, F.G. 2010. Variation in wood anatomical traits of Pinus sylvestris L. between Spanish regions of provenance. In Trees Structure and Function, 24(6): 10171028. DOI: 10.1007/s00468-010-0471-4. NURMI, J. 2007. Recovery of logging residues for energy from spruce (Picea abies) dominated stands. In Biomass Bioenergy, 31(6): 375380. DOI: 10.1016/j.biobioe.2007.01.011. OKAI, R., BOATENG, O. 2007. Analysis of sawn lumber production from logging residues of branchwood of Aningeria robusta and Termina liaivorensis. In European Journal of Forest Research, 126(3): 385390. DOI: 10.1007/s10342-006-0157-z. OLUWADARE, A.O., ASHIMIYU, O.S. 2007. The Relationship between Fibre characteristics and Pulpsheet Properties of Leucaena leucocephala (Lam.) De Wit. In Middle-East Journal of Scientific Research, 2(2):6368. ÖHMAN, M., BOMAN, C., HEDMAN, H., NORDIN, A., BOSTRÖM, D. 2004. Slagging tendencies of wood pellet ash during combustion in residential pellet burners. In Biomass and Bioenergy, 27: 585596. PITMAN, R. 2006. Wood ash use in forestry – a review of the environmental impacts. In International Journal of Forest Research, 79(5): 563588. https://doi.org/10.1093/forestry/cpl041. SAHOTA, T.S. 2007. Lime or wood ash? Which one should apply? In Northwest Link, pp 67. SETH, R.S. 1995. The effect of fiber length and coarseness on the tensile strength of wet webs: a statistical geometry explanation. In Tappi J., 78(3): 99102. SETH, R.S. 2011. Fibre quality factor in papermaking – II The importance of fibre coarseness, In Symposium U-Materials Interactions Relevant to the Pulp, Paper and Wood Industries, pp 143197. DOI: https//doi.org./10.1557/PROC-197-143. SHMULSKY, R., JONES, P.D. 2011. Forest Products and Wood Science an Introduction sixth edition, John Wiley and Sons Ltd. Publisher, UK. SJÖSTRÖM, E. 1993. Wood Chemistry: Fundamentals and Applications, second edition, Academic Press, San Diego, 1993. SUANSA, N.I., AL-MEFARREJ, H.A. 2020. Branch wood properties and potential utilization of this variable resource. In BioResources, 15(2): 479491. ŠPANIČ, N., JAMBREKOVIČ, V., KLARIČ, M., 2018. Basic chemical composition of wood as parameter in raw material selection for biocomposite production. In Cellulose Chem. Technol., 52(34): 163169.

VURDU, H. 1977. Anatomical characteristics of stem, branch and root wood in European black alder (Alnus glutinosa L. Gaertn). Iova State University Capstones, Theses and Dissertation. TARELKIN, Y., HUFKENS, K., HAHN, S., VAN DENBULCKE, J., BASTIN, J.-F., ILONDEA, B.A., DEBEIR, O., VAN ACKER, J., BEECKMAN, H., DE CANNIÉRE, C. 2019. Wood anatomy variability under contrasted environmental conditions of common deciduous and evergreen species from central African forests. In Trees, 33(3): 893909. DOI:10.1007/s00468-019-01826-5. WANG, J. 1998. Important achievements in the field of paper fiber morphology in the 20th century. In Pap. Paper Making, 4: 79. DOI:10.13472/j.ppm.1998.04.002. WANG L., DIBDIAKOVA J. 2014. Characterization of ashes from different wood parts of norway spruce tree. In Chemical Engineering Transactions, 37: 3742. DOI: 10.3303/CET1437007 VESTERINEN, P. 2003. Wood ash recycling state of the art in Finland and Sweden. In VTT Research Report/PRO2/6107/03/31.10.203 www.cti2000it/solidi/wWoodAshReport%20VTTpdf WATSON, P., BRADLEY, M. 2009. Canadian pulp fibre morphology: Superiority and consideration for end use potential. In The Forestry Chronicle, 85(3): 401408. ZHAO, X., GUO, P., ZHANG, Z., WANG, X., PENG, H., WANG, M. 2018. Wood density and fibre dimensions of root, stem and branchwood of Populus unsuriensis Kom. Trees. In BioResorces, 13(3): 7026-7036. DOI: 1015376/biores.13.3.7026-7036. ZHAO, X., GUO, P., PENG, H., ZHAO, P., YANG, Y., ZHANG, Z. 2019a. Potential of pulp production from whole-tree wood of Betula platyphylla Roth based on wood characteristics. In BioResources, 14(3): 10. DOI: 10.15376/biores.14.3.7015-7024. ZHAO, X., GUO, P., ZHANG, Z., PENG, H. 2019b. Anatomical features of branch wood and stem wood of Betula costata Trautv. from natural secondary forest in China. In BioResources, 14(1): 19801981. DOI: 1015376/biores.14.1.1981-1991. ZOBEL, B.J., BUIJTENEN, J.P.V. 1989. Wood variation: Its causes and control, Berlin: SpringerVerlag. ZOBEL, J., JACKSON, B.J. 1995. Genetics of Wood Production. Berlin Heidelberg: Springer- Verlag. ACKNOWLEDGEMENTS This publication is the result of the project implementation: Centre of Excellence of Forest-based Industry, ITMS: 313011S735) supported by the Research & Development Operational Programme funded by the ERDF.

ADDRESSES OF AUTHORS Ing. Vladimír Ihnát, PhD. Ing. Mária Fišerová, PhD. Ing. Elena Opálená Mgr. Albert Russ Ing. Štefan Boháček, PhD. Pulp and Paper Research Institute Dúbravská cesta 14 841 04 Bratislava Slovak Republic ihnat@vupc.sk fiserova@vupc.sk opalena@vupc.sk russ@vupc.sk bohacek@vupc.sk

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 31−38, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.03

MICROCELLULOSE AS A MODIFIER FOR UF AND PF RESINS ALLOWING THE REDUCTION OF ADHESIVE APPLICATION IN PLYWOOD MANUFACTURING Jakub Kawalerczyk – Dorota Dziurka – Radosław Mirski – Joanna Siuda – Ján Sedliačik ABSTRACT The cellulosic derivatives has become widely investigated modifiers for urea-formaldehyde (UF) and phenol-formaldehyde (PF) adhesives in recent years. As microcellulose is more cost-effective and commercially available than nanocellulose, the aim of the presented study is to investigate the possibility of reducing the amount of applied UF and PF resins due to their microcellulose reinforcement. Plywood was manufactured with different amounts of modified resin and compared to the reference one containing rye flour or tannin based filler. The panels were tested in terms of shear strength and formaldehyde release (in the case of plywood glued with UF resin). The studies have shown that the addition of microcellulose (MFC) allowed reducing the amount of applied resin by 6% while maintaining the equally good bonding strength. Moreover, this reduction led to a decrease in the formaldehyde release by 9%. Key words: microcellulose, UF and PF resins, reinforcement, plywood, shear strength, formaldehyde.

INTRODUCTION Cellulose is an almost inexhaustible polymeric natural raw material with unique structure and properties, the highly functionalized, linear stiff‐chain homopolymer formed by D‐glucose building blocks (KLEMM et al. 2005). Significant amounts of cellulose are produced each year from various crop sources such as wood fibres, annual plants, flax, sisal, jute, hemp etc. (EICHHORN et al. 2001). Moreover, besides the plants it is also distributed through nature in algae, animals, minerals, fungi (HEINZE 2016). Nowadays, among the various cellulosic derivatives, microcellulose and nanocellulose (NCC) are mostly studied for possible ways of their application (HALDAR and PURKAIT 2020). Microcrystalline cellulose (MCC) was discovered in 1955 by Battista and Smith (THOORENS et al. 2014). Usually it is obtained by a partial hydrolysis of amorphous regions in cellulosic chains, which leads to formation of particles characterized by a diameter of approx. 50 μm and length of 100-1000 μm (TRACHE et al. 2016). Microcellulosic particles have some desirable features such as non-toxicity, biodegradability, tremendous surface area, high chemical reactivity and biocompability that create the opportunities for applications in many 31

industrial areas (AYRILMIS et al. 2016, VINEETH el al. 2019, PAŽITNÝ et al. 2019). The example of promising use for cellulosic derivatives is a wood-based materials industry. Many studies concerning the introduction of cellulosic particles as a modifier for the adhesives have been already done. The modification significantly influence the properties of resins such as a viscosity and their curing behaviour (TSCHURTSCHENTHALER 2012, VEIGEL et al. 2012). Moreover, studies performed by VEIGEL et al. (2011) showed that the nanocellulose addition caused a toughening of urea-formaldehyde (UF) adhesive-wood bond. VEIGEL et al. (2012) also determined how the modification with NCC affects the properties of particleboards and oriented strand boards (OSB). Studies have shown that their mechanical performance was significantly enhanced. However, the investigations performed by PAWLAK and BORUSZEWSKI (2018) indicate that the addition of MFC to UF resin in 3-layer lightweight particleboards manufacturing had no significant effect on their mechanical properties i.e. modulus of rupture, modulus of elasticity and internal bond. The example of widely used wood-based material having advantageous properties resulting from its layer structure is plywood (BEKHTA et al. 2016, KAWALERCZYK et al. 2019a). ZHANG et al. (2011) introduced silanized NCC to UF resin in plywood manufacturing process. The modification resulted in the improvement of bonding quality and formaldehyde emission from produced panels. Similar study conducted by KAWALERCZYK et al. (2020a) determined the effect of NCC and MFC addition to UF resin. Based on the results it was concluded that application of both modifying derivatives led to the improvement in plywood shear strength. However, the results varied depending on the type of used cellulosic derivative and the NCC was more effective. A similar dependence can be observed in studies concerning phenol-formaldehyde (PF) adhesive. Both MFC and NCC introduction to PF resin resulted in the improvement in plywood properties, however, the particles in the dimensional range of nanotechnology had a more advantageous effect (KAWALERCZYK et al. 2019b, KAWALERCZYK et al. 2020b). The improvement in the strength properties of plywood can result in the opportunity to reduce the resin spread rate, which is the least amount of resin necessary to produce a quality material (BEKHTA and MARUTZKY 2008). An interesting concept of the veneer surface modifications and their effect on the reduction in adhesive consumption became recently investigated. Studies have shown that both DBD (dielectric barrier discharge) plasma treatment (CAO et al. 2018) and the veneer compression (BEKHTA and MARUTZKY 2008) can lead to reduction in UF resin spread rate. The enhancement in glue line strength due to the adhesives modification can also reduce their consumption in plywood manufacturing. The introduction of fumed nano-SiO2 to MUPF (melamine-urea-phenol-formaldehyde) allowed decreasing the amount of applied resin by 30% (DUKARSKA and CZARNECKI 2016). Similar studies performed by KAWALERCZYK et al. (2020c, 2021) showed that the addition of NCC can reduce the UF and PF resins spread rate by 30% and by 20%, respectively. This paper is a continuation of the studies on cellulosic particles incorporation to the plywood manufacturing process. The effect of NCC addition differs from the addition of MFC. Moreover, when compared to the NCC, the MFC can be used as more cost-effective and commercially available modifier for the widely used UF and PF resins. Thus, the aim of presented study was to investigate the effect of microcellulose application on the possible reduction of adhesive spread rate in plywood manufacturing process.

MATERIALS AND METHODS The PF and UF resins were purchased from Silekol (Kędzierzyn-Koźle, Poland) with the following characteristics summarized in Table 1. 32

Tab. 1 Properties of applied resins. Parameter Viscosity (mPa·s) Solid content (%) Gel time at 100 °C (s) Gel time at 130 °C (s) pH

UF resin 650 69 69 – 8.1

PF resin 471 48 – 190 12.5

Ammonium nitrate (20 wt%) was included in the UF adhesive composition as a hardener. In order to adjust the viscosity, the rye flour and the tannin based filler (UT-10) containing chalk and mimosa tannins were added to UF and PF resins, respectively. Microfibrillated cellulose commercially named as ARBOCEL (Rettenmaier GmbH, Poland) with an average particle sizes of 6 - 12 µm was applied as a modifier. Plywood panels were manufactured using rotary cut birch (Betula) veneer sheets with an average density of 560 kg/m3, a moisture content of 6 ± 1%, an average thickness of 1.5 mm and dimensions of 320 × 320 mm. In the case of UF resin, the MFC was introduced in a state of 10% aqueous suspension prepared with the use of magnetic stirrer (700 rpm, 10 min). Due to the lower solid content, there was no necessity to apply cellulosic particles in wet state to the PF resin, and therefore they were added directly to this adhesive type. The amounts of MFC included in both compositions were selected based on the results of previously conducted research (KAWALERCZYK et al. 2019b, KAWALERCZYK et al. 2020a). Tables 2 and 3 present the compositions of prepared adhesive mixtures. The experimental and reference variants differed in the amount of water contained in the formulations due to the fact that the control variants were prepared according to industrial regulations. Tab. 2 Composition of UF resin mixture. Variant label Reference Experimental

Quantity (g/100 g of solid resin) MFC suspension

Rye flour

Water

0 10

15 4

15 0

Total solution weight of hardener 2 2

Tab. 3 Composition of PF resin mixture. Variant label Reference Experimental

Quantity (g/100 g of solid resin) MFC powder UT-10 0 20 5 17

After the addition of fillers the adhesives were mixed at 1000 rpm for 2 min with the use of CAT-500 homogenizer to attain proper level of homogenization. The reference mixtures were spread on the veneer in the amount of 170 g/m2 and these variants were labelled as 170REF. The experimental adhesives containing MFC were applied on the surface of the veneers in the amount of 170, 160, 150, 140, 130 g/m2 (the calculations were made in relation to the total adhesive mixture mass). The veneer sets for 3-layered plywood were hot pressed with the unit pressure of 1.4 MPa, temperature of 120 °C in case of UF resin and 140 °C in case of PF resin for 4 minutes. The manufactured plywood panels were tested in terms of bonding quality according to EN 314-1 (2004). In case of UF resin the shear strength was determined in dry state and after soaking in water for 24 h. Plywood glued with PF resin were tested after soaking in water for 24 hours and after a pre-treatment consisting of boiling in water for 4 h followed 33

by drying in laboratory oven for 16 h at 60 °C, another boiling in water for 4 h and cooling in water for 1 h at 20 °C. The assessments involved 11 samples of each variant. In addition to bonding quality, the formaldehyde emission of plywood bonded with UF resin was investigated with a flask method according to EN 717-3 (1996). The results were analysed with the use of multivariate statistical analysis ANOVA. Tukey test was carried out in order to distinguish the homogeneous groups on a significance level of α = 0.05 using Statistica 13.0 software.

RESULTS AND DISCUSSION The results of shear strength test of plywood bonded with UF resin are presented in Fig. 1. The samples soaked in water were characterized by lower strength values due to the hydrolysis of the UF resin in the presence of water. As expected based on the previous results the addition of MFC resulted in the increase of bonding quality in comparison with the flourfilled adhesive applied in the same amount. The bonding quality was improved by 14% and by 8% in case of panels tested in dry conditions and after soaking in water, respectively. As the amount of applied adhesive decreased, the shear strength also decreased. The results of statistical analysis showed that the addition of MFC allowed to reduce the resin spread rate by 10 g/m2 while maintaining as good bonding strength as the reference plywood. Further decrease in the amount of applied resin to 130 g/m2 resulted in a deterioration of bonding quality by up to 24% in dry state and by 27% after soaking.

Fig. 1 Shear strength of plywood glued with UF resin (a, b, c, d letters mark homogenous groups in the HSD Tukey test; F[x,y] = z,p where, F is Roland Fisher's test method, x is number of degrees of freedom, y is number of tests, z is value of F test, p is probability level).

Fig. 2 presents the results of bonding quality investigated in plywood glued with PF resin. Similarly as in panels with UF resin the introduction of MFC led to the improvement in bonding strength. The shear strength was improved by 10% and 6% in the case of plywood tested after soaking and after boiling, respectively. The reinforcing effect of MFC allowed the reduction in adhesive application by 10 g/m2 and manufacturing plywood with equally good bonding quality as the reference one. BEKHTA and MARUTZKY (2006) explained that the further reduction in shear strength might result from the insufficient glue quantity to cover the veneer surface and maintain the required glue line thickness. The improvement in bonding quality can be associated with the chemical bonding between the hydroxyl groups of cellulosic derivatives and the functional groups of resin, which consequently allow obtaining a highly cross-linked structure. Moreover, studies have 34

shown that the introduction of cellulosic particles can improve resin morphology. The structure of cellulose-reinforced adhesive was more solid, less porous and had significantly less microcracks (KAWALERCZYK et al. 2021). Since water participates in the crosslinking reactions during the resin condensation, the lowered amount of water in experimental variant could also have an influence on the results in case of the UF resin-bonded plywood. The addition of MFC instead of NCC resulted in less efficient reduction of adhesive spread rate. The reason could be an increased chemical reactivity of particles within the dimensional range of nanotechnology. Furthermore, the nanoparticles are characterized by a tremendous surface area which favours a chemical bonding process.

Fig. 2 Shear strength of plywood glued with PF resin (a, b, c, d letters mark homogenous groups in the HSD Tukey test; F[x,y] = z,p where, F is Roland Fisher's test method, x is number of degrees of freedom, y is number of tests, z is value of F test, p is probability level).

The application of MFC allowed reducing the amount of resin by 6% in the case of both UF and PF resin which is less effective also when compared with the addition of nanoSiO2 (DUKARSKA and CZARNECKI 2016). The adhesives reinforcement with MFC seems to be also slightly less effective than the veneer surface modification with compression or DBD plasma (BEKHTA and MARUTZKY 2006, CAO et al. 2018). However, the reduction of adhesive consumption resulted in a decrease of formaldehyde release (Fig. 3).

Fig. 3 Formaldehyde emission from plywood glued with UF resin (a, b, c, d, e letters mark homogenous groups in the HSD Tukey test; F[x,y] = z,p where, F is Roland Fisher's test method, x is number of degrees of freedom, y is number of tests, z is value of F test, p is probability level).

On the basis of the formaldehyde release results it can be concluded that the introduction of MFC itself has not caused a decrease in the amount of emitted formaldehyde which is similar to the effect observed by AYRILMIS et al. (2016) and KAWALERCZYK et al. (2020a). Studies performed by ZHANG et al. (2011) suggest that modification of cellulosic particles with 3-aminopropyltriethoxysilane (APTES) can contribute to the more effective scavenging abilities. The reduction of adhesive application by 6% (10 g/m2) led to a decrease in formaldehyde release by 9%. As the resin spread rate decreased, the amount of emitted formaldehyde also decreased. The further reduction in the amount of applied resin to 130 g/m2 resulted in lowering the investigated emissions by 13% which was expected since the adhesives are the main sources of emitting formaldehyde.

CONCLUSIONS The introduction of MFC to UF and PF resins led to the statistically significant improvement in bonding quality. The MFC addition allowed to reduce the amount of applied resin by 6% while maintaining the equally good strength of the reference, rye flourcontaining variant. The microcellulose-reinforcement itself did not cause the reduction in formaldehyde release. However, the reduction in adhesive application by 6% resulted in a decrease in HCHO emission by 9%. Further reduction in resin spread rate led to a significant deterioration in glue line strength. REFERENCES AYRILMIS, N., LEE, Y.K., KWON, J.H., HAN, T.H., KIM, H.J. 2016. Formaldehyde emission and VOCs from LVLs produced with three grades of urea-formaldehyde resin modified with nanocellulose. In Building and Environment, 97: 8287. DOI: 10.1016/j.buildenv.2015.12.009 BEKHTA, P., BRYN, O., SEDLIAČIK, J., NOVÁK, I. 2016. Effect of different fire retardants on birch plywood properties. In Acta Facultatis Xylologiae Zvolen, 58(1): 5966. DOI: 10.17423/afx.2016.58.1.07 BEKHTA, P., MARUTZKY, R. 2008. Reduction of glue consumption in the plywood production by using previously compressed veneer. In European Journal of Wood and Wood Products. 65: 8788. DOI: 10.1007/s00107-006-0142-8 CAO, Y., ZHOU, X., CHEN, M., CHEN, W., YU, P., THIPHUONG, N. 2018. Enhancing resin efficiency in plywood production via DBD plasma treatment and atomized air spray of UF resin. In Holzforschung, 72(12): 10571062. DOI: 10.1515/hf-2018-0015 DUKARSKA, D., CZARNECKI, R. 2016. Fumed silica as a filler for MUPF resin in the process of manufacturing water-resistant plywood. In European Journal of Wood and Wood Products, 74: 514. DOI: 10.1007/s00107-015-0955-4 EICHHORN, S.J., BAILLIE, C.A., ZAFEIROPOULOS, N., MWAIKAMBO, L.Y., ANSELL, M.P., DUFRESNE, A., ENTWISTLE, K.M., HERRERA-FRANCO, P.J., ESCAMILLA, G.C., GROOM, L., HUGHES, M., HILL, C., RIALS, T.G., WILD, P.M. 2001. Review: Current international research into cellulosic fibres and composites. In Journal of Materials Science, 36: 21072131. DOI: 10.1023/A:1017512029696 HALDAR, D., PURKAIT, M.K. 2020. Micro and nanocrystalline cellulose derivatives of lignocellulosic biomass: A review on synthesis, applications and advancements. In Carbohydrate Polymers, 250, 116937. DOI: 10.1016/j.carbpol.2020.116937 HEINZE, T. 2016. Cellulose: Structure and Properties. In Cellulose Chemistry and Properties: Fibers, Nanocelluloses and Advanced Materials. Springer, Heidelberg, New York. p. 152. ISBN 978-3319-26015-0.

KAWALERCZYK, J., DZIURKA, D., MIRSKI, R., GRZEŚKOWIAK, W. 2019a. The effect of veneer impregnation with a mixture of potassium carbonate and urea on the properties of manufactured plywood. In Drewno, 62(203): 107115. DOI: 10.12841/wood.1644-3985.281.12 KAWALERCZYK, J., DZIURKA, D., TROCIŃSKI, A. 2019b. Wpływ modyfikacji żywicy fenolowoformaldehydowej z wykorzystaniem mikrocelulozy na właściwości wytworzonej sklejki. In VI Ogólnopolska Konferencja Młodych Naukowców Przyroda-Las-Technologia, Poznań, Poland. KAWALERCZYK, J., DZIURKA, D., MIRSKI, R., SZENTNER, K. 2020a. Properties of plywood produced with urea-formaldehyde adhesive modified with nanocellulose and microcellulose. In Drvna Industrija, 71(1): 6167. DOI: 10.5552/drvind.2020.1919 KAWALERCZYK, J., DZIURKA, D., MIRSKI, R., SIUDA, J., SZENTNER, K. 2020b. The effect of nanocellulose addition to phenol-formaldehyde adhesive in water-resistant plywood manufacturing. In BioResources, 15(3): 53885401. DOI: 10.15376/biores.15.3.5388-5401 KAWALERCZYK, J., DZIURKA, D., MIRSKI, R. 2020c. The possible reduction of phenol-formaldehyde resin spread rate by its nanocellulose-reinforcement in plywood manufacturing process. In Annals of WULS - SGGW. Forestry and Wood Technology, 111, 2126. DOI: 10.5604/01.3001.0014.6420 KAWALERCZYK, J., DZIURKA, D., MIRSKI, R., SIUDA, J. 2021. The reduction of adhesive application in plywood manufacturing by using nanocellulose-reinforced urea-formaldehyde resin. In Journal of Applied Polymer Science, 138(7): 49834. DOI: 10.1002/app.49834 KLEMM, D., HEUBLEIN, B., FINK, H.P., BOHN, A. 2005. Cellulose: fascinating biopolymer and sustainable raw material. In Angewandte Chemie International Edition, 44(22): 33583393. DOI: 10.1002/anie.200460587 PAWLAK, D., BORUSZEWSKI, P. 2018. Influence of addition of microfibrillated cellulose (MFC) on selected properties of low-density particleboard. In Annals of WULS - SGGW. Forestry and Wood Technology, 102: 139148. PAŽITNÝ, A., RUSS, A., BOHÁČEK, S., STANKOVSKÁ, M., IHNÁT, V., ŠUTÝ, Š. 2019. Various lignocellulosic raw materials pretreatment processes utilizable for increasing holocellulose accessibility for hydrolytic enzymes. Part II. Effect of steam explosion temperature on beech enzymatic hydrolysis. In Wood Research, 64(3): 437447. THOORENS, G., KRIER, F., LECLERCQ, B., CARLIN, B., EVRARD, B. 2014. Microcrystalline cellulose, a direct compression binder in a quality by design environment – A review. In International Journal of Pharmaceutics, 473(12): 6472. DOI: 10.1016/j.ijpharm.2014.06.055 TRACHE, D., HUSSIN, M.H., CHUIN, C.T., SABAR, S., FAZITA, M.R.N., TAIWO, O.F.A., HASSAN, T.M., HAAFIZ, M.K.M. 2016. Microcrystalline cellulose: Isolation characterization and biocomposites application – A review. In International Journal of Biological Macromolecules 93(Part A), 789804. DOI: 10.1016/j.ijbiomac.2016.09.056 TSCHURTSCHENTHALER, G. 2012. Die synthese eines cellulose-nanofibrillenverstarkten harnstoffformaldehyd-leimes. Master Thesis, BOKU, Vienna, 57 p. VEIGEL, S., MULLER, U., KECKES, J., OBERSRIEBNIG, M., GINDL-ALTMUTTER, W. 2011. Cellulose nanofibrils as filler for adhesives: Effect on specific fracture energy of solid wood-adhesives bonds. In Cellulose, 18(5): 1227. DOI: 10.1007/s10570-011-9576-1. VEIGEL, S., RATHKE, J., WEIGL, M., GINDL-ALTMUTTER, W. 2012. Particleboard and oriented strand board prepared with nanocellulose-reinforced adhesive. In Journal of Nanomaterials, 158503. DOI: 10.1155/2012/158403 VINEETH, S.K., GADHAVE, R.V., HADEKAR, P.T. 2019. Nanocellulose applications in wood adhesives – Review. In Open Journal of Polymer Chemistry, 9(4): 6375. DOI: 10.4236/ojpchem.2019.94006 ZHANG, H., ZHANG, J., SONG, S., WU, G., PU, J. 2011. Modified nanocrystalline cellulose from two kinds of modifiers used for improving formaldehyde emission and bonding strength of ureaformaldehyde resin adhesive. In BioResources, 6(4): 44304438. ACKNOWLEDGEMENT This research was funded by the National Centre for Research and Development, BIOSTRATEG3/ 344303/14/NCBR/2018.

This research was supported by the Slovak Research and Development Agency under the contracts No. APVV-18-0378 and APVV-19-0269.

AUTHOR’S ADDRESS Mgr Jakub Kawalerczyk prof. UPP dr hab. Dorota Dziurka prof. UPP dr hab. Radosław Mirski Dr Joanna Siuda Poznań University of Life Sciences Faculty of Forestry and Wood Technology Department of Wood-Based Materials ul. Wojska Polskiego 38/42 60-627 Poznań Poland dorota.dziurka@up.poznan.pl Prof. Ing. Ján Sedliačik, PhD. Technical University in Zvolen Department of Furniture and Wood Products T.G. Masaryka 24 960 01 Zvolen Slovakia jan.sedliacik@tuzvo.sk

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 39−53, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.04

CREEP LIFE PREDICTION BY THE BASIC MODELS OF DEFORMATION-DESTRUCTION KINETICS OF WOOD-BASED COMPOSITES Sergiy Kulman – Liudmyla Boiko – Yaroslav Bugaenko – Ján Sedliačik ABSTRACT Existing models of kinetics of deformation-destruction are related mainly to the description of the transition from the undamaged state of the material into a destroyed stage. The purpose of this study was to create the basic model of deformation-destruction kinetics describing this process in the form of several successive transitions of separate structural elements (SE), in a material that deforms from one rheological state to another one in time. A formal kinetics apparatus is involved to describe this process, which allows, knowing the speed of the transition of the SE from one rheological state to another one, to predict the time to reach the critical concentration of destroyed SE. It is established that the process of deformationdestruction can be considered a process of gradual transition of the SE, first an elastic state into viscoelastic and then destroyed. The change in the concentration of one or another SE can be determined experimentally by measuring the quantities that correlate with the parameters of deformation-destruction. For the first time, a two-stage nonlinear kinetic model of resource loss with the creep of composite materials based on wood was proposed. The application of the method of basic deformation diagrams in conjunction with the twostage description of the process of accumulation of damage can increase the accuracy of the prediction of allowable time for different load patterns during creep. Key words: particleboard, model of kinetics, kinetics of deformation, criteria of destruction, two-stage model of deformation-destruction.

INTRODUCTION One of the first theories of strength was formulated by GRIFFITH (1921). According to this energy theory, strength was defined as the stress under which the condition of equality of two energies is fulfilled: the energy expended on the formation of a new surface of a growing crack, and the elastic energy released during a growth of crack. According to this theory, destruction is interpreted as a critical event, which follows a critical tension. This interpretation is not confirmed by experience, because it is known that the accumulation of molecular and supramolecular defects occurs long before the moment of destruction. Therefore, destruction develops over time and is not a critical event. The introduction of a temporary scale of the fracture process led to the creation of a kinetic theory of strength. Temporal dependencies were proposed by ALEKSANDROV (1941) who describes the relaxation properties of viscoelastic deformation of solids, in the form of generalized Maxwell equations. The clearest physical interpretation and development of these ideas was 39

provided by professor S. M. Zhurkov, who is the founder of the kinetic theory of strength (REGEL et al. 1974). In this, the theory destruction is considered as a temporary process of accumulation of molecular and supramolecular defects. The durability of a stressed bodies is defined as a fundamental parameter of strength and reflects the average rate of destruction at all structural levels: molecular, supramolecular and macroscopic. A fundamental form of kinetic strength theory is the Zhurkov equation. Moreover, the Zhurkov equation, which is created to describe the mechanism of destruction at the molecular level, is automatically transferred to describe the mechanism of destruction of mezzo and macro-levels, and does not take into account how the relaxation processes occur at these levels. In addition, this equation assumes that the process passes in only one stage. Namely, the connections between the elements are considered to be either not destroyed, and are under the action of thermo-force load or destroyed after a time determined by the Zhurkov formula. The initial state of a structure element is a state in which there is no external mechanical effect. In this case, all the individual elements of the structure of the target are in thermodynamic equilibrium. Elastic steady state (E) of a structure element is a state, where structure element is in an elastic state for some time. The viscoelastic steady state (VE) of a structure element is a state, where structure element is in a viscoelastic state, that is, a state of relaxation of internal stresses. Destructed steady state (D) of a structural element is a state, where destroyed structural element is formed from a viscoelastic, whose stresses at the time of destruction are redistributed between those elements of the structure that are presently in elastic and viscoelastic states. We shall consider a solid body not in the form of a continuous continuum, but in the form of a solid structure, consisting of separate supporting elements of the structure, interconnected in a certain way. Then the action of external loading on the surface of the body will cause its internal elements of the structure, due to internal heterogeneity, to experience different in magnitude and direction of internal stresses (KULMAN and BOIKO 2015). Elements of structure whose stresses are greater than the tensile strength will be immediately destroyed. After their destruction, external efforts will be redistributed between the remaining undamaged elements of the structure so that each of them will bear a new load on the part of neighbouring SE. The behaviour of polymers over time is usually described in the form of combinations of primary rheological bodies, such as the elastic body of Hook, the viscous body of Newton, or combinations thereof (Maxwell or Voigt bodies). By combining these primary rheological bodies, one can in one way or another create a rheological model, whose behaviour in time (kinetics) will accurately describe the behaviour of the real body under the influence of external loads. It is generally recognized that during a period of constant creep, a solid body, under the action of external temperature-force loading, behaves as consistent in an elastic, viscoelastic or plastic state. At the same time, it is assumed that the total deformation of the body consists of three different parts by nature: instantaneously reversible (elastic), highly elastic, reversible in time, i.e., relaxing in time (viscoelastic) and irreversible (plastic) (REGEL 1964). In this case, total deformation is understood as the total deformation of the whole body, and not its individual parts or elements of structure (AKYÜZ et al. 2019, YAPICI et al. 2016). We take as the main hypothesis of the deformation that: first, the body is deformed and destroyed individually in separate elements of the structure. Namely, only in those elements in which the local stress first reaches the limits of proportionality, and then the limits of strength. In this case, the individual element of the structure is first elastically deformed, and then passes in series in a viscoelastic or plastic state, and then collapses. Secondly, a single destroyed element of the structure is surrounded for some time by non40

destructive elements of the structure that are in an elastic or viscoelastic state. In-third, the body will be considered destroyed if the number of destroyed elements of the structure exceeds certain predefined values (POŠTA et al. 2016). Thus, the kinetics of the destruction of the body under the action of thermomechanical loading will consist of different variants of the sequence of transition of the elastic elements of the structure (E) into the viscoelastic elements of the structure (VE), and then in the destroyed elements of the structure (D): E→VE→D. Within the process of destruction, the number and, consequently, the concentration of elements of the structure that are in different rheological states are constantly changing. The changes concentration of a particular element of the structure can be determined experimentally by measuring values that correlate with the fracture parameters of a particular type of body deformation. In this case, the total number of structural elements in different states, according to the law of conservation of masses at each moment must remain constant. Consideration of the fracture process, from the point of view of combining cracks, for example, smaller ones into larger ones, does not change the number of structural elements, but only examines them in different states. The purpose of this study is to create basic model of kinetics of the process of scattered fracture based on a multistage transition of structural elements from one rheological state to another by successive or more complex transformations. To create a kinetic model of the relaxation behaviour of a polymer under the action of thermomechanical action, several different states of its individual structure element (SE) was taken into consideration. Individual elements of a structure are understood to mean local body volumes, whose local stresses are different. For a crystalline polymer under load, these can be regions both crystalline and amorphous.

MATERIAL AND METHODS Materials Three commercially produced structural particleboards bonded with urea formaldehyde resin (UF) provided by Kronospan UA Ltd. were used for this study: melamine faced particleboard (MF PB) according to EN 14322; veneered faced particleboard by oak (VF PB) according to EN 316, EN 622-5 and particleboard P2 (P2 PB) according to EN 312 type P2; EN 13501-1: class D-s1, d0. For each type, two regular-size (2750 mm × 1830 mm) of boards with thicknesses of 18 mm were cut into 450 mm × 50 mm pieces. Panels before cutting and all specimens were conditioned at 20 °C and 65% RH, to the moisture content of about 5%. Static 3-point bending tests were carried out in the special test machine with temperature-controlled chamber (BOIKO et al. 2013). Specimens were prepared and cut according to ASTM D 1037-12. Loading force and deflection were measured, and MOR and MOE were calculated according to Section 9 in ASTM D 1037-12. Investigated temperatures were 20 °C and 60 °C. Specimens were preheated in the chamber until they reached equilibrium with the target temperature. The preheating times were determined from preliminary experiments by an embedded thermocouple, and the prediction model was developed in a previous study (KULMAN and BOIKO 2016). The mechanical properties of reference samples were tested in the chamber at the target temperature, results are shown in Table 1. One hundred fifty specimens were cut from each type of board. Ten test pieces were prepared for determination the modulus of elasticity (MOE) and modulus of rupture (MOR) before main testing.

Tab. 1 Standard properties of particleboard. Board type

Density Thickness (kg/m3) (mm) MF PBa 757 ± 7 18.1 ± 0.1 VF PBb 792 ± 8 18.5 ± 0.1 P2 PBc 733 ± 6 18.1 ± 0.1 a MF PB – Melamine Faced Particleboard. b VF PB – Veneered Faced Particleboard. c P2 PB – Particleboard according to EN 312, type P2.

MOR (MPa) 17.1 ± 1.1 20.5 ± 1.9 16.2 ± 0.6

MOE (MPa) 2 110 ± 29 2 520 ± 15 2 020 ± 22

Tests according to the scheme of three-point bending according to ASTM D 1037-12 loads equal to 75% of the maximum allowable. The nature of the behaviour of the system under the action of constant load was described by the movement of the midpoint in time until rupture. Model hypothesis and rationale The option of sequential destruction in two elementary stages, which occur simultaneously in the body for its deformation-destruction (DD) is considered. The scheme of transformation of structural elements in the process is given in the following form: E→VE→D. The rate of destruction is defined as the change in the number of destroyed structural elements N per unit time, attributed to the unit volume of the body. We first assume that destruction can occur in any body volume (homogeneous destruction). Unlike heterogeneous destruction, when it can occur at the boundaries of phase distribution. The rate of destruction is written: r = ± dN/Vdt (for homogeneous destruction), r = ± dN/Sdt (for heterogeneous destruction). Derivative sign means, spent or accumulated (formed) or a certain amount of SE in different states in the process of destruction. If, during homogeneous destruction, the volume of the system remains constant (closed system), then: dN/V = dC. Hence, the rate of destruction is related to the concentration of elements of the structure in time: r = ± dC/dt. We apply the stoichiometric law of equivalents to the kinetics of DD, which states that the ratio the number of SE, that change their state, is equal to or a multiple of the molar quantities equivalents. In our case, the equivalent is some part of the volume of the body, which, when destroyed, becomes one of the destroyed structural elements. That is, this element is formed only as a result of DD. We also apply the law of multiple relations to the quantitative description of the DD process. According to the law of multiple relations, in the event of destruction, the mass of one type of SE is the mass of another type of SE, which are referred to as small integers. The law of multiple relations is confirmed by the discreteness of the process of destruction, as well as by the fact that all destroyed elements of one homogeneous substance have the same concentration and a strictly defined specific mass (volume). In our description of the DD process, we will rely on two concepts: first, the law of mass action as the law of a simple process (its elementary stage); secondly, the complexity of the mechanism of the DD process, consisting of one or another set of successive or parallel elementary stages. The law of the active masses has been applied far beyond chemical kinetics in socalled "models of development" and has been widely used in biology and ecology, economics, neurophysiology and genetics. The fundamental concept of chemical kinetics is the mechanism of reaction. In a broad sense, it is the interpretation of experienced data about the complex process. Such a mechanism should identify the individual stages and stages of the process, describe the characteristics of the intermediate products, certain energy levels 42

of the process, etc. A narrow understanding of the mechanism as a set of stages can also be applied. If the stages are assumed to be simple, then they consist of elementary reactions as a kinetic law, which is accepted as the law of mass action or the law of acting surfaces - for catalytic reactions. This is the concept of mechanism that operates in formal kinetics, which studies kinetic models – systems of differential and algebraic equations corresponding to the mechanism of the process. The two concepts outlined above define the development of two lines that should complement each other. Namely, it is a study of the kinetic regularities of the elementary act of fracture and the design of the theory of kinetics of complex joint deformations. Mathematical model of deformation-destruction process Let us write down the process of the DD of a body in the form of a change in the number of structural elements that are in different states during the whole time before destruction. The exponential dependence of the change in the concentration of submicrocracks, microcracks and macrocracks on creep gives reason to consider the process of delocalized destruction as a kinetic analogue of the first order process (REGEL et al. 1974, PETROV et al. 1993, YAPICI et al. 2016). The main variables that characterize the state of the system, determine the substances Ai, that is in different rheological states. Denote the number SE in different states as Ni; n – is the vector of quantities components. Denote the concentration: Ci ≡ Ni/V. Each stage of the system is matched by its speed Ws(C,T). The velocity of the stage is intense and is defined as a function of intense quantities – concentration and temperature. The kinetics equations have the following in coordinate form: dNi  V  siWs  C , T  , dt s

i  1,, n

(1)

where γsi is the stoichiometric vector with components  si   si   si ; s – stage;  si ,  si – non-negative integers are stoichiometric coefficients. In the absence of autocatalysis, as in our case, this vector completely determines the stoichiometric equations of the stage. For each material, there are a priori restrictions on vectors γsi – linear conservation laws (balance ratios). If Nsi – the number of structural elements that are in a certain rheological state and k is a species in the molar volume of a substance Ai, then for any s and k: ∑𝑖 𝛼𝑠𝑖 𝑁𝑘𝑖 = ∑𝑖 𝛽𝑠𝑖 𝑁𝑘𝑖 𝑜𝑟 ∑𝑖 𝛾𝑠𝑖 𝑁𝑘𝑖 =0

(2)

The balance relations (2) give rise to linear conservation laws for system (1), that is, for any k: i ki Ni  const , means:

d  ki Ni  V s,isi kiWs  0 dt i

(3)

Record the speed of each stage of the deformation-destruction process using the law of mass action. At the same time, the general scheme of phenomenological kinetics was used, which consists in recording the equations of the dynamics of state change, based on the general concepts and dependence of the processes of formation and consumption of an individual SE system:

Ws C, T   Ws C, T  Ws  C,T 

(4)

Ws  C, T   ks T  Cisi

(5)

Ws  C, T   ks T  Cisi

(6)

Ws  C, T  – function of appearance a structural element in a state Ci (E, VE or D);

Ws  C, T  – the function of disappearance a structural element in a state Ci (E, VE or D); ks/  T  – rate constants of the transition of the SE from one rheological state to another in

the s-stage. From the law of conservation of masses follows, that changes in the concentrations of quantities, characterizing the properties of the material, or their changes during DD, satisfy the ordinary differential equations system (ODE) of the nth order of the form: dCi   siWs  C , T  , dt s

i  1, , n

(7)

System (7) is defined in some bounded region of the phase space S: (0 ≤ Ci ≤ bi, i = 1,…, n) – simplex process area. The boundaries are determined from the balance of the number of elements of the structure that are in different rheological states. To create a DD model, first of all, we will compile a list of objects that are involved in this process. Objects by analogy with the study of DD processes can be taken as a state of a separate structural element of the material, which largely determines its behaviour over time: elastic, viscoelastic, destroyed. Thus, the state of the material at certain points in time will be determined by the number of structural elements that are in a particular rheological state. Denote the objects that enter the material at a certain point in time after its loading, A1,…, Аn. Let us assume for the kinetic model the mechanism of the DD process in the form of a list of its elementary stages. Each stage is determined by its stoichiometric equation:

 s1 A1   sn An   s1 A1   sn An

(8)

where s – stage; αsi, βsi – non-negative integers are stoichiometric coefficients. The mechanism of DD is interpreted as follows. Each stage corresponds to one elementary transition, each elementary transition goes to one clock cycle, in which only those SE are involved and in the quantities specified in the stoichiometric equation. The elementary act time is much less time between them. We will consider loading material as a structure consisting of a set of separate structural elements that are under constant external loading. Then the state of its individual structural element at different points in time can be described as follows: A1 ≡ A – the elastic state of the structural element (E); A2 ≡ B – viscoelastic state (VE); A3 ≡ C – destroyed state (D). Then the following kinetic model of the DD process will correspond to the transformation scheme (3): dCA  t  (9)  k1CA  t  dt dCB  t  (10)  k1CA  t   k2CB  t  dt dCC  t   k2CB  t  dt

(11)

with initial conditions, t = 0: CA(0) = 1; CB(0) = CC(0) = 0, and boundary conditions: CA(t) + CB(t) + CC(t) = 1, where CA(t) ∈ (1…0) – the current concentration of structural elements that are in an undamaged, elastic state at time t; CB(t) ∈ (0…1) – the current concentration 44

of structural elements that are in a non-destructive, i.e. viscoelastic state at time t; CC(t) ∈ (0…1) – the current concentration of the structural elements that are in a destroyed state at time t; k1 – the rate constant of the transition of structure elements from elastic state to viscoelastic, s-1; k2 – the rate constant of the transition of the elements of the structure from the viscoelastic state to the destroyed, s-1; Considering the curve of long load, it can be determined that its initial section contains elements of the structure only in the elastic state, which after removal of the load return to the non-deformed state. During this period of elastic deformation, there is no transition of structural elements to a viscoelastic state. Therefore, the beginning of the kinetics of the DD is taken to be the end time of the elastic deformation and the beginning of the site of viscoelastic deformation, i.e. the time of the beginning of the process first unsteady and then steady creep. Throughout the process of creep, there is a decrease in the number of nondestructive structural elements, i.e., those in elastic or viscoelastic states and an increase in structural elements in the destroyed state. To describe the kinetics of deformation, it is possible to compare the change in the concentration of SE in the elastic state and the destroyed state with the change in the magnitude of the deformation over time. In this case, the rate of change of the concentration over time of elastic SE will be proportional to the change in the rate of deformation due to the transition of elastic SE in viscoelastic, and the decrease in the concentration of elastic SE will be proportional to the increase in the absolute deformation of the DD process, i.e.: CA(t) ∝ ε(t). Since the change in the SE concentrations in one state or another can be given as a first order kinetic dependence (REGEL et al. 1974, PETROV et al. 1993), we can write: dCA  t   k1 CA  t  or, having gone to proportional deformation: dt d e t  (12)  k1  e  t  dt We integrate (12) after separation of variables (time, deformation): d e t    k1 t ,  e t 

t2 d e  t      1   e  t   t k1dt , ln   12   k1t2  t1  , k1   t2  t1  ln   12  . 1 1

2

(13)

As we can see, when the deformation changes ε1 and ε2 by the same number of times the value of k1 does not change. This allows for the creation of a kinetic model of deformation-destruction to replace concentrations proportional to their values – deformations, stresses, acoustic or electromagnetic emission pulses, quantities of matter, etc. The magnitude inverse of the first order reaction rate constant, τe = 1/k1, has a dimension of time and characterizes with creep the average life expectancy of SE, which are in an elastic state before their transition to a viscoelastic state. Similarly, the inverse of the rate constant of the transition of the viscoelastic state of the SE to the destroyed determines the average life expectancy of the structural elements that are in the viscoelastic state: τve = 1/k2. Since time is included in Eq. (13) as a difference rather than a relation as a deformation, we introduce a dimensionless time to eliminate the dimension factor: τi = ti / tcr. At the same time, total creep deformation can be represented as the sum of total elastic and viscoelastic deformations, i.e. deformations due to the destruction of elastic and viscoelastic SE:  e   ve   cr (14) te  tve  tcr (15) 45

Because according to Eqs. (12), (14):

cr cr 1 1  t1 tei   t1 tvei   1 k1 k2 tcr

1 ˙ 1 ˙  e +  ve   cr , then k1 k2

(16)

The expression for k2 is obtained by solving the system of two Eqs. (13), (16):   ln  2   1  k2  (17)  1  t1  t2  tcr *ln    2  The solution of the system of differential equations (9), (10), (11), for example by the operating method, allows to obtain values of current concentrations of SE, which are in elastic, viscoelastic and destroyed states: CA  t   CA  0 ek1t

CB  t   CA  0  (

(18)

k1 k1 e  k2t  e  k1t ) k1  k2 k1  k2

(19)

  k1 k2 (20) CC  t   CA  0  1  e  k2 t  e k1t  k1  k2  k1  k2  Moreover, given the initial conditions: t = 0; CA(0) = 1; CB(0) = CC(0) = 0, and boundary conditions: t ∈ (0…1): CA(t) + CB(t) + CC(t) = 1, in the coordinates of dimensionless time we get: 1

CA  t  dt  CB  t  dt  CC t  dt  1

(21)

therefore, k1-1 + k2-1 = 1. So in real time coordinates k1-1 + k2-1= tcr. The method involves determining the characteristic parameters of the damage and predicting the time of reaching the critical mark of the measure of damage. While as a preventive comparison of the degree of damage is a scalar value ψ which is equal to the ratio of the current concentration of destroyed elements of the structure to the current concentration of non-destructive elements of the structure. The resource is considered to be exhaustive value the extent of damage exceeding (KULMAN 2020): CC  t  (22)   f  1 CA  t   CB  t  where ψf – a measure of permissible damage; CA(t), CB(t), CC(t) – are determined by   1 formulas (18), (19), (20), respectively; k1  ln  1  – rate constant of transition of t1  t2   2  structure elements from elastic state to viscoelastic creep diagram, s-1; t1 – the time of the end of the section of elastic deformation and the beginning of the section of viscoelastic deformation, with; t2 – the time of measurement of deformation at the site of constant creep; ε1 – deformation at time t1; ε2 – deformation at time t2; k2 = α–1k1 – the rate constant of the transition of the elements of the structure from the viscoelastic state to the destroyed, s-1; α – parameter characterizing the rheological properties (features) of the material, the degree of its inelasticity. In this case, the rate constants of the transition of local structure elements from one state to another, depending on the load and temperature, are determined by the

long-term strength chart, which establishes the dependence of the change in the rate of deformation in time. The method is implemented as follows: a basic deformation diagram is constructed at isothermal creep of a part under constant load at a fixed temperature in the coordinates: ε absolute deformation, μm; t - time, s. In this case, the time to fracture and the maximum deformation at the time of fracture are determined. The moment of time of the end of the section of elastic deformation is fixed, as well as the magnitude of the deformation at this moment t. After the base diagram is constructed, the moment is selected t2 = tcr / (2…3) and at this point the deflection value is determined. Further, formulas (13) and (17) determine the value k1 and k2, as well as the value α: (23)

α = k1/k2

By measuring the change in the deformation of a part over time, a control chart of deformation is constructed. The moment of time is chosen t2k at the site of constant creep and absolute deformation at that moment is recorded ε2k. From Eq. (16), there is a value k1k, and the Eq. (23) k2k = k1k/α. Using Eqs. (17), (18), (19), it is determined the change in the current concentrations of SE, which are in elastic, viscoelastic and fractured states. And on the basis of Eq. (22) the estimated time is determined tψ reaching the limit value of the damage parameter ψ.

RESULTS AND DISCUSSION The experimental factor levels and test results of studies of the long-term strength particleboard at constant load are shown in Table 2. Tab. 2 Experimental factors levels and test results for particleboards. Board type

Temperature (°C)

MF PB

12.8

20 60

In all εcr 2.46 3.95

VF PB

Test resultsa

Stress level (MPa)

15.4

20 60

1.50 1.77

0.59 0.54

39.3 30.5

20 60

4.91 1.57

1.63 0.51

33.2 32.1

P2 PB a

Test conditions

12.15

Displacement, mm visco elastic elastic % εe εve 0.79 32.1 1.67 1.33 33.7 2.62

Time displacement, h visco elastic elastic % te tve 98.4 77.4 28.70 73.3 87.3 10.70

In all tcr

67.9 66.3

127.1 84.00

0.91 1.23

60.7 69.5

477.82 190.71

380.9 140.0

79.7 73.4

96.92 50.74

20.3 26.6

3.28 1.07

66.8 67.9

78.77 35.75

69.6 24.2

88.4 67.8

9.17 11.51

11.6 32.2

The average test values for each group of 20 samples are presented.

% 22.6 12.7

Experimental studies of the long-term strength curve of PB, have found that the material behaves in a complex viscoelastic manner during the creep. And the curve of its deformation contains the recurring sections, characteristic for the curves of elastic and viscoelastic deformation (Fig. 1).

Fig. 1 Basic and control creep diagrams. Dependences of the average absolute deformation for one group of the samples VF PB by stress level SL = 15.4 MPa and temperature T = 20 °C for basic diagram, and by stress level SL = 17.7 MPa and temperature T = 20 °C for control (predicted) diagram.

The nature of the deformation-destruction curves over time (Fig. 1 and 2) indicates that the process is non-stationary. The deformation process at all its stages, both at the subcritical (stationary creep process) and at the closed (active fracture process) behaves nonlinearly and has a stepped character. In Fig. 2, there is shown that the general curve of dependence of deformation in time, consists of separate sections, which have different repetitive strain rates.

t,h

Fig. 2 Fragments by different time periods in creep curve for one of the sample VF PB by stress level SL = 15.4 MPa and temperature T = 20 °C.

Table 2 shows that the amount of viscoelastic deformation is on average 33% of the total amount of deformation. In this case, the time of viscoelastic deformation (behaviour) PB occupies about 80% of the total deformation time. And the value of both the magnitude of viscoelastic deformation and its time is in a very narrow range, which indicates the similarity of the deformation process and the destruction of PB under different conditions of its loading. Analysing the deformation curves over time, we can say that the general deformation curves up to fracture periodically repeats cycles of elastic and viscoelastic behaviour. This confirms our assumption that the process is multi-stage (GÜNTEKIN and AYDIN 2016, VAN BLOKLAND et al. 2019, SHARAPOV et al. 2019). According to the graph of the basic deformation diagram (basic long-term strength curve) Fig. 1 we define such quantities: t1 = 0.003 × 3600 = 10.8 [s], ε1 = 80 [μm], t2 = 200 × 3600 = 7.2 × 105 [s], ε2 = 730 [μm], tcr = 477 × 3600 = 1.717 × 106 [s]. By Eqs. (13), (17), (23) we define: k1 = 0.307 × 10-5 s-1, k2 = 0.719 × 10-6 s-1; α = 4.273. A basic model diagram is shown in Fig. 3, constructed in the coordinates of the “concentrations of SE in different states – time”.

The results of basic tests of six groups of samples, as well as the results of calculations of the kinetic basic parameters are presented in Table 3. Tab. 3 Basic test results and calculated basic kinetic parameters. Board type

Test conditions Stress level (MPa)

MF PB

Creep life tcr (s)

k1 (s-1)

k2 (s-1)

20 60 20 60 20 60

457560 ± 7250a 302400 ± 3500 1717000 ± 9500 686556 ± 5600 283572 ± 850 128700 ± 780

1.16E-05 1.74E-05 3.07E-06 7.94E-06 1.88E-05 4.53E-05

2.69E-06 4.08E-06 7.19E-07 1.78E-06 4.34E-06 9.38E-06

4.325 4.261 4.273 4.453 4.331 4.831

15.4

P2 PB aThe

Temperature (°C)

12.8

VF PB

Test results

12.1

confidence interval is indicated at p = 0.05 level.

The results of control tests of six groups of samples, as well as the results of calculations of the control kinetic parameters are presented in Table 4. Tab. 4 Control test results, calculated control kinetic parameters and predicted time to failure. Board type

Test conditions Stress level (MPa)

MF PB

14.7

VF PB

17.7

P2 PB

13.9

aThe

Temperature (°C)

k1k (s-1)

k2k (s-1)

20 60 20 60 20 60

1.43E-05 1.82E-05 7.99E-06 1.19E-05 2.50E-05 5.95E-05

3.30E-06 4.27E-06 1.87E-06 2.67E-06 5.78E-06 1.23E-05

confidence interval is indicated at p = 0.05 level. results based on average kinetic coefficients.

Test results Creep life tcr (s) Predict by Control diagram Eq. (22) 338933 ± 7500a 372827b 262957 ± 5600 289252 670000 ± 9800 660000 504821 ± 5500 458928 234357 ± 6200 213052 99000 ± 860 98000

bPrediction

A multi-stage description of the kinetics of deformation-fracture allows us to take into account the change in the rheological state of the material during its deformation. Comparing the results of the model time before the destruction of the control samples in each test group (Tab. 4. Creep life predict by Eq. (22)) with the actual time before the destruction of the control samples (Tab. 4. Creep life control diagram), we can conclude that it is quite high convergence of these quantities. This suggests that the proposed method for predicting longterm creep strength can be used to improve the accuracy of predicting the performance of controlled mechanical systems (MELZEROVÁ et al. 2016).

CONCLUSIONS Based on the analysis of the results of theoretical and experimental studies of the process of deformation-destruction of wood composites, we can draw the following conclusions: 1. A two-stage nonlinear kinetic model of resource loss due to the creep of wood-based composites is proposed for the first time.

2. The use of formal kinetics methods for modelling the physicochemical processes that occur during deformation-destruction allows to design the multi-stage kinetic models. 3. The use of the method of basic deformation diagrams in combination with the two-stage description of the process of accumulation of damage, allows to increase the accuracy of the prediction of allowable time under different load schemes during creep. REFERENCES AKYÜZ, I., ERSEN, N., TIRYAKI, S., BAYRAM, B.Ç., AKYÜZ, K.C., PEKER, H. 2019. Modelling and comparison of bonding strength of impregnated wood material by using different methods: artificial neural network and multiple linear regression. In Wood Research, 2019, 64(3): 483–498. ALEKSANDROV, A.P. 1941. Relaxation in polymers. Doctoral thesis. Russia, 1941, 252 p. ASTM D 1037-12. Standard Test Methods for Evaluating Properties of Wood-Base Fiber and Particle Panel Materials. BOIKO, L., GRABAR, I., KULMAN, S. 2013. Durability particleboards in furniture. Osvita Ukrainy, Ukraine, 2013, 210 p. GRIFFITH, A.A. 1921. The phenomena of rupture and flow in solids. Philosophical Transactions of the Royal Society of London, A, 221: 163–198. GÜNTEKIN, E., AYDIN, T.Y. 2016. Prediction of bending properties for some softwood species grown in Turkey using ultrasound. In Wood Research, 2016, 61(6): 993–1002. KULMAN, S. 2011a. Kinetics of the trials of composite materials base of wood. In Zbіrnik naukovih prats Podіlskogo Sovereign Agrarian and Technical University. Special issue up to the VI scientificpractical conference. Kam’yanets – Podilsky, 2011, p. 196–206. KULMAN, S. 2011b. Nonlinear Effects of Deformation and Ruin of Composite Materials Base of Wood. In Science Newsletter NUBIP of Ukraine / Seriya “Lisіvnitsvo and decorative garden. Part 1, No. 164. 2011, p. 250–255. KULMAN, S. 2013. The non-dynamic model of deformation and the ruin of compositional materials based on wood. In Science Newsletter NUBIP of Ukraine / Seriya “Lisіvnitsvo and decorative garden. Part 2, No. 185. 2013, p. 312–319. KULMAN, S., BOIKO, L. 2015. Kinetic model of long durability of porous composite materials based on wood. In Collection of scientific papers. "Modern building materials from metal and wood." Odessa: LLC Vneshklamservis, 2015, 134 p. KULMAN, S. 2020. A method for predicting creep life. UA Patent 120865. MELZEROVÁ, L., KUCÍKOVÁ, L., JANDA, T., ŠEJNOHA, M. 2016. Estimation of orthotropic mechanical properties of wood based on non-destructive testing. In Wood Research, 2016, 61(6): 861–870. PETROV, V.A., BASHKAREV, V.I., VETTERGEN, V.I. 1993. Physical basis for predicting the durability of structural materials. St. Petersburg: Polytechnicak, 1993, 475 p. POŠTA, J., PTÁČEK, P., JÁRA, R., TEREBESYOVÁ, M., KUKLÍK, P., DOLEJŠ, J. 2016. Correlations and differences between methods for non-destructive evaluation of timber elements. In Wood Research, 2016, 61(1): 129–140. REGEL, V.R., SLUTSKER, A.I., TOMASHEVSKY, E.E. 1974. Kinetic nature of the strength of solids. Monograph: Science, 1974, 560 p. REGEL, V.R. 1964. High-molecular compounds. No. 6, 1964, 395 p. SHARAPOV, E., BRISCHKE, C., MILITZ, H., SMIRNOVA, E. 2019. Prediction of modulus of elasticity in static bending and density of wood at different moisture contents and feed rates by drilling resistance measurements. In European Journal of Wood and Wood Products, 2019, 77(5): 833–842. VAN BLOKLAND, J., OLSSON, A., OSCARSSON, J., ADAMOPOULOS, S. 2019. Prediction of bending strength of thermally modified timber using high-resolution scanning of fibre direction. In European Journal of Wood and Wood Products, 2019, 77(3): 327–340. YAPICI, F., ŞENYER, N., ESEN, R. 2016. Comparison of the multiple regression, ANN, and ANFIS models for prediction of MOE value of OSB panels. In Wood Research, 2016, 61(5): 741–754.

ACKNOWLEDGEMENT 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. The authors thank VEGA project No. 1/0556/19.

AUTHOR’S ADDRESS Assoc. Prof. Sergiy Kulman, PhD. Assoc. Prof. Liudmyla Boiko, PhD. Assoc. Prof. Yaroslav Bugaenko, PhD. Zhytomyr National Agroecological University, ZNAU Department of Wood Processing Blvd Stary 7 10008 Zhytomyr Ukraine sergiy.kulman@znau.edu.ua sergiy.kulman@gmail.com Prof. Ing. Ján Sedliačik, PhD. Technical University in Zvolen Department of Furniture and Wood Products T.G.Masaryka 24 960 01 Zvolen Slovakia jan.sedliacik@tuzvo.sk

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 55−71, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.05

THE COLOR OF BEECH WOOD MODIFIED IN AIR, PARAFFIN OR POLYETHYLENE GLYCOL, AND AFTER FOLLOWING WEATHERING IN XENOTEST Miroslav Repák – Ladislav Reinprecht ABSTRACT The thermal treatment of European beech (Fagus sylvatica L.) wood specimens was performed in three different media – air, paraffin or polyethylene glycol 6000 (PEG 6000) – during 1, 2, 3 or 4 hours, applying either only heating at a temperature of 100 °C or modification process at the temperatures of 190 °C and 210 °C. Following, the thermally treated specimens were artificially weathered in Xenotest in an accordance with the modified EN 927-6 (2018) for 6-weeks. The color parameters L*, a*, and b* of beech specimens were measured in their original state, after thermal treatment, and finally after weathering. The total color difference E*ab values of the air-thermally, paraffin-thermally and PEGthermally modified beech specimens ranged from 23.6 (for air-thermal mode 190 °C/1h) up to 55.8 (for PEG-thermal mode 210 °C/4h), evidently in connection with their apparent darkening with L* from – 22.9 up to – 54.2. Due to the following exposure of beech specimens in Xenotest, the highest total color difference E*ab from 20.1 to 36.4 had the PEG-thermally modified ones, which evidently lightened with L* from 18.6 to 32.6. At weathering the PEG-thermally modified beech wood specimens were the best resistant against creation of cracks in surfaces, in contrast to the air-thermally modified specimens. Key words: beech wood, paraffin, polyethylene glycol, thermal modification, weathering, color.

INTRODUCTION European beech (Fagus sylvatica L.) is one of the commercial most used hardwoods in Central Europe. Due to its easy workability and bendability it is mainly used for the furniture, plywood, decorative veneer and floor manufacturing (KLEMENT et al. 2020). However, at changed weather conditions it has low dimensional stability and in its surface are created cracks. According to the Standard EN 350 (2016), beech wood belongs to non-durable species, so it can be easily attacked by decaying fungi and insects, and without convenient chemical or physical treatments it cannot be used for outdoor building structural elements. Beech wood products are in outdoor exposures degraded also by ultraviolet (UV) radiation and other environmental factors, e.g., water, oxygen and dust particles in air. The main reason of photodegradation of wood is the presence of light-absorbing chromophoric groups (a-carbonyl, biphenyl and ring conjugated double bond structures) in lignin. The absorption of light by chromoforms leads to the formation of free radicals, which react with oxygen and are responsible for the color changes of wood (HON and SHIRAISHI 2000). 55

The resistance of wood against aging processes can be increased by its treatment with biocides and anti-weathering coatings containing UV absorbers or pigments (REINPRECHT 2016). However, today an environmentally very acceptable method for increasing the aging resistance of wood is its thermal modification (ESTEVES and PEREIRA 2009). Due to thermal modification of wood, usually at temperatures from 160 °C to 260 °C, its molecular structure is changed in association with degradation of hemicelluloses, creation of hemicelluloseslignin linkages, and extinction of some hydroxyl groups (TJEERDSMA et al. 1998, SRINIVAS and PANDLEY 2012, ANDOR and LAGAŇA 2018, CAI et al. 2018). These changes have a positive effect on the water resistance, dimensional stability and biological durability of the thermally modified wood. A range of changes in the molecular structure and partly also in the anatomical structure (e.g. composition and thickness of the individual cell wall layers) and geometry structure (e.g. volume and cracks) of thermally modified woods, connected with following changes in their physical, mechanical and biological properties, is related: (a) to the type of used heating medium, e.g. air, nitrogen, steam, plant oil, or wax; (b) to the thermal modification technology, e.g., regulation of the temperature, pressure, and time; (c) to the tree species used and its initial moisture content (TJEERDSMNA et al. 1998, HILL 2006, YILDIZ et al. 2006, ESTEVES and PEREIRA 2009, KOCAEFE et al. 2015, REINPRECHT 2016). In addition, heat treated wood has uniform dark-brown color in its entire cross-section, while more intense color changes occur due to higher modification temperatures and prolonged thermal modification time (BEKHTA and NIEMZ 2003, SRINIVAS and PANDLEY 2012, DZURENDA and DUDIAK 2020, CIRULE et al. 2021). Darkening of the thermally modified wood is generally attributed to the decomposition of hemicelluloses and the chemical changes in extractives (SUNDQVIST and MORÉN 2002, ESTEVES et al. 2008). Thermal modification of wood can be realised in various media – e.g., in air, nitrogen, natural oils, or waxes (REINPRECHT and REPÁK 2019). Paraffin and other non-polar waxes have hydrophobic properties. They are often used to protect wood against water, as its water sorption kinetic is reduced and dimensional stability is improved (CHEN et al. 2020, REPÁK and REINPRECHT 2020). Combination of wax treatment and heat treatment of wood has even more significant effect on increasing its hydrophobicity (REINPRECHT and REPÁK 2019, YANG and LIU 2020, YANG et al. 2020, CHEN et al. 2020, ZHANG et al. 2020). The hydrophobic treatment of wood with waxes reduces its equilibrium moisture content, increases its dimensional stability and improves its resistance to creation of cracks – because waxes create a barrier in cell walls that slows down the process of absorption of water molecules to wood (WANG and COOPER 2005, AWOYEMI et al. 2009, LESAR and HUMAR 2011, DUBEY et al. 2012, REINPRECHT and REPÁK 2019). The thermal modification of wood has also a positive effect on the formation of cracks in its surfaces, when high temperatures cause depolymerization of hemicelluloses, increase in the proportion of crystalline cellulose and crosslinking of lignin – which reduces the proportion of free -OH groups in wood and the wood has a better dimensional stability. Polyethylene glycols, in contrary to paraffin, are polar macromolecules having hydrophilic properties. They are mainly used for dimensional stabilization of archaeological waterlogged wooden artefacts (HOCKER et al. 2012, MAJKA et al. 2018). Additional effect of the outdoor weathering on the color stability of heat-treated wood is important for practice and was examined in some previous studies (e.g., AYADI et al. 2003, NUOPPONEN et al. 2004, TEMIZ et al. 2006, DUBEY et al. 2010, YILDIZ et al. 2011, HUANG et al. 2012). Thermal modification of wood increased its weathering resistance in comparison to untreated wood (AYADI et al. 2003, NUOPPONEN et al. 2004, TEMIZ et al. 2006, DUBEY et al. 2010). On contrary, some of studies showed that thermal modification of wood decreases its resistance to weathering (YILDIZ et al. 2011, HUANG et al. 2012).

The aim of the experiment was to determine the color parameters of the European beech wood after its thermal treatments in three different media – air, paraffin or polyethylene glycol 6000 (PEG 6000), and also to study the color stability and creation of cracks in the thermally treated wood after its exposition to artificial weathering in Xenotest.

MATERIALS AND METHODS Wood The European beech (Fagus sylvatica L.) wood specimens with dimension of 50 mm × 50 mm × 10 mm (longitudinal × tangential × radial), characterized with a good quality, i.e., without growth anomalies, cracks or biological defects, were prepared from the sawn timber from the National Forest Centrum in Zvolen, naturally seasoned to a moisture content of 13.5 ± 2%. The specimens were dried at 103 ± 1 °C to the oven-dry state in the kiln Memmert UNB 100 (Memmert GmbH + Co.KG, Schwabach, Germany), subsequently cooled in desiccators to a temperature of 20 ± 2 °C, and weighed with an accuracy of 0.001 g (m0). Media for thermal treatment of wood Three different media for the thermal treatment of beech wood specimens were used – air, paraffin and polyethylene glycol 6000 (PEG 6000). The clear paraffin wax (MOL, Hungary) with a melting point of 61 ± 1 °C. The PEG 6000 (HiMedia, Laboratories Pvt. Ltd., Mumbia, India) with a melting point of 58 ± 3 °C, and a molecular weight from 5000 to 7000. Thermal treatments of wood in the air, paraffin or PEG 6000 The thermal treatment of beech wood specimens was performed during 1, 2, 3 or 4 hours, either only by their heating at 100 °C or by their thermal modifications at 190 °C and 210 °C (Fig. 1). The thermal treatments were performed at atmospheric pressure in a kiln Memmert UNB 100 (Memmert GmbH + Co.KG, Schwabach, Germany), using one of the following media – (1) air, (2) paraffin, (3) PEG 6000. The paraffin and PEG 6000 (both solid media up to app. 60 °C) were firstly melted in stainless steel containers during 1 h at a temperature of 100 °C. In the second phase, the specimens were inserted into the melt-liquid medium and impregnated 1 h at a temperature of 100 °C. In the third phase, lasting 1 h, the temperature of the air or the liquid medium was either stable 100 °C or it increased continuously to 190 °C or 210 °C. In the fourth phase, a temperature of the used modification medium was maintained at 100 °C, 190 °C or 210 °C for 1, 2, 3 or 4 hours. In the last fifth phase, the beech wood specimens were cooled directly in the used medium app. to 75 °C, then taken out from steel containers and their surfaces cleaned with filter papers from the liquid-melts of paraffin or PEG 6000, and cooled in desiccators to a temperature of 20 ± 2 °C (Fig. 1). Beech specimens treated at temperatures of 100 °C, 190 °C or 210 °C (i.e., all with a moisture content of 0%) were weighed with an accuracy of 0.001 g (mt0). Then they were 14 days conditioned to a moisture content of 10 ± 2% at a temperature of 20 ± 2 °C and a relative air humidity of 60 ± 3%. The weight percent gain (WPG) values of paraffin and PEG 6000 into the thermally treated beech wood specimens were determined by equation 1: WPG = /(mt0 – m0) : m0/  100 (%)

(1)

Temperature [°C]

250 4th phase

200 150 2nd phase

100 50 0

4 Time [h]

Fig. 1 Phases of the thermal treatment of beech wood in air, paraffin or PEG 6000.

Artificial weathering of wood The artificial weathering (6 cycles, each lasting 1-week – i.e., totally lasting 42 days) of all beech wood specimens (reference and thermally treated) was performed in the Q-SUN Xe1-S Xenotest (Q-Lab Corporation, Westlake, OH, USA). It took place in accordance with the modified version of the standard EN 927-6 (2018), when each 1-week of weathering consisted from 24 h conditioning of specimens at 45 °C and then from 48 subcycle steps— each lasting 3 h (2.5 h UV-radiation and then 0.5 h water spraying). At the artificial weathering, the following modifications to the standard EN 927-6 (2018) were made: xenon lamps instead of fluorescent UV lamps; - irradiance at 340 nm set to 0.55 W·m2.nm1 instead of 0.89 W·m2·nm1; - the temperature on the black panel at 50 °C instead of 60 °C. Color of wood and cracks in wood The color analyses of beech wood specimens 50 mm × 50 mm × 10 mm were made firstly in the original state, then after thermal treatment, and finally after artificial weathering, always in the same eight places of two replicates. The color measurements were performed with the Color Reader CR-10 (Konica Minolta, Japan), having a CIE 10° standard observer, CIE standard illuminate D65, sensor head with a diameter of 8 mm (i.e. the measuring area was 50 mm2), and a detector with 6 silicon photocells. The colorimetric parameters of each beech wood specimen were analysed according to the CIEL*a*b* color system (CIE 2007). A larger value of L*, a*, or b* means a lighter, redder, or yellower colour, respectively. From the relative color changes L, a, and b, namely differences between color coordinates of the thermally treated and original specimens, and of the artificially weathered and thermally treated specimen, the total color difference E*ab was calculated for each beech wood specimen by equation 2: ∗ 𝐸𝑎𝑏 = √𝐿∗2 + 𝑎 ∗2 + 𝑏 ∗2

(2)

Cracks in wood surfaces were determined visually after the thermal treatments and also after the artificial weathering, using magnifier with 10X magnification. Rating of cracks was as follows: [0] no cracks, [1] small cracks with width of 0.1 – 0.3 mm, [2] medium cracks with width of 0.3 – 0.6 mm, [3] large cracks with width of  0.6 mm Statistical Analyses The statistical software STATISTICA 12 (StatSoft, Inc., Tulsa, OK, USA) was used for analysing the gathered data. Determined were the basic statistical characteristics of color parameters, i.e., the arithmetic means and standard deviations. Duncan test analysed for individual color parameters the differences caused at weathering in Xenotest between the thermally treated and reference specimens, on the significance levels of: a ≥ 99.9%, b ≥ 99%, c ≥ 95%, or without significance d < 95%. 58

RESULTS AND DISCUSSION Weight percent gain (WPG) values of paraffin and PEG 6000 The average WPG values of paraffin into the thermally treated beech wood specimens ranged from app. 15% for mode 210 °C/1 h up to app. 20% for several modes at 100 °C and 190 °C (Fig. 2). The PEG 6000 uptake into beech specimens was higher, and it ranged from app. 19% for mode 210 °C/4 h up to 32.4% for mode 100 °C/3h (Fig. 2). The main reason of lower WPG values determined at 210 °C are probably current significant thermal degradation processes in wood structure which are associated with decrease of its weight. 100 °C

WPG (%) 40 30 20 10 0

Paraffin

PEG 6000

190 °C

WPG (%) 40 30 20 10 0

Paraffin

PEG 6000

210 °C

WPG (%) 40 30 20 10 0

Paraffin

PEG 6000

Fig. 2 The WPG values of paraffin and PEG 6000 into beech wood specimens determined after 1, 2, 3 and 4 hours of their thermal treatments at 100 °C, 190 °C and 210 °C.

Color of the thermally treated beech wood It is generally known that color of thermally treated woods changes more intensively with increasing temperature and longer modification time (BEKHTA and NIEMZ 2003, SRINIVAS and PANDLEY 2012, CIRULE et al. 2021). This fact also was confirmed by the present experiments (Tab. 1, Fig. 3, Figs. 6a8a).

Tab. 1 The color changes (ΔL*, Δa*, Δb*) of beech wood specimens due to their air-thermal, paraffinthermal and PEG-thermal treatments. Notes: The arithmetic means were determined from 8 values. The standard deviations are in parentheses. Average color coordinates of the reference (original) beech wood specimens: L* = 78.8, a* = 7.7, b* = 15.8.

Mode of beech wood thermal treatment 100 °C/1 h 100 °C/2 h 100 °C/3 h 100 °C/4 h 190 °C/1 h 190 °C/2 h 190 °C/3 h 190 °C/4 h 210 °C/1 h 210 °C/2 h 210 °C/3 h 210 °C/4 h

ΔL* – Due to thermal treatment In air

In paraffin

In PEG 6000

0.9 (0.23) 1.1 (0.18) 1.4 (0.39) 1.2 (0.39) 23.5 (0.80) 29.0 (2.62) 29.5 (0.97) 32.0 (0.61) 33.2 (0.74) 35.7 (1.01) 38.0 (1.75) 44.7 (3.04)

12.6 (1.84) 11.0 (0.95) 11.3 (0.57) 12,0 (1.04) 28,5 (1.58) 29.9 (1.47) 32.2 (4.20) 39.6 (2.53) 35.4 (3.42) 34.7 (2.04) 43.3 (4.54) 45.8 (0.98)

10.8 (0.68) 12.3 (0.24) 11.6 (0.86) 11.8 (0.43) 22.9 (1.39) 33.8 (1.86) 43.2 (1.20) 38.5 (1.44) 35.4 (1.70) 43.5 (2.28) 51.2 (0.41) 54.2 (0.48)

Δa* – Due to thermal treatment In air

In paraffin

In PEG 6000

0.1 (0.27) 0.0 (0.25) 0.3 (0.98) 0.0 (0.26) 1.4 (0.47) 2.0 (1.81) 1.9 (1.38) 2.6 (0.97) 0.1 (1.05) 0.7 (1.10) 2.9 (0.99) 1.5 (0.49)

5.9 (0.66) 5.7 (0.54) 5.6 (0.45) 5.3 (0.50) 5.7 (1.73) 5.0 (0.63) 6.4 (1.15) 6.3 (1.05) 4.0 (0.57) 1.6 (0.90) 3.0 (1.05) 0.3 (0.93)

5.7 (0.57) 5.0 (3.39) 6.2 (0.67) 5.7 (0.50) 5.5 (0.43) 5.8 (1.13) 4.0 (0.94) 3.0 (1.23) 5.9 (0.55) 4.1 (0.28) 1.6 (0.58) 3.8 (0.68)

Δb* – Due to thermal treatment In air

In paraffin

In PEG 6000

0.3 (0.14) 0.3 (0.23) 0.4 (0.15) 0.4 (0.18) 1.3 (0.97) 1.6 (0.75) 1.1 (0.67) 0.1 (0.72) 1.4 (0.68) 3.9 (0.77) 7.6 (0.75) 7.6 (0.48)

6.7 (0.27) 7.5 (0.34) 8.0 (0.58) 7.1 (0.53) 5.0 (1.15) 3.6 (0.54) 4.2 (0.56) 1.2 (1.19) 0.8 (0.40) 1.0 (0.36) 3.8 (0.82) 7.8 (0.96)

4.8 (0.17) 4.9 (0.28) 5.8 (0.52) 5.3 (0.53) 5.0 (0.81) 3.5 (0.76) 2.3 (0.70) 4.3 (0.69) 2.6 (0.44) 4.3 (0.65) 11.9 (0.55) 12.9 (0.95)

ΔE*ab

Thermal treatment of beech wood in air 100 °C 190 °C 210 °C

60 50 40 30 20 10 0

1h ΔE*ab

Thermal treatment of beech wood in the melt of paraffin 100 °C 190 °C 210 °C

60 50 40 30 20 10 0

1h ΔE*ab

Thermal treatment of beech wood in the melt of PEG 6000 100 °C 190 °C 210 °C

60 50 40 30 20 10 0

Fig. 3 The total color differences (ΔE*ab) of beech wood specimens due to their air-thermal, paraffinthermal and PEG-thermal treatments.

The paraffin-thermal and PEG-thermal treatments of beech wood usually caused greater changes in its color comparing to the air-thermal treatment. This was typical mainly for the lower heating temperature of 100 ° C, when the total color difference (ΔE*ab) for the air-thermally treated specimens ranged only from 1 to 1.7, however, for the paraffinthermally and PEG-thermally treated specimens from 13.1 to 15.5 (Fig. 3). It indirectly 61

means that paraffin and PEG 6000, both having a waxy consistency, significantly affected the color of beech wood specimens already at 100 °C in connection with their darkening, i.e., without their thermal modification at 190 °C or 210 °C (Tab. 1, Figs. 7a and 8a). At using the modification temperatures of 190 °C and 210 °C, the ΔE*ab values were apparently higher from 23.6 to 55.8, however, without an evident depending on the medium (air, paraffin, or PEG 6000) used. Generally, the ΔE*ab values evidently increased with increasing the modification temperature from 190 °C to 210 °C, and with prolonging the modification time from 1 to 4 hours (Fig. 3). Color changes of the thermally treated beech wood due to weathering The color stability of the thermally treated wood is important for interior and mainly for outdoor exposures (REINPRECHT 2016). The color changes of beech wood specimens at the 6-weeks artificial weathering in the Xenotest were significantly influenced by the mode of thermal treatment (Tab 2, Fig. 4, Figs. 5-8). Due to weathering, the brightness changes (ΔL*) of the air-thermally, paraffinthermally and PEG-thermally treated beech wood specimens were specific (Tab. 2, Figs. 58). The reference specimens and specimens heated in paraffin at 100 °C due to weathering darkened, with ΔL* from + 1 to – 19 (Tab. 2, Figs. 5 and 7). In contrast, lighter shades after weathering obtained specimens heated at 100 °C in the melt of PEG 6000, and also specimens thermally modified at 190 °C or 210 °C in all used media. Mostly lightened specimens which were thermally modified in PEG 6000, with ΔL* from + 18.6 to + 32.6 (Tab. 2, Fig. 8). Due to weathering, the reference and usually also the air-thermally treated beech wood specimens showed a positive change of the chromatic coordinate Δa*, i.e., their shade turned to red (Tab. 2, Figs. 5 and 6). In contrast, the surfaces of paraffin-thermally modified specimens reached negative values of Δa*, i.e., their shade turned to green (Tab. 2, Fig. 7). The surfaces of PEG-thermally modified specimens achieved after weathering different changes of Δa*, i.e., specimens modified at 210 °C for 3 and 4 hours obtained redder shade, but the other ones changed color with a shift to green (Tab. 2, Fig. 8). Beech wood specimens thermally modified at 190 °C or 210 °C achieved after weathering in the Xenotest a positive change of the chromatic coordinate Δb*, which means that their shade turned to yellow – the most at PEG-thermally modified specimens (Tab. 2, Figs. 6-8). In contrast, the change in Δb* was due to weathering usually negative for specimens heated at a lower temperature of 100 °C in the melt of paraffin as well as in the melt of PEG 6000 with a color shift to blue (Tab. 2, Figs. 7 and 8). The total color difference of specimens heated at 100 °C in the melts of paraffin or PEG 6000 were due to artificial weathering usually the same (E*ab approximately 15) with those of the reference ones (E*ab equal 14.8) (Fig. 4). Specimens thermally modified in PEG 6000 at 190 °C or 210 °C obtained due to weathering in the Xenotest apparently higher values of E*ab from 20.1 to 36.4, compared to those ones thermally modified in air with E*ab from 5.6 to 13.2, and also to those ones thermally modified in paraffin with E*ab from 7.2 to 15.7 (Fig. 4). This result means that beech wood thermally modified at 190 °C or 210 °C in the melt of PEG 6000 had apparently the less color stability at action of UV light and water during the artificial weathering.

Tab. 2 The color changes (ΔL*, Δa*, Δb*) of the air-thermally, paraffin-thermally and PEG-thermally treated beech wood specimens due to their 6-weeks artificial weathering in Xenotest. Notes: Arithmetic means were determined from 8 values. Standard deviations are in parentheses. The Duncan test was performed in relation to reference specimens with significances: a = 99.9%, b = 99%, c = 95%, d < 95%.

Mode of beech wood thermal treatment 100 °C/1 h 100 °C/2 h 100 °C/3 h 100 °C/4 h 190 °C/1 h 190 °C/2 h 190 °C/3 h 190 °C/4 h 210 °C/1 h 210 °C/2 h 210 °C/3 h 210 °C/4 h

Reference

Mode of beech wood thermal treatment 100 °C/1 h 100 °C/2 h 100 °C/3 h 100 °C/4 h 190 °C/1 h 190 °C/2 h 190 °C/3 h 190 °C/4 h 210 °C/1 h 210 °C/2 h 210 °C/3 h 210 °C/4 h

Reference

Mode of beech wood thermal treatment 100 °C/1 h 100 °C/2 h 100 °C/3 h 100 °C/4 h 190 °C/1 h 190 °C/2 h 190 °C/3 h 190 °C/4 h 210 °C/1 h 210 °C/2 h 210 °C/3 h 210 °C/4 h

Reference

ΔL* – Due to following artificial weathering Treated in air

Treated in paraffin

Treated in PEG 6000

-13.1 (2.89) d -13.5 (3.28) d -12.9 (3.33) d -12.0 (3.45) d 9.7 (1.1) a 11.7 (1.7) a 6.1 (2.4) a 3.8 (0.9) a 5.2 (1.3) a 3.8 (1.1) a 4.2 (1.6) a 9.8 (0.7) a

0.6 (1.44) a -9.4 (1.15) c -10.0 (1.59) c -19.0 (1.15) b 13.8 (0.60) a 10.4 (2.22) a 9.3 (0.77) a 11.1 (1.19) a 15.1 (1.12) a 10.8 (0.80) a 6.5 (0.98) a 7.5 (2.76) a

8.3 (2.01) a 15.1 (1.40) a 12.1 (0.68) a 10.4 (1.03) a 18.6 (2.96) a 22.5 (1.14) a 30.4 (0.54) a 32.6 (2.45) a 20.6 (3.58) a 32.3 (2.97) a 25.2 (1.53) a 30.1 (2.71) a

-13.1 (2.87) Δa* – Due to following artificial weathering Treated in air

Treated in paraffin

Treated in PEG 6000

1.4 (1.03) d 1.1 (0.93) d 1.4 (0.78) d 1.3 (1.11) d 0.4 (0.83) d 0.4 (0.79) d 0.3 (1.14) d 2.0 (0.25) d 2.3 (0.75) d 3.0 (0.47) b 0.9 (0.58) d -1.3 (0.46) d

-4.9 (0.51) a -2.2 (0.52) a -5.7 (0.48) a -1.4 (0.38) a -4.0 (0.47) a -2.7 (0.84) a -4.5 (0.29) a -0.9 (0.19) a -2.6 (0.34) a -3.0 (0.59) a -2.3 (0.46) a 0.0 (0.45) c

-6.0 (3.19) a -8.6 (0.16) a -8.9 (0.32) a -8.7 (0.08) a -4.7 (2.08) a -4.6 (0.60) a -3.4 (0.19) a -5.2 (0.76) a -2:1 (1.06) a -1.3 (0.90) b 4.2 (1.35) b 8.5 (0.39) a

1.2 (0.86) Δb* – Due to following artificial weathering Treated in air

Treated in paraffin

Treated in PEG 6000

6.9 (2.29) d 6.9 (2.15) d 7.1 (1.76) d 6.6 (1.82) d 5.7 (0.92) d 5.0 (2.12) d 1.3 (0.62) a 3.5 (0.44) b 6.1 (1.31) d 6.5 (0.51) d 4.9 (1.29) d 7.7 (0.96) d

-2.5 (0.60) a -5.0 (0.94) a -10.4 (0.76) a -4.4 (0.89) a 0.0 (0.83) a 0.2 (0.57) a 3.5 (0.89) b 3.1 (0.55) a 3.1 (0.49) a -0.8 (1.07) a 2.1 (0.55) a 4.5 (0.89) b

1.0 (0.16) a -1.7 (0.88) a -7.3 (0.91) a -5.5 (0.98) a 4.1 (0.71) c 3.6 (1.39) c 5.0 (0.89) d 6.5 (1.47) d 8.1 (1.20) d 9.4 (1.10) c 14.5 (0.72) a 18.6 (0.57) a

6.7 (2.23)

ΔE*ab

Artificial weathering of the air-thermally treated beech wood 100 °C 190 °C 210 °C

40 30 20 10 0

1h ΔE*ab

Artificial weathering of the paraffin-thermally treated beech wood 100 °C 190 °C 210 °C

40 30 20 10 0

1h ΔE*ab

Artificial weathering of the PEG-thermally treated beech wood 100 °C 190 °C 210 °C

40 30 20 10 0

Fig. 4 The total color differences (ΔE*ab) of the air-thermally, paraffin-thermally and PEG-thermally treated beech wood specimens due to their 6-weeks artificial weathering in Xenotest.

Weathered beech wood

Original beech wood

Fig. 5 The color of the original beech wood specimen (a) and the color change together with the longitudinal cracks created at the 6-weeks weathering in Xenotest (b).

190/1 a)

190/2 a)

190/3 a)

190/4 a)

190/1 b)

190/2 b)

190/3 b)

190/4 b)

210/1 a)

210/2 a)

210/3 a)

210/4 a)

210/1 b)

210/2 b)

210/3 b)

210/4 b)

Fig. 6 The colors of the beech wood specimens thermally modified in air at 190 °C and 210 °C during 1– 4 hours (a), and the color changes together with the longitudinal cracks created at the 6-weeks weathering in Xenotest (b).

100/1 a)

100/2 a)

100/3 a)

100/4 a)

100/1 b)

100/2 b)

100/3 b)

100/4 b)

190/1 a)

190/2 a)

190/3 a)

190/4 a)

190/1 b)

190/2 b)

190/3 b)

190/4 b)

210/1 a)

210/2 a)

210/3 a)

210/4 a)

210/1 b)

210/2 b)

210/3 b)

210/4 b)

Fig. 7 The colors of the beech wood specimens thermally treated in paraffin at 100 °C, 190 °C and 210 °C during 1 – 4 hours (a), and the color changes (for the hardest modification modes “210 °C/24h” as well as the longitudinal cracks) created at the 6-weeks weathering in Xenotest (b).

100/1 a)

100/2 a)

100/3 a)

100/4 a)

100/1 b)

100/2 b)

100/3 b)

100/4 b)

190/1 a)

190/2 a)

190/3 a)

190/4 a)

190/1 b)

190/2 b)

190/3 b)

190/4 b)

210/1 a)

210/2 a)

210/3 a)

210/4 a)

210/1 b)

210/2 b)

210/3 b)

210/4 b)

Fig. 8 The colors of the beech wood specimens thermally treated in PEG 6000 at 100 °C, 190 °C and 210 °C during 1 – 4 hours (a), and the color changes created at the 6-weeks weathering in Xenotest (b).

Cracks created in the thermally treated beech wood at weathering The outdoor weathering of wood is often connected with creation of cracks in its surfaces. During artificial weathering in the Xenotest, the cracks formed in surfaces of the reference specimens (Tab. 3, Fig. 5b), and also in surfaces of the air-thermally and paraffin-thermally treated specimens (Tab. 3, Figs. 6b and 7b). 67

TIRALOVÁ and MAMOŇOVÁ (2005) found out that during thermal modification of beech wood at 205 °C were created in its surfaces only microscopically visible cracks. Similarly, in the present experiment no macroscopic cracks occurred in the thermally treated beech wood specimens (Figs. 6a8a). It indirectly means, that due the following cyclic action of UV radiation and water in the Xenotest the potentially created microscopic (i.e., macroscopically invisible) cracks in the surfaces of the air-thermally modified beech wood and less often also in the paraffinthermally modified beech wood specimens gradually increased and stayed visible by a human eye (Tab. 3, Figs. 6b and 7b). The paraffin-thermally modified specimens were characterized by a slightly better resistance to the formation of surface cracks than the airthermally ones. It can be attributed to the hydrophobic nature of paraffin and an increase in the water resistance of the surface of the beech wood treated with paraffin. At the same time is valid the fact that during the thermal modification of wood in the environment of oils and waxes (i.e., without the presence of air) oxidation reactions in the wood components slowdown, which also results in the reduction of cracks in the wood at its thermal modification (HILL 2006). From the point of view of the weathering effect on the creation of cracks in the paraffin-thermally modified beech wood, the modification process at 190 °C lasting for 1, 2, 3 and 4 hours appears the most optimal as no cracks were detect (Tab. 3, Fig. 7b). Due to weathering of the PEG-thermally modified beech wood specimens no cracks formed at all. It means that PEG 6000 had an even more positive effect on the elimination of cracking the beech wood in its surfaces (Tab. 3, Fig. 8b). Tab. 3 The cracks in the air-thermally, paraffin-thermally and PEG-thermally treated beech wood specimens due to their 6-weeks artificial weathering in Xenotest. Note: Determined from 2 specimens.

Cracks in wood surfaces created at the artificial weathering (0-3)

Modes of beech wood thermal treatment 100 °C/1 h 100 °C/2 h 100 °C/3 h 100 °C/4 h 190 °C/1 h 190 °C/2 h 190 °C/3 h 190 °C/4 h 210 °C/1 h 210 °C/2 h 210 °C/3 h 210 °C/4 h Reference

Treated in air

Treated in paraffin

Treated in PEG 6000

1 2 1 1 1 1 1 1 0 1 1 2

1 0 1 0 0 0 0 0 1 2 1 2

0 0 0 0 0 0 0 0 0 0 0 0

CONCLUSIONS 

The thermal modification of beech wood at 190 °C and 210 °C lasting from 1 to 4 h in three different media – air, paraffin, PEG 6000 – caused always its apparent darkening when L* ranged from – 22.9 up to – 54.2.



Following exposure of the thermally treated beech wood to artificial weathering in the Xenotest, i.e., with presence of UV light and water, reflected in its different color changes. The highest total color difference E*ab from 20.1 to 36.4 had the PEGthermally modified (at 190 °C and 210 °C) beech wood, which evidently lightened with L* from 18.6 to 32.6. On contrary, darkening was determined only for air-thermally and paraffin-thermally heated (at 100 °C) beech wood. The PEG-thermal modification occurred as the best prevention of crack formation in beech wood surfaces.

REFERENCES ANDOR, T., LAGAŇA, R. 2018. Selected properties of thermally treated ash wood. In Acta Facultatis Xylologiae Zvolen, 60(1): 51–60. doi: 10.17423/afx.2018.60.1.06. AYADI, N., LEJEUNE, F., CHARRIER, F., CHARRIER, B., MERLIN, A. 2003. Colour stability of heat treated wood during artificial weathering. In Holz als Roh- und Werkstoff, 61(3): 221–226. doi: 10.1007/s00107-003-0389-2. AWOYEMI, L., COOPER, P. A., UNG, T. Y. 2009. In-treatment cooling during thermal modification of wood in soy oil medium: soy oil uptake, wettability, water uptake and swelling properties. In European Journal of Wood and Wood Products, 67(4): 465–470. doi: 10.1007/s00107-009-0346-9. BEKHTA P., NIEMZ P. 2003. Effect of high temperature on the change in color, dimensional stability and mechanical properties of spruce wood. In Holzforschung. 57(5): 539–546. doi: 10.1515/HF.2003.080. CAI, C., ANTIKAINEN, J., LUOSTARINEN, K., MONONEN, K., HERAJARVI, H. 2018. Wetting-induced changes on the surface of thermally modified Scots pine and Norway spruce wood. In Wood Science and Technology, 52(5): 1181–1193. doi:10.1007/s00226-018-1030-1. CHEN, J., WANG, Y., CAO, J., WANG, W. 2020. Improved water repellency and dimensional stability of wood via impregnation with an epoxidized linseed oil and carnauba wax complex emulsion. In Forests, 11(3): 271–283. doi: 10.3390/f11030271. CIE. 2007. Colorimetry - Part 4: CIE 1976 L*a*b* COLOUR SPACE. CIE DS 0144.3/E:2007, CIE Central Bureau, Vienna, Austria. CIRULE, D., SANSONETTI, E., ANDERSONE, I., KUKA, E., ANDERSONS, B. 2021. Enhancing thermally modified wood stability against discoloration. In Coatings, 11(1): 81–93. doi: 10.3390/coatings11010081. DUBEY, M. K., PANG, S., WALKER, J. 2010. Color and dimensional stability of oil heat treated radiata pinewood after accelerated UV weathering. In Forest Products Journal, 60(5): 453–459. doi: 10.13073/0015-7473-60.5.453. DUBEY, M. K., PANG S., WALKER J. 2012. Changes in chemistry, color, dimensional stability and fungal resistance of Pinus radiata D. Don wood with oil heat-treatment. In Holzforschung, 66(1): 49–57. doi: 10.1515/HF.2011.117. DZURENDA, L., DUDIAK, M. 2020. Changes in wood tree species Fagus sylvatica L. and characteristics of the thermal process of modifying its color with saturated water steam. In Advances in Ecological and Environmental Research, 5(4): 142-156. ISSN 2517-9454. EN 350. 2016. Durability of wood and wood-based products. Testing and classification of the durability to biological agents of wood and wood-based materials. European Committee for Standardization: Brussels, Belgium. EN 927-6. 2018. Paints and varnishes. Coating materials and coating systems for exterior wood Exposure of wood coatings to artificial weathering using fluorescent UV lamps and water. European Committee for Standardization: Brussels, Belgium. ESTEVES, B., DOMINGOS, I., PEREIRA, H. 2008. Pine wood modification by heat treatment in air. In BioResources, 3(1): 142-154. doi: 10.15376/biores.3.1.142-154. ESTEVES, B., PEREIRA, H. 2009. Wood modification by heat treatment: A review. In BioResources, 4(1): 370-404. doi: 10.15376/biores.4.1.370-404.

HILL, C. A. S. 2006. Wood modification – chemical, thermal and other processes. Chichester: John Wiley  Sons, Ltd., UK, 2006, 260 p. HOCKER, E., ALMKVIST, G., SAHLSTEDT, M. 2012. The Vasa experience with polyethylene glycol: A conservator's perspective. In Journal of Cultural Heritage, 2012, 13 (3): 175–182. doi: 10.1016/j.culher.2012.01.017. HON, D. N. S., SHIRAISHI, N. 2000. Wood and celulosis chemistry. Boca Raton: CRC Press, LLC, Florida, USA, 928 p. HUANG, X., KOCAEFE, D., KOCAEFE, Y., BOLUK, Y., PICHETTE, A. 2012. Study of the degradation behavior of heat-treated jack pine (Pinus banksiana) under artificial sunlight irradiation. In Polymer degradation and stability, 97(7): 1197–1214. doi: 10.1016/j.polymdegradstab.2012.03.022. KLEMENT, I., VILKOVSKÝ, P., VILKOVSKÁ, T. 2020. The effect of contact-drying on physical properties of European beech (Fagus sylvatica L.). In Forests, 11(8): 890–900. doi: 10.3390/f11080890. KOCAEFE, D., HUANG, X., KOCAEFE, Y. 2015. Dimensional stabilization of wood. In Current Forestry Reports, 1(3): 151-161. doi:10.1007/s40725-015-0017-5. LESAR, B., HUMAR, M. 2011. Use of wax emulsion for improvement of wood durability and sorption properties. In European Journal of Wood and Wood Products, 69(2): 231–238. doi: 10.1007/s00107010-0425-y. MAJKA, J., ZBOROWSKA, M., FEJFER, M., WALISZEWSKA, B., OLEK, W. 2018. Dimensional stability and hygroscopic properties of PEG treated irregularly degraded waterlogged Scots pine wood. In Journal of Cultural Heritage, 2018, 31: 133–140. doi: 10.1016/j.culher.2017.12.002. NUOPPONEN, M., WIKBERG, H., VUORINEN, T., SIRKKA, L. M., JÄMSÄ, S., VIITANIEMI, P. 2004. Heat-treated softwood exposed to weathering. In Journal of Applied Polymer Science, 91(4): 2128– 2134. doi: 10.1002/app.13351. REINPRECHT, L. 2016. Wood deterioration, protection and maintenance. Chichester: John Wiley  Sons, Ltd., UK, 357 p. REINPRECHT, L., REPÁK, M. 2019. The impact of paraffin-thermal modification of beech wood on its biological, physical and mechanical properties. In Forests, 10(12): 1102–1115. doi: 10.3390/f10121102. REPÁK, M., REINPRECHT, L. 2020. Physico-mechanical properties of thermally modified beech wood affected by its pre-treatment with polyethylene glycol. In Acta Facultatis Xylologiae Zvolen, 62(1): 67–78. doi: 10.17423/afx.2020.62.1.06. SRINIVAS, K., PANDEY, K. K. 2012. Effect of Heat Treatment on Color Changes, Dimensional Stability, and Mechanical Properties of Wood. In Journal of Wood Chemistry and Technology, 2012, 32(4): 304–316. doi:10.1080/02773813.2012.674170. SUNDQVIST, B., MORÉN, T. 2002. The influence of wood polymers and extractives on wood colour induced by hydrothermal treatment. In Holz als Roh- und Werkstoff, 60(5): 375–376. doi: 10.1007/s00107-002-0320-2. TEMIZ, A., TERZIEV, N., JACOBSEN, B., EIKENES, M. 2006. Weathering, water absorption, and durability of silicon, acetylated, and heat-treated wood. In Journal of Applied Polymer Science, 102(5): 4506–4513. doi: 10.1002/app.24878. 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. TJEERDSMA, B. F., BOONSTRA, M., PIZZI, A., TEKELY, P., MILITZ, H. 1998. Characterisation of the thermally modified wood: molecular reasons for wood performance improvement. In Holz als Rohund Werkstof, 56(3): 149–153. doi: 10.1007/s001070050287. YANG, L., LIU, H. -H. 2020. Effect of a Combination of moderate-temperature heat treatment and subsequent wax impregnation on wood hygroscopicity, dimensional stability, and mechanical properties. In Forests, 11(9): 920–928. doi: 10.3390/f11090920. YANG, L., HAN, T., FU, Y. 2020. Effect of heat treatment and wax impregnation on dimensional stability of Pterocarpus macrocarpus wood. In Wood Research, 65(6): 963–974. doi: http://doi.org/10.37763/wr.1336-4561/65.6.963974.

YILDIZ, S., GEZERB, D., YILDIZ, C. 2006. Mechanical and chemical behaviour of spruce wood modified by heat. In Building and Environment, 41(12): 17621766. doi:10.1016/j.buildenv.2005.07.017. YILDIZ, S., YILDIZ, U. C., TOMAK MAIL, E. D. 2011. The effects of natural weathering on the properties of heat-treated alder wood. In BioResources, 6(3): 2504–2521. WANG, J. Y., AND COOPER, P. A. 2005. Effect of oil type, temperature and time on moisture properties of hot oil-treated wood. In Holz als Roh- und Werkstoff, 63(6): 417422. doi: 10.1007/s00107-005-0033-4. ZHANG, J. -W., LIU, H.,-H., YANG, L., HAN, T. -Q., YIN, Q. 2020. Effect of moderate temperature thermal modification combined with wax impregnation on wood properties. In Applied Sciences, 10(22): 82318242. doi: 10.3390/app10228231. 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. APPV-17-0583.

ADRESSES OF AUTHORS Ing. Miroslav Repák Technical University in Zvolen Faculty of Wood Sciences and Technology T. G. Masaryka 24 960 01 Zvolen Slovak Republic xrepak@tuzvo.sk Prof. Ing. Ladislav Reinprecht, CSc. Technical University in Zvolen Faculty of Wood Sciences and Technology T. G. Masaryka 24 960 01 Zvolen Slovak Republic reinprecht@tuzvo.sk

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 73−84, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.06

EXPERIMENTAL RESEARCH ON THE CUTTING FORCE DURING LONGITUDINAL MILLING OF SOLID WOOD AND WOOD-BASED COMPOSITES Valentin Atanasov ABSTRACT Experimental results concerning the influence of some technological factors on the cutting force in longitudinal milling of oak (Quercus petraea), tropical tree species – meranti (Shorea leprosula), koto (Pterygota macrocarpa) and wood-based composite materials such as medium density fibreboard (MDF) and plywood widely used in the furniture manufacturing are presented in the paper. In addition, the results are compared with previously performed studies under the same conditions for tree species like beech (Fagus sylvatica L.) and white pine (Pinus sylvestris L.). Two factors are studied: feeding speed and uncut chip thickness, as it is found that the cutting speed has a lower impact and the optimum spindle revolutions are approximately 6000 rpm. The results show that the forces in plywood milling (maximum value ≈ 46.9 N) significantly exceed those of solid tree species such as beech (Fagus sylvatica L.) and oak (Quercus petraea) (maximum values ≈ 27.7 and 22.4 N). On their basis, regression equations that allow the analytical determination of the target function are derived. Key words: Wood shaper, milling, cutting force, solid wood, wood-based composites, MDF, plywood.

INTRODUCTION In the manufacture of solid wood furniture, mainly because of its very good physical and mechanical characteristics and beautiful texture, tree species such as oak and tropical tree species of meranti find great application. Other widely used tree species in the industry, also with good physical and mechanical parameters but with lower value, are beech and white pine (PETKOV and PANAYOTOV 2016). Recently, within the Republic of Bulgaria, the tropical species like koto find application as well. They are characterized by their beautiful and lighter wood which is not so popular and still more difficult to be recognized. On the other hand, plywood furniture is the basis of many design experiments and appears in response to a search for an urban and contemporary vision of the interior. This material allows the creation of complex shapes that have been provoking designers and furniture builders from around the world for years (SIMEONOVA 2015). Another versatile composite material that is used to make cabinet furniture, armchairs, beds, and so on, are medium density fiberboards. To obtain furniture from the above-mentioned materials, it is necessary to process them mechanically. One of the methods of machining is cutting and the milling takes a 73

significant part of it. Much of the processing is done on the single-spindle moulders, also called universal milling machines, which can be used for various operations. When designing the spindles of the milling machines, it is necessary to determine the forces and moments that affect them. They are due to cutting, unbalanced parts of the elements, tensioning the belt, forces of the weight of the parts, etc. On their basis, the material for the construction is selected, the bending moments in the individual sections, their diameters, the bearings and engine for the cutting mechanism are selected, the technological resistances for the feed, the feed and cutting power, etc. In addition, cutting forces are directly related to forced spatial vibrations, which are often responsible for the occurrence of dangerous resonance (OGUN and JACKSON 2017, VUKOV and GOCHEV 2020). Their determination can also help to improve the reliability of both the elements and the entire machine, even at the stage of its design (TODOROV and KAMBEROV 2017, TODOROV et al. 2018). There is a large number of studies in the literature relating to the vibrations in the cutting mechanism and the surface quality of single-spindle moulders. For example – with а cutting tool in idle mode, with a cutting tool in operating mode, without a cutting tool in idle mode, spatial vibrations caused by unbalance of the cutting tool as well as such caused by cutting force, etc. (KOVATCHEV 2014, VITCHEV 2019, VITCHEV et al. 2020, VUKOV et al. 2018, VUKOV et al. 2020). There are also ones for cutting forces at longitudinal cutting of beech and white pine (ATANASOV et al. 2018, GOCHEV et al. 2018, GOCHEV et al. 2017), as well as other power-energetic indicators of wood moulders for operating of heat-treated (or not) beech wood, white pine, poplar (Populus tremula L.), birch (Betula pendula Roth.), oak, meranti, koto and wood-based composites (KUBŠ et al. 2016, KRAUSS et al. 2016, BARCIK et al. 2008; DURKOVIC et al. 2018, ATANASOV and KOVATCHEV 2018a; ATANASOV and KOVATCHEV 2018b, ATANASOV and KOVATCHEV 2019, KVIETKOVA 2015). Recently, studies of the power-energetic indicators of other woodworking machines were also carried out (KOVAC and MIKLES 2010, KOPECKY et al. 2014, ORLOWSKI and OCHRYMIUK 2017, CHUCHALA and ORLOWSKI 2018, CHUCHALA et al. 2020, CHUCHALA et al. 2021a, CHUCHALA et al. 2021b). There are also many comparative studies related to the processing of different tree species with various types of woodworking machines (SYDOR et al. 2021). The aim of this study is to determine the cutting forces in longitudinal milling of solid wood – koto, meranti, oak, and composite materials such as plywood and MDF and to compare the results with previously conducted under the same conditions.

THEORETICAL BACKGROUND Cutting forces are a result from the interaction of the cutting tool with the material. For simplification, they are reduced to a tangential force, called cutting force (Fc), and a radial force (Ft) which may be insertion or repulsion (positive or negative value). It is known from the Cutting theory that the force Fc is variable – since the thickness of the chip changes. The momentary force Fcm acting on the angle range φ is determined by the formula (ATANASOV and KOVATCHEV 2019) 𝐹𝑐𝑚 = 𝑘𝑐 𝑎𝑝 𝑓𝑧 𝑠𝑖𝑛𝜑, where kc is the specific cutting resistance, N·m2; ap – axial depth of cut (cutting width), m; fz – feed per tooth, m; φ – kinematic angle of encounter, rad.

(1)

With sufficient accuracy for practice, the following dependence is indicated in the literature – obtained by averaging the moment forces for one revolution (KOLEDA et al. 2019) 𝑃

𝐹𝑐 = 𝑣𝑐 = 𝑐

𝑘𝑐 𝑎𝑝 𝑎𝑒 𝑣𝑓 𝑣𝑐

(2)

where Pc is the cutting power, W; vc – cutting speed, m.s-1; vf – feed speed, m.s-1; ae – uncut chip thickness, m. For radial force it is mentioned that it is equal to the cutting force multiplied by a coefficient which takes into account the cutting edge wear (VLASEV 2007). Some of the parameters mentioned in the three formulas are graphically represented in the Figure 1A, B, C. From the above dependencies it can be seen that the following things have an influence on the cutting force – kinematics of the process, the area of milling, physical-mechanical characteristics of the tree species, the type of cutting (edge 90° to grain, traveling parallel to grain – 90°-0°; edge 90° go grain, traveling 90° to grain – 90°-90°; edge parallel to grain traveling 90° to grain – 0°-90°), the degree of cutting edge wear, etc.

Fig. 1 Basic cutting against the feed parameters.

MATERIAL AND METHODS Experimental studies were conducted in the laboratory of Woodworking Machines at the University of Forestry Sofia. For this purpose, a single-spindle moulder was used. Before commencing the experiments, geometric accuracy of the machine was determined according to the standard BDS 3780-84 – by measuring radial and axial oscillation of the spindle, deviation from the flatness of the working surface of the table, deviation from the straightness of the working surfaces of the guide line and deviation from the perpendicularity of the spindle axis to the working table. The values are reported with a measuring clock, a control line, a gauge block set, in accordance with the standard. The tolerances in the parameters can also be seen in KOVATCHEV (2014). Some more important technical parameters of the machine are presented in Table 1.

Tab. 1 Basic parameters of the machine. Element Worktable dimensions 1. Type 2. Power 3. Revolutions per minute 4. Power supply voltage 1. Type of the belt 2. Diameter of the drive pulley 3. Diameter of the driven pulley 1. Material 2. Bearings 3. Vertical motion 4. Mounting diameter

Symbol Value Worktable 860/720/840 Hm, Bm, Lm Electric motor Asynchronous, AC current 3000 N 2880 nе 3 x 380 V/50 Hz Drive gear V-ribbed belt, unit PK 190 D1 90 D2 Spindle Steel 45 2 pcs. radial, single row, ball bearings 95 mv 30 dw

Units mm

W min-1

mm mm

The cutting tool used in experimental studies is a groove cutter which is brand new and was only used in these experimental studies. Since the number of the experiments is not large, it can be assumed that the degree of teeth wear does not significantly affect the results and this factor is not taken into account. Some of its linear and angular parameters can be seen in Table 2. In addition, they are schematically represented in Figures 1 A, B, C. Tab. 2 Basic parameters of the cutting tool. Element 1. Type 2. Material 3. Number of cutting blades 4. Thickness of the cutting blades 5. Working diameter 6. Diameter of the attachment hole 7. Rake angle 8. Sharpening angle 9. Flank angle

Symbol

Value Groove cutter Body – Structural steel/ Plates – HW 6 z 12 s 140 D 30 d 20 γ 58 β 12 α

Units

pcs. mm mm mm ° ° °

The test specimens are presented in Figure 2. They were all selected without any defects from the sapwood of the logs (where it is needed). Some of them were processed in preliminary experiments, the goal of which was to determine the levels of variation of factors. For the tree species of meranti and koto (Fig. 2A, B) they had a cross-section (C × B) 50 × 50 mm and a length L = 1000 mm. In beech and white pine they had the same crosssection (C × B) but a length L = 1520 mm (Fig. 2D, E). The oak wood had a size of 30 × 60 mm (C × B) (Fig. 2C) and length L = 1000 mm, and the composite materials were with a cross section of ≈ 20 × 60 (C × B) and a length L = 1200 mm (Fig. 2F, G). Plywood was made from beech veneer sheets with a thickness of 1.2 mm. Two-component ureaformaldehyde adhesive from DYNEA, Hungary was used. Its consumption was 150 g.m2. The pressing of the plywood was carried out on a multistage press Vecciato VALTER – Italy. The pressing temperature was 110 °C, the duration was 15 minutes and the pressure was 1.3 N.mm2 (SIMEONOVA 2015). On each test specimen, the density was calculated by weight measurement with electronic scales (RADWAG WLC 1/A2 – Poland) and the volume with а calliper and tape measure. In addition, the moisture content of the solid wood was also measured using a moisture meter (Lignomat – Germany). As the results for the cutting forces in longitudinal milling of beech and white pine were already presented in previous scientific 76

works, they were used only for comparative analysis (ATANASOV et al. 2018, GOCHEV et al. 2017, GOCHEV et al. 2018).

Fig. 2 Test samples: А) meranti; B) kоtо; C) oak; D) beech; E) white pine; F) MDF; G) plywood.

In this study, the cutting speed was not changed. Since in the above mentioned studies it was proved that its optimal value in terms of the dynamic behavior of the machine is vc = 44.3 m.s1. This value was obtained by the revolutions of the electric motor, the diameters of the drive D1 and the driven pulley D2 of the belt drive, sliding coefficient of the belt ε and the cutting diameter D (Table 1 and Table 2). The cutting force was obtained by the first part of formula 2, by preliminary calculation of the efficiency coefficient of the cutting mechanism and the cutting power (GOCHEV et al. 2017, KUBŠ et al. 2016) 𝑃𝑐 η = (1 − Pc idle ) 100, (3) load

where Pcidle is input power of the cutting mechanism in idle condition, W; Pcload – input power of the cutting mechanism in load condition, W. Pcload − Pcidle

𝑃𝑐 = (

100

) η.

(4)

In order to measure the input power of the cutting mechanism, the multipurpose device US301EM (Unisyst Engineering Ltd. – Bulgaria) was used. It allows reporting of current, voltage, power factor, active, reactive, full power – for each phase and total. This device is connected to the machine by means of 3 current (CNC® CURRENT TRANSFORMER) and 3 voltage (UNITRAF AD Ltd 220/100 V) transformers – according to the requirements of the manufacturer. The reported values, through software to the device, are automatically imported into Microsoft Excel. Their arithmetic mean was calculated as well. The levels of the factors studied were selected as a result of preliminary experiments. For the feed speed factor they were vf = 2, 6 and 10 m.min-1 and the uncut chip thickness ae = 4, 8 and 12 mm. The visualisation of the experimental studies is shown in Figure 3.

Fig. 3 Scheme of the experimental studies.

The obtained results were processed by the regression analysis method. In this case, a planned two-factor experiment was carried out. Due to their large volume, the individual steps are not mentioned in this report, but they can be seen in the specialized literature on mathematical modelling and optimization of technological objects – like VUCHKOV and STOYANOV (1986). The total number of combinations of the factors is 9. Factor’s levels in explicit and encoded form can be seen in Table 3. Furthermore, for the purpose of verifying the results, additional experiments with levels corresponding to the middle of the factor space were performed – X1 = 0 (6 m·min-1), X2 = 0 (8 mm). The software products QstatLab5 and Microsoft Excel were used to perform the calculations. Through these, regression equations (second degree polynomial) were obtained. They can be used to analytically determine the influence of the factors on the respective target function by entering the coded levels (1, 0, +1). Tab. 3 Experimental matrix. № 1. 2. 3. 4. 5. 6. 7. 8. 9.

X1 (vf) +1 +1 –1 –1 0 0 +1 0 –1

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

X2 (ae) +1 –1 +1 –1 0 +1 0 –1 0

ae, mm 12 4 12 4 8 12 8 4 8

RESULTS AND DISCUSSION Measurements of the machine accuracy indicate that the machine meets the requirements of the standard BDS 3780-84. This means that it can be used for current experiments and will not be affected by additional adverse side effects. In Table 4, the arithmetic mean values of the density and moisture content measurement of the test specimens are shown. It shows that solid wood materials have a moisture content of 1213% and in terms of density they can be divided into two groups – with low density (meranti, koto, white pine) and with high density (beech and oak). In practice, the density of the particular composite materials does not vary within such a wide range – i.e. it does not have such influence as in solid wood. 78

Tab. 4 Density and moisture content of the test samples. № 1. 2. 3. 4. 5. 6. 7.

Density ρ, kg. m-3 490 510 720 650 450 585 735

Tree species/ Composites

Meranti Koto Oak Beech White pine MDF Plywood

Moisture content W, % 13 12 12 13 12 -

The following regression equations were obtained after processing the experimental results. They can be used to calculate the cutting forces in Newtons: – Meranti, – Koto, – Oak, – MDF, – Plywood,

Fc=5.189+3.452vf+6.046ae-0.068vf2+2.843ae 2+2.617vfae;

(5)

Fc=5.685+3.474vf+4.918ae+0.271vf2+0.767ae2+2.369vf ae;

(6)

Fc=5.979+3.181vf+6.949ae+0.609vf2+3.858ae 2+1.827vf ae;

(7)

Fc=5.821+3.271vf+4.580ae+1.083vf2+0.496ae2+2.752vfae;

(8)

Fc=21.772+9.002vf+13.379ae -1.399vf2-1.670ae2+5.821vf ae.

(9)

In order to verify the results according to the requirements for carrying out a planned regression analysis, the Fisher criterion was also computed and compared to its table value. Thus, it has been proven that the equations are adequate and can be used to analytically determine the influence of the input factors on the specific target function. Accordingly, the calculations are performed using the encoded factor values (-1; 0; 1). Furthermore, intermediate levels can be used as well – for example -0.75; 0.25, etc. It can be seen from the equations that for all materials studied, the coefficient of regression is larger next to the milling area factor. Hence, it has a greater impact on the cutting force. This trend is more pronounced in meranti and oak, since the difference in coefficients of both factors for these tree species is approximately double. Figure 4 shows graphical results after solving the above regression equations – influence of the feed speed on the cutting force at different uncut chip thicknesses. In this case, the levels of factor variation are presented in an explicit form.

Fig. 4.1 Influence of feed speed at different uncut Fig. 4.2 Influence of feed speed at different uncut chip thicknesses when milling meranti. chip thicknesses when milling koto.

Fig. 4.3 Influence of feed speed at different uncut chip thicknesses when milling oak.

Fig. 4.4 Influence of feed speed at different uncut chip thicknesses when milling MDF.

Fig. 4.5 Influence of feed speed at different uncut chip thicknesses when milling plywood.

It is clearly seen from Figure 4.5 that plywood composite material requires much greater forces in its processing. The reason for this is the technology used to produce it. In this case the cutting will be more complicated and will be longitudinal for some layers and end-grain cutting for the other layers. Furthermore, the large amount of adhesive also requires larger forces. A general trend for all materials is that at the lowest uncut chip thickness ae = 4 mm, the feed rate has a minimal impact. For example, the difference between the measured values for an uncut chip thickness ae = 4 mm at feed speeds of 2 and 10 m.min1 is 1.04 and 1.67 N for MDF and meranti. This proves the minimal influence of the variable factor at the lowest uncut chip thicknesses. Furthermore, the figures show that, except for plywood milling, at the lowest uncut chip thickness, the cutting force reaches around 5 N only in the oak – the highest density wood. At the average values of the more significant factor of 8 mm, it is noted that for all materials at a feed speed of 10 m.min-1, the cutting force is approximately 10 N. Again an exception to this trend is the plywood, which even at a feed speed of 2 m.min1, the cutting power is about 11.4 N, and at the highest speed it delivers 29.4 N, which is more than the maximum force obtained with all other materials. Figure 5 presents graphically the results for the influence of feeding speed on the cutting force at an uncut chip thickness of 12 mm – for the materials under consideration. In addition, for the purpose of comparative analysis, the polynomial curves for previous tests for the cutting power of beech and white pine are shown as well.

Cutting force Fc, N

50 45 40 35 30 25 20 15 10 5 0

Meranti Koto Oak Beech* White pine* MDF Plywood

5 6 7 8 . Feed speed vf, m min-1

*Results for beech and white pine were added for easier comparison (ATANASOV et al. 2018, GOCHEV et al. 2018)

Fig. 5 Influence of feed speed on cutting force at an uncut chip thickness of 12 mm.

It can be seen from Figure 5 that over its entire length the cutting force curve of plywood significantly exceeds that of all other materials. Its maximum value is 46.9 N. During the experiments, it was clearly felt that the electric motor loaded significantly above its nominal power – 3 kW, which is not desirable. Cutting forces at longitudinal cutting of hard wood such as beech and oak are also relatively high and brought to exceed the rated power of the motor, but to a significantly lesser degree. The values at vf = 10 m.min-1 and ae = 12 mm are 22.4 and 27.7 N. Furthermore, it is noted that a feed speed of about 4 m.min1 the beech cutting force has a lower value than that of oak. From the results follow that beech and oak can be referred to a group of materials with high cutting forces. The other four milling materials – meranti, koto, MDF and white pine, have maximum cutting forces of 20.1, 17.5, 18.0 and 18.8 N, so the first tree species (meranti) can be applied to a group having average cutting forces and the rest to a group with small cutting forces. The figure also shows that curves for both koto and MDF are approximately identical. The reason for the low forces in longitudinal milling of koto is its low density (ρ = 510 kg.m3) and relatively homogeneous and devoid of many defects wood. MDF milling results can be considered a bit unexpected, since the technology for the production of these boards requires the addition of glue. The presence of an adhesive adversely affects the cutting capabilities of the tools and, for longer periods of cutting time, MDF results may have been higher. The relationship between the cutting power and the cutting force allows a comparison with other studied wood materials. It should not be overlooked that the levels of variation of the factors under consideration and the conditions of the experiments are very different. For example, we can assume that thermally treated summer oak (Quercus robur) and silver birch (Betula pendula Roth.) can be referred to materials having medium cutting forces (KOLEDA et al. 2019, KVIETKOVA 2015). This creates preconditions for future research.

CONCLUSIONS Based on the study conducted, the following conclusions and recommendations can be drawn:

1. Regression equations that can be used for the analytical determination of the cutting force for widespread tree species and wood-based composites – oak, meranti, koto, MDF and plywood are proposed. 2. The experimental results allow the examined tree species to be divided into four groups for the relevant ranges of variation of the factors studied – requiring very large forces for their cutting (plywood), with large cutting forces (beech and oak), with medium cutting forces (meranti) and with low cutting forces (white pine, koto, MDF). 3. It is not advisable to process plywood at the highest values of the factors considered, as the electric motor is heavily loaded. It is recommended to use machines with more powerful electric motors in heavier cutting modes. If this is not appropriate, a significant reduction in the feed speed or milling is recommended and carried out in several passes through the machine. Engine overload is also observed when milling beech, but to a lesser extent, which is not dangerous for shorter periods of time. The optimal engine load is obtained by operating the oak with the highest levels of factors. 4. The results obtained for the cutting force can be used in designing the spindle, the cutting mechanism, the bearing determination, the cutting mechanism with an electric motor, the determination of feed resistances, etc. REFERENCES ATANASOV, V., KOVATCHEV, G. 2018a. Determination of the cutting power in processing some deciduous wood species. In 8th Conference on Hardwood Research and Utilisation in Europe: Sopron, 8: 5355. ATANASOV, V., KOVATCHEV, G. 2018b. Study of the cutting power in longitudinal milling of oak wood. In ICWST 2018: Zagreb, 2733. ATANASOV, V., KOVATCHEV, G. 2019. Determination of the cutting power during milling of wood-based materials. In Acta Facultatis Xylologiae Zvolen, 61 (1): 93102. DOI:10.17423/AFX.2019.61.1.09. ATANASOV, V., GOCHEV, ZH., VUKOV, G., VITCHEV, P., KOVATCHEV, G. 2018. Influence of some factors on the cutting force in milling of solid wood. In Chip and Chipless Woodworking Processes, 2018: 915. BARCIK, ST., PIVOLUSKOVA, E., KMINIAK, R. 2008. Effect of Technological Parameters and Wood Properties on Cutting Power in Plane Milling of Juvenile Poplar Wood. In Drvna industrija, 59(3): 107112. BDS 3780:1984 – BULGARIAN STATE STANDARD, Woodworking equipment. Single-spindle moulders. Standards of accuracy and stability. CHUCHALA, D., OCHRYMIUK, T., ORLOWSKI, K., 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. DOI: 10.15376/biores.15.1.1844-1860. CHUCHALA, D., ORLOWSKI, K. 2018. Forecasting values of cutting power for the sawing process of impregnated pine wood on band sawing machine. In Mechanik, (8–9): 766–768. DOI: 10.17814/mechanik.2018.8-9.128. CHUCHALA, D., ORLOWSKI, K., SINN, G., KONOPKA, A. 2021a. Comparison of the fracture toughness of pine wood determined on the basis of orthogonal linear cutting and frame sawing. In Acta Facultatis Xylologiae Zvolen, 63(1): 7583. DOI: 10.17423/afx.2021.63.1.07.

CHUCHALA, D., SANDAK, A., ORLOWSKI, K., SANDAK, J., EGGERTSSON, O., LANDOWSKI, M. 2021b. Characterization of Arctic Driftwood as Naturally Modified Material. Part 1: Machinability. In Coatings 2021, 11, 278. DOI: 10.3390/coatings11030278. DURKOVIC, M., MLADENOVIC, G., TANOVIC, L., DANON, G. 2018. Impact of feed rate, milling depth and tool rake angle in peripheral milling of oak wood on the cutting force. In Maderas. Ciencia y Tecnología, 20(1): 25–34. GOCHEV, ZH., VUKOV, G., VITCHEV, P., ATANASOV, V., KOVATCHEV, G. 2017. Modeling and experimental study of the processes in longitudinal milling of solid wood. In Theme № 22, Sofia: NIS / LTU, 76 pp. GOCHEV, ZH., VUKOV, G., ATANASOV, V., VICHEV, P., KOVACHEV, G. 2018. Study on the Power – Energetic Indicators of a Universal Milling Machine. In 8-th International Scientific and Technical Conference Innovations in Forest Industry and Engineering Design. Sofia, 2018, (1): 18–24. KOLEDA, P., BARCIK, ST., NASCAK, L., SVOREN, J., STEFKOVA, J. 2019. Cutting Power during Lengthwise Milling of Thermally Modified Oak Wood. In Wood Research, 64(3): 537– 548. KOPECKY, Z., HLASKOVA, L., ORLOWSKI, K. 2014. An Innovative Approach to Prediction Energetic Effects of Wood Cutting Process with Circular-Saw Blades. In Wood Research, 59(5): 827–834. KOVAC, J., MIKLES, M. 2010. Research on Individual Parameters for Cutting Power of Woodcutting Process by Circular Saws. In Journal of Forest Science, 56(6): 271–277. KOVATCHEV, G. 2014. Dynamics of the cutting mechanism of the milling machine with bottom of the spindle. Ph.D Thesis, Sofia: University of Forestry, 195 pp. KRAUSS, А., PIERNIK, M., PINKOWSKI, G. 2016. Cutting Power during Milling of Thermally Modified Pine Wood. In Drvna industrija, 67(3): 215222. KUBS, J., GAFF, M., BARCIK, ST. 2016. Factors Affecting the Consumption of Energy during the Milling of Thermally Modified and Unmodified Beech Wood. In Bio Resources, 11(1): 736–747. KVIETKOVA, M. 2015. The effect of thermal treatment of birch wood on the cutting power of plain milling. In Bio Resources, 10(4): 19302126. OGUN, PH., JACKSON, M. 2017. Active vibration control and real-time cutter path modification in rotary wood planning. In Mechatronics, (46): 2131. DOI: 10.1016/j.mechatronics.2017.06.007. ORLOWSKI, K., OCHRYMIUK, T. 2017. A newly-developed model for predicting cutting power during wood sawing with circular saw blades. In Maderas. Ciencia y Tecnología, 19(2): 149162. DOI: 10.4067/S0718-221X2017005000013. PETKOV, T., PANAYOTOV, P. 2016. Correlation between physical and mechanical properties of the wood. In 8th International Scientific and Technical Conference Innovations in Forest Industry and Engineering Design. Sofia, 2016, (1): 1320. SIMEONOVA, R. 2015. Strength and deformation characteristics of corner joints of structural elements made of plywood. Ph.D Thesis, Sofia: University of Forestry, 166 pp. SYDOR, М., MIRSKI, R., STUPER-SZABLEWSKA, K., ROGOZIŃSKI, T. 2021. Efficiency of Machine Sanding of Wood. In Applied Sciences 11(6): 2860. DOI: 10.3390/app11062860. TODOROV, G., KAMBEROV, K. 2017. Virtual prototyping of drop test using explicit analysis. In 43rd International Conference Applications of Mathematics in Engineering and Economics AIP Conf. Proc. 1910, 020012-1–020012-8. DOI:10.1063/1.5013949. TODOROV, G., KAMBEROV, K., KYURKCHIEV, G. 2018. Parametric optimisation of flywheel design. In Journal of the Balkan Tribological Association, 24(3): 390399. VITCHEV, P. 2019. Evaluation of the surface quality of the processed wood material depending of the construction of the wood milling tool. In Acta Facultatis Xylologiae Zvolen, 61(2): 8109. DOI: 10.17423/afx.2019.61.2.08. 83

VITCHEV, P., GOCHEV, ZH., VUKOV, G. 2020. The influence of some factors on the vibrations generated by woodworking spindle moulder machine when processing specimens from beech wood. In Acta Facultatis Xylologiae Zvolen, 62(2): 99−107. DOI: 10.17423/afx.2020.62.2.09. VLASEV, V. 2007. Exercise manual of woodworking machines. Sofia: Publishing House at University of Forestry, 78 pp. VUCHKOV, I., STOYANOV, S. 1986. Mathematical modeling and optimization of technological objects. Sofia: State Publishing House Technique, 341 pp. VUKOV, G., ATANASOV, V., SLAVOV, V., GOCHEV, ZH. 2018. Investigation of spatial vibrations of a wood milling shaper and its spindle, caused by cutting force. In 5th PTF BPI 2018 at the TUM School of Life Sciences Weihenstephan: Freising/Munich, 2018, 144152. VUKOV, G., GOCHEV, ZH. 2020. Investigations of the space vibrations of a woodworking shaper. In Drewno, 63(206): 121136. VUKOV, G., SLAVOV, V., VITCHEV, P., GOCHEV, ZH. 2020. Forced spatial vibrations of a wood shaper, caused by the wear of the cutting tool. In 10th International Scientific and Technical Conference Innovations in Forest Industry and Engineering Design: Sofia, 2020, 81–91. AUTHOR'S ADDRESS Chief Assist. Prof. Valentin Atanasov, PhD, University of Forestry, Faculty of Forest Industry Kliment Ohridski Blvd. 10, 1797 Sofia, Bulgaria vatanasov_2000@ltu.bg

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 85−92, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.07

METHOD OF COMPUTER TEMPLATE ADJUSTMENT FOR WOOD LASER ENGRAVING Maria Zykova – Valeria Kasimiova – Mikhail Chernykh – Vladimir Štollmann – Galina Evstafieva ABSTRACT To achieve the identity of wood-engraved replicas with a pseudo-volume original, it is necessary to adjust the computer model. The method of selecting the model tone range ensuring the reproduction of the original light-and-dark transitions in the replica is introduced in the paper. The method is based on the selection of boundary values of the model black color saturation, taking into account the brightness threshold value perceptible to a human eye. Two methods for determining the boundary values of the model tone saturation were compared – visual by expert assessments and instrumental by scanning replicas and processing scans in Photoshop. The model black color saturation recommended for beech wood in CMYK color space was 2% (lower boundary value) up to 33% (upper boundary value). Key words: beech wood, computer model, laser engraving, perception.

INTRODUCTION Changing the color by thermal exposure (thermodification) expands the possibility of using wood in design, decorative arts, production of interior items and furniture, carpentry (DZURENDA 2013, 2018, GEFFERT et al. 2017). The thermal effect can be carried out by different technologies with different types of energy. Engraving and cutting are performed applying a laser beam to the wood. Laser engraving of wood and wood materials (BAKCIKOWSKI et al. 2004, 2006, YAKIMOVICH et al. 2016, CHERNYKH et al. 2018) allows obtaining a variety of images of high aesthetic value. The color of wood in the laser beam action area changes from natural to brown depending on the heating temperature, which, in turn, depends on the pulse power of radiation P, movement speed of laser head V, saturation of tone R of the computer model developed from the original image (photo, drawing). Pseudo-volume images with light-anddark transitions created by the tone gradient are the most difficult for identity reproduction on wood. The identity of engraved images determines their aesthetic perception and value for a consumer. The most approximation of the replica to the original is achieved when displaying both the lightest and darkest areas of it, which allows providing volume to the engraved image to show small elements. When developing the model, the original is transformed into black-and-white format. The saturation of its color in CMYK color space can vary from 0 up to 100%, and the black color saturation of the engraved image – from the lower threshold value Kf (equal to the 85

natural saturation of wood black color) to the upper threshold value Kt (the tone limit), significantly less than the maximum possible saturation of the original black color, i.e. one hundred percent. The correct display of the original light-and-dark transitions in the replica is possible only in the narrow range of the power-to-speed ratios of engraving without adjusting the model (YAKIMOVICH et al. 2016). The determination of these ratios requires the production of a series of prototypes in the range of possible values of P and V. Therefore, in most cases, the model adjustment is required. The simplest and quickest is the adjustment (and simultaneous approval by the customer) using a computer template attached to the software of some laser units (PHOTOGRAPHV, YSTO CROUP). However, when using this method, high aesthetic properties of replicas are not achieved due to the difference between the computer template and the real work piece in color and texture due to the inevitable fluctuation of these properties from work piece to work piece within the tree species. It seems more promising to use the method of model adjustment based on threshold values of Kf and Kt (CHERNYKH et al. 2015, YAKIMOVICH et al. 2016), which can be found in two ways: using a spectrophotometer and scanning a halftone wedge engraved on the work piece (GOST 24930-84 (State Standard), followed by the scan computer processing in Photoshop. However, even in these cases, the task of the model adjustment cannot be considered completely solved. A consumer visually evaluates the aesthetic value of the engraved product and the desire to purchase it. Perception of one tone brightness by a human eye is limited by a so-called threshold brightness value (MIKOV 2007), so the visually perceived range of black color saturation can differ from the one measured instrumentally, and the range of black color saturation set by the model values of Kf and Kt may need to be clarified. To test the hypothesis, the study was conducted.

MATERIALS AND METHODS The study was carried out using replicas of geometric ornaments of the same pattern engraved on beech wood. Sixteen replicas were engraved from sixteen black-and-white originals. The size of each replica was 20 × 20 mm. For the study convenience, all replicas were concentrated in 100 × 100-mm field on the sample with dimensions in the plan of 130 × 130 mm. The moisture content of the wood was 12%, the front surface of the sample was grinded before engraving. The drawing of each original contained an internal element in the form of a circle, a middle element in the form of a square with four rays radiating from the middle of the square sides, and four external elements in the form of angles (Fig. 1). The originals were made on white paper using laser printer and had the same black saturation of the external elements equal to 100% in CMYK color space, the same saturation of the internal elements equal to 0% and different black saturation of the middle elements. The latter was changed as follows: 3, 5, 7 and 10% in the area of light tones and then up to 70% in increments of 5%. The selected pattern was convenient for comparing the consistently changing tone saturation of the middle elements with the constant tone saturation of the internal and external elements.

Fig. 1 Image original.

The engraving was performed with Trotec Speedy 100R laser engraver with the power of P = 25 W, speed V of up to 2.85 m/s and resolution R = 125-1000 dpi. All replicas were engraved in the same mode P = 14 W, V = 0.7 m/s and R = 500 dpi. To process the results, two methods were used – instrumental measurement of the black saturation of elements on replica scans in Photoshop and method of expert assessments using the method similar to the one in paper (CHERNYH et al. 2012). The difference between the tone of middle elements from lighter internal elements, on the one hand, and darker external elements, on the other, was compared by the method of expert assessments. If an expert believed that the tone of the middle engraved element of the replica in question seemed darker than the tone of the internal (non-graved) element, i.e. answered in the affirmative to the first question of the table, then the expert entered one into the survey table. If the expert thought that the middle element of the replica did not darken as a result of engraving, and was perceived as equal to the tone of the internal element, the expert entered zero into the table. In case of doubt, the expert entered 0.5 into the table. Similarly, the tone of the middle element of the replica was compared with the tone of the external ones. All experts had a high level of color perception, passed or are currently undergoing art training at universities. Two groups of fifteen experts were interviewed at different times. The results of each group were averaged. Comparing the instrumental and visual assessments of the replicas, it can be concluded that it is necessary to take into account the threshold value of a human eye brightness when designing a computer model of the engraved value.

RESULTS AND DISCUSSION The replicas obtained by engraving are shown in Figure 2. The numbers on the left of the replicas indicate the black color percentages of the model middle element corresponding to the replica.

Fig. 2 Replicas laser-engraved on beech wood.

The results of expert assessment of the replicas are presented on the graphs in the coordinates “black color percentage of the model middle element Km – perception W” (Fig. 3). In table 1 line 1 demonstrates the comparison of the middle element tone of the replicas with the tone of their external elements, line 2 – the middle element tone of the replicas with the internal element tone.

Fig. 3. Results of expert assessment of the tone of replica element: 1 – comparison of the tone of middle and external elements, 2 – comparison of the tone of middle and internal elements.

Line 1 can be divided into 2 specific sections – ab and bc. With low saturation of the middle element tone of the model (0≤ Km≤ 10%), corresponding to the section ab, all the interviewed experts, as expected, noted that the middle element tone of the replica was lighter than the tone of its external elements. At the same time, the value of perception W – the ratio of the number of affirmative answers to the number of interviewed experts – equaled one. With further increase in the black color percentage of the middle element of the models, the number of experts who gave an affirmative answer and the perception of W decreased, and at Km ≈33% (in point B of the graph) the perception became equally probable (W=0.5), 88

when the number of experts who gave an affirmative answer equaled the number of those who gave a negative answer. Even with the growth of Km, the perception decreased, approaching zero at Km≈50% or more.

Lighter than Darker than light Perception option of the replica dark middle element tone

Tab. 1 The perception of the tone contrast of the elements of the replicas by experts Replica number 1

Km , % 3

0.7

0.89

0.97

0.84

0.63

0.35

0.36

0.22

0.18

0.06

0.05 0.05 0.02

In the instrumental assessment, the tones of the middle and external elements also converged with the growth of Km (Fig. 4). But their equality in the zone of dark tones occurred at higher values of Km (approximately 50-60%) than in the visual assessment. The discrepancy was caused by the influence of the threshold value of human eye brightness. A human eye, unlike the instrument, does not distinguish tones similar in brightness Kn, Ks, Kv, %.

Fig. 4 Black color percentage of replica elements: 1 – middle, 2 – external, 3 – internal.

In the graph of comparison of the middle element tone with the internal element one (line 2 in Fig. 3) two sections – od and de can also be distinguished. When the tone saturation of the model middle element corresponded to Km values of 7% or more, all experts noted that the middle element looked darker than the internal one, therefore, the perception of W in de section equaled one. The left branch od of graph 2 proceeded from the origin of coordinates. In point 0 the perception equaled zero because the replica middle part with zero saturation of its tone in the model cannot look darker than the internal one. At 3% of the model black color, the average perception value W was 0.62, at 5% – 0 .85, at 7% – 1. Point A of the equally probable perception (W=0.5) in section od corresponded to Km≈ 2%. The black color percentage of the internal element of the replicas measured using Photoshop varied from 0 up to 3%, the fluctuation of values can be explained by the texture influence. Thus, the discrepancy in the results of visual and instrumental assessments in the area of light tones was insignificant and within the experiment error margin. Nevertheless, even in the area of light tones, we can talk about the influence of the brightness threshold value perceived by a human eye, since only at 7% of the model tone saturation all experts noted the difference between the replica tone and the natural tone of wood. The availability of the brightness threshold perceived by a human eye is essential in the design of wood-engraved pseudo-volume images for their identity. For beech wood, the rational black color gradient range of the model is from 2 to 33%. The obtained data are consistent with the results of previous work (CHERNYKH et al. 2018). In this case, the lightand-dark transitions of the original are distinguishable in the replica by most consumers, and the image on the replica is perceived as three-dimensional.

CONCLUSION When engraving pseudo-volume images on wood with a laser, it is necessary to adjust the model to display light-and-dark transitions of the original in the replica in order to achieve identity. The highest aesthetic value of products with engraved images is achieved when the model is adjusted according to the threshold values of black color percentage in CMYK system of the natural tone of wood, on the one hand, and the tonal limit for laser processing, on the other. This method allows taking into account the tone and texture features of the real workpiece. Visual perception of light-and-dark transitions engraved on beech wood is limited by the brightness threshold value perceived by a human eye. A visually perceivable brightness range of a human is narrower than the range determined using replica scanning and processing in Photoshop. In the zone of light tones on beech wood, the difference in ranges, determined visually and instrumentally, is insignificant and within the experiment error margin. In the zone of dark tones on laser-treated wood, the difference reaches 20-30% of black color. When designing the model of pseudo-volume laser-engraved images on wood, the model black color range should be 2≤Km≤33%. REFERENCES BARCIKOWSKI, S., KOCH, G., ODERMANTT, J. 2006. Charakterisation and Modification of the heat affected zone during laser material processing of wood composites. In Holz als Roh und Werkstoff, 64: 94103. BARCIKOWSKI, S., OSTENDORF, A., BUNTE, J. 2004. Laser cutting of wood Composites- Evaluation of cut quality and comparison to conventional wood cutting techniques. In Application of Laser and Optics. Pp. 1823.

CHERNYKH, M., DRYUKOVA, E., USOLTSEVA, A. et al. 2015. Laser engraving of raster images on charred materials. In Design. Materials.Technology. 201, 54(39): 7477. CHERNYKH, M., E. KARGASHINA, V. STOLLMANN 2013. Assessing the impact of aesyhetic properties characteristics on wooddecorativeness. In Acta Facultatis Xylologiae Zvolen. 2013. 55(1): 1326. CHERNYKH, M., KARGASCHINA, E., STOLLMANN, V. 2018. The use of wood veneer for laser engraving production. In Acta Facultatis Xylologlae Zvolen, 60(1): 121127, 2018, DOI: 10.17423/afx.2018.60.1.13 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. Sfarbenie bukoveho dreva v procese termickej upravy sytou vodnou parou [The color of beechwood in the process of internal treatment with water vapor]. In Acta Facultatis Xylologiae Zvolen, 56(1): 1322, ISSN 1366-3824. DZURENDA, L. 2018. Colour modification of Robinia Pseudoacacia L. during the processes of heat treatment with saturated water steam. In Acta Facultatis Xylologlae Zvolen, 60(1): 6170, 2018 DOI: 10.17423/afx.2018.60.1.07 GEFFERT, A., VYBOHOVA, E., GEFFERTOVA, J. 2017.Characterization of the changes of colour and some wood components on the surface of steamed beech wood. In Acta Facultatis Xylologiae Zvolen, 59(1): 4957, ISSN 1366-3824, doi: 10.17423/afx.2017.59.1.05. GOST 24930-81. Tone wedge for facsimile facilities. M.: Standard publishers, 1984, 5 p. MIKOV I.N. 2007. Technologies of automated engraving of artworks / I.N. Mikov, V.I. Morozov. М.: World of mountain book, 2007. 346 p. PHOTOGRAV. The Laser Engraving Power Tool [online]. Document Version 2021/10/04T:15:40:00z 2021. [cit.2021-10-04]. Available online: http://www.photograv.com/ aspent2/ipsAnd1ricks.aspx/. PLATON: Might of a portrait. How To Creat Strong High Key Portraits Inspired by Platon`s Portrait Photography Style [online]. Document Version 2021/10/04T:15:40:00z 2021 [cit.2021-10-04]. Available online: http://www.newsinphoto.ru/ iskusstvp/pleton-mogushestvo-portreta/. YAKIMOVICH, B. et al. 2016. Influence ar selected laser parameters on quality of images engraved on the wood. In Acta Facultatis Xylologiae Zvolen, 2016, 58(2): 4550. YSTO GROUP. Laser Cut 5.0/5.1/5.3 [online]. Document Version 2021/10/04T:15:40:00z 2021 [cit. 2021-10-04]. Available online: http://www.ysto.ru/articles/56-software-for-tools/16-lasercutrusifikator/. NEWSINPHOTO.

AUTHORS’ ADDRESSES Galina Evstafieva, Assoc. Prof. North-Eastern Federal University in Yakutsk Department of Processing Technology of Precious Stones and Metals Kulakovskogo St. 48 Yakutsk, 677 000 todkim@mail.ru Maria Zykova Valeria Kasimova Mikhail Chernykh, Prof., DSc. Kalashnikov Izhevsk State Technical University Department of Industrial and Artistic Processing of Materials Studenchaskaya 7 Izhevsk, 426069 frau.zyckowa2017@yandex.ru kasimova_lera1@mail.ru rid@istu.ru

Vladimir Stollmann, Assoc. Prof. 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

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 93−102, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.08

ANALYSIS OF THE THERMAL BRIDGE OF WOOD-ALUMINUM WINDOW INSTALLATION POSITION Roman Nôta – Zuzana Danihelová ABSTRACT Low-energy houses and passive houses are characterised by having building envelopes with low thermal resistance. Windows of such buildings are required to meet higher demands in terms of heat thermal as well. The window installed in the wall creates a structural thermal bridge, whose size is strongly affected by the thermal performance of the structural components as well as by the position of the window within the opening in the envelope. The objective of the paper was to analyse the impact of the window installation position in the envelope with various construction types and to determine the best possible place for installing the window. Mainly the position of the lower edge is discussed since this part demonstrates the worst values in terms of thermal performance. Due to the presence of the metal windowsill, the window frame is protected to a minimum extent by the outside insulation of the envelope. Therefore, the lower edge indicates the most significant thermal bridge. It was assumed that the most effective position of the window lower edge within the window opening, which has the lowest value of the linear thermal transmittance of the thermal bridge, is in places of the smallest deformation of the thermal field caused by window installation. The calculations showed that the proportion of the deformation size of the thermal field and the value of the linear thermal transmittance of window installation (Ψinstall.), in other words, their position in the structure, depend also on the curve direction of the temperature distribution in the envelope construction. Key words: thermal bridges, linear thermal transmittance, window, envelope structure.

INTRODUCTION Energy efficiency of buildings is closely associated with the number and size of thermal bridges. Currently, the thermal bridges in the building envelope can be avoided by proper insulation of a hight quality. In the case of building envelopes based on wood, this can be achieved also by creating a sufficiently tight envelope preventing humid air from penetrating into the layer of thermal insulation, which would subsequently cause a decrease in the efficiency of thermal insulation. OSB4 boards as the airtight layer are considered the best replacement of foils with high diffusion resistance (SEDLÁK et al. 2020). Outside walls based on silicate building materials do not require the airtightness to such extent. If the building envelope does not contain any significant thermal bridges in terms of geometry (corners) or material (e.g. load-bearing elements made of sandwich constructions), or the thermal bridges are sufficiently eliminated, the building indicates a good energy efficiency without the occurrence of adverse water vapour condensation on its interior surface. 93

The installation of windows and doors is the most significant combined thermal bridge (geometry and material), which virtually cannot be eliminated (O´GRADY et al. 2018, BARNES et al. 2013). The size of the thermal bridge depends on various factors, e.g. construction (NÔTA 2016) and material for the window production or the type of wood in case of wood windows (AHN, PARK 2020). When installing the windows, the most suitable method of eliminating the thermal bridge is to sufficiently overlay the window frame by thermal insulation used for the outside of the envelope (IGELI et al. 2014, CAPPELLETTI et al. 2011). The weakest point in the entire window system is the sill. Overlaying of the window frame at this point is difficult due to the need of creating a construction system for draining the water, either rainwater or condensed water, in the decompression cavity of the window construction. Water most often flows down on the metal (aluminium) overlaying of the envelope – outside windowsill, which is partially connected with the construction of the window and thus creates a thermal conductor. The thermal performance of wood-aluminium windows is partially balanced by the aluminium window parts, and in the case of windows designated for passive houses with lower value of Uf, the impact on the window thermal performance is low. For determining the thermal performance of installed window, compared to the window itself, the relation for calculating the Uw value (EN ISO 10077-1:2019) is modified by adding the linear thermal transmittance of window installation Ψi [W/(m.K)] and relation for calculating the thermal transmission through installed window Uwi is created: Uwi =

Uw ∙Aw + ∑ li ∙Ψi Aw

(1)

where Aw is the overall window surface and li is the length of window groove for sill installation (AUTHORS 2020).

THEORETICAL – EXPERIMENTAL PART The value of Ψi was determined for 5 various wall constructions and one window construction. The wall thickness was constant. The wall constructions were divided as follows:  2x wooden construction: sandwich construction (WC1) a log cabin construction (WC2),  2x brick construction: aerated concrete blocks (BC1) a brick wall with insulation (BC2),  1x polystyrene construction system with reinforced concrete (PCS). Wood-aluminium window with additional insulation designated for passive buildings was used in all models. Calculation of Ψi-value was carried out according to the methodology “B.C. Reference Procedure for Using THERM to Determine Window Performance Values for Use with the Passive House Planning Package. The “BC Reference Procedure” published in September 2019 is the first methodology using LBL THERM software to be recognized by the Passive House Institute for use in certifying Passive Houses to the International Passive House Standard“ (AUTHORS 2019). It was conducted by modelling in computer programme THERM 7.6 (HUIZENGA et al. 2017). Boundary conditions for the calculation were according to the standard STN 73 0540.

Reference temperature:

internal external

θi = 20.0 °C (293.15 K) θe = -12.0 °C (261.15 K) Rsi = 0.13 (m2.K) /W Rse = 0.04 (m2.K)/W

Reference surface resistance: internal external

Tab. 1 Wall compositions, Coefficient of thermal conductivity of wall materials and U-value of walls. Layers Thickness [m] λ [W/(m.K)] Uwall [W/(m2.K)] WC1 – sandwich construction Exterior mineral plaster 0.008 0.800 Fiberboard 0.060 0.050 Mineral Fiber insulation 0.300 0.045 * I beam (OSB+ Picea Abies (L.)) 0.118 OSB 0.018 0.130 Fame Cavity – simplified + Wooden grate 0.040 0.208 ** (Picea Abies (L.)) SDK 0.0125 0.150 WC2 – log cabin construction Log Cabin (Picea Abies (L.)) 0.191 0.110 Mineral Fiber insulation 0.200 0.043 * Wooden grate (Picea Abies (L.)) 0.146 Frame Cavity Slightly Ventilated + Wooden 0.028 0.306 ** grate (Picea Abies (L.)) Wooden siding (Picea Abies (L.)) 0.019 0.110 BC1 – aerated concrete blocks Exterior mineral plaster 0.008 0.800 Aerated concrete blocks 0.420 0.130 0.293 Interior mineral plaster 0.010 0.700 BC2 – brick wall with insulation Exterior mineral plaster 0.008 0.800 Mineral fiber insulation 0.120 0.034 0.175 Bricks 0.300 0.155 Interior mineral plaster 0.010 0.700 PCS – polystyrene construction system Exterior mineral plaster 0.008 0.800 Polystyrene EPS 70Z 0.210 0.039 Reinforced concrete 0.140 1.430 0.134 Polystyrene EPS 70Z 0.070 0.039 Interior mineral plaster 0.010 0.700 * Wooden supporting grate (wooden beam λ=0.11 or I beam: OSB 9mm and wood 40x60 mm) 2pieces per meter and mineral fibre insulation λ=0.034. ** Equivalent thermal conductivity (λeq) of air cavities was determined according to the algorithms in the software program THERM, modelled using the ISO 15099 (Thermal performance of windows, doors and shading devices – Detailed calculations) cavity Model

Calculations were carried out using data from various positions of window installation within the window opening in the wall, and the positions were gradually moved by 10 mm. The first selected interior position was a place where the distance of the 0°C isotherm of glazing was 268 mm from the exterior (the overall thickness of the walls was ca. 438 mm). Altogether, 21 positions of window installation in the window opening were used in the model calculations (24 with BC2 and PCS). In modelling, the impact of the hardware used for fitting the window, especially for attaching to the insulation, was neglected. The exterior windowsill was modelled from 2 mm sheet aluminium (λ = 160W/(m.K)), the insulation in the groove for sill installation from polyurethane foam (λ = 0.024W/(m.K)) and the interior sill was from particleboard (λ = 0.11W/(m.K)).

The Uf = 0.7 W/(m2.K) was adopted from NÔTA, 2020. The window labelled in the publication as ALP2 was used as the model window. It is a wood-aluminium window construction designated for passive buildings. The used model of insulation glazing (IGU) was three panes insulated glazing unit with the thickness of 48 mm (4-18-4-18-4) with Ug = 0.6W/m2K. The Ψg-value was 0.033 W/mK for 2Box Model of the spacer profile Chromatech Ultra F (BUNDESVERBAND FLACHGLAS E.V. 2016).

WW1-1

WW1-10

WW1-20

Fig. 1 Outside-end (WC1-1), central (WC1-10) and inside-end (WC1-20) window installation in the window opening of the wall WC1.

Δbθ0

For each of the model situations the size of thermal field deformation caused by the window position was determined. The size of the deformation was expressed numerically using the 0°C isotherm deviation distance measured in the places of equal distribution of the thermal field in the construction of the envelope and in IGU. This value was labelled as Δbθ0 (Fig. 2).

Fig. 2 Size of the thermal field deformation – 0°C isotherm deviation (Δbθ0)

RESULTS AND DISCUSSION The individual calculated Ψi values from models of window installations were used to plot a graph illustrating the change in terms of the widow position within the window opening (Fig. 3 – 7). The course of Ψi was compared to the value of thermal field deformation in order to compare a possible dependence between this deformation and the size of the linear thermal transmittance. The course of Ψi illustrated graphically is in accordance with studies by e.g. MISIOPECKI et al. 2017 a HØYDAL 2019. 96

Fig. 3 Course of Ψi values and Δbθ0 in WC1.

Fig. 4 Course of Ψi values and Δbθ0 in WC2.

Fig. 5 Course of Ψi values and Δbθ0 in BC1.

Fig. 6 Course of Ψi values and Δbθ0 in BC2.

Fig. 7 Course of Ψi values and Δbθ0 in PSC.

The graphs indicate that the greater the thermal field deformation is, the higher the Ψi value is. Although it was assumed that if the thermal field deformation is minimal, the Ψi value will be minimal as well (Δbθ0 = 0 ꓥ Ψi = min), this assumption was confirmed only in the case of construction BC2. In other cases, the minimum values of Ψi, with the lowest value of Δbθ0, were shifted by 20 and 40 mm into the interior – see Tab. 2 and Fig. 8. It represents a relative shift of 5.75 % and 11.49 % respectively, regarding the thickness of the envelope. This shift was labelled as ΔbΨθ. Tab. 2 ΔbΨθ for individual envelope constructions. 11.49%

ΔbΨθ [mm] 40.0 20.0 40.0 0.0 20.0

[%] 11.49 5.75 11.49 0.00 5.75

ΔbΨθ [%]

Wall construction WC1 WC2 BC1 BC2 PSC

11.49% 5.75%

5.75% 0.00%

WC1

WC2

BC1

BC2

PCS

Fig. 8 ΔbΨθ for individual envelope constructions.

When assessing the discrepancy between the assumption and results, the position of Δbθ0 = 0 and Ψi = min was shown graphically by the temperature distribution in the envelope construction. This procedure was selected due to the ability of analysing the temperatures and thermal field deformation, since the temperature of the construction or 0°C value was selected as the major parameter for assessment. When comparing the temperature distribution in individual wall compositions, where the distances are the same as per the calculations, i.e. WC1 vs. BC1 and WC2 vs. PCS, where the ΔbΨθ is zero, it can be seen that the value of ΔbΨθ decreases with the increasing value of curve direction (m) of temperature development (Fig. 9, 10 and 11). This curve direction was determined at the point where the minimum value of Ψi and zero value of Δbθ0 intersect.

Fig. 9 Temperature distribution with marking min Ψi a min Δbθ0 (A) – WC1, B) – BC1)

A)) A

B) B)

Fig. 10 Temperature distribution with marking min Ψi a min Δbθ0 (A) – WC2, B) – PCS)

Fig. 11 Temperature distribution with marking min Ψi a min Δbθ0 (BC2)

Thermal transmittance is the main way of thermal conduction in materials that the envelope consists of and values Δbθ0 = 0 and Ψi = min. are in solid substances. At the steady 99

state (at which the individual situations were modelled), the temperatures at individual points of a body are determined only by their position, and the temperature curve is represented by a straight line. The line direction can be derived from the density of the heat flow rate. For the steady state and θi(e) = θsi(se) conditions it can be expressed as: m=

∆θs

(2)

where Δθs the difference between the temperatures on the layer surfaces and d is the thickness of this layer (HALAHYJA et al. 1985). In the present models, the minimum values of the linear thermal transmittance of window installation are located in the wall layer where the 0°C isotherm occurs. The values of directions in the curve parts where Δbθ0 = 0 and Ψi = min occurred are provided in Tab. 3 and Fig. 12. 12%

WC1 WC2 BC1 BC2 PCS

m 84.00 109.20 71.90 165.80 110.00

ΔbΨθ [%] 11.49 5.57 11.49 0.00 5.57

ΔbΨθ [%]

Tab. 3 Values of curve direction of temperature distribution in the envelope vs. ΔbΨθ.

8% 4% 0%

100

120

Slope

140

160

Fig. 12 Dependence of m on ΔbΨθ

The course of curve direction dependence of the temperature distribution in the construction and the size of the thermal field deformation (ΔbΨθ) indicates a certain dependence between these two values. However, from already published and available studies and the present models, this dependence cannot be determined precisely.

CONCLUSIONS AND FUTURE WORK The present models compared the dependence of window installation position in the window opening and the values of linear thermal transmittance of window installation. Subsequently, the course of transmittance values was compared with the size of the thermal field deformation in the detail of window installation in order to assess their dependence. One widow construction and five wall constructions (two wood-based walls, two walls composed of brick components, one cast-in-place reinforced concrete wall) were used in the research. It was assumed that the minimum deformation of the thermal field caused by the thermal bridge of window installation will be accompanied by a lower value of linear thermal transmittance. However, this assumption was confirmed only in one of the five studied models. After plotting the minimum values of Ψi and thermal field deformation into the temperature distribution in the construction, a possible dependence between the lowest value of Ψi and thermal performance of the construction was indicated. The difference between the minimum value of Ψi and minimum thermal field deformation decreases with an increasing absolute value of the curve direction of the temperature distribution in the construction. However, due to the low number of studied models, this dependence cannot be determined precisely. Therefore, it is inevitable to verify the dependence in a study with a higher number of models. 100

REFERENCES AUTHORS 2020. Information, Criteria and Algorithms for Certified Passive House Components: Sun Protection and Window installation Systems, Version 2.0, 2020-07-03, on line: <https://passivehouse.com/03_certification/01_certification_components/02_certification_criteria/0 1_transparentcomponents/01_transparentcomponents.html>, Passive House Institute, Darmstadt, 2020 AUTHORS 2019, B.C. Reference Procedure for Using THERM to Determine Window Performance Values for Use with the Passive House Planning Package. On-line: <https://www.fenbc.org/resource_details.php?id_resource=3>, Fenestration Association of BC, 2019. AHN, N, PARK S, 2020. Heat transfer analysis of timber windows with different wood species and anatomical direction. In Energies, vol. 13, MDPI, Basel, Switzerland, 2020, ISSN 1996-1073 BARNES B., PAGÁN-VÁZQUEZ, A., LIESEN, R., YU, J., ALEXANDER, N., et al. 2013. Window related thermal bridges. In Thermal Performance of the Exterior Envelopes of Whole Buildings XII International Conference, Proceedings of a meeting held 1-5 December 2013, Clearwater, Florida, USA, ISBN 9781510827837 BUNDESVERBAND FLACHGLAS E.V. Data sheet Psi values for windows, based on determination of the equivalent thermal conductivity of spacers by measurement, for the Chromatech ultra F (Nr. W 16 3-10/2018 CAPPELLETTI, F., GASPARELLA, A., ROMAGNONI, P., BAGGIO, P.,. 2011, Analysis of the influence of installation thermal bridges on windows performance: The case of clay block walls. In Energy and Buildings, 43: 14351442, Elsevier B.V, 2011, ISSN 0378-7788, DOI: 10.1016/j.enbuild.2011.02.004 IGELI, R. VAVROVIČ, B., ČEKON, M., PAULOVIČOVÁ, L., 2014. Thermal bridges minimizing through window jamb in low energy buildings. In Advanced Material Research, vol 899, pp 6669, Trans Tech Publications, Switzerland, ISSN 1662-8985, DOI: 10.4028/www.scientific.net/AMR.899.66 HALAHYJA, M. et al. 1985, Stavebná tepelná technika, akustika a osvetlenie, Bratislava ALFA, 2019, Kuldebroer ved vindusinnsetting - Termal Bridges for Window-to-Wall Connections, Masteroppgave, Norwegian University of Life Sciencest, 2019 HUIZENGA, CH. et al. 2017. THERM Fine Element Simulator v7.6.1.0: Program description. A PC program for analyzing the two-dimensional heat transfer through building products. Berkeley. California: University of California 2017. MISIOPECKY, C. , BOUQUIN, M., GUSTAVSEN, A., JELLE, B.P. 2017, Thermal modeling and investigation of the most energy-efficient window position. In Energy & Buildings, vol. 158, pp. 10791086, Elsevier B.V, 2018, ISSN 0378-7788, DOI: https://doi.org/10.1016/j.enbuild.2017.10.021 NÔTA, R. 2020, Okná na báze dreva: tvarové riešenie profilu vo vzťahu k tepelnotechnickým vlastnostiam (Wood-based windows: shape solution profile with respect to thermal properties) Zvolen: Technická univerzita vo Zvolene, 2020. ISBN 978-80-228-3228-1 NÔTA, R. 2016, Thermal performance of wood aluminum and wooden windows. In Acta Facultatis Xylologiae Zvolen, 58(1): 8394 ISSN 1336-3824. DOI: 10.17423/afx.2016.58.1.10 O´GRADY, M., LECHOWSKA, A. A., HARTE, A. M. 2018, Application of infrared thermography technique to the thermal assessment of multiple thermal bridges and windows. In Energy & Buildings, vol. 168, pp. 347362, Elsevier B.V, 2018, ISSN 0378-7788, DOI: https://doi.org/10.1016/j.enbuild.2018.03.034 SEDLÁK, P., BEDNÁR, J., BÚRYOVÁ. D. 2020. Air permeability of OSB and its influence to heating energy costs. In Sustainability of forest-based industries in the global economy: proceedings of scientific papers: Vinkovci, Croatia, September 28th-30th 2020, p. 279-284/; Zagreb: WoodEMA, i.a.: University of Zagreb, Faculty of Forestry, 2020. ISBN 978-953-57822-8-5 STN 73 0540-2/2012, Thermal protection of buildings. Thermal performance of buildings and components. Part 2: Functional requirements, ÚNMS SR, 2012 STN EN ISO 10077-1/2019, Thermal performance of windows, doors and shutters – Calculation of thermal transmittance – Part 1: General, ÚNMS SR, 2019

101

AUTHOR´S ADDRESS Ing. Roman Nôta, PhD. Department of Furniture and Interior Design Technical University in Zvolen T.G. Masaryka 24 960 01 Zvolen nota@tuzvo.sk Mgr. Zuzana Danihelova, PhD. The Institute of Foreign Languages Technical University in Zvolen T.G. Masaryka 24 960 01 Zvolen

102

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 103−116, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.09

RESISTANCE OF SPRUCE WOOD (Picea abies L.) TREATED WITH A FLAME RETARDANTS AFTER THE RADIANT HEAT EXPOSURE Iveta Čabalová – Martin Zachar – Michal Bélik – Žaneta Balážová ABSTRACT The aim of the paper was to determine the resistance of spruce wood (Picea abies L.) treated with a flame retardant after exposure to radiant heating. The treatment was performed with a typical flame retardant HR prof and an aluminium coating used to protect metal objects at high temperatures – Striebrenka, an atypical retardant not commonly used in wood protection. The samples treated with retardants as well as untreated (control) ones were heatloaded by radiant heating using a radiation panel for 30 min. Changes in the basic chemical composition, especially the content of extractive substances, holocellulose, cellulose, hemicelluloses and lignin in two layers removed below the charred layer (layer 1 - thickness up to 20 mm. and layer 2 - thickness from 20 to 40 mm) and thickness of the charred layer were evaluated. During the thermal loading of the control sample, we observed the temperature of 300°C in time of 860 seconds on thermocouples (M 1.0). The average thickness of the charred layer was 15.41 mm. The limit temperature of 300°C was not reached on any thermocouple (M 2.0 – M 5.0) in all samples. The average thickness of the charred layer of the sample treated with HR prof was 14.9 mm and with Striebrenka 15.18 mm. Following the results, it can be stated that the chemical composition of the timber was changed as a result of the chemical treatment of wood by radiant heating. However, the largest changes were recorded in the sample not coated with any retardants at all. A decrease in the saccharides component of wood, especially hemicelluloses (from 13.57 to 20.25%) was observed in the case of this sample. When comparing the chemical changes of samples treated with flame retardants, it can be seen that the HR prof retarder was more effective in terms of saccharides and the Striebrenka retardant in terms of lignin protection. In contrast, during the thermal degradation of treated samples, a decrease in lignin (from 7.66 to 8.10%, HR prof retardant) and holocellulose (from 5.83 to 8.3%, Striebrenka retardant) content compared to the original sample was observed. Key words: spruce wood, radiant heating, flame retardant, extractives, lignin, cellulose, hemicelluloses.

INTRODUCTION Due to the properties of spruce wood and its availability, it is widely employed tree species in constructions (e.g. solid timber floors, windows, doors, furniture, toys, telegraph and other poles, structural material for trusses and ceilings). It is one of most often-used tree species in Slovakia represented by 22.5 % (ZELENÁ SPRÁVA 2019). However, the wood is a 103

combustible material and the process of its heat treatment results in degradation. Degree of degradation depends on time and type of heat loading (OČKAJOVÁ et al. 2020). Flame retardants or protective coatings improving its fire resistance are used to protect the wood. The nature of these products is chemical or physical. When applying to wood, stability of heat treated wood increases significantly. There are chemical and physical flame retardants, however, the difference is evident. While chemical flame retardants are focused on reducing the flame spread covering the material with a foam or producing the inert gasses into the fire, the physical retardants are able to absorb or to deflect heat released by fire. The fact that physical retardants are especially produced of solid material covering wood, thus the microclimate resulted in biodegradation of material is not created can be considered a disadvantage (ZAIKOV, LOMAKIN 2002). During the process of heat treatment, firstly, the wood is dried, i.e. the redundant moisture evaporates, later, the significant chemical changes occur (KAČÍKOVÁ et al. 2013, KUČEROVÁ et al. 2016, KAČÍK et al. 2017, GAŠPARÍK et al. 2018, LUPTÁKOVÁ et al. 2018, GAFF et al. 2019a, b). At a temperature higher than 250°C, carbonization process starts and CO2 and other pyrolysis products are emitted. When the temperature is increasing, the main chemical components of wood, especially cellulose and hemicelluloses are affected by pyrolysis and oxidation reactions and inflammable gases are products resulting from their decomposition. Polysaccharides are substances supporting combustion most. During the thermochemical reactions, lignin, which is thermally more stable, contributes to coal formation, in a similar way to hemicelluloses or cellulose (NUOPPONEN et al. 2005). Flammable gasses formation useful in protecting wood from thermochemical reactions is stopped when the product formation such as coal increases. Moreover, the development of the reactions is slowed. In our other paper we followed the thermal loading of untreated spruce wood by radiant heating and the results show the both lignin and extractives content increased in light and dark brown layer but extractive content in charred layer decreased. Cellulose content (determined by Seifert method) in each layer increased because of its carbonization and crosslinking but total content of saccharides dropped, the most in the charred layer. The strength of spruce wood is decreased by thermal loading and cellulose depolymerisation (ČABALOVÁ et al. 2013). A large part of lignin is resistant to the effect of thermal exposure. Its degradation starts at lower temperature, similar to polysaccharides, but the carbonization results in a product resistant to the heat and flammable only with difficulty. During slow heating and at the presence of atmospheric pressure, there are conditions suitable for condensation reactions. The amount of volatile products is relatively low in comparison to the extract of solid component, coal. Phenols are the most important products forming during the thermolysis. Radicals with different structures and stability result from the application of heat. Various degradation and condensation reactions are joined (BREBU, VASILE 2010). At the temperatures ranging between 100180°C, plasticisation happens (NUOPPONEN et al. 2003, KAČÍKOVÁ et al. 2008). The phase of stabilized combustion occurs after forming the charred layer on the exposed surface of samples. Heat release stays almost constant. It is followed by the phase of burning the overheated sample described with substantial pyrolysis through the whole sample (RANTUCH et al. 2015). The depth of charred area and the speed of carbonization is one of the most important fire properties of wood and wood products. The standard (Eurocode 5) is used to analyse the depth of charred area, which is considered the most reliable figure in terms of assessing the flame spread. In accordance with the standard STN EN 1995-1-2 (2004), the depth of charred area is defined as a distance between the outer surface of the original element and the position of the line between the charred layer and the rest of the cross-section. The line of the charred 104

layer of timber construction is specified in the standard as the place with a temperature of 300°C. As it is mentioned by FONSECA and BARREIRA (2009), the charred layer is the dividing line between thermally degraded and non-degraded wood bounded by a black and a brown wood layer and it is characterised by a temperature of 300°C. According to the statement of WHITE and NORDHEIM (1992), the charred layer corresponds to the temperature of 288°C. FINDORÁK et al. (2016) agree with the mentioned statement, i.e. rapid thermal decomposition of wood (in the case of short-term exposure) starts just below the temperature of 300°C. The aim of the paper was to observe the effectiveness of two flame retardants, first standard flame retardant based on the chemical principles and the second atypical following the physical principals, used for the protection of thermal loaded spruce wood.

MATERIAL AND METHODS Samples Four samples of spruce wood (Picea abies L. Karst.) (harvested in the east part of Slovakia in Dobšiná, trunk diameter approximately of 35 cm, age of 85 years) were cut into the shapes of blocks with the dimensions of 150 × 150 × 1,000 mm (thickness × width × length). The wood surface was treated by 80 grit sandpaper. The retardant was applied by painting to the air-dry wood (moisture of 8 %) in two coats in a time interval at least of 40–60 minutes. The overall coating thickness was 300 g/m2 in line with the instructions of producer, which is our case 81 g of retardant per sample. The weight of applied retardant was less than 1% of the weight of the test sample. Flame retardants used HR prof It is a fire protection paint restricting ignition and spread of flame of timber constructions, staircases, cassette ceilings, wood flooring and other products made of wood and cellulose. It can be used in interior as well as exterior spaces. Once dry it is not broken down by water or humidity. At the high temperatures, there is no smoke emission and charred dust dispersion. Therefore, it is not harmful to the environment. When material treated with HR prof is exposed to the temperatures higher than 1700 ºC, it turns black with smoke and the flame stops spreading. Striebrenka (Silver paint) Aluminium paint is determined to protect derusted metal items against oxidation during heat exposure up to the temperature of 600 °C. Properties: dispersion of sodium silicate solution, heat-resistant inorganic fillers, organic ingredients and aluminium powder. It is used to protect metal items against corrosion at high temperatures up to 600 °C. According to the instructions of the producers, the clean surface must be covered with 1-3 layers of undiluted retardant. During the first exposing the treated unit to the heat, thermal shock (due to fast emission of volatile components) is to be avoided to give the volatile components time to penetrate slowly with no impact on the surface integrity. The reflection of thermal radiation is the principle of Striebrenka; therefore, it does not act as a chemical retardant and it is not used as a flame retardant. Due to its properties it belongs to the fourth class – retardants with physical characteristics. Most of retardants in this group are of solid state, therefore there is a chance to create a microclimate after their application to wood. On the contrary, Striebrenka is a liquid, it must be dried before thermal loading.

105

Radiant heating Three samples were thermally loaded, namely the sample treated with HR prof, with the Striebrenka retardant and control (untreated) sample of original wood. A ceramic radiated panel used for model testing to evaluate the building elements in terms of fire protection was used as the source of radiant heating. The heat source can be characterized by the following data: the dimensions of the radiation zone – 480 × 280 mm, maximum performance of the radiation zone – 50.5 kW∙m-2, the total reached temperature of the radiation zone – 935°C. The radiating flow (30.9 kW∙m-2) must supply the body of the radiant with propane at the constant flow of 13 l∙hour-1 in order to achieve the requested performance. The support stand was placed 30 cm from the radiation panel. The thermoselement of K type K (Ni-Cr-Ni) to measure the temperature on the surface of the block were placed there as well. Second thermos-element of K type measuring the temperature of the environment during the experiments (T = 25.0 ± 0.7 °C) was located in the same room. The first thermocouple (M0.0) was placed on the side directly exposed to the radiant heat from the radiant panel. Another five thermocouples were inserted into drilled holes (75 mm deep below the surface, i.e. in the centre of the sample) on the top side of the wooden block (M1.0 – M5.0) at a distance of 10 mm from each other. During measurement, the thermocouples on the top side of the sample were covered with mineral wool. The last thermocouple was freely placed in the testing laboratory to measure the ambient temperature. The samples were gradually placed on the stand in front of the radiation panel (at a distance of 20 cm) affected by the radiation heat for 30 minutes per each sample. The procedure was monitored and recorded by digital measurement device (Almemo 2290-8). The recorded data were evaluated. Size and depth of charred area The thickness of the charred layer or the depth of charring is defined as the thickness of the layer of material at the surface of the timber cross-section that burned out and charred (thickness added to the original cross-section) and lost the ability to transfer stresses due to degradation of mechanical properties. The material in this layer is charred, with no strength or even fallen off. The measurement was conducted using a laboratory measuring instrument and a digital calliper. Firstly, the length of the charred layer was measured. The charred layer must be removed to measure its depth. After removing the charred layer, the depth was measured in nine predetermined points (starting in the centre of the sample and continuing at a distance of 100 mm from the centre). Sampling for the purpose of chemical analysis Sawdust must be used to determine the chemical composition of wood. Two layers were removed by a circular saw. The thickness of the first layer was up to 20 mm, dark brown coloured wood (layer 1) and the thickness of the second one, light brown coloured wood, was from 20 to 40 mm (layer 2) below the charred layer. Chemical composition of wood Samples were disintegrated into sawdust and fractions 0.5 mm to 1.0 mm in size were used for the chemical analyses. The extractive content was determined in a Soxhlet apparatus with a mixture of ethanol and toluene (2:1) according to ASTM D1107-96 (2007). The lignin content was determined according to SLUITER et al. (2012), the cellulose content was determined according to the method by SEIFERT (1956) and the holocellulose content according to the method by WIESE et al. (1946). Hemicelluloses were calculated as a difference between the holocellulose and cellulose content. Measurements were performed on four replicates per sample. The results were presented as oven-dry wood percentages.

106

RESULTS AND DISCUSSION When samples were exposed to radiant heat from the radiant panel, the changes to various extent could be monitored in each of them. In the case of the sample treated with no retardant, so-called control sample, the most significant changes occurred.

Temperature (°C)

Changes in a temperature when loading by radiant heat The sample with HR prof The changes in temperatures in individual thermocouples are presented in the graph (Fig. 1). The temperature on the thermally loaded side of the sample rose substantially and there was a gradual fall to the depth (10–50 mm). In the case of this sample, an increase in a temperature of the first two thermocouples M0.0 and M1.0 placed on wood close to the radiant panel could be seen. The final value of the temperature was 175 °C during the first 60 s. Furthermore, the temperature rose slowly to 190.6 °C (maximum). In the case of the thermocouple M1.0, rapid increase in a temperature during the measurement was observed. The temperature values measured in this place were 479 °C. There was no fluctuation in temperatures, they increased steadily after the third minute. In the case of the thermocouple M2.0, the value was lower than in the case of M1.0. The thermocouple M2.0 was placed 20 mm far from the loaded side. The temperature recorded increased very slowly to the value of 170.4 ° at the end of measurement. The temperatures recorded in the case of other thermocouples were not higher than 150 °C during the measurement. Following the data, it can be stated that there was the biggest load of the sample with the high temperature in the depth of 10 mm from the loaded side of wood. 600

M0.0

500

M1.0

400

M2.0

300

M3.0

200

M4.0

100

M5.0 M6.0

Time (s) Fig. 1 Changes in temperatures during the thermal loading, the sample treated with HR prof (M0.0, M1.0, M2.0, M3.0, M4.0, M5.0, M6.0 - thermocouples).

In the case of the sample treated by retardant HR prof, on thermocouple M 1.0, the temperature of 300°C was reached in time of 990 sec which correspond to the thickness of charred layer 10 mm below the thermal loaded surface. On other thermocouples (M2.0 – M5.0) (Fig. 1), the limit temperature 300°C was not reached in time up to 1800 sec. The sample with Striebrenka The temperatures when the sample was thermally loaded under the same conditions are mentioned in the graph (Fig. 2). However, the development of temperatures was different. 107

600 M0.0

Temperature (°C)

500

M1.0

400

M2.0

300

M3.0

200

M4.0 M5.0

100 1800

1700

1600

1500

1400

1300

1200

1100

1000

900

800

700

600

500

400

300

200

100

M6.0 0

Time (s) Fig. 2 Changes in temperatures during the thermal loading, the sample treated with Striebrenka (M0.0, M1.0, M2.0, M3.0, M4.0, M5.0, M6.0 - thermocouples).

A significant increase in the temperatures in the case of the thermocouples M0.0 and M1.0 in comparison to other thermocouples could be monitored. The temperature recorded by the thermocouple M0.0 placed on the surface of loaded side rose from the beginning of the measurement. From the beginning to the 10th minute, the temperature recorded in this place was lower than 200 °C. Later, the temperature started to increase to the value of 328 °C after 28 minutes. From this point, the temperature started to fall down to the value of 238°C. There were differences in the development of temperatures recorded by the thermocouple M1.0. The temperature of the material after 5 minutes from the beginning of the thermal loading was 100 °C. Since then, the temperature increased steadily without significant differences. After 18 minutes, there was a radical change in a temperature, from 263 °C to 377 °C during 3 minutes. From this point, the temperature increased steadily for next 24 minutes when another, albeit not so radical, change occurred. The temperature value was 436 °C and continued rising until the end of the measurement, to the value of 567 °C. The temperatures recorded by the thermocouple M2.0 were significantly lower. At the end of the measurement the temperature recorded by the thermocouple was approaching the value of 200 °C. The temperatures recorded by other thermocouples were lower with growing tendency during the measurement. In case of sample treated by Striebrenka, on thermocouple M 1.0, the temperature of 300°C was reached in time of 1 170 sec which correspond to the thickness of charred layer 10 mm below the thermal loaded surface. On other thermocouples (M2.0 – M5.0) (Fig. 2), the limit temperature 300°C was not reached in the time interval up to 1800 sec. Control sample The sample was used as a control sample in order to determine the impact of a retardant during the thermal loading of wood. In the case of this sample, the faster increase in temperature as well as higher temperatures comparing to treated wooden blocks occurred. The most significant change was observed in the case of the first two thermocouples M0.0 and M1.0. The temperature recorded by the thermocouple M0.0 in one minute was 100 °C. Sharp rise in temperature lasted for two minutes and the achieved temperature was 235 °C. Since then the temperature was rising steadily for 21 minutes when the sample started burning. In this point, a significant change (Fig. 3) and subsequent increase in temperature can be seen. The flames were visible for 3 minutes. At the end of flame burning, a substantial fall in values of measured temperatures was monitored. The temperature plummeted from 108

the value 506°C to 362 °C. The same fluctuation in temperatures was similarly recorded by the thermocouple M1.0. In the case of this thermocouple, there was a steady increase during the first 20 minutes. At this point, there is a significant change from 336 °C to the value of 441.7 °C, and the flames could be seen as well. In the case of this thermocouple, there was a steady rise in a temperature during the time of burning. The temperature achieved was 553.5 °C. After burning, the temperature decreased significantly to the value of 384 °C. Since then the temperature rose steadily to the value of 424°C. Furthermore, there was a falling tendency to the temperature of 351 °C (the end of measurement). In the case of other thermocouples, slower increase in the temperature was recorded.

600 M0.0

Temperature (°C)

500 M1.0 400

M2.0

300

M3.0 M4.0

200

M5.0 100

M6.0

Time (s) Fig. 3 Changes in temperatures during the thermal loading, control sample (M0.0, M1.0, M2.0, M3.0, M4.0, M5.0, M6.0 - thermocouples).

In the case of untreated sample, on thermocouple M 1.0, the temperature of 300°C was reached in time of 860 sec which correspond to the thickness of charred layer 10 mm below the thermal loaded surface. On other thermocouples (M2.0 – M5.0) (Fig. 3), the limit temperature 300°C was not reached in time up to 1800 sec. Size and depth of charred area The sample with HR prof Following the nine measured data, it was found out that the average depth of charred area was 14.9 mm, up to maximum 19.4 mm and minimum 9.3 mm (Table 1). Total carbonisation of the sample occurred on the area of 416 × 150 mm (in the middle of the sample). Significant change in colour of wood (black) could be to a distance of 413 mm from the centre of the sample on both sides, of the wooden block on the loaded side. During the thermal loading, a visible protective layer similar to bitumen emerged on the wood surface as a result of the effect of a retardant. This protective layer reduced the effect of radiant heating in the sample and at the same time the spread of charred area

109

Fig. 4 Sample treated with flame retardant HR prof after thermal loading.

The sample with the retardant Striebrenka After painting, a light grey matt coat was formed. The average depth of the charred area was 15.18 mm, up to maximum 20.2 mm and minimum 8.2 mm that can be compared to the effect of the retardant HR prof (Table 1). Only approximate measurement of the area of charred layer was carried out because it was symmetrical with the dimensions of 458.3 mm ×150 mm (in the middle of the sample). In the place with the most intensive loading, the protective layer of Striebrenka broke off and subsequently, wood degradation and carbonisation occurred exactly in this place. Striebrenka degradation emerged on the edges of wooden block and spread towards the centre. When Striebrenka broke off completely, burning the wood without flames occurred and the material began to smoulder.

Fig. 5 Sample treated with Striebrenka after thermal loading.

Control sample with no treatment The sample was treated with no retardant, thus the effect of fire and the spread of flame was natural and slowing down with nothing. The average depth of charred layer was 15.4 mm. It is more in comparison to the samples treated with a retardant, however, the differences in the nine measured points were not significant so much. Maximum depth was 17.9 mm and minimum one was 10.1 mm (Table 1). It can be stated that burning the control sample was steadier comparing to treated samples. The area of charred layer was 569.6 × 150 mm (in the middle of the sample). In the case of this sample, the surface of material was getting black, therefore it can be assumed that during further heat loading this area would spread. The change in colour (black) occurred on the area with the length of 750 mm, i.e. it is less than in the case of the retardant HR prof used. Surface carbonisation started earlier than in the case of treated samples. An increase in the temperature in shorter time was seen. In the two thirds of loading time, the beginning of burning the already charred surface was observed. The flame could not be seen; it was not spread to other parts of the surface or deeper to the sample. It was only in the place exposed to heat. The flame could be seen for 204 seconds and then it went out. 110

Fig. 6 Control sample, with no treatment, after thermal loading. Tab. 1 Thickness of the charred layer of samples loaded for 30 minutes. Treatment

Type of treatment

Control sample

HR prof Striebrenka

Treated sample

Thickness of charred layer (mm) min max average 10.10 18.40 15.41 ± 0.25 9.30 19.40 14.90 ± 0.36 8.20 20.20 15.18 ± 0.79

According to the findings of HADVIG (1981) and BABRAUSKAS (2005), the speed of carbonisation is the topic in the centre of attention of structural engineers who want to determine the load-bearing capacity of timber beams or columns during post-flashover. During their research on the samples of Nordic spruce, the oven was run at the temperature ranging between 920–1 070 °C. The depth of the charred layer resulting from the experiment was 7–9 mm (when the time of loading was 300 s), 20–21 mm (when the time of loading was 1200 s) and 30–32 mm (when the time of loading was 2400 s). It can be compared to the samples in our experiment. According to DÚBRAVSKÁ et al. 2019, charred layer in the test samples exposed to thermal loading corresponded to the parametric fire curve. In the case of 120 thick test sample, 20 mm charred layer appeared in the 22nd minute. In the case of 240 mm thick test sample, 20 mm charred layer appeared in 29th minute. According to SU et al. 2019, charred layer of the wood stud specimens estimated based on their experiment was 15 mm when the time of loading was 30 min. It can be compared to the samples in our experiment (on control sample average thickness of charred layer was 15.41 ± 0.25 mm). Chemical analysis of wood during the thermal treatment Chemical composition of original wood, wood treated with the retardant HR prof, Striebrenka and the control sample after thermal loading are mentioned in Table 1. The analysis of original wood was carried out due to the opportunity for comparing the changes in chemical composition before and after the thermal treatment. The values of basic chemical composition in the case of original sample show the extent of wood degradation treated and untreated with retardants and at the same time, the effect of the retardants used for the protection. Following the table, it can be seen that in the case of the control sample there was a decrease in all wood components. In the layer 1, an increase in average number of extractives from the value 2.12 to 3.55 occurred. An increase of these components after thermal degradation of wood was mentioned by other authors as well (KUČEROVÁ et al. 2011, ČABALOVÁ et al. 2013). Their amount is higher because of degradation products of thermal decomposition of lignin and carbohydrates in wood (KUČEROVÁ et al. 2011). During thermal degradation of wood, degradation of thermally least stable wood components, especially hemicelluloses was 111

monitored. It was confirmed in our experiment as well. As a result of radiant heating, these components decreased by 20.25 % (layer 1) and by 13.57 % (layer 2). A decrease in hemicelluloses after thermal degradation of spruce wood with radiant heating was mentioned in our further work (ČABALOVÁ et al. 2014). Together with a decrease in these wood components, there was a decrease in thermally more stable wood components, especially cellulose and lignin. The average values of cellulose decreased by 13.88 % (layer 1) and by 9.63 % (layer 2) and in the case of lignin, there was a decrease by 12.26 % (layer 1) and by 8.34 % (layer 2). At the temperatures over 250 °C, intensive decomposition of cellulose can be seen and at a temperature over 300 °C, the decomposition of lignin occurs (ČABALOVÁ et al. 2013). Following the graph in Fig. 3, it can be seen that in the case of material loaded, the achieved temperatures were higher than 500 °C in the thermocouples M0.0 a M1.0. Following the chemical analysis of the sample treated with the retardant HR prof after thermal loading, it can be stated that in comparison to original wood sample there was less decrease in the main chemical components of wood than in the control sample. Especially in the case of cellulose, the mentioned decrease in both investigated layers was slight. During thermal loading, cellulose stability comparing to hemicelluloses is evident. It is in compliance with other published works (WINDEISEN, WEGENER 2009, POLETO et al. 2012). Due to thermal treatment, there was a decrease in average amount of saccharides (holocelullose) by 3.06 % (layer 1) and only by 2.39 % (layer 2). In the case of main wood components, the most stable component – lignin suffered from the most significant degradation. There was a decrease in lignin by 8.10 % (layer 1) and by 7.66 % (layer 2) comparing to original wood. The thermocouple M0.0, placed on the loaded side, recorded the maximum temperature of 181.9 °C and the temperature recorded by the thermocouple M1.0 was up to 479 °C. The effect of the retardant is evident as the highest temperature achieved in the case of the control sample was in the thermocouple M0.0 – 506.6 °C and in the thermocouple M1.0 – 538.6°C. The degradation of sample treated with the retardant HR prof occurred to a lesser extent in comparison to the control untreated sample. Lignin degradation could be compared but the saccharide stability was evident. Tab. 2 Chemical analysis of untreated wood and wood after thermal degradation. Sample and treatment Original wood sample*

Layer

Extractives (%)

Lignin (%)

Holocellulose (%)

Cellulose (%)

Hemicelluloses (%)

2.12 ± 0.02

29.12 ± 0.27

78.32 ± 0.73

41.63 ± 0.06

36.69 ± 0.79

1 3.55 ± 0.01 25.55 ± 0.06 65.11 ± 1.50 35.85 ± 0.45 2 3.24 ± 0.07 26.69 ± 0.04 69.33 ± 0.02 37.62 ± 0.01 1 2.04 ± 0.02 26.76 ± 0.01 75.92 ± 0.39 40.67 ± 0.57 HR prof 2 1.88 ± 0.05 26.89 ± 0.06 76.45 ± 1.64 41.45 ± 0.44 1 2.23 ± 0.09 28.89 ± 0.16 71.82 ± 2.02 37.21 ± 0.10 Striebrenka 2 2.57 ± 0.03 29.24 ± 0.01 73.75 ± 1.12 37.78 ± 0.92 *original sample from the wood trunk, without any surface and heat treatment

29.26 ± 1.06 31.71 ± 0.02 35.25 ± 0.96 35.00 ± 2.08 34.61 ± 1.92 35.97 ± 0.20

Control sample

Chemical composition of thermally loaded wood sample treated with physical retardant called Striebrenka is given in Table 1 as well. Following the results of chemical analysis, it can be seen that in comparison to original wood, the amount of lignin stayed the same, however, the degradation of saccharides (decrease of 8.5% in layer 1 and 8.9% in layer 2) comparing to the sample treated with the retardant HR prof. occurred. Nevertheless, the effect of the retardant is evident when the results were compared to the control sample. The amount of the chemical components of wood (Extractives, Lignin and Holocellulose) ranged between 94–108% what is in compliance with the statement of 112

FENGEL and WEGENER (2003), They mentioned that the amount of the components of the analysed wood can exceed 100%. At the same time, during thermal degradation of wood, an increase in the amount of cellulose determined by SEIFERT (1956) and lignin determined by SLUITER et al. (2012) as a result of condensation reactions of lignin and saccharides was observed (KAČÍK et al. 2006).

CONCLUSION In this study, the effect of thermal loading by radiant heat source on spruce wood samples treated with a HR prof retardant and Striebrenka was investigated. While the retardant HR prof is a common retardant based on chemical principles, the Striebrenka retardant is not used to protect wood and its nature is physical. During and after the thermal loading, the degree of its charring (extent and depth) was evaluated. It was found out that once a continuous carbon layer was formed, the degradation of the wood slowed down. This is due to the well-known fact that the carbonized layer itself acts as a flame retardant for specific time. As the temperature on the surface of the sample rose, the coating foamed and increased its volume, creating a kind of protective barrier that slowed down the degradation of the wood. The results related to the thickness of charring layer by reaching a critical temperature of 300 °C, specified also in Eurocode 5, which was determined by thermocouples placed in the cross section of the sample and further manually measured (at nine places on the sample surface) were compared. A significant deviation between those values was found. HR prof and Striebrenka has a demonstrable wood protective function due to its prohibitive capability to reach a critical temperature of 300 ° C, in the depth of 10 mm below the surface of the sample. However, when the thickness of the average charred layer is manually measured, it reached a value of 15.41 mm (control sample), 15.18 mm (Striebrenka) and 14.90 mm (HR prof). During the thermal loading of wood, the changes in its properties and the degradation of main wood components were monitored. Following the changes of chemical components as a result of thermal loading:  the amount of extractives in the control sample increased from the value 2.12 to 3.55 by the products resulting from the lignin and saccharide degradation;  a decrease in lignin (from 8.34 to 12,26%), cellulose (from 9.63 to 13.88%) and hemicelluloses (from 13.57 to 20.25%) in the control sample was observed;  a decrease in lignin (from 7.66 to 8.10%) in the case of the sample treated with HR prof occurred but the percentage of saccharides in wood stayed the same comparing to the original wood sample;  the proportion of lignin treated with the retardant Striebrenka did not change in comparison to untreated wood;  a decrease in the amount of cellulose (from 9.25 to 10.62%) of sample treated with Striebrenka and the control sample (in the layer 2) is similar. HR prof retardant was more effective in terms of saccharides and the Striebrenka retardant in terms of lignin protection. REFERENCES ASTM D1107-96. 2007. Standard test method for ethanol-toluene solubility of wood, ASTM International, West Conshohocken, PA. BABRAUSKAS, V. 2005. Charring rate of wood as a tool for fire investigations. In Fire Safety Journal, 40(6): 528554. DOI: 10.1016/j.firesaf.2005.05.006

113

BREBU, M., VASILE, C. 2010. Thermal degradation of lignin – A review. In Cellulose Chemistry and Technology, 44(9): 353363. ČABALOVÁ I., KAČÍK, F., KAČÍKOVÁ, D., ORAVEC, M. 2014. The influence of cross-section of spruce wood specimens on saccharides changes at thermal loading (in Slovak). In Acta Facultis Xylologiae Zvolen, 56(2): 8186. ČABALOVÁ, I., KAČÍK, F., KAČÍKOVÁ, D., ORAVEC, M. 2013. The influence of radiant heating on chemical changes of spruce wood (in Slovak). In Acta Facultatis Xylologiae Zvolen, 55(2): 59−66. DÚBRAVSKÁ, K., ŠPILÁK, D., TEREŇOVÁ, Ľ., ŠTEFKOVÁ, J. 2019. Charring layer on a crosslaminated timber panel construction. In Acta Facultatis Xylologiae Zvolen, 61(2): 109−119, DOI: 10.17423/afx.2019.61.2.11 FENGEL, D., WEGENER, G. 2003. Wood: Chemistry, Ultrastructure, Reactions, Verlag Kessel, Remagen, Germany. FINDORÁK, R., FRÖHLICHOVÁ, M., LEGEMZA, J., FINDORÁKOVA, L. 2016. Thermal degradation and kinetic study of sawdusts and walnut shells via thermal analysis. In Journal of Thermal Analysis and Calorimetry, 125:689–694. DOI:10.1007/s10973-016-5264-6 FONSECA, E.M.M.A., BARREIRA, L.M.S. 2009. Charring rate determination of wood pine profiles submitted to high temperatures. In Safety and Security Engineering III. Polytechnic Institute of Bragança, Portugal. WIT Transactions on the Built Environment, 108, 2009: 449–457. GAFF, M., KAČÍK, F., GAŠPARÍK, M., TODARO, L., JONES, D., CORLETO, R., MAKOVICKÁ OSVALDOVÁ, L., ČEKOVSKÁ, H. 2019a. The effect of synthetic and natural fire-retardants on burning and chemical characteristics of thermally modified teak (Tectona grandis L. f.) wood. In Construction and building materials, 200: 551558. DOI: 10.1016/j.conbuildmat.2018.12.106 GAFF, M., KAČÍK, F., SANDBERG, D., BABIAK, M., TURČÁNI, M., NIEMZ, P. HANZLÍK, P. 2019b. The effect of chemical changes during thermal modification of European oak and Norway spruce on elasticity properties. In Composite structures, 220: 529538. DOI: 10.1016/j.compstruct.2019.04.034 GAŠPARÍK, M., MAKOVICKÁ OSVALDOVÁ, L., ČEKOVSKÁ, H., POTŮČEK, D. 2017. Flammability characteristics of thermally modified oak wood treated with a fire retardant, In Bioresources. 12(4): 84518467. HADVIG, S. 1981. Charring of Wood in Building Fires. Lyngby : Technical University of Denmark. KAČÍK, F., LUPTÁKOVÁ, J., ŠMÍRA, P., EŠTOKOVÁ, A., KAČÍKOVÁ, D., NASSWETTROVÁ, A., BUBENÍKOVÁ, T. 2017. Thermal analysis of heat-treated silver fir wood and larval frass. In Journal of Thermal Analysis and Calorimetry, 130(2): 755762. DOI: 10.1007/s10973-017-6463-5 KAČÍK, F., KAČÍKOVÁ, D., BUBENÍKOVÁ, T. 2006. Spruce wood lignin alterations after infrared heating at different wood moistures. In Cellulose Chemistry and Technology, 40(8): 643648. KAČÍKOVÁ D., KAČÍK F., BUBENÍKOVÁ T., KOŠÍKOVÁ B. 2008. Influence of fire on spruce wood lignin changes. In Wood Research, 53(4): 95−103. KAČÍKOVÁ, D., KAČÍK, F., ČABALOVÁ, I., ĎURKOVIČ, J. 2013. Effects of thermal treatment on chemical, mechanical and colour traits in Norway spruce wood. In Bioresource Technology, 144: 669674. DOI: 10.1016/j.biortech.2013.06.110 KUČEROVÁ, V., KAČÍKOVÁ, D., KAČÍK, F. 2011. Alterations of extractives and cellulose macromolecular characteristics after thermal degradation of spruce wood. In Acta Facultatis Xylologiae Zvolen, 53(2): 7783. KUČEROVÁ, V., LAGAŇA, R., VÝBOHOVÁ, E., HÝROŠOVÁ, T. 2016. The effect of chemical changes during heat treatment on the color and mechanical properties of fir wood, In BioResources 11(4): 90799094. DOI: 10.15376/biores.11.4.9079-9094 LUPTÁKOVÁ, J., KAČÍK, F., EŠTOKOVÁ, A., KAČÍKOVÁ, D., ŠMÍRA, P., NASSWETTROVÁ, A., BUBENÍKOVÁ, T. 2018. Comparison of activation energy of thermal degradation of heat sterilised silver fir wood to larval frass regarding fire safety. In Acta Facultatis Xylologiae Zvolen, 60(1): 1929. DOI: 10.17423/afx.2018.60.1.03 NUOPONNEN, M., VUORINEN, T., JÄMSÄ, S., VIITANIEMI, P. 2003. The effect of a heat treatment on the behaviour of extractives in softwood studied by FTIR spectroscopic methods. In Wood Science and Technology, 37: 109115.

114

NUOPPONEN, M., VUORINEN, T., JAMSÄ, S., VIITANIEMI, P. 2005. Thermal modifications in softwood studied by FT-IR and UV resonance Raman spectroscopies. In Journal of Wood Chemistry and Technology, 24: 1326. DOI: 10.1081/WCT-120035941. OČKAJOVÁ, A., KUČERKA, M., KMINIAK, R., KRIŠŤÁK, Ľ., IGAZ, R., RÉH, R. 2020. Occupational Exposure To Dust Produced When Milling Thermally Modified Wood. In International Journal of Environmental Research and Public Health, 17: 1478. DOI: 10.3390/ijerph17051478 POLETTO, M., ZATTERA, A. J., SANTANA, R.M.C. 2012. Thermal decomposition of wood: Kinetics and degradation mechanisms. In Bioresource Technology, 126: 712. DOI: 10.1016/j.biortech.2012.08.133 RANTUCH, P., KAČÍKOVÁ, D., MARTINKA, J., BALOG, K. 2015. The influence of heat flux density on the thermal decomposition of OSB (in Slovak). In Acta Facultatis Xylologiae Zvolen, 57(2): 125−134. DOI: 10.17423/afx.2015.57.2.13 125. SEIFERT, V.K. 1956. Űber ein neues Verfahren zur Schnellbestimmung der Rein-Cellulose (About a new method for rapid determination of pure cellulose). In Das Papier 10(13/14): 301306. SLUITER, A., HAMES, B., RUIZ, R., SCARLATA, C., SLUITER, J., TEMPLETON, D., CROCKER, D. 2012. Determination of Structural Carbohydrates and Lignin in Biomass (NREL/TP-510-42618). In National Renewable Energy Laboratory, Golden, CO. STN EN 1995-1-2 (Eurokód 5). 2004. Navrhovanie drevených konštrukcií (všeobecné pravidlá a navrhovanie konštrukcií na účinky požiaru). SU, H. C., TUNG, S. F., TZENG, C. T., LAI, C. M. 2019. Variation in the charring depth of wood studs inside wood-frame walls with time in a fire. In Wood Research, 64(3): 449-460. WHITE, R.H., NORDHEIM, E.V. 1992. Charring rate of wood for ASTM E 119 exposure. In Fire Technology, 28: 5–30. WINDEISEN, E., WEGENER, G. 2009. Chemical characterization and comparison of thermally treated beech and ash wood. In Material Science Forum, 599: 143158. DOI: 10.4028/www.scientific.net/MSF.599.143 WISE, L.E., MURPHY, M., D'ADDIECO, A.A. 1946. Chlorite holocellulose, its fractionation and bearing on summative wood analysis and on studies on the hemicelluloses. In Paper Trade Journal, 122(2): 3544. ZAIKOV, G.E., LOMAKIN, S.M. 2002. Ecological issue of polymer flame retardancy. In Journal of Applied Polymer Science, 86: 2449–2462. DOI: 10.1002/app.10946.| ZELENÁ SPRÁVA 2019. Available on: https://www.mpsr.sk/zelena-sprava-2019/123---14927/ ACKNOWLEDGEMENT This work was supported by the Slovak Scientific Grand Agency (Contract No. VEGA 1/0397/20 (50%) and No. VEGA 1/0387/18 (20%) and the Slovak Research and Development Agency under the contract No. APVV-16-0326 (30%).

ADDRESSES OF AUTHORS Doc. Ing. Iveta Čabalová, PhD. Ing. Michal Bélik Technical University in Zvolen Department of Chemistry and Chemical Technologies T. G. Masaryka 24 960 01 Zvolen Slovakia cabalova@tuzvo.sk xbelik@is.tuzvo.sk

115

Doc. Ing. Martin Zachar, PhD. Technical University in Zvolen Department of Fire Protection and Safety T. G. Masaryka 24 960 01 Zvolen Slovakia zachar@tuzvo.sk Mgr. Žaneta Balážová, PhD. Technical University in Zvolen The Institute of Foreign Languages T. G. Masaryka 24 960 01 Zvolen Slovakia balazova@tuzvo.sk

116

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 117−130, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.10

ASSESSMENT OF SELECTED TYPES OF THE STRUCTURAL ENGINEERED WOOD PRODUCTION FROM THE ENVIRONMENTAL POINT OF VIEW Rozália Vaňová – Jozef Štefko ABSTRACT According to recent findings, the construction industry contributes to almost 40% of global CO2 emissions. Life Cycle Assessment is one of the methods assessing the anthropogenic impact on the environment. This paper compares the production of selected types of structural wood-glued laminated timber (glulam), plywood and three-layer solid wood panel (SWP) in terms of their environmental impact. The results of this analysis are important especially for designers and architects, who can reduce the environmental footprint by choosing materials already in the design phase and thus be part of the building eco-design creation. This direction is becoming increasingly popular in the world and, in addition, it increases market competitiveness. The selected structural wood products were compared using the IMPACT 2002+ method and SimaPro 8.5.2 software was used for the assessment. Characterization and single score results were assessed. Glulam for indoor use achieved the lowest total environmental impact. On the other side, plywood for outdoor use was found the worst one of the assessed products. The two most affected impact category for all evaluated wood products was the impact on Human health and Ecosystem quality, respectively. Generally, plywood products present a considerable environmental burden, therefore it could be replaced by SWP. The study also showed the importance of production technology in the environmental context. Key words: environmental impact, structural wood, eco-design, life cycle assessment, glulam, plywood, three-layer solid wood panel.

INTRODUCTION Globally, the construction industry is showing an increase in emissions and energy consumption. In 2018, it was responsible for 36% of final energy consumption and 39% of CO2 emissions, of which 11% came from the production of building materials and 28% of emissions were produced in connection with operational energy consumption (IEA 2019). Modern energy-efficient buildings limit the growth of energy consumption in both residential and non-residential buildings. Nevertheless, further actions to reduce energy performance still need to be carried out (Directive 2010/31/EU). Therefore, greenhouse gas emissions reduction steps should be one of the top environmental policy priorities for the construction sector. Awareness of the need for greater use of renewable materials, that do not only reduce resource depletion but also address a range of other environmental issues is gradually 117

increasing (SOTAYO et al. 2019). In the field of sustainable construction, structural engineered wood and wood-based building systems represent an irreplaceable position as they bond carbon and thus mitigate the climate change (PAROBEK et al. 2019, PALUŠ et al. 2020). Adoption of new regulations and the excellent physical, environmental and economic properties brought attractiveness of structural engineered wood products amongst architects, as well as slow replacement of mineral-based building materials use with sustainable ones (HILDEBRANDT et al. 2017). At present, massive wood constructions made of crosslaminated timber (CLT), glued laminated timber (glulam) and laminated veneer lumber (LVL) come to the fore as a substitute for concrete and steel (CRAWFORD, CADOREL 2017). Architects and designers often do not know the relationship between the choice of construction material and its impact on the environment. Assessment of alike impacts can bring new architectural practices and introduce environmental awareness in this area. LCA – Life Cycle Assessment LCA is a voluntary environmental management tool that systematically assesses the environmental aspects of products or services at all stages of their life cycle. It has a precisely defined structure (Fig. 1) and is standardized within the series of ISO 14040 standards (ISO 14040; ISO 14044).

Interpretation

Goal and scope definition

Inventory analysis

Impact assessment Fig. 1 LCA structure (ISO 14040).

The product life cycle is generally divided into stages (Table 1). It represents the cycle of materials and energy throughout individual life stages - from the acquisition of raw materials, production of construction materials, through the construction itself, the use of the building to its disposal and eventual recycling. The evaluation includes all material and energy inputs to the examined system and the corresponding output flows. Tab. 1 Life cycle stages of buildings (EN 15804).

Maintenance

Repair

Replacement

Refurbishment

Operational energy use

Operational water use

Deconstruction and demolition

Transport

Waste processing for reuse,

recovery or recycling Waste disposal

118

Recycling potential

Use

Recovery

Construction process

Reuse

Transport to site and on site

End of life stage

Manufacturing

Use stage

Benefits and loads beyond the system boundaries

Transport to manufacturer

Construction stage

Raw material supply

Product stage

The results of LCA are used generally to support product decision-making activities, such as the identification of hotspots in product systems, product development, product comparison, green procurement and market requirements (HAHNEL et al. 2021; LIIKANEN et al. 2019). Moreover, LCA serves as a source of information for the development of ecodesign by allowing comparison of different variants of considered product (PAJCHROWSKI et al. 2014). The results of the analysis can also be used to select appropriate product processing technologies and to introduce these technologies into the perspective of the product-related chain. LCA is increasingly used at the strategic level for business development, policy and education (UNEP/SETAC 2005). Structural wood Wood becomes one of the key building materials in the field of sustainable construction. New progressive wood-based materials are still being added to the market. Structural wood is a type of excellent quality wood (KRETSCHMANN 2013, DINWOODIE 2000). It meets the set technical requirements resulting from the valid technical standards (STN EN 1995, PORTEOUS KERMANI 2013). According to its technological composition, structural wood is divided into solid wood, modified wood and wood composites, that involves laminated structural wood, veneer-based structural wood, agglomerated structural wood and combined structural wood (Fig. 2). Composite represents wood-based product made by gluing wood material with a non-wood material. The individual types of structural wood also differ from each other by the production technology, which can be decisive in assessing the environmental impacts of construction materials. Energy consumption in the stage of material manufacturing, the need for thermal energy in drying operations and the amount and type of adhesive used are some of numerous factors in the environmental footprint assessment of structural wood. Structural wood Solid wood Softwood

Modified wood

Hardwood

Wood composites Laminated structural wood KVH Duo/Triobalken Glulam

Veneer-based structural wood

Agglomerated structural wood

Plywood

Wood-fiber boards

Laminated veneer lumber

Wood-particle boards Orinted strand boards

CLT

Sawdust and sawdust-chip boards

Combined structural wood Wood-cement boards Woodgypsum boards Cork boards Bark boards

Mulch boards

Fig. 2 Structural wood typology (STN EN 14080).

It is well known that wood has much lower negative impact on the environment compared to conventional construction materials. SAADE et al. (2020) compared wood frame buildings with their concrete variant and concluded that Global Warming Potential (GWP) for concrete building was higher than for the wood-based building in every aspect. At the 119

same time, they claimed that the result of the LCA study is deeply dependent on the decisions made and scenarios created during the modeling of the life cycle of the investigated object. Nowadays, CLT panels, LVL, glulam and fibreboards are becoming increasingly popular construction materials, recording annual demand growth rates of 2.5% to 15% (HILDEBRANDT et al. 2017). Moreover, KVH (Konstruktionvollholz), a finger-jointed solid timber; and Duo/Triobalken, a solid timber made of two to five planks glued together parallel to the fibres, can also be used for load-bearing wood structures. However, only data for structural engineered wood available in the ecoinvent database were included in this paper. The study by POMMIER et al. (2016) evaluated 3 types of wood - hardwood, softwood and marine pine (Pinus pinaster) - using the ReCiPe Midpoint method (H). The latter had the lowest environmental impact values. The research also included the area of plywood production technology. The traditional plywood production process requires very thorough drying of the wood before gluing. Heat consumption in this process could be reduced by applying a new technology of gluing green wood and vacuum forming plywood (POMMIER et al. 2016, ENQUIST et al. 2014). Glulam is one of the most popular construction materials globally. Its favorable environmental impact has also been confirmed by several studies. HASSAN AND JOHANSSON (2018) proved that glulam beams produce less CO2 emissions than their steel variant. A similar conclusion was reached by SATHRE AND GUSTAVSSON (2009) when comparing wooden frames in construction with reinforced concrete materials. BRANDNER et al. (2016) found that central production of prefabricated products reduces costs compared to conventional construction techniques. Generally, when assessing the environmental impacts of buildings, the manufacturing technology of construction materials is poorly described, leading to wider range of uncertainty of the base data. More and more studies show that the production of materials can have even greater impact on the environmental performance of buildings than the operation of them (MITTERPACH et al. 2018, CRAWFORD et al. 2017, DODOO et al. 2012, PETROVIC et al. 2019; HAFNER, SCHÄFER 2017). Therefore, it is important to know and select alternative construction methods that can help to reduce the environmental effects of production associated with the production of the material (CRAWFORD, CADOREL 2017). Moreover, one of the most important features of glued wood-based materials is the type of glue used and its amount. Different types of adhesives have different environmental impacts (POMMIER et al. 2016). From the environmental point of view, particularly in terms of human health and toxicity, polyurethane-based adhesives prove to be the most suitable, as they do not emit VOCs (ENQUIST et al. 2014, POMMIER, ELBEZ 2006). The aim of this study is to compare the production of five types of structural engineered wood used in wood-based constructions - glulam for outdoor and indoor use, three-layer solid wood panel and plywood for indoor and outdoor use – from the environmental point of view.

MATERIALS AND METHODS The analysis was carried out using SimaPro 8.5.2 software and IMPACT 2002+ evaluation method was chosen (JOLLIET et al. 2003). This method provides characterization results as 15 midpoint impact categories whereas single score results are conjugated to 4 endpoint categories (Fig. 3). Characterization factors convert Life Cycle Inventory (LCI) result to the common unit of category indicator which creates impact categories as an outcome of physical, chemical and biological processes with the assessed system.

120

Fig. 3 Overall scheme of the IMPACT 2002+ framework (JOLLIET et al. 2003).

Single score is a weighting step within impact assessment which give weight to the different environmental impacts. Single scores are given in Pt units, representing one thousandth of the yearly environmental load of one average European inhabitant. Background data of analyzed products were found in the ecoinvent v3.5 databases and are used by researchers to analyze and facilitate calculations of the environmental impacts of products and services (WERNET et al. 2016). The selected functional unit was 1 m3 of the product. System boundaries from raw material acquisition to manufacture of the structural engineered wood were investigated („from cradle to gate“). Inventory analysis Glued laminated timber - Glulam Glulam is a type of structural engineered wood produced by gluing lamellas to a length parallel to the direction of the fibers. It can be used as a beam, column or as a roof structure. A vast advantage is its high strength and ability to create an arch (STARK et al. 2002). Tab. 2 Selected structural engineered wood. Analysed wood product Glulam for indoor use Glulam for outdoor use Plywood for indoor use Plywood for outdoor use Three-layer solid wood panel

Input raw material

Type of adhesive

Amount of adhesive (kg.m-3)

Softwood board, unplaned, dried - spruce

11.36

64.76

PVAC

7.32

Hardwood veneer - beech Softwood board, unplaned, dried - spruce

In our study, glulam is made from unplaned, dried (u=20%) softwood board. Glulam for indoor and outdoor use differs from one another in the type of adhesive used – ureformaldehyde (UF) and melamine-formaldehyde (MF), respectively. All types of selected structural wood are presented in Table 2. 121

Plywood Plywood is one of the oldest wood-based composite materials. It consists of an odd number of layers of veneers laid perpendicular to the direction of the fibers. In construction, plywood is used in technically demanding formwork. Other applications include scaffolding work platforms, tiling, roofing elements, wood-based structures, facades and floors (STARK et al. 2002). As with the previous product, the plywood is divided into indoor and outdoor use and the same types of adhesives are used in the production – UF and MF, respectively. However, PF is often used in construction plywood. The input material for the production of plywood is hardwood veneer. Manufacturing of selected wood-based construction materials is similar but there is a difference in some processes. Wood processing begins with debarking of logs, which are then cut into either veneers or other sawmill products (in the production scheme glulam referred to as laths, boards and beams). In the case of glulam production (Fig. 4), the lumber is first joined to length and then the individual lamellas are glued to each other parallel to the direction of the fibers. In the production of plywood (Fig. 5), the veneers are laid and glued to each other perpendicular to the direction of the fibers. This is followed by cutting and surface treatment (BLASS et al. 1995). Three-layer solid wood panel – SWP The SWP is a special form of plywood with relatively thick layers - lamellas. The production of SWP is like the production of glulam. The lamellas are dried, planed, joined and glued into long slats and glued together into a block, which is cut into boards and eventually planed. The thermal energy for drying the slats is produced mainly by industrial residual wood (WERNER et al. 2007).

Emissions to air, water and soil

Waste heat

Logging

Waste wood

Pressing

Cutting

Sawnwood gluing

Storing

Log transport

Planing Storing

Debarking

Sawing/Cutti ng

Boards

Sorting

Boards gluing

Laths

Drying

Pressing

Beams Glulam

Electricity

Heat

Adhesi ve

Water

Fig. 4 Simplified glulam production scheme.

122

Surface treatment

Other auxiliary materials

Emissions to air, water and soil

Waste heat

Waste wood

Pressing

Formatting

Prepressing

Planing

Gluing

Storing

Logging

Log transport

Debarking Storing Hydrothermal treatment

Sorting

Cutting

Inputs

Drying

Veneers

Electricity

Legend

Plywood

Heat

Outputs

Adhesive

Processes

Water

Other auxiliary materials

Products

Fig. 5 Simplified plywood production scheme.

RESULTS Prior to the analysis of selected types of structural wood, a preliminary analysis of the environmental impacts of the production and processing of wood types used (Table 2) and their variants was performed; planed softwood board and softwood veneer compared to unplaned softwood board and hardwood veneer, respectively. Due to the analysis (Table 3, Fig. 6) planed board has significantly higher environmental impact than its unplaned variant. The most affected was the Ecosystem quality impact category, reaching its top in aquatic ecotoxicity values for all the analyzed wood types. Human health was hit the worst in Ionizing radiation impact category. Higher environmental impact of hardwood veneer against softwood veneer are bound to different chemical nature and physical characteristics of hardwood regarding more intense energy consumption in the manufacturing stage.

123

Tab. 3 Midpoint characterization of selected sawnwood. Impact category

Category indicator

Carcinogens Non-carcinogens Respiratory inorganics Ionizing radiation Ozone layer depletion Respiratory organics Aquatic ecotoxicity Terrestrial ecotoxicity Terrestrial acid/nutri Land occupation Aquatic acidification Aquatic eutrophication Global warming Non-renewable energy Mineral extraction

kg C2H3Cl eq kg C2H3Cl eq kg PM2.5 eq kBq C-14 eq mg CFC-11 eq g C2H4 eq t TEG water t TEG soil kg SO2 eq m2 org. arable kg SO2 eq g PO4 P-lim kg CO2 eq GJ primary MJ surplus

Softwood board, unplaned, dried (u=20%) 0.90 1.03 0.11 0.64 9.28 66.06 4.23 2.04 2.10 313.80 0.40 10.88 68.06 1.03 1.99

Softwood board, planed, dried (u=20%) 1.43 2.66 0.23 1.06 11.11 79.20 13.71 5.40 3.27 336.75 0.64 21.11 101.45 1.48 3.19

Hardwood veneer

Softwood veneer

0.39 0.37 0.03 0.19 3.99 111.30 1.63 1.04 0.65 194.76 0.12 3.16 23.51 0.37 0.46

0.27 0.27 0.02 0.15 3.23 70.32 1.28 0.80 0.53 155.75 0.09 2.36 18.83 0.29 0.34

Environmental impact (mPt)

80 70

9.76

10.25

6.79

6.87

30 20 10 0

30.21 2.42 2.37

26.32 24.05

16.15

11.62 Softwood board, unplaned, dried (u=20%)

Softwood board, planed, dried (u=20%)

3.10 Hardwood veneer

1.93 1.90 12.90 2.45 Softwood veneer

Sawnwood Human health

Ecosystem quality

Climate change

Resources

Fig. 6 Single score of selected sawnwood.

Following, a preliminary analysis of used adhesives was performed. It is clear from Table 4 and Fig. 7 that the most appropriate type of adhesive is polyvinyl acetate (PVAC). In this respect, the SWP should achieve the best environmental ratings. On the contrary, MF resin had the worst effect on the environment, and thus products for outdoor use should have worse environmental profile than products intended for indoor use. Ecosystem quality impact category was the least affected in all cases.

124

Tab. 4 Midpoint characterization of selected adhesives. Impact category Carcinogens Non-carcinogens Respiratory inorganics Ionizing radiation Ozone layer depletion Respiratory organics Aquatic ecotoxicity Terrestrial ecotoxicity Terrestrial acid/nutri Land occupation Aquatic acidification Aquatic eutrophication Global warming Non-renewable energy Mineral extraction

Category indicator g C2H3Cl eq g C2H3Cl eq g PM2.5 eq Bq C-14 eq mg CFC-11 eq g C2H4 eq kg TEG water kg TEG soil g SO2 eq m2 org.arable g SO2 eq g PO4 P-lim kg CO2 eq MJ primary MJ surplus

MF 163.30 82.70 5.66 26.77 0.65 1.84 298.08 77.95 131.20 0.04 27.73 0.95 4.54 89.81 0.31

UF 120.86 53.93 3.11 14.63 0.42 1.44 191.85 48.32 64.87 0.03 14.65 0.59 2.64 59.87 0.20

PVAC 93.84 36.14 2.55 25.27 0.31 2.37 198.84 31.11 31.00 0.02 9.71 0.70 2.11 66.13 0.11

2.0

Environmental impact (mPt)

1.8 1.6 1.4

0.59

1.2 1.0 0.8 0.6 0.4 0.2 0.0

0.46

0.40

0.06

0.44

0.27 0.04

0.66

0.21 0.02

0.38

0.30

UF Adhesive

Human health

Ecosystem quality

Climate change

PVAC

Resources

Fig. 7 Single score of selected adhesives.

Applying the previous results to compare selected types of structural engineered wood (Table 5, Fig. 8), it could be stated that:  Plywood had the highest environmental impact, mainly due to the high consumption of glue. Another reason is the use of hardwood veneer.  Products for outdoor use generally had higher values, due to the use of MF glue.  Glulam for indoor use had the lowest environmental impacts at up to three endpoints – Human health, Climate change and Resources.  Human health impact category was the most affected in all cases emerging from Ionizing radiation.  Despite using PVAC as the best adhesive option the SWP board eventually reached the third place in the total environmental impact chart (Fig. 8). 125

Tab. 5 Midpoint characterization of selected structural engineered wood. Impact category

Category indicator

kg C2H3Cl eq kg C2H3Cl eq kg PM2.5 eq kBq C-14 eq mg CFC-11 eq kg C2H4 eq t TEG water t TEG soil kg SO2 eq m2 org.arable kg SO2 eq g PO4 P-lim kg CO2 eq GJ primary MJ surplus

Glulam, indoor use 4.16 10.70 0.66 3.99 29.40 0.20 69.80 26.20 8.16 452.04 1.58 70.61 235.89 4.02 7.50

Glulam, outdoor use 4.63 11.00 0.69 4.13 31.90 0.21 70.60 26.40 8.88 445.91 1.72 74.50 256.03 4.33 8.83

Plywood, indoor use 11.40 25.00 1.20 8.74 58.20 0.57 171.00 62.60 14.10 480.30 2.71 159.73 391.81 8.10 18.03

Plywood, outdoor use 14.20 26.90 1.37 9.53 72.70 0.59 178.00 64.50 18.40 481.08 3.56 182.64 515.06 9.79 25.50

SWP 4.57 9.16 0.73 4.52 36.40 0.24 50.50 19.30 9.46 745.64 1.99 76.11 335.43 5.33 8.06

350 66.20

Environmental impact (mPt)

300 53.39 250

39.59

200 150

26.34 23.83

28.56

100

51.96

51.71

50 0

52.03

76.01

33.84

25.88 71.43 133.64

151.78 77.38

74.12

71.16 Glulam, indoor use

35.12

77.55

Glulam, outdoor Plywood, indoor Plywood, outdoor use use use Structural engineered wood

Human health

Ecosystem quality

Climate change

SWP

Resources

Fig. 8 Single score of selected structural engineered wood.

The results clearly showed that plywood production significantly burdens the environment, compared to other types of analyzed structural wood. This is mainly related to the manufacturing technology and the amount of glue consumed. In this case SWP could be considered environmentally more suitable variant compared to plywood. The two most affected impact category for all evaluated wood products was the impact on Human health and Ecosystem quality, respectively.

126

DISCUSSION The production of SWP is similar to the production of glulam, so there are slight differences between them. Most of the negative environmental impacts of SWP come from the technology used, since the PVAC adhesive had the lowest environmental impact amongst used adhesives. Within all the assessed wood-based products glulam for indoor use achieved the lowest total environmental footprint, and thus appears to be the most suitable design variant. Wood raw material choice also affects the environmental impact of the product. Diverse technology of processing and production of structural wood are related mainly to the different electricity consumption as was showed in the preliminary analysis of selected sawnwood (Table 3, Fig. 6). From this point of view, planed softwood boards were the worst assessed. The lowest environmental impact refers to softwood veneer, however, hardwood veneer is more common. However, production technology can have considerable effect on overall environmental assessment (POMMIER et al. 2016). Structural engineered wood-based materials have not been compared in this way, thus this study is unique. The studies performed concerned either a comparison of different production technologies (ENQUIST et al. 2014, POMMIER, ELBEZ 2006), environmental impacts of wood-based and conventional buildings (HASSAN, JOHANSSON 2018, SATHRE, GUSTAVSSON 2009), or products containing wood-based materials as a part of the whole (GONZALEZ-GARCIA et al. 2012). Due to the different functional unit and different evaluation methods, it is not possible to compare these results in a relevant way. In general, the lesser adhesive used and the lesser energy and technologically demanding manufacture is, the better the engineered wood products perform from the environmental point of view. The available data do not describe the production technology in detail, so some data may differ slightly from the actual ones. Errors in calculations may have occurred due to insufficient data, the use of estimated values from the available literature, different types of technology used and outdated data. The data for this study are typical for European countries. The authors of the laminated board databases for outdoor and indoor use state that some data for glulam are derived from data for SWP. The type of plywood is unspecified. Emission data are limited. Transport is not included in the calculations (WERNER et al. 2007). Moreover, it is important to note, that none of selected wood products do not count with planing despite manufacturers do. The results of the analysis should widen the knowledge of architects and people working in construction industry of the environmental context of the use of selected types of structural engineered wood-based materials.

CONCLUSIONS LCA is a tool for quantifying the environmental performance of products, considering the entire life cycle, from the production of raw materials to the final disposal of products, including the recycling of materials (GOEDKOOP et al. 2013). The performed study deals with the comparison of selected types of structural engineered wood contained in the ecoinvent v3.5 database. SimaPro 8.5.2 software was used and the IMPACT 2002+ evaluation method was chosen for the analysis. The aim of this paper was to analyze environmental impact of selected structural engineered wood and to deepen the knowledge of architects and people working in construction industry of the environmental context of the use of selected types of structural engineered wood-based materials. 127

The study proved considerable environmental burden of plywood production, mainly due to the high consumption of glue. Plywood achieved the highest environmental impact amongst the selected types of structural engineered wood. Generally, products for outdoor use showed higher values, due to the use of MF glue, which was the worst rated of the adhesives used. The SWP paradoxically had higher environmental impact than glulam, despite the use of PVAC glue. It follows that the resulting environmental impact of the SWP is significantly influenced by the manufacturing technology. The lowest values were reached by glulam for indoor use. Also, technical operations like planning and drying raise environmental burdens. Several wood-based construction materials exist. However, world databases have this information only to a limited extent. This can be a problem when creating an LCA, as even a small detail in the form of newer technology can be decisive in the overall environmental assessment. If the studies are to be comparable with each other, the database needs to be extended to include new production processes and advanced products, considering the technology used, the geographical area, the electricity sources and many other related parameters. This is the only way to achieve the highest possible reliability of the environmental assessment and the lowest inaccuracy in subsequent calculations. REFERENCES BLASS, H.J., AUNE, P., CHOO, B.S., GÖRLACHER, R., GRIFFITHS, D.R., HILSON, B.O., RACHER, P., STECK, G. 1995. Timber Engineering: Step 1, 1st ed.; Deventer: Salland De Lange. 1995, 624 s. BRANDNER, R., FLATSCHER, G., RINGHOFER, A., SCHICKHOFER, G., THIEL, A. 2016. Cross Laminated Timber (CLT): Overview and Development. In European Journal of Wood and Wood Products, 2016, 74: 331351. DOI: 10.1007/s00107-015-0999-5. CRAWFORD, R.H., BARTAK, E.L., STEPHAN, A., JENSEN, C.A. 2017. Evaluating the life cycle energy benefits of energy efficiency regulations for buildings. In Renewable and Sustainable Energy Reviews 2017, 63: 435451. DOI: 10.1016/j.rser.2016.05.061. CRAWFORD, R.H., CADOREL, X. 2017. A Framework for Assessing the Environmental Benefits of Mass Timber Construction. In Procedia Engineering, 2017, 196: 838846. DOI: 10.1016/j.proeng.2017.08.015. DINWOODIE, J. M. 2000. Timber: Its nature and behavior, 2nd ed.; CRC Press, 2000. 272 s. DODOO, A., GUSTAVSSON, L., SATHRE, R. 2012. Lifecycle primary energy analysis of conventional and passive houses. In International Journal of Sustainable Building Technology and Urban Development, 2012, 3: 105111. DOI: 10.1080/2093761X.2012.696320. EN 15804:2012 + A1:2013 Sustainability of construction works - Environmental product declarations - Core rules for the product category of construction products. ENQUIST, B., STERLEY, M., SERRANO, E., OSCARSSON, J. 2014. Green-Glued Products for Structural Applications. In Materials and Joints in Timber Structures: Recent Developments of Technology, 2nd ed.; AICHER, S., REINHARDT, H.-W., GARRECHT, H., Eds.; RILEM Bookseries, 2014. p. 4555. EUROPEAN PARLIAMENT, COUNCIL OF THE EUROPEAN UNION Directive 2009/125/EC of 21 October 2009 establishing a framework for the setting of ecodesign requirements for energy-related products. Official Journal of the European Union L, 285. 2009, s. 1035. EUROPEAN PARLIAMENT, COUNCIL OF THE EUROPEAN UNION Directive 2010/31/EU of the 19 May 2010 on the Energy Performance of Buildings. Official Journal of the European Communities. Brussels: European Parliament and the Council of the European Union; 2010. 23 s. GOEDKOOP, M., OELE, M., LEIJTING, J., PONSIOEN, T., MEIJER, E. 2013. Introduction to LCA with SimaPro, version 5.1; Netherlands: PRé Consultants, Netherlands, 2013; 80 s. [cit. 2020-01-16] Available online: <https://www.pre-sustainability.com/download/SimaPro8 IntroductionToLCA.pdf>. GONZALEZ-GARCIA, S., LOZANO, R.G., BUYO, P., PASCUAL, R.C., GABARRELL, X., RIERADEVALL I PONS, J., MOREIRA, M.T., FEIJOO, G. 2012. Eco-innovation of a wooden based modular social

128

playground: application of LCA and DfE methodologies. In Journal of Cleaner Production, 2012, 27, s. 2131. DOI: 10.1016/j.jclepro.2011.12.028. GUIDICE, F., LA ROSA, G., RISITANO, A. 2019. Product Design for the Environment: a Life Cycle Approach, 1st ed.; CRC Press, Florida, 2019. 520 s. HAFNER, A., SCHÄFER, S. 2017. Comparative LCA study of different timber and mineral buildings and calculation method for substitution factors on building level. In Journal of Cleaner Production, 2017, 167, s. 630642. DOI: 10.1016/j.jclepro.2017.08.203. HAHNEL, G., WHYTE, A., BISWAS, W. K. 2021. A comparative life cycle assessment of structural flooring systems in Western Australia. In Journal of Building Engineering, 2021, 35, 102109. DOI: 1 0.1016/j.jobe.2020.102109. HASSAN, O.A.B., JOHANSSON, CH. 2018. Glued laminated timber and steel beams: A comparative study of structural design, economic and environmental consequences. In Journal of Engineering, Design and Technology, 2018, 16 (3); DOI: 10.1108/JEDT-12-2017-0130. HILDEBRANDT, J., HAGEMANN, N., THRÄN, D. 2017. The contribution of wood-based construction materials for leveraging a low carbon building sector in Europe. In Sustainable Cities and Society, 2017, 34: 405418. DOI: 10.1016/j.scs.2017.06.013. IEA. 2019. Global Status Report for Buildings and Construction 2019. IEA, Paris. [cit. 2020-01-16] Available online: <https://www.iea.org/reports/global-status-report-for-buildingsand-construction-2019>. ISO 14040:2006 Environmental management – Life cycle assessment – Principles and framework. ISO 14044:2006 Environmental management – Life cycle assessment – Requirements and guidelines. JOLLIET, O., MARGNI, M., CHARLES, R., HUMBERT, S., PAYET, J., REBITZER, G., ROSENBAUM, R. 2003. IMPACT 2002+ A New Life Cycle Impact Assessment Methodology. In International Journal of Life Cycle Assessment, 2003, 324(6): 324–330. DOI: 10.1007/BF02978505. KRETSCHMANN, D.E. 2013. Mechanical Properties of Wood. In Wood Handbook - Wood as an Engineering Material, 1st ed.; ROSS, R., Eds.; Madison, WI: U.S. Department of Agriculture Forest Service, Forest Products Laboratory, 2013. 44 s. LIIKANEN, M., GRÖNMAN, K., DEVIATKIN, I., HAVUKAINEN, J., HYVÄRINEN, M., KÄRKI, T., VARIS, J., SOUKKA, R., HORTTANAINEN, M. 2019. Construction and demolition waste as a raw material for wood polymer composites – Assessment of environmental impacts. In Journal of Cleaner Production, 2019, 225, s. 716727. DOI: 10.1016/j.jclepro.2019.03.348. MITTERPACH, J., ILEČKOVÁ, R., ŠTEFKO, J. 2018. Life cycle impact assessment of construction materials of a wood-based building in an environmental context, In Acta Facultatis Xylologiae, 60(1): 147157. DOI: 10.17423/afx.2018.60.1.16. PAJCHROWSKI, G., NOSKOWIAK, A., LEWANDOWSKA, A., STRYKOWSKI, W. 2014. Wood as a building material in the light of environmental assessment of full life cycle of four buildings. In Construction and Building Materials, 2014, 52: 428436. DOI: 10.1016/j.conbuildmat.2013.11.066 PALUŠ, H., PAROBEK, J., MORAVČÍK, M., KOVALČÍK, M., DZIAN, M., MURGAŠ, V. 2020. Projecting climate change potential of harvested wood products under different scenarios of wood production and utilization: study of Slovakia. In Sustainability 2020, 12(6): 2510. DOI: 10.3390/su12062510. PAROBEK, J., PALUŠ, H., MORAVČÍK, M., KOVALČÍK, M., DZIAN, M., MURGAŠ, V., ŠIMO-SVRČEK, S. 2019. Changes in Carbon Balance of Harvested Wood Products Resulting from Different Wood Utilization Scenarios. In Forests, 2019, 10(7): 590. DOI: 10.3390/f10070590. PETROVIC, B., MYHREN, J. A., ZHANG, X., WALLHAGEN, M., ERIKSSON, O. 2019. Life Cycle Assessment of Building Materials for a Single-family House in Sweden. In Energy Procedia, 2019, 158: 3547–3552. DOI: 10.1016/j.egypro.2019.01.913. POMMIER, R., ELBEZ, G. 2006. Finger-jointing green softwood: evaluation of the interaction between polyurethane adhesive and wood. In Wood Material Science and Engineering, 2006, 1: 127137. DOI: 10.1080/17480270701217269. POMMIER, R., GRIMAUD, G., PRINCAUD, M., PERRY, N., SONNEMANN, G. 2016. LCA of EVP Engineering Veneer Product: plywood glued using a vacuum moulding technology from green veneers. In Journal of Cleaner Production, 2016, 124: 383–394. DOI: 10.1016/j.jclepro.2016.02.130.

129

PORTEOUS, J., KERMANI, A. 2013. Structural timber design to Eurocode 5, 2nd ed.; Oxford: Blackwell, 2013. 640 s. ISBN: 978-1-118-59729-3. SAADE, M.R.M., GUEST, G., AMOR, B. 2020. Comparative whole building LCAs: How far are our expectations from the documented evidence? In Building and Environment, 2020, 167; art. 106449. DOI: 10.1016/j.buildenv.2019.106449. SATHRE, R., GUSTAVSSON, L. 2009. Using wood products to mitigate climate change: External costs and structural change. In Applied Energy, 2009, 86(2): 251257. DOI: 10.1016/j.apenergy.2008.04.007. SOTAYO, A., BRADLEY, D., BATHER, M., SAREH, P., OUDJENE, M., EL-HOUJEYRI, I., HARTE, A.M., MEHRA, S., O’CEALLAIGH, C., HALLER, P., NAMARI, S., MAKRADI, A., BELOUETTAR, S., BOUHALA, L., DENEUFBOURG, F., GUAN, Z. 2019. Review of State of the Art of Dowel Laminated Timber Members and Densified Wood Materials as Sustainable Engineered Wood Products for Construction and Building Applications. In Developments in the Built Environment, 2019, Art. 100004. DOI: 10.1016/j.dibe.2019.100004. STARK, N.M., CAI, Z., CARLL, C. 2002. Wood-Based Composite Materials Panel Products, GluedLaminated Timber, Structural Composite Lumber, and Wood–Nonwood Composite Materials. In Wood Handbook - Wood as an Engineering Material, 1st ed.; ROSS, R., Eds.; Madison, WI: U.S. Department of Agriculture Forest Service, Forest Products Laboratory. 26 s. ISBN-13:9781484859704. STN EN 14080:2013 Timber structures. Glued laminated timber and glued solid timber. Requirements. STN EN 1995-1-1 + A1: 2008. Eurocode 5: Design of timber structures - Part 1-1: General - Common rules and rules for buildings. UNEP/SETAC 2005. Life Cycle Initiative. Life Cycle Approaches: The road from analysis to practice. 2005. 89 s. [cit. 2020-01-16] Available online: <https://www.lifecycleinitiative.org/wpcontent/uploads/2012/12/2005%20-%20LCA.pdf>. WERNER, F., ALTHAUS, H.J., KÜNNIGER, T., RICHTER, K., JUNGBLUTH, N. 2007. Life Cycle Inventories of Wood as Fuel and Construction Material. Final report ecoinvent 2000 No.9.; Swiss Centre for Life Cycle Inventories, Dübendorf, CH., 2007. WERNET, G., BAUER, C., STEUBING, B., REINHARD, J., MORENO-RUIZ, E., WEIDEMA, B. 2016. The ecoinvent database version 3 (part I): overview and methodology. In International Journal of Life Cycle Assessment, 2016, 21(9): 1218–1230. DOI: 10.1007/s11367-016-1087-8. ACKNOWLEDGEMENTS This work was supported by the grant APVV-17-0206 Ultra-low-energy Green Buildings Base on Wood as a Renewable Raw Material.

AUTHORS’ADDRESS Ing. Rozália Vaňová prof. Ing. Jozef Štefko, CSc. Technical University in Zvolen Faculty of Wood Sciences and Technology Department of Wood Structures T. G. Masaryka 24 960 01 Zvolen Slovakia xvanova@is.tuzvo.sk stefko@tuzvo.sk

130

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 131−142, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.11

CHRISTMAS TREE IGNITION BY SPARKLERS Miroslava Nejtková  Martin Podjukl ABSTRACT The paper deals with the potential of sparklers to ignite a Christmas tree. The conclusions from the experimental measurements when four fir trees decorated with Christmas ornaments were tested and burning sparklers were used for the ignition of a Christmas tree are presented in the paper. The measured temperatures were recorded by means of thermal imagers throughout the measurement process. The maximum temperatures during combustion are presented. Based on the measurements, it can be stated that sparklers are a possible source of ignition. There was always a flare through an easily ignitable Christmas decoration, not from the contact of the sparkler with a needle or branch itself. During the experiment, there was a total ignition of a decorated tree in only two cases. Key words: sparklers, Christmas fir tree, ignition source, flash point, thermal imaging camera.

INTRODUCTION Although hanging sparklers on a Christmas tree has become almost a tradition in some households, the incidence of fires caused by sparklers is very low. Recent years have shown that in apartments, family houses or other types of housing, fewer than ten cases are registered every year. Over the last decades, the highest number of events of this type was recorded in 2019, when a total of 7 fires ignited by a burning sparkler were registered (NEDĚLNÍKOVÁ 2020). No one has been killed as a result of a fire ignited by a sparkler in the Czech Republic so far. Christmas tree fires are rare in other parts of the world as well, yet they can still cause serious damage to health or even death. For example, according to estimates based on the US statistics, an average of 160 Christmas tree fires occurred annually in the USA in the period of 2013–2017. The fires resulted in 10 deaths and 20 injuries, with property damage in excess of USD 17 million (TU, NG 2019). On average, German insurance companies report 10,000 claims each year in connection with fires of Advent wreaths, Christmas trees and fireworks, with total damages reaching EUR 32 million (GDV 2018). Although the officially reported numbers of fires caused by sparklers in the Czech Republic are very low, there have not been any experiments to date to determine the fire causes through verification of the distance and conditions under which a burning sparkler can ignite modern, often plastic, decorative material used on Christmas trees. The aim of the experimental measurements was to confirm or refute the hypothesis that a burning Christmas sparkler can ignite a Christmas tree. 131

The secondary aims of the experiment were to measure the temperature during sparkler combustion, to determine the duration of combustion, and to assess the potential of the flame to spread to the surrounding flammable objects. Sparklers as a source of ignition In order to initiate the combustion process, interaction between a flammable substance, an oxidizing agent and an ignition source is necessary. Any flammable substance can be ignited by a source that reaches a defined minimum ignition temperature and is able to deliver the required amount of energy to the flammable substance over a given time interval. This time interval must be longer than the induction period of the respective flammable set (DAMEC 1998) Sparklers are classified in the F1 fireworks category (CSN EN 15947-2 (668300) 2019). This category includes products that present a very low hazard, produce negligible noise and are intended for indoor use, including residential buildings. The fire properties of the Christmas sparklers used in the experiment were not measured. However, records from 1996 are available, when Christmas sparklers (MOI – DG FRS CR. 1996) were tested according to the ČSN 640149 testing method in a certified testing laboratory. Christmas sparklers with the lengths of 10 cm, 28 cm, and 70 cm were tested. The flash point of these sparklers was 410 °C, the flash time 102 seconds, the ignition temperature 415 °C, and the ignition time 103 seconds. According to the results (RESEARCH INSTITUTE OF INDUSTRIAL CHEMISTRY 1997), the surface of the sparklers consisted of a mixture of flammable material (deposit) composed of 50% barium nitrate, 6% pyro aluminium, 5% kaolin, 25% steel and cast iron sawdust, and 10% dextrin. The fire properties of wood are described in the literature (FIRE PROTECTION UNION 1980). Wood consists of a mixture of organic substances with a cellulose content of 50%. The ignition temperature of wood is in the range of 330–470 °C. At temperatures up to 100 °C, chemically bound water evaporates from the wood; above 200 °C, the wood decomposes. At the temperatures above 270 °C, the wood is carbonized and forms a selfigniting mass. The flammability and ignition susceptibility of wood depends on the moisture content, content of resins and wax, hardness of the wood and the size of the wood particles. The ignition temperature of pine is 399 °C, spruce ignites at 397 °C, and fir at 270 °C. The flash point of fir-needles and natural forests is 270–350 °C (DEHANN, ICOVE 2011). Before conducting the Christmas tree ignition experiment, ancillary measurements were performed in the second half of 2019 in order to clarify the behaviour of sparklers according to the specifications below. Sparklers of various lengths (16 cm, 28 cm, 70 cm, 90 cm) were used for the measurements. The experiment determined the maximum surface temperature of burning sparklers in various positions, the maximum distance of flying burning sparks from vertically and horizontally placed sparklers, the potential of burning sparks to ignite test specimens, the potential of burning sparklers to ignite test specimens by direct contact or by placing a burning sparkler in/on a specimen, and the potential of a sparkler to ignite test specimens immediately after the sparkler had been extinguished. Several specific conclusions can be drawn from the measurements: - shorter sparklers reach higher absolute burning temperatures than longer sparklers. - the maximum temperature of burning sparklers measured using a thermal imaging camera ranged from 600 °C to 1,000 °C. - the temperature of the burnt part of the sparkler drops very quickly to less than 100 °C. - burning sparkler particles fell to a maximum distance of 20 cm from the vertical axis of the burning sparkler. - for 16 cm long sparklers, maximum temperatures between 800 °C and 1,000 °C were measured. The average sparkler burn time was 40 seconds. 132

for 28 cm long sparklers, maximum temperatures between 800 °C and 1,000 °C were measured. The average sparkler burn time was 90 seconds.

EXPERIMENTAL PART Experimental setup The experiment was performed in the Experimental and Training Laboratory of the Fire Cause Investigation Laboratory of the Population Protection Institute. For the experiment, Christmas trees of the Caucasian fir (Abies nordmaniana) variety were used, which had been used as Christmas decorations in households during the Christmas holidays. In January 2020, the trees were stored outdoors for 7 days. Subsequently, the trees were stored in a closed, unheated room for seven days before Experiments 1 and 2, and for fourteen days before Experiments 3 and 4. The experiments were performed in a testing unheated room at 5 °C. The testing room is equipped with ventilation unit and installed filters, but it is not equipped with heating. For this reason, the ambient temperature during the experiment did not correspond to normal temperatures in household rooms. A total of four trees were selected for the experiments and decorated with Christmas decorations. The height of the fir trees was in the range of 1.8-2 m. The tested materials included traditional Christmas tree ornaments made of straw as well as contemporary ornaments that are commonly available in supermarkets. The decorations consisted of plastic balls, decorative garlands and bows. A 2 m long Christmas garland was placed on the trees. The number of other decorations intentionally differed from one measurement to another. In all cases, 16 cm long sparklers were hung on the fir trees. Their average weight was 1.46 g with an average layer length of the pyrotechnic mixture of 10 cm. All measurements were recorded using thermal imaging cameras (FLUKE Ti32 and Ti400) and video cameras, and digital still photographs were also taken. Thermal imaging recordings were evaluated using specialised software (smartView 3.15, smartView 4.3). The Thermal imaging camera FLUKE Ti400 was located 5,5 m from the fir. Experimental measurements procedure Contact thermometers or non-contact thermometers can be used to measure temperature. Contact thermometers are mostly used for stationary objects. However, since the flame temperature of the burning sparkler changed over a very short time period (thermocouples require a certain amount of time for the actual measurement), it was not possible to use thermocouples to measure the temperature change curve. For these reasons, a thermal imaging camera was used to measure the temperature of the samples continuously. Spectral emissivity depends on the wavelength, temperature and material. Thermal imaging cameras support manual adjustment of emissivity (e.g. value of 1 – black body; wood (70 °C) 0.94; plastic PE, PP, PVC (20 °C) 0.94; wall (40 °C) 0.93; colour (90 °C) 0.92–0.96 (FLUKE CORPORATION 2013). For the purposes of our experiment, emissivity was set to 0.95. Fluke Ti32 thermal imaging camera temperature measurement range allows temperature measurements up to+ 650°C. The Ti 400 thermal imaging camera took radiometric video with a temperature measurement range up to +1200°C. These thermal imaging cameras scan at 320x 240 (76800 px) infrared resolution with accuracy +- 2°C and 2% at nominal temperature of 25°C. The distance of measured object from the thermal imaging camera was measured by the integrated laser distance meter.

133

RESULTS AND DISCUSSION Evaluation of experimental measurements Experiment no. 1 The Christmas tree was decorated with a Christmas garland on one half of the tree, and 7 Christmas ornaments (balls and bows) on the other one. Two sparklers, each 16 cm in length, were placed on the tree with the lower part of the sparklers touching the Christmas ornaments and garlands. Evaluation of experiment no. 1 – Two sparklers were lit. They burned down within 40 seconds. The Christmas tree garland was not broken and the tree branches did not burn off. The tree did not ignite. Experiment no. 2 The Christmas tree was decorated with Christmas garlands only. Only one sparkler was placed on the tree. The central part of the sparkler was touching the garland. (The experimental measurements showed that this specific Christmas garland ignited within 3 seconds of placement of the burning sparkler on the garland; the maximum surface temperature measured was 863 °C. Evaluation of experiment no. 2 – After the burning sparkler came into contact with the garland, the garland ignited and subsequently, the flames spread on both sides of the garland. The maximum temperature measured on the tree with the burning garland was 584 °C. The garland was completely burned within 3 minutes of the lighting of the sparkler, but due to the radiant heat it emitted and the gradually igniting needles, heat accumulated in the upper part of the tree, which preheated the remaining needles and branches. At 03:30 minutes, the upper part of the tree ignited and the flames began spreading in all directions. During the fourth minute the temperature reached 905 °C. During the fifth minute the entire tree was in flames. Temperatures of around 1,000 °C were measured. The flames extinguished themselves during the sixth minute. Only glowing remnants of the needles were visible. Fig. 1 – Fig. 3 are thermographic images of the burning tree in experiment no. 2. Fig. 4 shows a spreading of the fire at 210 seconds. Experiment no. 3 The tree was decorated with various Christmas ornaments including several Christmas garlands wound around the tree and tinsel that had been subsequently placed evenly around the tree. 15 Christmas decorations in the shape of a ball were used for decoration. The balls were made of polystyrene. There was a layer of paint on the surface of the polystyrene ball. Balls with different surfaces were used - 5 pieces with a matte surface, 5 pieces with a glossy surface and 5 pieces with a roughened surface with glitter. The average weight of the ornaments was 8.25 g. The tree was still decorated with fringes (same material as the decorative chain). Three sparklers were hung on the tree. Two of the sparklers were touching the garland (the upper half of the first sparkler and the lower half of the second sparkler) and the third sparkler was hung above the Christmas ornaments but was not touching the ornaments. Evaluation of experiment no. 3 – The sparkler, which did not touch the Christmas garland, simply burned off. Both sparklers, touching the garland, spread flame along the length of the garland as they burned off. (Fig. 6). Parts of the Christmas ornaments located in the proximity of the garland were burned, but the flames did not spread to the needles or branches (Fig. 7). Temperatures of about 400 °C were measured on the burning garland. The fire extinguished itself during the seventh minute of the experiment.

134

Fig. 1a, 1b,1c - Experiment no. 2 – Spreading of the fire at 30 seconds, 90 seconds, 150 seconds.

Fig. 2a, 2b,2c - Experiment no. 2 – Spreading of the fire at 210 seconds, 240 seconds, 270 seconds.

Fig. 3a, 3b,3c - Experiment no. 2 – Spreading of the fire at 300 seconds, 330 seconds, 360 seconds.

Experiment no. 4 The tree was decorated with various Christmas ornaments, including several Christmas garlands wound around the tree and tinsel which had been subsequently placed evenly around the tree. Three sparklers were placed on the tree. One of the sparklers was placed so that its central part was touching the Christmas garland. Evaluation of experiment no. 4 – After the sparklers were lit, the Christmas garland was ignited by the sparkler which was touching it. One minute after the start of the experiment, the temperature of the burning tree was in the range of 400–760 °C. At 2 minutes and 30 seconds, flames covered approximately 75% of the tree. Three minutes after the start of the experiment the tree was completely consumed by flames. The height of the flames from the 135

floor was up toca. 4 m (the height of the tree including the stand was 2 m). At 3 minutes and 30 seconds only glowing needles remained on the tree and the temperature ranged from 290 to 450 °C. Over the next two minutes, the glowing of the needles gradually subsided and the temperature reached 250 °C. Fig. 8 – Fig. 10 are thermographic images of the burning tree in experiment no. 4. Fig. 5 shows spreading of the fire at 150 seconds.

Fig. 5 - Experiment no. 4 – Spreading of the fire at 150 seconds

Fig. 4 Experiment no. 2 - The spreading fire at 210 seconds

Fig. 6 Experiment no. 3 – Spreading of flames on the Christmas garland

Fig. 7 Experiment no. 3 – View of the burned garland and a partially burned ornament

136

Fig. 8a, 8b,8c - Experiment no. 4 – Spreading of the fire at 30 seconds, 60 seconds, 90 seconds.

Fig. 9a, 9b, 9c - Experiment no. 4 – Spreading of the fire at 120 seconds, 150 seconds, 180 seconds.

Fig. 10a, 10b - Experiment no. 4 – – Spreading of the fire at 210 seconds, 240 seconds.

137

Tab. 1 Maximum temperatures measured on the decorated Christmas tree. [°C] [s] Experiment no. 1 Experiment no. 2 Experiment no. 3 Experiment no. 4 [°C] [s] Experiment no. 1

15 260 260 205.9 260 135 x

30 240.8 539.1 419.6 487.8 150 x

45 129.8 518.8 391.8 537.2 165 x

60 65.0 540.5 352.9 569.2 180 x

75 37.7 464.4 315.5 735.5 195 x

Experiment no. 2 Experiment no. 3 Experiment no. 4

626.6 370.1 681.3

649.5 348.8 895.8

578.5 428.8 968.3

483.3 370.1 803.0

481.8 348.8 642.9

[°C] [s] Experiment no. 1 Experiment no. 2 Experiment no. 3 Experiment no. 4

255 x 927.0 308.4 357.2 432.4*

270 x 919.7 362.1 361.2 441.1*

285 x 971.4 384.4 343.3 441.2*

300 x 996.2 345.6 346.3

330 x 753.7 247.6 x

90 29.7 560.1 320.6 698.1 210 x

661.5 428.8 447.7 525.1* 360 375 x x 565.1 260.0 260.0 x x

105 x 541.8 291.7 784.1 225 x

120 x 501.6 308.3 705.1 240 x

830.4 353.8 395.1 471.2* 390 x

872.0 278.2 376.4 447.1* 405 x x 120.1 x

231.8 x

*The highest temperature was measured on the fallen material around the tree

The authors of the paper focused on searching information about fires, their ignition and investigation of fires caused by sparklers. Detailed information on pyrotechnics can be found in the literature, but specifically the issue of sparklers is not addressed (except for the burning temperature). The book Kirk’s Fire Investigation (DEHANN, ICOVE 2011) describes the principles of how wood burns and the characteristics of pyrotechnics and explosives, but does not include details about the properties of sparklers. The publication NFPA 921 (TECHNICAL COMMITTE ON FIRE INVESTIGATIONS 2020), which addresses fire cause investigation, does not contain information about sparklers as an ignition source. It mentions e.g. the principles of combustion, fire triangle, spreading of forest fires, and fireworks as an ignition source. The bachelor’s thesis Analysis of Temperature Changes During Launching of Fireworks (MEJZLÍK 2018) gives a summary of measurements of maximum surface temperatures and heat flow to selected fireworks. The measurements were conducted using both contact (Type K thermocouple) and non-contact methods. The objectivity of the temperature measurements was negatively affected by the selection and placement of a Type K thermocouple, and also by the selection of a thermal imaging camera that only depicts the maximum measured temperature. Furthermore, it was not possible to take a detailed thermographic image from any measurement location. The publication Propellants, explosives, pyrotechnics (SABATINI, NAGORI 2011) lists the properties of various substances and compounds that are used as ingredients in pyrotechnics and their role in activation of pyrotechnics. The NFPA (AHRENS 2020) has also developed fire safety conditions for the use of fireworks and other pyrotechnics, e.g. safe handling of pyrotechnics and related injury statistics and also states that the temperature of sparks from pyrotechnics is 1200° F (648° C). The association NFPA and the Institut für Schadenverhütung have created videos that show flames from a burning Christmas tree spreading to objects in the room. These videos also show the time elapsing as the flames spread. However, the fires in these cases were started by different ignition sources. In case (NFPA 2018), the likely ignition source was a short-circuit on an electrical Christmas garland (the authors did not clearly state the ignition source). In the case (NIST 2017), the source was a candle that came loose from the 138

candleholder as it burned down and then fell on wrapped presents under the tree. Video (NFPA 2007) shows the spreading of flames under a Christmas tree that was watered daily and then the spreading of flames under a Christmas tree with dry needles. The videos do not show the temperatures during the fires. Wood is a thermally degradable and combustible material. Applications range from a biomass providing useful energy to a building material with unique properties. Wood products can contribute to unwanted fires and be destroyed as well. Flame spread is the sliding movement of the flaming ignition point over the surface of a solid combustible. The flame provides the ignition source as well as mixing of volatiles and air. The quasisteady time response of the material to heat flux distribution from the flame and external sources in reaching the surface temperature of rapid volatilizations determines the rate of flame spread. Rate of flame spread generally decreases with increases in density, moisture content, surface emissivity, surface temperature at ignition, and thermal conductivity (WHITE, DIETNBERGER 2001). The article (ZACHAR et al. 2017) dealing with comparison of the activation energy requires for spontaneous ignition and flash point of the Norway spruce wood and thermowood specimens presents following conclusions. The results obtained within particular experiments showed that the spontaneous ignition temperature of Norway spruce wood specimens varied in the temperature range of 429 °C to 472 °C. The results achieved by the Thermowood specimens were comparable to Norway spruce wood specimens results (temperature range of 418 °C to 462 °C). The minimum flash point temperature was reached at a temperature of 346 °C in the case of the Thermowood specimens. The Norway spruce wood specimen flash point occurred at the temperature about 9 °C higher, i.e. 355 °C. The activation energy needed to reach the spontaneous ignition was very similar for both Norway spruce specimen kinds tested. The papers dealing with flash point of the fir we did not find. Evaluation of experiments In the experiments, thermographic images were used to measure the surface temperature of burning sparklers and the temperature after they had burned, along with the temperature of the combustion of ornaments and the Christmas tree. Four sets of measurements were taken. The highest temperatures measured on the burning Christmas tree are listed in Table no. 1. Within experiment No. 1, when the sparklers burned off, the Christmas garland, burning for one minute, extinguished itself. The temperatures reached in experiment No. 2 after the sparkler ignited reached temperatures of up to 600 °C in the first two minutes. This temperature oscillated for the next minute and a half. In the 3rd minute and 45 seconds, there was a total ignition of the tree. In the 5th minute after the sparkler ignited, complete ignition of the tree was achieved. For the next minute and a half, the tree extinguished itself. In experiment no. 3, the Christmas garland gradually burned along with the adjacent needles. There was not a massive amount of falling, burning needles. In experiment no. 3, temperatures of approximately 400 °C were measured, which lasted for almost the entire first 5 minutes of the experiment. In contrast, the highest temperatures in experiment no. 4 were measured only until the third minute. In experiment no. 4, the temperatures were higher (almost 1,000 °C), the ignition of the tree was also more forceful and the height of the flames was up to 4 m from the floor. After lighting a sparkler, the temperature rose to almost 550 °C in one minute. After the first minute, the temperature oscillated around 710 °C. The highest temperature of almost 1000 °C was reached within the 2nd minute and 45 seconds. This was followed by burning of the entire tree and subsequent self-extinguishing. From the 3rd minute and 30 seconds, the highest surface temperatures were measured on flown material and fallen on the ground around the tree. At 139

this time, only a part of a twig burned locally on the tree (450 °C), the average temperature measured on the tree was 164 °C. The average temperature in the fourth minute was 104.5 °C. Based on the data from the experiment, it can be assumed that when a Christmas tree is placed on a carpet in e.g. an apartment in a housing estate, the flames could reach the ceiling of the room. Without a doubt, the flames would spread extremely quickly to the surrounding furnishings, Christmas presents, etc. The maximum surface temperature of burning sparklers and emitted sparks can be as high as 1,000 °C. However, because the induction period is very brief, neither the tree nor the ornaments actually ignite. The proof of this statement is the fact that although the temperatures measured on burning sparklers in two cases of this experiment were actually higher than the ignition temperature of needles or Christmas ornaments, the Christmas tree did not ignite.

CONCLUSION The experiments confirmed the hypothesis that burning sparklers can be a potential ignition source for fires on decorated Christmas trees. Connection between a larger number of decorations on the tree and a larger fire were proven. In the case of experiment No. 2, only one sparkler was used. The tree was decorated only with a Christmas garland, without any other Christmas decorations, but it still caused a fire. In experiment No. 3, the Christmas tree was decorated with both garlands and Christmas ornaments. Three sparklers were used, the tree did not ignite, but only a partial burning of the Christmas garland. In all cases where the flame spread, it initially spread along the Christmas garland. If the sparkler was loosely hung on the tree (it did not touch the ornaments), it only burned out. Also if the sparkler hung on the tree and touched the needles, the flames did not spread to the tree. The sub-objective of the measurement was met because a large amount of data containing temperature values over time was obtained during the experimental measurement, and it is possible to return to these measured values at any time and re-measure at any point in the burning tree. These measured values can also be used for the purpose of fire cause investigation. Sparklers are undoubtedly a source of high ignition energy that exceeds 1,000 °C under ideal conditions. Nonetheless, this temperature is not maintained over of the entire surface of the sparkler, but only in the part of the sparkler that is actually burning. As a result, the burning surface of the sparklers affects surrounding materials for only a very brief period of time. This fact means that sparklers are capable of igniting only highly flammable materials, e.g. dry pinecones, foam, some types of textiles, etc., upon contact. Ignition of a Christmas tree or its needles, if sufficiently dry, could also be caused by a sparkler. The experiments, however, demonstrated that ornaments on a decorated Christmas tree ignite more readily than needles. But in these cases ignition occurs only when the ornament and the sparkler are actually touching. In the case of straw ornaments, however, the energy of sparks that have flown away is sufficient for ignition. Therefore, sparklers are a dangerous pyrotechnic, but the safety risk is low. Sparklers may be categorized as an ignition source with the potential to ignite common, highly flammable substances, but only when the given substance is in immediate contact with a burning sparkler. Sparks alone do not have this potential, or rather, their potential is limited to highly flammable materials (such as dry hay) or flammable gases.

140

REFERENCES AHRENS, M. 2020. Fireworks Fires Supporting Tables. National Fire Protection Association [online]. 2020 [cit. 2020-09-20]. Available online: < https://www.nfpa.org/News-and-Research/Dataresearch-and-tools/US-Fire-Problem/Fireworks-fires-and-injures> CSN EN 15947-2 (668300): 2019. Pyrotechnic articles – Fireworks, Categories F1, F2 and F3 – Part 2: Categories and types of firework. Czech Office for Standards, Metrology and Testing. DAMEC, J. 1998. Explosion Prevention. Association of Fire and Safety Engineering. 1998. ISBN: 80-86111-21-0 DEHANN, J. D., ICOVE. D. J. 2011. Kirk’s Fire Investigation. 7th ed. ISBN-13: 978-0-13-508263-8. FIRE PROTECTION UNION. 1980. Tables of flammable and hazardous substances, Fire Protection Union, Czechoslovak Socialist Republic. 1980 FLUKE CORPORATION. 2013. User manual [on-line] Rev. 2 [cit.2020-09-20] Available online: <https: //www.termokamery.cz/out/pictures/wysiwigpro/n%C3%A1vody/navody%20Fluke%20Ti.pdf> GDV. 2018. [online] Kerzen häufigste Ursachen für Brände in Haus und Wohnung [online] [cit.2020-09-20] Available online:<https://www.gdv.de/de/medien/aktuell/adventskraenze-undweihnachtsbaeume-haeufige-ursache-fuer-braende--42072> MEJZLÍK, L. 2018. Analysis of Temperature Changes During Launching of Fireworks. Ostrava, 2018, 59 pp. On-line Thesis. Faculty of the Safety Engineering VSB – Technical University of Ostrava. MOI – DG FRS CR. 1996. Fire Protection Technical Institute. Report no. 2595 on Technical Fire Properties Tests. 1 p. 1996 NEDĚLNÍKOVÁ, H. 2020. Statistical yearbook 2019 [online] [cit.2020-09-20] Available online: <https://www.hzscr.cz/clanek/statisticke-rocenky-hasicskeho-zachranneho-sboru-cr.aspx> NFPA. 2007. Christmas Tree Fire [on-line] 2017 [cit.2020-09-20] Available online: < https://www.youtube.com/watch?v=RNjO3wZDVlA> NFPA. 2018 Winter holiday fire facts [on-line] [cit.2020-09-20] Available online: <https://www.nfpa.org/Public-Education/Fire-causes-and-risks/Seasonal-fire-causes/Winterholidays/Holiday-fires-by-the-numbers> NIST. 2017. Christmas Tree Fire:Watered Tree vs. Dry Tree. 2017 [on-line] [cit.2020-09-20] Available online: < https://www.youtube.com/watch?v=pZBzPAHrMmQ> RESEARCH INSTITUTE OF INDUSTRIAL CHEMISTRY. 1997. Synthesia a.s. 1997. Report no. 57/97 on Sparkler Tests. 7 p. SABATINI, NAGORI, J. A. 2011. Propellants, explosives, pyrotechnics. 2011. pp. 37378. TECHNICAL COMMITTE ON FIRE INVESTIGATIONS. 2020. NFPA 921 Guide for Fire and Explosion Investogations [online]. 2020 [cit. 2020-11-20]. Available online: <https://www.nfpa.org/codes-andstandards/all-codes-and-standards/list-of-codes-and-standards/detail?code=921> TU., Y., NG, J. 2019. Fireworks Annual Report Fireworks-Related Deaths, Emergency DepartmentTreated Injuries and Enforcement Activities During 2 [online] U.S. Consumer Product Safety Commission [cit.2020-09-20] Available online: <https://www.cpsc.gov/s3fspublic/Fireworks_Report_2018.pdf?5kZ4zdr9jPFyhPmeg3MoL35mGX8fB0s7> WHITE R. H., DIETNBERGER 2001. Encyclopedia of Materials: Science and Technology pp. 97129716, 2001. ISBN: 0-08-0431526 ZACHAR, M., MAJLINGOVA, A., SISULAK, S. et al. 2017. Comparison of the activation energy required for spontaneous ignition and flash point of the norway spruce wood and thermowood specimens. In Acta Facultatis Xylologiae Zvolen. 2017, 59(2): 7990. DOI: 10.17423/afx.2017.59.2.08

141

ADDRESSES OF AUTHORS Ing. Miroslava Nejtková MoI – DG FRS CR Population Protection Institute Na Lužci 204 533 41 Lázně Bohdaneč Czech Republic Ing. Martin Podjukl MoI – DG FRS CR Kloknerova 26 pošt. příhr. 69 148 01 Praha 414 Czech Republic

142

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 143−152, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.12

INFLUENCE OF CARBON ACCOUNTING ON ASSESSMENT OF WOOD-BASED PRODUCTS Rozália Vaňová ABSTRACT As the population grows, the number of products and services increases escalating the burden on the environment. Carbon dioxide is the largest contributor to global warming among all greenhouse gases. Life Cycle Assessment defines specific impacts of anthropogenic activities through multiple calculation methods, the majority of which are not identical. Carbon uptake accounting can substantially affect the perception of wood products in the overall assessment. Different approaches to the inclusion of greenhouse gases to global warming and their impact on the production of selected wood-based products – glued laminated timber, dimensional timber, solid structural timber, oriented strand board, particleboard, and light, medium and high-density fibreboard are shown in the paper. Dimensional timber achieved the lowest emissions proving the easier the manufacture the least the product burdens the environment. However, glulam seems to be the best carbon sink when carbon uptake is taken into account. Fibreboards were ranked the worst by the majority of methods. Key words: carbon balance, structural timber, life cycle assessment.

INTRODUCION In order to reduce negative impacts of human activities on the Earth ecosystems, sustainability becames one of the most important pillars of the global building agenda enhanced by the intense worldwide interest in carbon dioxide (CO2) emission reduction (WOODARD and MILNER 2016). In 2019, building construction and operations accounted for the largest share of global total final energy consumption (35%) and energy-related CO2 emissions (38%) (UNEP 2020). Therefore, attention was drawn on manufacturing and usage of sustainable construction materials and buildings (RÉH et al. 2021, VANOVA et al. 2021). Life Cycle Assessment (LCA) method is an analytic tool evaluating the impact of human activities on the environment using various impact categories based on material and energy balances of input and output flows of the system under study (ISO 14040). One of the most common categories used in the evaluation is global warming or climate change, driven by large amount of greenhouse gases (GHG) emissions; particularly CO2, methane (CH4) and nitrous oxide (N2O); in the atmosphere (MONTZKA et al. 2011). Wood-based products are one of the key aspects of climate change impact mitigation (SKULLESTAD et al. 2016). Wood represents one of the earliest construction material (RYBNÍČEK et al. 2020). As it grows, wood sequesters CO2 from the air and stores it until the combustion or decomposition. Therefore, wood biomass can be considered carbon 143

neutral over time (COWIE et al. 2019, PAROBEK et al. 2019). However, the perception of wood from the environmental point of view remains inconsistent. Two types of carbon sources exist – fossil and biogenic. Fossil carbon emissions comprise vast amount of substances of which CO2 is the most represented. Transport and fossil-based electricity and heat production contribute to majority of global CO2 emissions. Efforts are currently being made to reduce fossil fuels and increase the share of renewable energy sources (PARASCHIV and PARASCHIV 2020). The treatment of biogenic carbon emissions and removals is a challenging issue in environmental assessment (BRANDÃO et al. 2013, LEVASSEUR et al. 2012). Biogenic carbon can be stocked in biological matter and soil. Wood represents a natural source of CO2. Thanks to photosynthesis, it is possible to remove carbon in the air by its incorporating into the organic matter. However, combustion of biomass forms the main source of biogenic CO2 (RODIN et al. 2020). Therefore, extending the service life of wood-based products is one of strategies to improve resource efficiency (CARUS and DAMMER 2018). This can be done through incorporating circular economy practises, e.g. recycling of waste wood. Though, allocation of burdens and benefits of recycling materials throughout their sequence of applications is rather unclear as a consequence of shifting from one life cycle stage to another (DJURIC ILIC et al. 2018). In most LCA studies the related climate change effect is not taken into account: biogenic CO2 is either not considered or biogenic CO2 emissions are assumed to balance out carbon uptake during biomass growth. Emission and removal of biogenic CO2 in wood biomass usually occur at different points in time. Uneven approaches to carbon life cycle assessment complicate the expression of related global warming and climate change (GARCIA et al. 2020). As a consequence, wood-based products or even whole buildings are considered more or less environmentally beneficial (HOSSAIN and POON 2018, HÄFLIGER et al. 2017, PIEROBON et al. 2019, SAADE et al. 2020, ZIEGER et al. 2020). The choice of calculation method can affect the overall assessment due to different scores of substances (SAFARI and AZARIJAFARI 2021, SARTORI et al. 2021). This study compares production burdens of selected wood products by different GHG emissions calculation methods in order to distinguish between carbon captures and emissions that consequently affect overall environmental impact of these products.

MATERIALS AND METHOD For the assessment purposes, 8 wood-based products were chosen (Table 1). LCA methodology was applied considering the cradle-to-gate assessment (ISO 14044). Hence, all operations from resource extraction to the factory gate were accounted. Data were taken from an international Life Cycle Inventory (LCI) database (WERNET et al. 2016) covering average global production activities extrapolated from existing regional datasets. Global datasets reflect the global average based on international data. The composition and share of specific datasets on the overall product database as well as other specific information is described in Table 1. Functional unit was set to 1 m3 of a particular product. Analysis was carried out by SimaPro software, version 9.1.1.1 (PRÉ CONSULTANTS 2016). Products were assessed due to global warming potential (GWP) by several calculation methods – CML-IA (GUINÉE 2002), EDIP (HAUSCHILD and POTTING 2003), Environmental Footprint (EF) (FAZIO et al. 2018), EPD (EPD INTERNATIONAL AB 2019), ILCD (JOINT RESEARCH CENTRE 2010), IMPACT 2002+ (JOLLIET et al. 2003), ReCiPe (HUIJBREGTS et al. 2017), BEES (LIPPIATT 2007), TRACI (BARE 2011) and IPCC (INTERNATIONAL PANEL ON CLIMATE CHANGE 2014). 144

All these methods are part of the SimaPro software and serve to identify specific environmental impacts. Each calculation method transforms input and output material and energy flows within the system under study into the GWP impact of GHG emissions expressed as carbon dioxide equivalent (CO2 eq). The calculation methods use different characterization factors for GHG emissions to compute the GWP impact of a product. Tab. 1 Selected wood-based products specification. RoW – rest of the world, GB – Great Britain, EwS – Europe without Switzerland, PF – phenolic resin, UF – urea formaldehyde resin, PMDI – polymeric methylene diphenyl diisocyanate, VWE – virgin wood (eucalyptus species), WP – wet process, WDP – wet and dry processes. Product

Density

Geographic area of a dataset

Glued Laminated Timber for indoor use (Glulam)

625 kg/m3

Solid Structural Timber (SST) Dimensional Timber (DT)

625 kg/m3 475 kg/m3

Canada Europe RoW Europe Canada Switzerland RoW

Oriented Strand Board (OSB)

640 kg/m3

Particle Board (PB)

650 kg/m3

High Density Fibreboard (HDF) Medium Density Fibreboard (MDF) Light Density Fibreboard (LDF)

920 kg/m3 750 kg/m3 200 kg/m3

Canada Europe RoW GB Europe RoW Europe RoW Europe RoW Canada EwS RoW Switzerland

Share of a dataset on the overall database 0.003 0.671 0.326 1 0.036 0.001 0.004 0.923 0.035 0.001 0.382 0.617 0.035 0.298 0.352 0.316 0.295 0.705 0.135 0.865 0.004 0.280 0.682 0.034

Used resin

Other specific data

PF UF

Kiln dried Air dried

UF none

Air dried Air dried Kiln dried Air dried Air dried Kiln dried -

PF PMDI UF

VWE -

VWE WP

MEF

PUR

WDP

RESULTS The amount of GHG emissions produced in the manufacturing stage of selected products were compared by 12 calculation methods (Table 2). CML-IA and EPD methods had the same basis resulting in equal GHG amounts; as well as EDIP and TRACI calculation methods. ILCD and IPCC including CO2 uptake were the only methods concerning CO2 in the air as a raw material stored in biological matter, thus CO2 values were negative indicating carbon removal. Slight discrepancies in the results between selected methods were caused by different inclusion of some emissions into the air, especially fossil ones.

145

Tab. 2 LCI results of selected wood-based products (kg CO2 eq; a - years); functional unit – 1 m3. Method CML-IA EDIP EF EPD ILCD IMPACT 2002+ ReCiPe BEES TRACI IPCC 100a IPCC 20a IPCC uptake

Glulam 235.78 234.93 241.40 235.78 1341.70 227.36 239.85 233.43 234.93 236.76 260.65 1339.89

SST 180.76 180.37 184.75 180.76 1199.10 175.32 183.44 179.28 180.37 181.34 196.08 1198.15

67.87 67.66 69.36 67.87 1186.27 66.44 68.74 67.35 67.66 68.10 73.37 1185.83

OSB 376.00 376.69 389.83 376.00 837.24 351.39 386.66 373.96 376.69 378.16 427.02 835.82

PB 365.09 361.42 375.75 365.09 642.98 343.74 373.87 358.54 361.42 366.98 427.32 637.55

HDF 1163.15 1152.98 1198.10 1163.15 249.60 1100.59 1190.92 1143.93 1152.98 1170.04 1355.77 232.62

MDF 761.90 757.06 781.96 761.90 340.51 727.72 778.06 751.90 757.06 766.41 869.66 349.73

LDF 87.80 87.42 89.69 87.80 0.96 84.72 89.23 86.87 87.42 88.14 96.90 1.66

To clarify the different approach of carbon accounting specific carbon emissions and captures were listed (Table 3). Almost all methods used similar pattern in carbon accounting each based on fossil carbon emissions and emissions from land transformation. Carbon storage in soil was also present except for ILCD method where it was replaced by CO2 uptake from the air and supplemented by biogenic carbon emissions and emissions from peat oxidation. IPCC uptake method contained all carbon emissions and removals mentioned except for the last one. Tab. 3 Carbon accounting in particular calculation methods (the value represents the contribution to climate change; negative sign refers to removal); LT – Land Transformation, PO – Peat Oxidation. Method CML-IA EDIP EF EPD ILCD IMPACT 2002+ ReCiPe BEES TRACI IPCC 100a IPCC 20a IPCC uptake

Fossil 1 1 1 1 1 1 1 1 1 1 1 1

Carbon emissions and removals LT Soil Biogenic Uptake 1 -1 0 0 1 -1 0 0 1 -1 0 0 1 -1 0 0 1 0 1 -1 1 -1 0 0 1 -1 0 0 1 -1 0 0 1 -1 0 0 1 -1 0 0 1 -1 0 0 1 -1 1 -1

PO 0 0 0 0 1 0 0 0 0 0 0 0

Table 4 shows the distribution of CO2 emissions and removals according to IPCC uptake method. Obviously, sequestrated carbon played a substantial role in the overall assessment. Emissions from land transformation were negligible. Fossil carbon emissions replicate CO2 emissions in methods non-considering carbon uptake reflecting mainly emissions associated with transport and fossil-based heat and electricity. Biogenic carbon emissions came from wood incineration within the particular life cycle. Fossil and biogenic carbon emissions were similar in most cases.

146

Tab. 4 Assessment of selected wood-based products using IPCC including CO2 uptake method (emissions to soil were assigned to fossil carbon); functional unit – 1 m3. Impact category Glulam SST DT OSB PB HDF MDF LDF

Fossil 234.62 179.33 67.51 376.35 357.22 1137.78 762.89 86.94

Carbon emissions and removals Biogenic Uptake LT 330.97 1.02 1906.50 335.14 0.95 1713.57 4.53 0.48 1258.36 374.75 1.04 1587.96 394.01 1.06 1389.83 553.40 16.57 1940.37 544.11 1.43 958.69 152.09 0.57 237.94

Total 1339.89 1198.15 1185.83 835.82 637.55 232.62 349.73 1.66

In order to determine the magnitude of the products impact two methods were selected – IPCC uptake and CML-IA – representing the inclusion and non-inclusion of carbon uptake (Table 5). In the first case, glulam was found the most beneficial product followed by SST and simple dimensional timber. During production of LDF and MDF fossil carbon prevailed which led to positive signs indicating environmental burden. The second approach rejecting carbon uptake considered DT as a product with the lowest impact on climate change. As the worst wood-based alternative HDF was shown. The sequence has changed for almost all products except PB that stayed at the 5th rank in both cases. Tab. 5 Ranking of selected wood-based products according to different carbon accounting (products were sorted from the most beneficial to the most burdensome); functional unit – 1 m3. Rank 1 2 3 4 5 6 7 8

Glulam SST DT OSB PB HDF LDF MDF

IPCC uptake

-1339.89 -1198.15 -1185.83 -835.82 -637.55 -232.62 1.66 349.73

DT LDF SST Glulam PB OSB MDF HDF

CML-IA

67.87 87.80 180.76 235.78 365.09 376.00 761.90 1163.15

So far, the calculations have considered the environmental load of 1 m3 of the product. As the density of the selected products varied mass allocation was performed (Fig. 1). With decreasing amounts of wood per unit weight, the content of sequestrated carbon declined leading to dominance of CO2 emissions into the air. According to the IPCC uptake method DT was the most environmentally beneficial product. On the other hand, MDF was found the worst. Pursuant to the CML-IA results highest emissions reported HDF followed by MDF. These products had the highest densities; thus they should act environmentally beneficial in terms of carbon uptake. As this is not the case, they will be considered below.

147

Fig. 1 GWP per 1 kg of a particular wood-based product (total weight of CO2 eq).

Comparison of IPCC uptake and CML-IA methods recorded the highest difference in CO2 emissions by dimensional timber reaching 2.64 kg CO2 eq/kg of product. Therefore, DT remained considered the most unstable product in terms of GWP due to different carbon accounting. However, DT was ranked the best according to both methods. In order to justify diverse amounts CO2 emissions from production, glulam, DT, HDF and MDF were chosen for a deeper evaluation of fossil emissions according to the IPCC uptake method (Fig. 2).

Fig. 2 Fossil carbon emission contribution per 1 m3; a – Glulam, b – DT, c – HDF, d – MDF; IPCC uptake method.

148

Dimensional timber production was technologically undemanding, corresponding to proportional distribution of fossil carbon emissions. Electricity and heat consumption took one fourth of overall emissions. Natural gas and diesel were bound to internal transport and associated processes. Transport referred to distances necessary for materials import. High share of other fossil-based emissions suggested numerous minor processes incorporated. MDF was ranked the worst by IPCC uptake method and the second worst by CML-IA method. Notably, 42% of the fossil carbon emissions were bound to energy consumption. Moreover, one fifth of the emissions created within the MDF production was directly linked to ammonia manufacture as an essential element of urea-formaldehyde adhesives. Only 6% and 9% were connected with natural gas and diesel; and transport, respectively. Most of the glulam production emissions (37%) were related to energy consumption, 22% were due to transport and about a quarter were associated to other minor processes. Ammonia production accounted for 6% of total fossil emissions. Fossil emissions from HDF production were the highest within all products, reaching 1137.78 kg CO2 eq/m3 HDF and 1.26 kg CO2 eq/kg HDF. Up to 65% of these emissions accounted for energy consumption. Notably, 6% was caused by hard coal mining operations.

DISCUSSION Product databases used for the analysis contained average data from all over the world, designed according to average production technologies and transport distances. So, they could be used for any geographical region. Specific production in specific areas might, therefore, report different values more or less burdensome than the stated GWP. CO2 uptake included all inputs within the scope of product manufacturing. Higher amounts of carbon also indicated wood losses during processing. Moreover, the choice of a calculation method affected dimensional timber as the simplest material the most. The higher the difference was, the higher the inaccuracies could have risen if the product was subsequently evaluated as a part of another system, for example wood-based building. This is the reason why negative emission results could occur according to (MONOKOVA and VILCEKOVA 2019). At the same time, this points out that simple production technology is less harmful to the environment than the production of complicated products containing several raw materials. Fossil emissions were largely influenced by the energy mix used (PARASCHIV and PARASCHIV 2020). That was evident predominantly in HDF production. Also, adhesives created a significant burden on the environment, particularly in the case of MDF manufacture, due to ammonia production by steam reforming. For both, bulk and mass allocation, results indicated the production of HDF to cause the greatest burden in terms of the highest fossil carbon emissions, in general. Carbon uptake caused negative carbon emissions and placed HDF two rank higher, leaving LDF and MDF behind. According to the IPCC uptake method, dimensional timber was the most environmentally beneficial product when mass allocation was applied. Bulk allocation reported glulam the best option. With decreasing amounts of wood per unit weight, the content of sequestrated carbon declined leading to dominance of CO2 emissions into the air. Insufficient identification of carbon emissions and removals treating during calculation could lead to discrepancies. A comparison of the individual studies should be therefore avoided.

149

CONCLUSION Sustainable, renewable and natural characteristics make wood a required construction material. However, its environmental impact is not uniform and depends on a calculation method in terms of life cycle approach. Carbon accounting in LCA is a controversial topic. Diverse calculation methods a create vast space for LCA practitioners to express the environmental impact of wood products. Neglecting of carbon uptake can omit significant positive impacts of wood-based products. It may happen that some studies show wood products as environmentally beneficial in the stage of their production. However, the decomposition of wood returns sequestrated carbon emissions back into the atmosphere, creating a virtual circle. To avoid inconsistencies, carbon emission and capture data should be mentioned for each LCA study. Therefore, the selection of a particular calculation method is an important step in assessment of the whole system. REFERENCES BARE, J. 2011. TRACI 2.0: the tool for the reduction and assessment of chemical and other environmental impacts 2.0. In Clean Technologies and Environmental Policy, 2011, 13: 687696. BRANDÃO, M., LEVASSEUR, A., KIRSCHBAUM, M. U. F., WEIDEMA, B. P., COWIE, A.L., JØRGENSEN, S.V., HAUSCHILD, M. Z., PENNINGTON, D. W., CHOMKHAMSRI, K. 2013. Key issues and options in accounting for carbon sequestration and temporary storage in life cycle assessment and carbon footprinting. In The International Journal of Life Cycle Assessment, 2013/01/01 2013, 18(1): 230–240. CARUS, M. DAMMER., L. 2018. The Circular Bioeconomy—Concepts, Opportunities, and Limitations. In Industrial Biotechnology, 2018, 14: 83–91. COWIE, A. L., BRANDÃO, M., SOIMAKALLIO, S. 2019. 13 - Quantifying the climate effects of forestbased bioenergy. In T.M. LETCHER ed. Managing Global Warming. Academic Press, 2019, p. 399– 418. DJURIC ILIC, D., ERIKSSON, O., ÖDLUND, L., ÅBERG, M. 2018. No zero burden assumption in a circular economy. In Journal of Cleaner Production, 2018/05/01/ 2018, 182: 352–362. EPD INTERNATIONAL AB. 2019. General Programme Instructions for the International EPD System. 2019, 80 p. [2021-11-05] Available at: <https://www.datocms-assets.com/37502/ 1611064110general-programme-instructions-v2-5.pdf>. FAZIO, S., CASTELLANI, V., SALA, S., SCHAU, E., SECCHI, M., ZAMPORI, L., DIACONU, E. 2018. Supporting information to the characterisation factors of recommended EF Life Cycle Impact Assessment methods: New methods and differences with ILCD. Luxembourg: Publications Office of the European Union, 2018. ISBN 978-92-79-76742-5, DOI:10.2760/671368. GARCIA, R., ALVARENGA, R. A. F., HUYSVELD, S., DEWULF, J., ALLACKER, K. 2020. Accounting for biogenic carbon and end-of-life allocation in life cycle assessment of multi-output wood cascade systems. In Journal of Cleaner Production, 2020/12/01/ 2020, vol. 275, 122795. DOI: 10.1016/j.jclepro.2020.122795. GUINÉE, J. B. 2002. Handbook on life cycle assessment: operational guide to the ISO standards. Boston: Kluwer Academic Publishers, 2002. ISBN 1402002289. HAUSCHILD, M., POTTING, J. 2005. Spatial Differentiation in Life Cycle Impact Assessment – The EDIP 2003 Methodology. Institute of Product Development Technical University of Denmark, 2005. HOSSAIN, M. U., POON, C. S. 2018. Comparative LCA of wood waste management strategies generated from building construction activities. In Journal of Cleaner Production, 2018/03/10/ 2018, 177: 387–397. HUIJBREGTS, M. A. J., STEINMANN, Z. J. N., ELSHOUT, P. M. F., STAM, G. VERONES, F., VIEIRA, M., ZIJP, M., HOLLANDER, A., VAN ZELM, R. 2016. ReCiPe2016: a harmonised life cycle impact assessment method at midpoint and endpoint level. In International Journal of Life Cycle Assessment, 2017, 22: 138–147.

150

HÄFLIGER, I.-F., JOHN, V., PASSER, A., LASVAUX, S., HOXHA, E., SAADE, M. R. M., HABERT, G. 2017. Buildings environmental impacts' sensitivity related to LCA modelling choices of construction materials. Journal of Cleaner Production, 2017/07/10/ 2017, 156: 805–816. DOI: 10.1016/j.jclepro.2017.04.052. INTERNATIONAL PANEL ON CLIMATE CHANGE. 2014. Climate Change 2013 – The Physical Science Basis. Eds: STOCKER, T.F., QIN, D., PLATTNER, G.-K., TIGNOR, M., ALLEN, S.K., BOSCHUNG, J., NAUELS, A., XIA, Y., BEX, V., MIDGLEY, P.M. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2014, 1535 p. [2021-11-05] Available at: <https://www.ipcc.ch/report/ar5/wg1/>. ISO 14040:2006, Environmental management. Life cycle assessment - Principles and framework. ISO 14044/A1:2006, Environmental management. Life cycle assessment - Requirements and guidelines. JOINT RESEARCH CENTRE. 2010. International Reference Life Cycle Data System (ILCD) Handbook - General guide for Life Cycle Assessment - Detailed guidance. First edition March 2010. Luxembourg: Publications Office of the European Union, 2010. EUR 24708. [2021-11-05] Available at: <https://eplca.jrc.ec.europa.eu/uploads/ILCD-Handbook-General-guide-for-LCA-DETAILEDGUIDANCE-12March2010-ISBN-fin-v1.0-EN.pdf>. JOLLIET, O., MARGNI, M., CHARLES, R., HUMBERT, S., PAYET, J., REBITZER, G., ROSENBAUM, R. 2003. IMPACT 2002+: a new life cycle assessment methodology. In. The International Journal of Life Cycle Assessment, 2003, 8: 324–330. DOI: 10.1007/BF02978505. LEVASSEUR, A., LESAGE, P., MARGNI, M., BRANDÃO, M., SAMSON, R. 2012. Assessing temporary carbon sequestration and storage projects through land use, land-use change and forestry: comparison of dynamic life cycle assessment with ton-year approaches. In Climatic Change, 2012/12/01 2012, 115(3): 759–776. DOI: 10.1007/s10584-012-0473-x. LIPPIATT, B. C. 2007. BEES 4.0: Building for Environmental and Economic Sustainability. Technical Manual and User Guide. Gaithersburg : NIST Interagency/Internal Report (NISTIR), National Institute of Standards and Technology, 2007. [2021-11-05] Available at: <https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=860108>. MONOKOVA, A., VILCEKOVA, S. 2019. Environmental impact analysis of five family houses in Eastern Slovakia through a life cycle assessment. In Journal of Civil Engineering, 2019, 14: 81–92. DOI: 10.1515/sspjce-2019-0009. MONTZKA, S., DLUGOKENCKY, E., BUTLER, J. 2011. Non-CO2 greenhouse gases and climate change. In Nature, 2011, 476: 43–50. DOI: 10.1038/nature10322. PARASCHIV, S., PARASCHIV, L. S. 2020. Trends of carbon dioxide (CO2) emissions from fossil fuels combustion (coal, gas and oil) in the EU member states from 1960 to 2018. In Energy Reports, 2020/12/01/ 2020, 6: 237–242. PAROBEK, J., PALUŠ, H., MORAVČÍK, M., KOVALČÍK, M., DZIAN, M., MURGAŠ, V., ŠIMO-SVRČEK, S. 2019. Changes in Carbon Balance of Harvested Wood Products Resulting from Different Wood Utilization Scenarios. In Forests, 2019, 10(7). DOI: 10.3390/f10070590. PIEROBON, F., HUANG, M., SIMONEN, K., GANGULY, I. 2019. Environmental benefits of using hybrid CLT structure in midrise non-residential construction: An LCA based comparative case study in the U.S. Pacific Northwest. In Journal of Building Engineering, 2019/11/01/ 2019, 26, 100862. PRÉ CONSULTANTS. 2016. Introduction to LCA with SimaPro. 2016, 80 p. [2021-11-05] Available at: < https://pre-sustainability.com/files/2014/05/SimaPro8IntroductionToLCA.pdf>. RÉH, R.; KRIŠŤÁK, Ľ.; SEDLIAČIK, J.; BEKHTA, P.; BOŽIKOVÁ, M.; KUNECOVÁ, D.; VOZÁROVÁ, V.; TUDOR, E.M.; ANTOV, P.; SAVOV, V. Utilization of Birch Bark as an Eco-Friendly Filler in UreaFormaldehyde Adhesives for Plywood Manufacturing. In Polymers 2021, vol. 13, 511. DOI:10.3390/polym13040511. RODIN, V., LINDORFER, J., BÖHM, H., VIEIRA, L. 2020. Assessing the potential of carbon dioxide valorisation in Europe with focus on biogenic CO2. In Journal of CO2 Utilization, 2020/10/01/ 2020, 41, 101219. RYBNÍČEK, M., KOČÁR, P., MUIGG, B., PEŠKA, J., SEDLÁČEK, R., TEGEL, W., KOLÁŘ, T. 2020. World's oldest dendrochronologically dated archaeological wood construction. In Journal of Archaeological Science, 2020/03/01/ 2020, 115, 105082. DOI: 10.1016/j.jas.2020.105082.

151

SAADE, M. R. M., GUEST, G., AMOR, B. 2020. Comparative whole building LCAs: How far are our expectations from the documented evidence? In Building and Environment, 2020/01/01/ 2020, 167, 106449. SAFARI, K., AZARIJAFARI, H. 2021. Challenges and opportunities for integrating BIM and LCA: Methodological choices and framework development. In Sustainable Cities and Society, 2021/04/01/ 2021, vol. 67, 102728. SARTORI, T., DROGEMULLER, R., OMRANI, S., LAMARI, F. 2021. A schematic framework for Life Cycle Assessment (LCA) and Green Building Rating System (GBRS). In Journal of Building Engineering, 2021/06/01/ 2021, vol. 38, 102180. SKULLESTAD, J. L., BOHNE, R. A., LOHNE, J. 2016. High-rise Timber Buildings as a Climate Change Mitigation Measure – A Comparative LCA of Structural System Alternatives. In Energy Procedia, 2016/09/01/ 2016, 96: 112–123. UNEP. 2020. 2020 Global Status Report for Buildings and Construction: Towards a Zero-emission, Efficient and Resilient Buildings and Construction Sector. Nairobi, 2020. [2021-11-05] Available at: <https://wedocs.unep.org/bitstream/handle/20.500.11822/34572/GSR_ES.pdf?sequence=3&isAllo wed=y>. VANOVA, R.; VLCKO, M.; STEFKO, J. Life Cycle Impact Assessment of Load-Bearing Straw Bale Residential Building. In Materials 2021, vol. 14, 3064. DOI: 10.3390/ma14113064. WERNET, G., BAUER, C., STEUBING, B., REINHARD, J., MORENO-RUIZ, E., WEIDEMA, B. 2016. The ecoinvent database version 3 (part I): overview and methodology. In The International Journal of Life Cycle Assessment, 2016, 21: 1218–1230. DOI: 10.1007/s11367-016-1087-8. WOODARD, A. C., MILNER, H. R. 2016. 7 - Sustainability of timber and wood in construction. In J.M. KHATIB ed. Sustainability of Construction Materials (Second Edition). Woodhead Publishing, 2016, pp. 129–157. ZIEGER, V., LECOMPTE, T., DE MENIBUS, A. H. 2020. Impact of GHGs temporal dynamics on the GWP assessment of building materials: A case study on bio-based and non-bio-based walls. In Building and Environment, 2020/11/01/ 2020, vol. 185, 107210. ACKNOWLEDGEMENTS The paper was supported by the Slovak Research and Development Agency under the contract no. APVV-17-0206 “Ultra-low Energy Green Buildings Based on Wood as a Renewable Construction Material”.

AUTHOR ADDRESS Ing. Rozália Vaňová Technical University in Zvolen T.G. Masaryka 24 96001 Zvolen Slovakia xvanova@is.tuzvo.sk

152

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 153−161, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.13

DENDROCHRONOLOGICAL RESEARCH ON OLD-AGED EASTERN WHITE PINE FROM THE KALUGA REGION – A NATURAL HERITAGE MONUMENT A.V. Cherakshev  D.E. Rumyantsev ABSTRACT The eastern white pine (Pinus strobus L.) is the North American species introduced in Europe in the 18th century. Although this species has high productivity within its native area, the eastern white pine trees have not become widespread in Russia due to their low resistance to blister rust (Cronartium ribicola J.C. Fisch). Mostly, they can be found in parks and botanical gardens. The eastern white pine tree identified in our research in the Dzerzhinsky Forest area in the Kaluga region (on the border with the Ugra National Park) has unique forest inventory parameters. Based on the research results, this tree has been listed in the Register of the Russian National Program “Trees as Natural Heritage Monuments”. Key words: old-aged trees, trees as natural monuments, coniferous introduced trees, Kaluga region, Pinus strobus L.

INTRODUCTION The Russian National Program “Trees as Natural Heritage Monuments” was launched in 2010 by the Council for Conservation of the National Natural Heritage of the Federation Council of the Russian Federal Assembly. The program was developed under the initiative of NPSA ZDOROVY LES and supported by the Federal Forestry Agency (Rosleskhoz) and Moscow State Forest University (UNIQUE RUSSIAN TREES 2019). As of May 2019, overall, 832 online applications from different Russian regions have been registered at the program website and 480 trees have been listed in the National Register of Veteran Trees. 216 trees from the Register were awarded by the certification board as Natural Heritage Monuments. Currently, trees from 76 Russian regions joining the program (out of 85 regions of the Russian Federation including the Republic of Crimea and the city of Sevastopol) are included in the National Register. The Register covers trees of about 43 species. The discussion to form the register started with the question concerning tree species with a natural area outside Russia. Can individual trees of such kinds of species be classified as the objects of cultural, natural and historical heritage? The affirmative opinion is because of many reasons. These trees are the result of artificial planting and are usually associated, in historical memory, with the activities of popular public opinion leaders of their time. Oldgrowth trees of exotic species in this territory, with high aesthetic properties, are interesting examples of silvicultural experiments. 153

The introduction of tree species exotic in Russia was a common and widely used feature of landscape architecture in the 19th and early 20th centuries. The Eastern White Pine imported to Europe by Lord Weymouth in 1705 became one of such introduced species. Following the landscaping trend of that period, a pine tree of the above-mentioned species was planted in the park on the Rasskazovo Estate (Arzhenka) owned by the Aseevs, the family of merchants and manufacturers. Formerly, alien pine species such as the eastern white pine, the Balkan pine and the black pine were largely introduced to many regions of the country (SUKACHOV 1934, TKACHENKO 1939, KAPPER 1954, JABLOKOV 1962, LAPIN 1975). Today, they can be mostly found in various parks, botanic gardens and even, in rare case, in forests. According to A. S. JABLOKOV (1962), the founder of the School for Selective Breeding and Reproduction of Forest Species, “If trees of especially high quality that meet the criteria for selection of good or best-of-normal units can be found among the planted trees, such trees should be registered by the scientific institutes, forest management bodies and landscape agencies, listed as protected trees and natural monuments and used for reproduction purposes to get highquality seeds”.

RESEARCH SUBJECT The first eastern white pine from the Tambov region was listed in the National Register under No 612 on March 23, 2017. This was the tree growing in the park of the Aseevs Estate in the town of Rasskazovo (UNIQUE RUSSIAN TREES 2019). In May of the same year, the team of experts from the Center for Wood Analyses (the leading expert of which was Andrey Cherakshev) travelled to the estate to perform a detailed inspection of the pine tree as a part of the activities under the program. As a result, the experts identified the main inventory parameters of the tree. Its height was 22 m and the trunk diameter was 78 cm. The tree did not have any signs of weakness or infestation with pests or diseases and was of higher aesthetic appeal in comparison with the local native tree species. Based on the age analysis of the wood samples, its age was identified to be 117 years old as of 2019. The obtained result confirms that the tree was planted in the early 20th century when the eastern white pine was prevalent as an introduced species in landscape architecture. Another eastern white pine was found during the research at one of our study sites in the Ozyorensky forest district of the Dzerzhinsky forest area in the Kaluga region. As far as the forest inventory parameters are concerned, this tree was larger and of higher aesthetic appeal than the pine described above. Its height was 32 m and its diameter was 90 cm as measured at the height of 1.3 m. Growing at the forest stand of the same species, the tree stood out of the neighbouring trees due to its great dimensions and excellent health and therefore it was decided to apply for this tree to join the Russian National Program “Trees as Natural Heritage Monuments” (UNIQUE RUSSIAN TREES 2019). The tree was included in the National Register of Veteran Trees under No 795. The certification board voted for this pine to become a Natural Heritage Monument Tree at their annual general meeting on April 18, 2019. The photo of the tree is shown in Figure 1.

154

Fig. 1 The white eastern pine from the Ozyorensky forest district of the Dzerzhinsky forest area, the Kaluga region, classified as a Natural Heritage Monument Tree (under registered No 795).

MATERIALS AND METHODS Methods of core preparing and ring width measuring were described in previous works (RUMYANTSEV 2010, RUMYANTSEV  CHERAKSHEV 2013). The Pressler borer was used for taking a core, Lintab-5 was used for measuring tree ring widths, and the Tsap-Win program was used to control the measurements by the method of cross dating chronologies. The age of the trees was identified based on the original method developed by A.V. Cherakshev (CHERAKSHEV et al. 2015, RUMYANTSEV  CHERAKSHEV 2020). This method ensures that tree age is accurately determined even in the cases when the drill does not come through the cross-section center but passes chordwise against the pith. Since it is practically impossible to get the auger precisely into the trunk center especially while inspecting trees of a significant diameter, a certain number of tree rings will fall out of the calculations, which results in obtaining a lower tree age. There are the age calculations of both eastern white pine trees described step-by-step. The parameters of the core samples collected from the eastern white pine trees are shown in Table 1. Tab. 1 Parameters of the core samples. Tree number according to the Register/Core sample number

Height of the core sampling, m

№612-1 №612-2 №795-1

1.15 1.1 1.3

155

Circumference at the height of core sampling, cm 246 247 281

Directions of the core sampling East – West North – South North – South

The following calculation steps were completed (based on the original method) to identify the tree age. 1. Determining the length of the wood section unavailable for study. If, when drilling, the auger does not come along the radius of the circle with the center at the pith but passes chordwise, it is required to measure the chord length b and the segment height a of the core sample (Figure 2). Consequently, an isosceles triangle was made and a circle around it which will be of an assumed extension of the tree ring on the core sample was drawn (Figure 3). The radius of the circle drawn around the isosceles triangle is calculated as follows: 𝐿2 =

𝑐2

(1)

√(2𝑐)2 − 𝑏 2

1 (2) 𝑐 2 = 𝑎2 + 𝑏 2 , 2 where L2 is a circle radius (length of the wood section unavailable for study), а is a segment height, and b is a chord length (Figure 3). In this case, the circle radius is the length of the wood section unavailable for study. Core sample

Segment height

Chord Fig. 2 The wood section and the chord represented on the core sample.

Fig. 3 The isosceles triangle and the circle drawn around it representing the extension of the tree ring.

2. Calculating the average width of the tree rings in the wood section unavailable for study: (3) 𝑀𝑥 = (𝑥1 + 𝑥2 + 𝑥3 + 𝑥4 + 𝑥5 )/5 , where Mх is the average tree ring width, х1…х5 are the widths of the last 5 tree rings at the closest proximity to the missing wood section. 3. Finding out the number of tree rings in the wood section unavailable for study: 𝐴1 = 𝐿2 /𝑀𝑥 ,

(4)

where А1 is an estimated number of the tree rings in the unavailable wood section, L2 is a circle 156

radius (length of the wood section unavailable for study), Mх is an average tree ring width. 4. Determining the tree age based on an individual radius at the height of core sampling: (5) 𝐴2 = 𝐴0 + 𝐴1 , where А2 is the tree age at the height of core sampling, А0 is a number of the tree rings identified in the core sample, А1 is an estimated number of the tree rings in the wood section unavailable for study. 5. Calculating the number of years required for a young tree to reach the height of core sampling: (6) 𝐴3 = 𝐻/𝑏 , where А3 is the age required for a tree to reach the height necessary to collect a core sample, H is the height of core sampling, b is a conditional linear growth. The conditional linear growth rate is in a positive correlation with biological and geographic conditions of the treeunder-study growing site and, as a rule, might be from 10 cm to 30 cm a year. 6. Eventually, we can calculate the biological tree age: (7)

𝐴4 = 𝐴2 + 𝐴3 ,

Similar to the investigated, the largest tree from the Kaluga region there are other trees of this species forming the linear tree planting and this confirmed the artificial nature of the planting origin. The cores of wood from 10 trees (including the largest) were taken, tree rings were measured and the chronology for eastern white pine stand was built and used for dendroclimatic analysis.

RESULTS AND DISCUSSION The age of the Natural Heritage Monument Trees described was calculated based on the data presented in Tables 1 and 2. The results are given in Table 3. Tab. 2 Input data for tree age calculations. Tree number Chord length Segment according to the b, height а, Register/Core mm mm sample number No 612-1 9.5 0.5 No 612-2 16.5 1.5 No 795-1 37 3.5

Average width of the Wood section Estimated number of tree 5-10 tree rings closest unavailable for rings in the wood section to the unavailable study, mm unavailable for study. section, mm 22.8 2.242 10 23.4 2.304 10 50.64 7.8 6

Tab. 3 Tree age calculations. Estimated number Conditional Age when the tree Number of tree of tree rings in the Core sample linear growth in reaches the height rings on the wood section number the early years of core sampling, core sample unavailable for of tree life, cm in years. study. No 612-1 99 10 20 6 No 612-2 99 10 20 6 No 795-1 84 6 20 7 Total age of Tree No 612 as of 2017: Total age of Tree No 795 as of 2018:

157

Total age, in years 115 115 97 115 97

Since the core samples were collected in different years, a certain number of years have to be added to the estimated age in order to get the current age of the trees. Thus, as of today, the age of the eastern white pine from the Tambov region is 119 years and of the pine from the Kaluga region is 101 years. Despite the insignificant age difference, the pine tree from the Kaluga region is greater than the pine tree from the Tambov region as far as their forest inventory parameters are concerned. The average chronology for tree ring width dynamics by the years for white eastern pine planting from the Kaluga region are shown in Fig. 4. The local minimum values and the local maximum values of this time series give information about years with favorable tree growth environmental conditions, and, on the contrary, about unfavorable ones.

average tree ring width, mm

6 5 4 3 2 1

1915 1919 1923 1927 1931 1935 1939 1943 1947 1951 1955 1959 1963 1967 1971 1975 1979 1983 1987 1991 1995 1999 2003 2007 2011 2015

year of tree ring forming Fig. 4 Average annual ring width values for the calendar years.

The analysis of the dynamics of radial growth shows that clear extremely wide annual rings were formed in 1937, 1962, 1982. Clear extremely narrow annual rings were formed in 1944, 1956, 1964, 1977, 2002, 2014. It may be related to phytocenotic events, or the effect of climatic factors. The correlation coefficient calculation between indexed average chronology and time series of climatic data was conducted for more detailed analysis. Each chronology was indexed by the division of ring width value on average ring value for five previous years. Indexing is needed for removing the effects associated with the presence of the age trends in the time series of radial growth. The results of calculations are shown in Fig. 5 and Fig.6. The significant values of the coefficients for 0.05 level are 0.27 and more. The calculations were conducted during the years 1966 2017.

158

0.4 0.3 correlation coefficent

0.2 0.1 0 -0.1 -0.2 -0.3 -0.4 Precipitation

Temperature

Fig. 5 The influence of climatic data values on tree ring formed during a calendar year of tree ring forming.

0.5

correlation coefficient

0.4 0.3 0.2 0.1 0 -0.1 -0.2 -0.3 -0.4 Precipitation

Temperature

Fig. 6 The influence of climatic data values on tree ring formed during a calendar year previous to tree ring forming.

The results of correlation analysis can be used to explain significant correlation coefficients observed for May (T5), October (T10) and December (T12) temperatures of the year previous to the year of tree ring forming. Also, the sum of rainfall in April (P4), May (P5) and August (P8) during the calendar year of tree ring forming are significant for total width of the annual ring. Such results make it possible to model the dynamics of the radial growth indices using the linear regression equation. As a result of the calculations, an equation of the following form was obtained: Y = 1.0889  0.0115 × Т5 + 0.0076 × Т12 + 0.0013 × P4  0.0021 × P5 + 0.0009 × P8 + 0.0365 × Т10 + 0.0135 × Т12. 159

The model is characterized by 51% determination coefficient. The graph in Figure 7 has a good synchroneity with real values and it is an even better indicator at a qualitative level. 1.6 1.4

radial growth index

1.2 1 0.8 0.6 0.4 0.2

1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

real value

model

Fig. 7 The comparison of results of radial growth indices modelling.

The results of an investigation differ from the results of other such investigations of this species (RUMYANTSEV, CHERAKSHEV 2013, CHIN et al. 2013, CHIN et al. 2018). It shows the role and importance of the dendrochronological method in identifying climatic factors limiting the cambial activity of a particular tree species. In the years with critical values of meteo-parameters, the monitoring of the health condition of the trees is needed.

CONCLUSION For the studied tree, the biological age was determined and climatic factors significant for its growth were identified. The short-term variability of radial increment fluctuations was modeled by a linear regression equation with a 51% determination coefficient. There are no doubts that the described trees are not the only two old-aged eastern white pines growing in the Russian Federation. Therefore, the research into veteran trees of the eastern white pine species and the Macedonian pine species throughout Russia will continue to list them in the National Register under the Trees as Natural Heritage Monuments Program because these introduced species are of a high natural, scientific, historical and cultural value. REFERENCES CHERAKSHEV, A.V., PALCHIKOV, S.B., ANCIFEROV, A.V. 2015. Determining the tree age using the LINTABTM scientific complex // Modern Scientific Potential, In Proc. 11th International Scientific and Practical Conference, 2015, pp. 42–45. (in Russian)

160

CHIN, S., CHUMACK, K., DAHL, T., DAVID, E.I. 2013. Growth-climate relationships of Pinus strobus in the floodway versus terrace forest along the banks of Red Cedar River, Michigan. In Tree ring research. 69(2): 3747. CHIN, S., ZALESNY, R.S., PARKER, W.C., BRISSETE, J. 2018. Dendroclimatic analysis of white pine (Pinus strobus L.) using long-term provenance test sites across eastern North America. In Forest Ecosystems, 5(18): 115. JABLOKOV, A.S. 1962. Selective breeding of wood species. Moscow: Selhozizdat, 1962, 487 p. (in Russian) SUKACHOV, V.N. 1934. Dendrology with the basics of forest geobotanics, Leningrad: Goslestehizdat, 1934, 614 p. (in Russian) KAPPER, O.G. 1954. Coniferous species, Moscow - Leningrad: Goslesbumizdat, 1954, 304 p. (in Russian) LAPIN, P.I. 1975. Aleksandrova M.S., Borodina N.A., Woody plants of the Main Botanic Garden of the USSR Academy of Sciences, Moscow: Nauka, 1975, 547 p. (in Russian) RUMYANTSEV, D.E. 2010. History and methodology of forestry dendrochronology, Moscow: MGUL, 2010, 109 p. (in Russian) RUMYANTSEV, D.E., CHERAKSHEV, A.V. 2013. Dendroclimatic diagnostics of STROBI pine health in the conditions of the MGUL Dendrological Garden // Forestry Bulletin, 2013, 7(99): 121127. (in Russian) RUMYANTSEV, D.E., CHERAKSHEV, A.V. 2020. Methodological approach to determine the tree age. In Ecology principals, 2020, (38): 104–117. (in Russian) TKACHENKO, M.E. 1939. General Forestry, Leningrad: Goslestehizdat, 1939, 746 p. (in Russian) UNIQUE RUSSIAN TREES 2019. Vol. 1, Moscow, 2019, 208 p. (in Russian)

AUTHOR’S ADDRESS A.V. Cherakshev Mytischi Branch of Bauman Moscow State Technical University Moscow Russia D.E. Rumyantsev Mytischi Branch of Bauman Moscow State Technical University Moscow Russia All-Russian Research Institute of Silviculture and Mechanization of Forestry Mytischi Moscow region dendro15@list.ru

161

162

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 163−173, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.14

MAPPING THE WOOD COLOUR PREFERENCES AMONG POTENTIAL CUSTOMERS Mariana Sedliačiková – Patrik Aláč – Mária Moresová – Ivan Sedliačik ABSTRACT The knowledge of the requirements and specifications of the current, as well as potential customers, is a necessary condition for the successful operation of enterprises in the market. The goal of this paper is to map and monitor the preference of Slovak customers for colour of wood products. The mapping of the interest in the wood colour among potential customers was carried out using the interrogating method in the form of a questionnaire. The research results were evaluated descriptively and graphically. Respondents stated that they were most affected by quality when buying wood products. Another result is that the respondents focus on domestic rather than tropical woody plants. Although tropical woods are becoming popular since they have a very attractive colour. Most respondents, specifically 80%, focus on the natural colour of wood. Respondents prefer mainly brown matt colours with a distinctive grain, followed by white and grey colours. A smaller number of respondents choose black and yellow colours of wood. Therefore, it is necessary to increase the supply of wood and furniture products with the demanded colours, it means the natural wood colours, grey, white and brown colours. The results of the research represent an opportunity for Slovak wood-processing and furniture manufacturing enterprises to adjust and flexibly react by the range of their products according to the needs and preference of potential customers which can bring them higher revenues and also help them overcome the current problems related to “COVID crisis”. Key words: wood colours, wood steaming, empirical research, customers` preferences.

INTRODUCTION One of the hardest tasks of an entrepreneur is to meet the requirements of the market which, in other words, means managing supply and demand. How to manage supply and demand is a question that leaves even the best business owners and operation managers flummoxed. Even though there are different methods for estimating preference through forecasts and increasing supply through manufacturing, the matching of supply and demand is never reliable or predictable. In these circumstances, a manager has to work with the best estimation that he has. Thus, there is a need for managing supply and demand. Meeting the needs and preferences of the customer while making a profit is the key to business success, and the process has become both harder and easier. It is easier in that modern technology makes it easier for businesses to identify, learn about and reach their market, but harder in that customers are becoming more discerning, less loyal and have increasingly high expectations. 163

If looking at the demand, there are several methods through which demand can be anticipated by both qualitative and quantitative methods. However, there are several instances where demand is not matching the supply. In some cases, there is an increased supply whereas preference is low. Thus, these instances require methods that push up the demand. The most basic one involves establishing price incentives (FERRI and TRAMONTANA 2020). The current authors previously discussed penetration pricing and break-even pricing. What these two models have in common is mainly the purpose of gaining fast market share. These two models can be an example of price incentives. But providing price incentives can be risky due to the customers’ perceptions regarding the quality of the product. Another method of pushing up demand is to provide complementary services. And here is where the family life cycle concept can be considered a relevant technique. KIRMAN et al. (2007) stated that by conducting a proper segmentation and taking into consideration the consumers from the different stages of the family life cycle, we can anticipate what the consumers need, besides what they have already been provided. Once the offered product/service accomplishes more of the needs of the customers, the demand is going to grow. According to CLAVERIA et al. (2020), some of the well-known strategies for pushing up the supply consists of using part-time employees, increasing customer participation, cross-training employees and scheduling work shifts. However, one of the major risks in increasing supply is the effect it has on the quality of the product. This is why the operations should be scalable. If in case, the supply is regularly not matching demand, then it is necessary to invest in a new factory or new methods to increase production. This is commonly known as the bottleneck effect. While the demand increases, but the supply doesn’t happen, a bottleneck is created which evokes stress at the customer. And the customer then easily shifts to another brand. Thus, managing supply and demand becomes critical. Customer behaviour is a set of patterns of customer thoughts and actions that are relevant to marketing in areas such as product design, pricing, promotion, customers experience and sales. Some examples of customer behaviour can be mentioned in the following. Needs represent the process by which customers decide they need something. Motivation is a customer´s fundamental drive. A customer uses the process of search to discover and find products (SHIOZAWA 2020). After that, purchasing decision comes when a customer decides which product to buy. Why customers stick to a product so that they make repeat purchases and why they leave from a particular producer responds to the socalled customer loyalty. Almost every customer has some level of price sensitivity, which means how he/she feels about prices in a particular product category. Then it results in customer´s perception of things such as brand, quality, reputation, value and risk. After the product utilization by a customer, it comes the phase of post-purchase evaluation of a particular product where a customer can share information with others in various product reviews, complaints and recommendations of further and repeated purchase. As presented by PUAH et al. (2017), meeting the customers' preference is a key performance to meet the strategic goal of an enterprise to reach a profit or to increase market share. Isolating the segment of the market appropriate for the product offering and knowing what type of person inhabits is the key to developing an in-demand product. Much money and time have been spent in the past by organizations trying to sell goods to a market before they have properly understood whether there was a serious need for it. Much data is gathered from current and potential customers now due to constant data capture through apps, websites and social networks so it is easier for a business to analyse who their customers are and what they want. Businesses need to understand their customer´s expectations and always 164

strive to meet or even exceed them. Generating preference for the product requires much more than simply releasing it onto the market. It is necessary to conduct research, determine what consumers´ needs are, establish an enterprise as a leader in the industry and repeatedly prove the product value. Wood colour is an important material property that has an influence on customers´ decision-making during the purchasing process. It seems that wood colour belongs among the most decisive purchasing criteria. (KADLEČEK 1989). The surface of the wood can be with its natural colour or it can be improved by painting. There are also external features that affect the change of wood colour like moisture content, humidity, UV radiation, thermal conditions etc. The Colour of wood differs not only according to various wood species but also according to various parts in the wood trunk. The most significant colour difference is between sapwood and heartwood (KUBOVSKÝ and KAČÍK 2010). By the steaming process (under demanded specific conditions - time, temperature pressure) of wood, colour can be unified or changed into the colour of tropical wood species. Steaming of wood also supports dimensional stability, wood softening and plasticizing (DZURENDA et al. 2020). According to the research of NGUYEN et al. (2020), the colour of wood is an important material property and the first sight that a customer could perceive. So, it can highly impact customer´s decision making. Wood processing and furniture making enterprises should meet various and specific demands of contemporary customers but at the same time, they have to focus on the demands of new potential customers. In order to satisfy market preference, enterprises have to focus on their good reputation, quality of their products, well-managed processes within the supply chain and material flow. At the same time, it is also necessary to take into account the effectiveness and efficiency of all these above-mentioned processes. The focus should be placed also on the latest trends and according to them to diversify the portfolio. According to STOJANOVIC et al. (2020), BUCKINGHAM et al. (2009), the market is typical by constant and running changes and enterprises should meet this variable demand. Competitiveness among enterprises depends also on technology equipment, on quality and skills of their employees and finally also on enterprise business policy. Related to the above mentioned, it is necessary to analyse customers´ preferences. The supply of wood products could be in natural colours of native wood. Much more trend is to adjust surface by finishing processes like penetrating, painting and bleaching (ANDAC 2020). The goal of this paper is to determine principal customers´ preferences of the enterprise supply. The authors analysed, customers preferences for various wood colours. According to the above mentioned assessment, for the business success, a diversified offer for various market segments should be prepared and flexible reactions on changing preferences of demand shall be performed. The specification and structure of the most demanded wood colours are the findings of the research.

METHODOLOGY The research methodology consisted of several steps. At the beginning, it was necessary to carry out a literary review of domestic and especially foreign authors based on the analysis of secondary sources. Then it was necessary to analyse the primary sources obtained by the empirical research, the method of questioning. The questionnaire was focused on mapping the preference of potential customers for colours of wood. Selected file of respondents was chosen on the base of random stratified choice (KOZEL 2006). Addressing and collecting data from respondents took place electronically, mainly due to 165

fast feedback and low complexity. The data of the questionnaire survey were evaluated descriptively, numerically and graphically. In the next step, the achieved results from the descriptive and graphic processing of the questionnaire survey, which was carried out on a research sample, i.e. potential customers in Slovakia, were evaluated. From the obtained data, information from customers was processed, which can help wood-processing enterprises in deciding what to focus on in the production of wood products, respectively, what customers are more interested in when buying, or how they should expand their offer. At this stage, the analysis of primary and secondary sources was done and the method of summarization were used. The questionnaire consisted of two parts, with a total of 20 questions. The first part of the questionnaire contained classification questions, which contained the basic characteristics of the respondents. The classification questions verified the gender, residence, age of the respondents, the region in which they live, their highest level of education as well as their current employment and their monthly income. The second part of the questionnaire consisted of questions to map the interest in the colours of wood. The intention was to find out which of the mentioned factors most influence the purchase of wood products, if the respondents could not choose anything from the given factors, they had the opportunity to express their opinion, or to mark the box “other”. The questions were also aimed at whether customers care about the type of woody plant when buying, or whether they prefer solid, chipboard, fibreboard or veneer wood. Other questions concerned the type of woody plants, what woody plants they buy most often, and what colour they prefer. When asked about the type of wood, respondents could mark more than one answer. They had a choice of 12 woody plants or an opportunity to express their opinion in the box “other”, if their woody plant was not in the offer. The authors also verified whether the respondents like more gloss or matt wood, wood with a distinctive grain or without a visible grain. The size of the examined sample was determined according to the following mathematical relation, which is intended for the calculation of the minimum number of respondents (KOZEL 2006): 𝑛≥

𝑧 2 × 𝑝 × (1 − 𝑝)

(1)

𝑐2

where: n – minimum number of respondents; z – reliability coefficient (at confidence level of 95% the variable z = 1.96); p – the proportion of the character (for unknown values it is substituted for p = 0.5); c – acceptable margin of error (significance level was set at 5%). Using the formula above, it was possible to calculate the minimum sample size for the survey (KOZEL 2006): 𝑛≥

1.962 × 0.5 × ( 1 − 0.5) 0.052

→

𝑛 ≥ 384

The calculation shows that the sample must consist of at least 384 respondents. The survey meets this condition since 570 respondents took part in the survey.

RESULTS AND DISCUSSION Up to 72.63% of women and 27.37% of men participated in the survey and all age groups were represented. However, 53.33% of respondents live in the town and 46.67% in the village. Other basic characteristics of the respondents are presented in follow table.

166

Tab. 1 Characteristics of the respondents. Questions Age of the respondent Permanent resident of the respondent

018 years old 1.93% 4655 years old 4.91% Bratislava region 9.12% Žilina region 11.93% primary education

Education

2.11% Employment of the respondent

Monthly income of the respondents

student 59.65% entrepreneur 1.93% max. 400 € 52.81% 1,001-1,200 € 6.14%

Answers 1925 years old 2635 years old 3645 years old 66.67% 19.12% 5.44% 66 and more years old 5665 years old 1.23% 0.70% Trnava region Trenčín region Nitra region 6.49% 8.95% 7.89% Banská Bystrica Prešov region Košice region region 36.67% 12.46% 6.49% secondary secondary education without education with university degree GCSE GCSE 3.51% 46.14% 48,24% an employee in the employee in the sole trader private sector public sector 25.44% 9.65% 1.05% unemployed retired other 1.40% 0.88% 0% 401600 € 601800 € 8011,000 € 9.65% 12.63% 12.28% 1,201 € and more 6.49% -

Source: authors.

Figure 1 shows that most respondents are affected by the quality of the products when buying wood products, 42.28% respondents think so. These results also correspond to the results of the authors GOLD and FUBIK (2009), VEISTEN (2002). The second factor that influences purchasing decisions according to the respondents is the price of products with a share of 26.67%. Furthermore, 19.12% of respondents stated that their purchase is influenced by the colour of the wood and 10.35% chose the type of woody plant. Respondents also had a choice of product gloss, but none of the respondents marked this option. 1.58% of the total number of respondents marked the option “others”. For example, 0.70 % of respondents stated that the appearance of wood influences them at the purchase and 0.35 % marked a combination of all offered options. The achieved results are supported by the research of several authors like DZURENDA and DUDIAK (2020), BAUMGARTNER et al. (2013) and DZURENDA (2014). 50% 0%

26.67%

product price

42.28%

10.35%

19.12%

product quality woody plant, from colour tone of which it is made wood

1.58%

other

Fig. 1 Factors influencing the purchase of wood products. Source: authors

The gender and residence of the respondents is compared based on the percentage of answers (Figure 2).

167

69.08% 30.92% 49.34% 50.66%

68.05% 31.95% 43.98% 56.02%

71.19% 28.81% 44.07% 55.93%

86.24% 13.76% 49.54% 50.46%

100% 0.00% 55.56% 44.44%

female male village town

100% 80% 60% 40% 20% 0%

product price product quality woody plant, colour of wood from which it is made

other

Fig. 2 Comparison of selected classification questions. Source: authors

5.44%

10.18%

2.98%

2.11%

47.02%

12.98%

20.53%

6.49%

2.98%

67.19%

11.05%

40.18%

80% 60% 40% 20% 0%

34.21%

Figure 2 shows that up to 100% of women expressed the option “other” and stated the above-mentioned options, while no man (0%) marked this option. The second largest proportion of women, – represented by 86.24%, indicated the option of the colour of the wood, while men accounted for 13.76%. Based on this, the colour of wood affects women the most when buying wood products, and men are most affected by the quality of products according to the Figure 2, with a share of 31.95%. Regarding the comparison of residence, most respondents from the town, 56.02%, chose the option of product quality. The respondents from the village reached the smallest percentage, namely 43.98% at this question. It is possible to state and it is also confirmed by the results of the authors KAUSTIA et al. (2008), SCHEER et al. (2020), ANTOV et al. (2020a), ANTOV et al. (2020b), ANTOV et al. (2021), that respondents from the town prefer the quality of products more than respondents living in the village. The evaluation of woody plants that most respondents buy is shown in Figure 3. The authors NYRUD et al. (2008), DZURENDA (2018) also evaluated these dependencies in their works and achieved similar results.

Fig. 3 The most frequently purchased woody plants. Source: authors.

Figure 3 shows the most frequently purchased woody plants. Respondents chose oak as the most frequently purchased wood species, which corresponds to 67.19%. The second most purchased wood species is Beech with a percentage of 47.02%, the third most preferred wood is cherry with 40.18% and finally spruce with 34.21%. This is followed by pine, which is purchased by 20.53% of respondents, mahogany with 12.98%, maple with 11.05% and alder with 10.18%. 6.49% of respondents buy birch, rosewood and teak had the same percentage, namely 2.98%. This is followed by acacia, which accounted for 2.11%. Similar results are confirmed and presented in the work of NORDVIK et al. (2010). Subsequently, the gender and residence of the respondents were compared through classification questions in order to find out which woody plant was the most popular among women and men, and then which trees are preferred by the respondents living in the town and village. SEDLIACIKOVA et al. (2020) also dealt with similar sociological factors. As it was possible to choose more options, a summary Figure 4 presents the results. 168

10%

20%

30%

40%

50%

60%

70%

80%

spruce

34.87%

65.13% 50.77% 49.23%

cherry

21.83% 45.41%

maple

26.98%

78.17%

54.59% 73.02%

52.38% 47.62%

rosewood birch pine mahagony beech acacia

Fig. 4 Comparison of selected classification questions. Source: authors

oak

25.85%

74.15%

51.44% 48.56%

17.65%

90%

teak alder other

female male village town female male village town female male village town female male village town female male village town female male village town female male village town female male village town female male village town female male village town female male village town female male village town female male village town

82.35%

52.94% 47.06%

18.92%

81.08% 62.16%

37.84%

64.96%

35.04% 43.59% 22.97% 43.24% 27.99%

56.41% 77.03% 56.76% 72.01%

51.49% 48.51%

66.67%

33.33%

66.67%

33.33%

76.47%

23.53% 41.18% 24.14%

58.82% 75.86%

53.45% 46.55%

29.03%

45.16%

70.97%

54.84%

Figure 4 presents the results, how is each woody plant preferred concerning gender and residence. Women most prefer rosewood, accounting for 82.35%, followed by birch with 81.08% and cherry with 78.17%. For men, pine was the most preferred, accounting for 35.04%, followed by spruce with 34.87%. The least preferred tree species for women was pine with 64.96% and for men, it was rosewood with 17.65%. Respondents who live in the town most often marked acacia, which reached a share of 66.67% and mahogany with a share of 56.76%. On the contrary, among the respondents living in the village, the preferred woods were spruce with 49.23% and beech with 48.51%. Even though the box “other” presented 54.84%, it included various woody plants. The least percentage in the town was reached in the box “other” and in the case of a specific woody plant, spruce was most represented, with 50.77%. Concerning village, acacia had the least percentage of answers, 33.33%. Also, authors like AXELSSON et al. (2007) and BECKER et al. (2005) claim that social inequalities and quality of life significantly influence the choice of various products. One of the most important questions in the second part of the questionnaire was the question of which colours the respondents prefer. CROSON and TREICH (2014) also examined these correlations in their research. There was a choice of five basic colours and of course the box “other” to express their preferred colour (Figure 5). 100% 0%

79.47%

43.51% white colour

13.16% yellow colour

brown colour

34.04%

23.51%

grey colour

black colour

1.23% other

Fig. 5 Preferred colours. Source: authors.

female male village town female male village town female male village town female male village town female male village town female male village town

80% 60% 40% 20% 0%

76.21% 23.79% 49.60% 50.40% 57.33% 42.67% 54.67% 45.33% 73.07% 26.93% 53.64% 46.36% 77.84% 22.16% 51.55% 48.45% 76.12% 23.88% 57.46% 42.54% 71.43% 28.57% 42.86% 57.14%

Figure 5 shows that the most preferred colour of wood according to the respondents was brown, confirmed by up to 79.47%. Respondents ranked white as the second most popular colour with a percentage of 43.51%, followed by a grey colour, which accounted for 34.04%. With a share of 23.51%, respondents focused on black colour of wood and 13.16% respondents prefer yellow colour. It can be seen that 1.23% of the respondents complemented their opinion in the box “other”. NIJDAM (2009) also presents similar results. The preferences concerning colours were studied in more detail. The results are shown in Figure 6.

white colour

yellow colour

brown colour grey colour black colour

other

Fig. 6 Comparison of selected classification questions. Source: authors.

Figure 6 shows that women prefer mainly grey colour, which represents the highest percentage of 77.84 % and men prefer yellow colour, namely 42.67%. Respondents living in the town opt for the black colour of the wood, which accounts for 57.46%. Respondents 170

from the village reached the biggest percentage share in the box “other” and regarding the specific colour, they prefer white colour with a share of 50.40%. Figure 6 shows that respondents focus on all offered colour shades, but women at least on yellow shades and men on grey shades.

CONCLUSION In the empirical survey, the research sample consisted of 570 respondents living in Slovakia, the aim of which was to map their interest in the colours of wood. The results showed that the quality and type of woody plant are very important for the respondents, even more than the price of these products. Within the comparison of classification questions, interesting results have been obtained. Women are most influenced by the colour of the wood, and this answer reached the highest percentage, namely 86.24%. On the other hand, men especially prefer the quality of wood products. It results that the quality of the products and the type of wood are essential for men, the quality also for women but they do not focus so much on the type of woody plant as on its colour. Respondents prefer mostly oak, followed by woody plants such as beech and cherry. Concerning tropical woods, respondents preferred mostly mahogany. Regarding colours, most respondents, up to 80% prefer the natural colour of the wood. Out of the modified colours of wood, the greatest preference is for a brown colour, followed by white and grey colour. There is less preference for black and yellow colours of wood. Again, the results were compared by gender, where it was found that women prefer grey, white and black colours of wood. For men, the order is yellow, brown and black colours of wood. The main goal of the paper was to map the preference for colours of wood at the potential customers in Slovakia. The results of the research can serve as a basis for companies operating in the wood-processing industry. It is important for the success of wood-processing enterprises that they take into account, first and foremost, the diverse and specific requirements, both of loyal and regular customers, but especially of potential customers. REFERENCES ANDAC, G.T. 2020. Consumer attitudes toward preference and use of wood, woodenware, and furniture: A sample from kayseri, Turkey. In BioResources 15: 2837. ANTOV, P., MANTANIS, G., SAVOV, V. 2020a. Development of wood composites from recycled fibres bonded with magnesium lignosulfonate. In Forests 11, 613. https://doi.org/10.3390/f11060613 ANTOV, P., SAVOV, V., NEYKOV, N. 2020b. Sustainable bio-based adhesives for eco-friendly wood composites a review. In Wood Research, 65: 5162. https://doi/10.37763/wr.1336-4561/65.1.051062 ANTOV, P., SAVOV, V., KRISTAK, L., RÉH, R., MANTANIS, G.I. 2021. Eco-friendly, high-density fiberboards bonded with urea-formaldehyde and ammonium lignosulfonate. In Polymers, 13: 220. https://doi/10.3390/polym13020220 AXELSSON, L.H., ANDERSSON, I., EDÉN, L., EJLERTSSON, G. 2007. Inequalities of quality of life in unemployed young adults: A population-based questionnaires study. In International Journal for Equity in Health, 6, 19. https://doi.org/10.1186/1475-9276-6-1 BAUMGARTNER, E., WIEBE, C.B., GEGENGURTNER, K. R. 2013. Visual and haptic representations of material properties. In Multisensory Research 26, 429455. https://doi/10.1163/22134808-00002429 BECKER, G.S., PHILIPSON, T.J., SOARES, R.R. 2005. The quantity and quality of life and the evolution of world inequality. In American Economic Review 95, 277291. DOI: 10.1257/0002828053828563

171

BUCKINGHAM, G., CANT, J.S., GOODALE, M.A. 2009. Living in a material world: How visual cues to material properties affect the way that we lift objects and perceive their weight. In Journal of Neurophysiology, 102: 31113118. https://doi/10.1152/jn.00515.2009 CLAVERIA, O., MONTE, E., TORRA, S. 2020. Economic forecasting with evolved confidence indicators. In Economic modelling 93, 576585. https://doi.org/10.1016/j.econmod.2020.09.015 CROSON, R., TREICH, N., 2014. Behavioral environmental economics: Promises and challenges. Environmental and Resource Economics 58, 335351. https://doi.org/10.1007/s10640-014-9783-y DZURENDA, L. 2014. Colouring of beech wood during thermal treatment using saturated water steams. In Acta Facultatis Xylologiae Zvolen, 56: 1322. DZURENDA, L. 2018. Colour modification of Robinia Ppseudoaccacia L. during the processes of heat treatment with saturated water steam. In Acta Facultatis Xylologiae Zvolen, 60: 6170. DOI: 10.17423/afx.2018.60.1.07 DZURENDA, L. 2018. Hues of acer Platanoides L. resulting form processes of thermal treatment with saturated steam. In Drewno, 61: 165176. DZURENDA, L., GEFFERT, A., GEFFERTOVA, 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: 25502559. DZURENDA, L., DUDIAK, M. 2020. The effect of the temperature of saturated water steam on the colour change of wood acer pseudoplatanus L. In Acta Facultatis Xylologiae Zvolen, 62: 1928. DOI: 10.17423/afx.2020.62.1.02 FERRI, P., TRAMONTANA, F. 2020. Autonomous demand multiple equilibria and unemployment dynamics. In Journal of Economic Interaction and Coordination, 115. https://doi.org/10.1007/s11403020-00306-1 GOLD, S., FUBIK, F. 2009. Consumer attitudes towards timber as a construction material and toward timber frame houses – selected finding of a representative survey among the German population. In Journal of Cleaner Production, 17: 303309. https://doi/10.1016/j.jclepro.2008.07.001 KADLEČEK, F. 1989. Ruční obrábění dřeva, 1 st edn. ed. Praha : SNTL. KAUSTIA, M., ALHO, E., PUTTONEN, V. 2008. How much does expertise reduce behavioral biases? The chase of anchoring effects in stock return estimates. In Financial Management, 37: 391412. https://doi.org/10.1111/j.1755-053X.2008.00018.x KIRMAN, A., RICCIOTTI, R.F., TOPOL, R.L., 2007. Bubbles in foreign exchange markets. In Macroeconomic Dynamics, 11: 102123. https://doi.org/10.1017/S1365100507060257 KOZEL, R. 2006. Moderní marketingový výzkum: nové trendy, kvantitativní a kvalitativní metody a techniky, průběh a organizace, aplikace v praxi přínosy a možnosti, 1 st edn. ed. Grada, Praha. KUBOVSKÝ, I., KAČÍK, F. 2010. Zmeny v hlavných zložkách dreva spôsobené ožarovaním povrchu CO2 laserom, 1 st edn. ed. Zvolen : Technická univerzita vo Zvolene. NGUYEN, T.T.H., TRAN, V.C., LI, S., LI. J. 2020. Effects of rosin-aluminum sulfate treatment on the leachability, color stability, and decay resistance of wood treated with a boron-based preservative. In BioResources, 15: 172186. NIJDAM, N.A. 2009. Mapping emotion to color, [WWW Document] Research paper. URL http://citeseerx.ist.psu.edu/viewdoc/download.pdf NORDVIK, E., SCHÜTTE, S., BROMAN, N.O. 2010. People`s perceptions of the visual appearance of wood flooring. A Kansei engineering approach. In Forest Products Journal, 59: 6774. https://doi/10.13073/0015-7473-59.11.67 NYRUD, A.Q., ROOS, A., RØDBOTTEN, M. 2008. Product attributes affecting consumer preference for residential deck materials. In Canadian Journal of Forest Research, 38: 13851396. https://doi/10.1139/X07-188 PUAH, CHIN-HONG, WONG, S., HABIBULLAH, M.S. 2017. Are business forecasts of the construction sector rational? Survey-based evidence from Malaysia. In Economic Research – Ekonomska Istrazivanja, 30: 858872. https://doi.org/10.1080/1331677X.2017.1305794 SCHEER, D., HOFFMANN, E. RUBIK, F. 2020. Mit Holz in die Zukunft? Eine Branche am Scheideweg. [WWW Document] Brochure. URL https://www.ioew.de/Broschuere.pdf

172

SEDLIACIKOVA, M., STROKOVA, Z., KLEMENTOVA, J., SATANOVA, A., MORESOVA, M. 2020. Impacts of behavioral aspects on financial decision-making of owners of woodworking and furniture manufacturing and trading enterprises. In Acta Facultatis Xylologiae Zvolen, 62: 165176. DOI: 10.17423/afx.2020.62.1.14 SHIOZAWA, Y. 2020. A new framework for analyzing technological change. In Journal of Evolutionary Economics, 30: 9891034. https://doi/10.1007/s00191-020-00704-5 STOJANOVIC, A., MILOSEVIC, I., ARSIC, S., UROŠEVIC, S. 2020. Corporate social responsibility as a determinant of employee loyalty and business performance. In Journal of Competitiveness, 12: 149166. https://doi/10.7441/joc.2020.02.09 VEISTEN, K. 2002. Potential demand for certified wood products in the United Kingdom and Norway. In Forest Science, 48: 767778. https://doi.org/10.1093/forestscience/48.4.767 ACKNOWLEDGEMENT The authors are grateful for the support of the Slovak Research and Development Agency, grants number APVV-18-0520, APVV-18-0378, APVV-17-0456, APVV-17-0583 and APVV-20-0004.

AUTHORS’ ADDRESSES Prof. Ing. Mariana Sedliačiková, PhD. Ing. Patrik Aláč, PhD. Ing. Mária Moresová, PhD. et PhD. Technical University in Zvolen Faculty of Wood Sciences and Technology Department of Business Economics T. G. Masaryka 24 960 01 Zvolen Slovakia sedliacikova@tuzvo.sk alac@tuzvo.sk maria.moresova@tuzvo.sk Ing. Ivan Sedliačik, PhD. Matej Bel University in Banská Bystrica Faculty of Economics Department of Finance and Accounting Tajovského 10 975 90 Banská Bystrica Slovakia ivan.sedliacik@umb.sk

173

174

ACTA FACULTATIS XYLOLOGIAE ZVOLEN, 63(2): 175−188, 2021 Zvolen, Technická univerzita vo Zvolene DOI: 10.17423/afx.2021.63.2.15

ANALYSIS OF INTRA-INDUSTRY MUTUAL TRADE IN THE FURNITURE MANUFACTURING INDUSTRY BETWEEN THE V4 AND THE EU-27 COUNTRIES USING THE GRUBEL-LLOYD INDEX Justyna Biernacka ABSTRACT The purpose of this paper is to present and evaluate the directions of changes in intra-industry trade of selected products in the furniture manufacturing industry. The study covers the results in the intra-industry mutual trade between the V4 and EU-27 countries in four main product groups, i.e. wooden furniture for offices, wooden furniture for kitchens, wooden furniture for bedrooms and wooden furniture excluding offices, kitchens and bedrooms. The research using the Grubel-Lloyd index shows the lowest values of intra-industry trade between Poland and the UE-27. The foreign trade of the Czech Republic, Slovakia and Hungary with the EU-27 was characterised as satisfying intra-industry trade results. In mutual trade between the V4 countries, the Grubel-Lloyd index values fluctuated depending on the analysed product group. The conducted research helps determine the position of foreign trade of each of the V4 countries, making it possible to develop a competitive strategy. Key words: intra-industry trade, Grubel-Lloyd index, furniture manufacturing industry, competitiveness, Visegrád Group.

INTRODUCTION The economic development of the European Union Member States is currently focused on the use of renewable resources. One such a natural material is wood – a common, renewable, natural, ecological and biodegradable raw material, valued for its physical and mechanical properties, used as a basic raw material in the wood industry, including the furniture manufacturing industry. The furniture manufacturing industry is an important component of the national economies of individual EU Member States, as well as of the EU economy as a whole. Among the European Union countries, Germany, Italy and Poland can be mentioned as the largest furniture producers. However, not only the largest producers contribute to the EU economy, so the role of smaller producers is also important (GRZEGORZEWSKA et al. 2021). Trade contacts between individual EU countries are a driving force for the development also for smaller representatives of this economic sector. The Czech Republic, Hungary, Poland and Slovakia are countries that share not only a geographical proximity, but also cultural and historical community. The V4 countries also share similar experiences in the field of economic transformation and the challenges of joining international structures (BRODZICKI 2011). Among the many benefits of membership 175

in the European Community, the very strong impulse in the development of trade turnover deserves attention. Trade has also been an important mechanism of integration into the European Union’s markets (KAWECKA-WYRZYKOWSKA et al. 2017). However, opening up the EU market also brings challenges, such as increasing competition (ŁAPIŃSKA 2017). There are forecasts, stating that furniture consumption will increase significantly in Europe mainly due to a high demand for luxury furniture products (ZION MARKET… 2018). The projected increase in demand for furniture may result in increased market competition. Therefore, furniture manufacturing companies are likely to face many challenges. Due to the growing importance and the need to build a competitive advantage of the product offered, the furniture manufacturing industry has become of interest to modern science. From the point of view of the national economy (as well as at an individual level), building a competitive advantage through constant development is a key factor. GRZEGORZEWSKA et al. (2020) recognises many factors influencing the competitiveness of enterprises and the furniture manufacturing industry as well dividing them into external and internal. According to SUJOVÁ et al. (2015), one of the most important factors is obviously the availability of wood, which is the basic raw material used in production, while the demand for products of the furniture manufacturing industry is factor that is just as important. Other factors influencing competitiveness of the furniture manufacturing industry worth mentioning are: economic and market conditions such as changes in prices of wood raw materials and fluctuations in exchange rates, as well as by changes and consumer preferences. The competitiveness of the furniture manufacturing industry is also influenced by drivers, such as economic growth, urbanisation, trends in housing and construction and family incomes (GRZEGORZEWSKA et al. 2021). This subject formed part of the work of WANAT and KLUS (2015) and - using the indicator method – also of RATAJCZAK et al. (2008). According to their research, the factors determining the competitiveness of the wood industry sector can be presented in the following groups: institutional (political and legal conditions), natural (environmental conditions and resources specific to the industry, as a natural economy based on wood - a renewable raw material), economic (the economic situation and market processes taking place in it), industrial (relations and interactions of entities - industry participants and sectoral state policy) as well as other conditions (social aspects). According to CVETANOVIĆ et al. (2019) the competitiveness is manifested as a company's ability to compete on domestic and foreign markets, as well as the ability to support business. In general, the researchers define competitiveness as the ability of companies, industries, regions or nations to generate high income and employment. In the literature, however, there is no generally accepted definition of the competitiveness of a sector. As a result, macro-scale competitiveness measures are used at the level of sector competitiveness research (SUJOVÁ et al. 2015). The foreign trade balance is a logical basis for measuring competitiveness, as it deals with the analysis of competitive advantage (CVETANOVIĆ et al. 2019). As there is no common concept of competitiveness in the literature, this results in there being no single measure of competitiveness of an entire economy or a sector. The most commonly used competitiveness measures are: a). the Grubel-Lloyd index, which analyses the share of intra-sectoral character of goods in foreign exchange; b). the Revealed Comparative Advantage index, which comes in several versions and can be used to assess the competitiveness of an entire economy as well as sectors thereof; c). the Michaely index, which measures the share of a given commodity group in total exports; d). the Contribution to Trade Balance index, which measures the share of given sectors in the national trade balance (CVETANOVIĆ et al. 2019). 176

These models have most often appeared in scientific research (MISALA 1985, GREENAWAY, THARAKAN 1986, CZARNY 2002, MOLENDOWSKI 2006, MOLENDOWSKI and POLAN 2015). Two main directions of using these indexes for analyses should be distinguished here: literally to measure the intensity of intra-industry exchange at the level of economies and selected industries and as a way to estimate competitiveness. Scientists have proposed many models for measuring the intensity of intra-industry trade, but most of them are based on Grubel and Lloyd’s formula and - despite the passage of time – the Grubel and Lloyd index is still highly appreciated by most researchers (ŁAPIŃSKA 2014). For this reason it was used in the calculations made for the purposes of this paper. More modern research in this field has been carried out by PLUCIŃSKI (1996) and WYSOKIŃSKA (1995), with the works by MISALA (1985) and MISALA and PLUCIŃSKI (2000) being considered the most complex. The Grubel-Lloyd index can be used to analyse aggregated goods as proposed by the authors of the formula: either on the basis of the similarity of goods from the point of view of the production process, or on the basis of the substitutability of goods from the point of view of the consumer (GRUBEL 1967, GRUBEL and LLOYD 1975). To avoid mistakes, it is enough to choose the appropriate classification group in foreign trade (the HS or the SITC classification). On the other hand, the increased fragmentation of production processes may cause problems when trying to define a specific sector of industry. The reason for this is the increase in the share of processed semi-finished goods in international trade (KAWECKAWYRZYKOWSKA et al. 2017). Contemporary international trade is increasingly based on intra-industry trade. The phenomenon of intra-industry trade occurs when exports and imports within the same industry take place at the same time (CZARNY 2002), most often concerns goods that are highly differentiated and substitutable for each other (KLIMCZAK 2016). The intra-industry trade tends to occur among geographically close rich countries that share a similar economic structure and level of development (OECD ECONOMIC GLOBALISATION INDICATORS 2010). The factors that are most often mentioned as key for the development of intra-industry trade are: processes of differentiation of final goods and differentiation of demand, similar tastes of consumers, similar costs of production factors. Other contributing development of intra-industry trade factors are the proliferation of technological products and processes and absence of obstacles in international trade (MISALA and PLUCIŃSKI 2000). As shown by empirical research and observation an increasing part of contemporary trade takes place between countries with a similar structure of production and consumption. Higher shares of intra-industry trade are observed between industrialized countries, with similar production factors. This situation is often described in the literature as "North-North" trade type. The growth of intra-industry trade is favoured by similar level of development of the economies and their size, as well as similar level of GDP per capita. In relations between countries with limited industrial production capacities, lower results in intra-industry trade should be observed – the trade will rather be inter-industry, using the complementarity of products, than we are dealing with “South-South” type of trade. On the other hand, in the case of trade relations between countries with different levels of industrial development, we can deal with disproportions in production potential and different consumer behaviours (BRODZICKI 2011, KLIMCZAK 2016). In contemporary studies on the subject, hypotheses about the development of intraindustry trade between developed and developing countries can be found; some of them predict a decline in the importance of this trade type as a result of its increased liberalization (THARAKAN and KERSTENS 1995, GREENAWAY and THARAKAN 1986). 177

The research in the field of intra-industry trade concerned entire economies, but also individual sectors, e.g. chemical and agri-food (ŁAPIŃSKA 2014, 2017). The part of the research papers concerned specific regions or selected countries (TALAR 2012, TOPOROWSKI 2013, WYSOKIŃSKA, 1995) or focused on the assessment of the wood processing industry competitiveness (SUJOVÁ et al. 2015, PAROBEK et al. 2016), or selected sectors of the industry (GRZEGORZEWSKA et al. 2020). With the competitiveness of the wood processing industry on a macro scale also dealt HAJDÚCHOVÁ and HLAVÁČKOVÁ (2014). In recent years, however, there were not many studies dealing with the issue of intraindustry trade measured by the Grubel-Lloyd index for the Visegrád Group countries, especially for specific groups of the furniture manufacturing industry products. The furniture manufacturing industry products are characterised by high added value, which makes them excellent exports products, that can maintain the growing trend of the foreign trade balance. Therefore, it seems reasonable to attempt to assess the size and development of intra-industry exchange for more precisely defined groups of furniture products. The aim of this paper is to analyse mutual relations in foreign trade balance and attempt to examine the competitiveness of the furniture manufacturing industry of four countries belonging to the Visegrád Group with the use of the Grubel-Lloyd index.

MATERIALS AND METHODS As the first step in the intra-industry trade research on selected segments of the furniture industry, share of imports of the V4 countries for each of the chosen products was analysed. The measure of the foreign trade structure of the country is the share of a country's import of a given industry sector or group of products compared to its export, as shown in the formula: 𝐼 𝑇𝑆 = 𝐸𝑖𝑗 ∙ 100% (1) 𝑖𝑗

where: TS – trade exchange structure index (%), Iij – import of ith group of products from jth country (USD), Eij- export of ith group of products to jth country (USD). If the calculated value exceeds 100, it means that the value of imported goods and services exceeds the value of goods and service sales to abroad consumers in a given country. The obtained results will make possible to indicate which countries are a net importer or net exporter in the analysed product group. The next step of the research was verifying the intra-industry trade value in each of the analysed countries for individual product groups, and then calculating the intensity of intraindustry exchange for a given product group in the analysed countries. The most commonly used method for determining the volume of intra-industry trade is the Grubel and Lloyd index, developed in 1975. This measure allows to determine the volume of intra-industry trade between country “j” for sector (or product group) “i” is defined as follows: 𝐺𝐿𝑉 = (𝐸𝑖𝑗 + 𝐼𝑖𝑗 ) − |𝐸𝑖𝑗 − 𝐼𝑖𝑗 | (2) where: GLV – Grubel-Lloyd by volume - index of intra-industry trade (USD), Eij- as above, Iij – as above. The volume of intra-industry trade is therefore equal to the total volume of trade within the industry (𝐸𝑖𝑗 + 𝐼𝑖𝑗 ) less net export or import |𝐸𝑖𝑗 − 𝐼𝑖𝑗 |. This quantity is expressed in 178

absolute terms. To calculate the intensity of intra-industry trade, the following model is most often used: |𝐸 −𝐼 |

(3)

𝐺𝐿𝑆 = 1 − (𝐸𝑖𝑗+𝐼𝑖𝑗) ∙ 100 𝑖𝑗

𝑖𝑗

where: GLS – Grubel-Lloyd by intensity - index of intra-industry trade (%), Eij- as above, Iij – as above. The Grubel-Lloyd index (3) varies between 0 and 100. For example, if a given country only imports or only exports goods or services within a sector, it means that intra-industry trade does not occur – then GLS index approaches 0. Similarly, if for a given country there are simultaneous exports and imports of goods and services belonging to the same industry sector, then the GLS index value approaches 100 and intra-industry trade is observed. The Grubel-Lloyd index (3) values were calculated for the Visegrád Group countries and additionally for the European Union countries in order to compare the results. The analysis of the intra-industry trade was carried out on the basis of the GrubelLloyd index (2) and (3) calculated at the level of 6-digit CN codes in 20152020 years for four groups of furniture products, namely: a CN 940330: wooden furniture for offices (excluding seats), b CN 940340: wooden furniture for kitchens (excluding seats), c CN 940350: wooden furniture for bedrooms (excluding seats), d CN 940360: wooden furniture (excluding offices, kitchens, bedrooms and seats). All analyses were performed using the LibreOffice package.

RESULTS AND DISCUSSION Table 1 summarises the values of the trade exchange structure index (1) obtained for the analysed countries of the Visegrád Group. Trade in wooden furniture for offices, kitchens, bedrooms and other wooden furniture with the European Union in the period 2015 to 2020 is analysed. Tab. 1 Trade exchange structure index values for selected wooden furniture products between Visegrád Group countries and the European Union in the period 20152020. Group of products

Wooden furniture for offices (excluding seats)

Wooden furniture for kitchens (excluding seats)

Partner

Czech Republic

Hungary TS [%]*

Poland

Slovakia

Year 2015 2016 2017 2018 2019

70.32 33.98 42.92 36.42 40.28

179.40 107.21 289.54 428.93 547.55

18.03 16.95 29.38 17.08 11.90

33.61 28.95 33.09 47.17 70.69

2020 2015 2016 2017 2018 2019 2020

52.20 226.07 210.97 257.10 284.90 281.89 277.60

204.33 105.14 136.64 133.73 154.57 170.53 148.81

11.18 18.03 16.95 29.38 17.08 11.90 11.18

66.70 14.37 24.82 49.00 66.73 78.45 222.63

179

Wooden furniture for bedrooms (excluding seats)

Wooden furniture (excluding offices, kitchens, bedrooms and seats)

2015 2016

61.97 75.39

47.22 41.51

8.54 6.12

74.72 71.62

2017 2018 2019 2020 2015 2016

61.30 67.19 66.11 78.15 92.27 96.84

46.40 56.97 69.99 55.60 107.38 128.08

6.63 5.44 6.18 6.94 8.25 7.40

77.43 66.57 57.88 62.16 34.01 35.06

2017 2018 2019 2020

98.87 106.46 171.75 124.58

164.20 160.80 192.09 321.26

7.48 7.86 7.34 9.16

34.88 34.18 43.85 65.54

* - Trade exchange structure index calculated as imports from the EU-27 to a partner country related to exports from a partner country to the EU-27.

The values of TS indicate that in foreign trade in the selected years and product groups there are countries whose imports exceed their exports. These are: in the group of wooden furniture for offices and classified to the group of other furniture - Hungary, and in the group of wooden furniture for kitchens - Hungary and the Czech Republic. The analysis also showed values of the index close to or exceeding TS=100 in the case of other wooden furniture for the Czech Republic. The lowest values of the measure are found in all the studied groups of furniture in Poland, which means that exports prevail in this country. The next stage of the research is to analyse the values of measures (2) and (3) for selected countries. The values of intra-industry trade indices (2) and (3) calculated for trade in wooden furniture between the Visegrád Group countries and the European Union in 20152020 on the basis of the available ITC database (ITC 2021) are summarised in Table 2. Tab. 2 Intra-industry trade Grubel-Lloyd indices for selected wooden furniture group of products between Visegrád Group countries and the European Union in the period 20152020. Group of products Wooden furniture for offices (excluding seats)

Wooden furniture for kitchens (excluding seats) Wooden furniture for

Partner Year

Czech Czech Hungary Poland Slovakia Hungary Poland Slovakia Republic Republic GLV [million USD] GLS [%]

2015 2016 2017

25.74 33.20 43.61

10.05 20.21 7.36

31.47 31.23 58.51

23.84 19.10 17.93

82.58 50.72 60.06

71.58 96.52 51.34

30.56 28.98 45.42

50.31 44.90 49.73

2018 2019 2020 2015 2016 2017

45.51 45.92 47.11 39.87 53.37 43.57

6.06 4.65 11.19 11.37 14.20 15.73

49.70 42.37 43.00 31.73 33.82 37.48

22.39 24.59 24.87 22.70 31.03 24.63

53.39 57.43 68.60 61.34 64.31 56.01

37.81 30.89 65.72 97.50 84.52 85.57

29.18 21.27 20.11 40.98 41.09 49.00

64.10 82.83 80.02 25.12 39.76 65.77

2018 2019 2020 2015 2016 2017

46.32 46.33 50.75 82.31 93.32 98.46

17.48 16.79 17.83 32.25 34.39 39.84

44.41 43.05 61.09 49.40 45.47 55.22

36.46 37.08 17.95 35.81 37.61 41.79

51.96 52.37 52.97 76.52 85.97 76.01

78.56 73.93 80.38 64.15 58.67 63.38

37.46 36.61 40.39 15.73 11.53 12.43

80.05 87.92 61.99 85.53 83.46 87.28

180

bedrooms (excluding seats)

2018 2019 2020

110.53 115.73 137.02

46.87 52.65 54.27

63.00 79.05 96.72

44.78 51.54 64.46

80.38 79.60 87.73

72.58 82.35 71.47

10.32 11.64 12.97

79.93 73.32 76.67

Wooden furniture (excluding offices, kitchens, bedrooms and seats)

2015 2016 2017 2018 2019 2020

246.16 286.36 307.36 311.29 291.33 436.55

112.09 113.92 94.32 111.33 97.22 64.89

211.21 218.28 238.14 297.71 289.30 382.18

114.33 127.04 141.69 145.68 148.86 204.93

95.98 98.40 99.43 96.87 73.60 89.05

96.44 87.69 75.70 76.69 68.47 47.48

15.24 13.78 13.92 14.57 13.68 16.78

50.76 51.92 51.72 50.95 60.97 79.18

The analysis of the data presented in Table 2 shows that the foreign trade of the Visegrád Group countries with the EU-27 in 2015-2020 is characterised by a large diversification of GLV values. This applies to all the analysed product groups. In the case of wooden furniture for offices, in most of the analysed years, the lowest values of intra-industry trade were recorded in Hungary - the values of the mentioned indicator for this country range from USD 7.36 million in 2019 (the lowest value obtained in the analysed period) up to USD 20.21 million in 2016. The highest value of the measure in the discussed product group can be observed in Poland (USD 58.51 million) in 2017; in other years, Poland records equally high values of intra-industry trade, but in the case of the last two years analysed, the highest values of the measure can be observed for the Czech Republic (over USD 45 million). The values of intra-industry trade calculated using the GLV for Slovakia remain at a similar level throughout the analysed period, amounting to USD 22 million on average. The wooden furniture for kitchens product group is characterised by similarly low levels of the GLV index for Hungary's foreign trade with the EU-27 countries. Throughout the analysed period, this country shows the lowest values of intra-industry trade with the EU-27 – according to the GLV index the values vary from slightly over USD 11 million to almost USD 18 million. According to the index, the Czech Republic had the highest values of intra-industry trade (2) with the EU-27 in this product group in 20152019 (approximately USD 46 million on average). Poland enjoyed the highest result in 2020 with an intra-industry trade index GLV value reaching USD 61 million. In the group of wooden furniture for kitchens, Slovak GLV index values for intra-industry trade with the EU-27 amount to slightly over USD 28 million on average. Another product group analysed using the GLV measure was wooden furniture for bedrooms. As in the previous two groups of furniture manufacturing products, the lowest values of intra-industry trade on the GLV index are observed for Hungary (on average, in the analysed period they slightly exceed USD 43 million). The values for Slovakia are slightly higher (reaching approximately USD 46 million on average). Contrary to the previous two groups products, the highest values occur in the Czech Republic (on average USD 106 million and an upward trend in value throughout the period). In this product group, in analysed period Poland received a GLV index value averaging USD 65 million. In the last of the studied product groups - wooden furniture, not classified elsewhere - the average highest GLV index value can be observed in the Czech Republic (averaged over all the analysed years: USD 313 million), and the lowest in Hungary (approximately USD 99 million). Poland’s average values of intra-industry trade according to the GLV index, are just slightly lower than those achieved by the Czech Republic (approximately USD 273 million), Slovakia achieved an average value of the GLV indicator in this product group at the level of USD 147 million. The second part of Table 2 summarises the results of the study into the intensity of intra-industry trade of the Visegrád Group with the EU-27 countries using the Grubel-Lloyd index (3). When analysing the results obtained in the group of wooden furniture for offices, 181

it can be noticed that Poland has the lowest intensity of intra-industry trade with the EU-27. The value of the GLS index decreased to the level of 20.11% in the last year studied. Over the analysed period, the averaged GLS index for Poland is less than 30%. This situation is caused by large shares of exports and small shares of imports in the country's trade. Other countries recorded similar average values on the Grubel-Lloyd index (3) (for the Czech Republic, Hungary and Slovakia: 62, 59 and 62% respectively). In the group of wooden furniture for kitchens, Hungary is characterised by a highest intensity of intra-industry trade. It can be said, that this product group demonstrates an almost perfect intensity of intra-industry exchange, with the GLS index for the country reaching an average score slightly over 83%. The calculation results for the Czech Republic and Slovakia are similar 56 and 60% respectively. Poland notes GLS index values in the range of 3749%. The results of the analysis for the GLS index for wooden bedroom furniture show a similar tendency to the results obtained in the calculations for the group of wooden office furniture. Again, the lowest intensity of intra-industry trade can be attributed to Poland, which records an average index value of less than 13% throughout the period under review. This result is understandable if we realise that the divergence between the volume of exports and imports continues to increase in this group of products. The average values of the GLS index for the Czech Republic, Hungary and Slovakia are at a similar level amounting to 81%, 69% and 81% respectively. In the last analysed group of wooden furniture not classified elsewhere, the lowest GLS results are also obtained by Poland. On average, values do not exceed 15% in the discussed period. Contrary to Poland, the other analysed countries of the Visegrád Group also recorded high indices of intra-industry trade in this product group, with the highest values for the Czech Republic (over 92%), followed by Hungary (over 74%) and Slovakia (approximately 58%). Summarising the obtained results, it can be stated that in trade with the EU-27 countries, Poland is a one-sided partner in the analysed groups of products of the furniture manufacturing industry, as exports prevail in its exchange. The values of the Grubel-Lloyd index (3) for mutual trade between the analysed V4 countries are presented in Tables 3 and 4. Tab. 3 Values of intra-industry trade Grubel-Lloyd index (3) of wooden furniture for offices and kitchen between Visegrád Group countries in the period 20152020 [%]. Partner Partner

Year

2016 2017 2018 2019 2020 2015

Poland Hungary Czech Republic Wooden furniture for offices (excluding seats) x 37.71 21.42 88.21 x 34.36 8.55 11.16 x 34.98 23.61 2.08 x 38.91 35.11 23.75 x 69.40 31.91 49.77 x 76.74 17.78 61.68 x 28.26 0.00 0.00

2016 2017 2018 2019 2020

73.40 65.35 50.00 24.67 34.20

2015

79.44 83.15

2015

Czech Republic

Hungary

Poland

2016

Slovakia

0.00 3.76 0.00 0.00 0.00 x x

182

x x x x x

1.79 0.00 0.00 0.26 5.37

0.00 0.00

3.95 10.09

2017 2018 2019 2020

Slovakia

Partner

2015 2016 2017 2018 2019 2020 Year

x x x x

76.66 51.80 33.44 21.20 x x x x x x

25.49 14.88 0.00 1.14 1.62

0.00 0.03 0.00

8.97 7.30 4.14

0.00 70.80 14.14 14.68 9.30 0.62

3.67 49.97 62.97 70.34 65.82 58.07

1.11 0.58 59.48 Wooden furniture for kitchens (excluding seats) Slovakia Poland Hungary Czech Republic Partner

Tab. 4 Values of intra-industry trade Grubel-Lloyd index (3) of wooden furniture for bedrooms and other wooden furniture between Visegrád Group countries in the period 20152020 [%]. Partner Partner

Czech Republic

Year 2015 2016 2017 2018 2019 2020

Hungary

Poland

Slovakia

2015 2016 2017 2018 2019

89.24 26.67 2.73 32.59 11.76

0.00 0.02 0.00 0.00 0.00

x x x x x

51.55 43.09 40.89 26.08 35.94

2020 2015 2016 2017 2018 2019

12.79 96.59 90.18 88.71 86.35 83.94

0.00 x x x x x

x 0.96 1.29 0.49 0.80 0.84

31.79 4.69 5.62 6.70 4.14 4.67

2020 2015 2016 2017 2018 2019

72.78 x x x x x

0.84 91.81 38.76 21.73 5.73 12.56

5.26 67.11 71.64 65.61 71.41 70.06

2020 Partner

Slovakia Poland Hungary Czech Republic Wooden furniture for bedrooms (excluding seats) x 59.91 9.71 45.01 x 63.57 4.28 20.67 x 61.02 12.11 21.51 x 81.25 9.55 26.77 x 84.35 25.03 34.17 x 69.52 56.57 27.64

Year

73.62 69.65 51.69 49.45 37.76

x 35.11 26.27 59.38 Wooden furniture for bedrooms (excluding offices, kitchens and bedrooms, and seats) Slovakia Poland Hungary Czech Republic Partner

183

The data analysis presented in Table 3 reveals that, in the group of wooden furniture for offices, the highest values of intra-industry trade are obtained in the Polish trade with Slovakia. The GLS index values, especially in 20152017, exceed 70%, reach in excess of 83% (2016). In the subsequent years, a decline in the intra-industry trade index between these countries is observed, and in the last year, analysed the Grubel-Lloyd index (3) is less than 22%. The structure of these partners’ exports and imports is helpful when explaining the values of the GLS index - in 2015 and 2016, imports in this commodity group from Slovakia to Poland accounted for 140 to 150% of exports, while in subsequent years, the value of imports plummeted to a level of less than 12% of the value of exports in the last analysed year. On the other hand, reverse tendency can be found from an analysis of intra-industry trade for Slovakia and the Czech Republic. The intra-industry exchange of wooden furniture for offices in 20152017 shows a stable level of 3538%, before recording significant increases in the next three years, up to approximately 77%. Therefore, a hypothesis can be made that the increase in intra-industry trade between Slovakia and the Czech Republic came about as a result of the decline in the value of intraindustry trade between Slovakia and Poland. In order to verify this hypothesis, in-depth studies of trade exchange between Poland, Slovakia and the Czech Republic in the area of the discussed furniture manufacturing industry product groups would be required. The intra-industry exchange in Slovakia and Hungary in the sector of wooden furniture for offices, apart from higher values in 2016, 2017 and 2018, caused by greater Hungarian exports, is at a level of up to 35%. As for the intra-industry exchange of wooden furniture for offices between Poland and the Czech Republic, the highest values of the GLS index were obtained in 2018 and 2019, though these values are still much lower than those obtained for intra-industry trade between Poland and Slovakia, especially when comparing the results obtained for these trading partners in the last years under analysis. The results of the study of trade exchange using the Grubel-Lloyd index (3) for Poland and Hungary showed that there is practically no intra-industry trade between the partners in the product group of wooden furniture for offices. The reason for the low values of the GLS index may be the negligible values of exports and imports in this product group. In contrast to the trade exchange of products from the group of wooden furniture for offices with Poland, Hungary achieves quite high results in this group of products for intraindustry trade with the Czech Republic. The value of the exchange between these countries is quite variable, ranging from 2% to slightly over 88%. In the foreign trade of wooden furniture for kitchens, the highest values of the GL S index are achieved by the exchange between Slovakia and the Czech Republic. The values of intra-industry trade for this product group range from approximately 50% in 2015. to over 70% in 2017. In the subsequent years, the Grubel-Lloyd GLS index for Slovakia's trade with the Czech Republic fluctuates around 6066%. When analysing the value of the GLS index in this product group between Slovakia and Hungary, the Grubel-Lloyd index (3) results are clearly significantly different from those calculated for these countries in the case of trade in wooden furniture for offices. The GrubelLloyd index (3) values of both countries in the first year analysed exceeds 70%, before dropping to between 10 and 20% in the following years, and finally, disappearing completely in the last analysed year. The reason for this situation is decreasing levels of export of kitchen furniture from Hungary to Slovakia on the one hand, and growing exports from Slovakia on the other. A similar situation occurs in relation to the trade between Slovakia and Poland in wooden furniture for kitchens. Initially, the value of intra-industry trade calculated using the 184

Grubel-Lloyd index (3) is at the level of several dozen per cent, before dropping dramatically from 2017 and in the last three analysed years. In the trade between these countries, there is also a slight export imbalance in favour of Poland. Equally low values of the Grubel-Lloyd GLS index are recorded for the trade in wooden kitchen furniture between Poland and the Czech Republic. The highest result obtained in this case is 10% for 2016. There is no inter-industry exchange in this product group between Poland and Hungary, or between Hungary and the Czech Republic throughout the analysed period. In the group of wooden furniture for bedrooms, the intra-industry trade measured by the Grubel-Lloyd index GLS records the highest results for trade between Poland and Slovakia. The intra-industry trade for wooden bedroom furniture for this pair of partners is mostly stable at level of 84 to 97%, only 2020 shows a visible decline in the index values. The next GLS index by value is found in trade between Slovakia and the Czech Republic. For these countries, the Grubel-Lloyd index (3) ranges from 60% to 85%. In turn, the lowest values of intra-industry trade are observed for this product group for the PolandHungary pairing. The values of the Grubel-Lloyd index for intra-industry trade in wooden bedroom furniture between Hungary and the Czech Republic reach a level from 21% to 45%, and between Hungary and Slovakia the level is from 3% to 33%. There is a negligible level of intra-industry trade between Poland and the Czech Republic. The last product group tested using the Grubel-Lloyd index (3) was the group of wooden furniture not classified into any of the aforementioned groups. In this group of products, the Czech Republic and Slovakia show the highest average values of intra-industry trade. The GLS index values for these countries remain stable at the 66 to 72% level throughout all the years analysed, except for the last year, in which the Grubel-Lloyd index results slightly decreased to around 60%. Intra-industry trade in Poland and Slovakia is characterised by similarly high results with the Grubel-Lloyd index values starting the period in excess of 70%, but with a gradual decline observed in subsequent years. Taking into account the average value of the GLS index calculated for the analysed countries of the Visegrád Group, similar results for trade between Hungary and the Czech Republic as well as Hungary and Slovakia are observed. Apart from the lowest values of intra-industry trade, again observed in the case of trade between Hungary and Poland throughout the analysed period, it is worth noting the small values of the index GLS for the exchange between the Czech Republic and Poland. For wooden furniture not classified in any of the other groups, the GLS values range from about 4% to about 7%.

CONCLUSIONS Recently, the furniture industry in Central and Eastern Europe has suffered in particular, due to the reduction in trade caused by the COVID-19 pandemics. The slowdown in global demand has also affected economies driven by domestic consumption – these countries have tended to fail to avoid the side effects of regional and global constraints. The analysis showed that Poland has experienced significant disproportions in foreign trade between export and import, especially in the three analysed groups of furniture, namely: furniture for kitchens, furniture for offices and other furniture (outside of those groups). Poland was a net exporter to the EU-27 – exports in these furniture groups exceeded imports several times. This situation occurred throughout the period under consideration. The furniture for bedroom exports to the EU-27 were characterised by a smaller, but significant advantage, amounting to three or four times the value of imports. This situation 185

was undoubtedly influenced by Poland’s position as one of the leading exporters of furniture - it is estimated that in 2019 about 90% of furniture production was exported. The analysis of the results shows that the imports from the EU in the analysed period has a constant upward trend for all discussed groups of furniture. The worst year for imports to Poland was 2019, while the best for exports in these groups of furniture products – was 2018. Such large disproportions in foreign trade are reflected in the Grubel-Lloyd index values - some of the lowest values in trade with the EU and with other analysed countries were reported by Poland. It seems that despite the difficulties related to the COVID-19 pandemic in 2020 and the decline in the furniture manufacturing industry's turnover by 35% compared to 2019, furniture exports from Poland have not changed significantly. The situation of the furniture industry in Poland was probably even improved by a positive impulse, generated by the renovation boom in 2020. An additional factor was an increase in furniture orders and the implementation of government anti-crisis shields. The Grubel-Lloyd index values calculated for trade between the EU-27 and the analysed countries confirmed that there is a good level of intra-industry trade. They did not show significant disproportions between imports and exports as in the case of Poland. The EU-27 foreign trade with the Czech Republic and with Slovakia was particularly balanced. In trade relations with the EU-27 Slovakia is also a net exporter for all furniture groups except kitchen furniture. In contacts between the individual partners from the Visegrád Group in the group of office furniture, the highest indicators of the GLS index were recorded in Poland's trade with Slovakia and in the Czech Republic's with Hungary. The geographical proximity testifies to the development of trade contacts between these countries. It seems that intra-industry trade between Poland and Hungary does not occur precisely because of the geographical distance, as well as the more limited size of the Hungarian market. In the group of kitchen furniture, the GLS index values helped confirm the existence of a well-developed intra-industry trade between close neighbours, such as the Czech Republic and Slovakia. In other cases, the intra-industry trade measured using the GLS index is unsatisfactory, or simply does not occur. The GLS index values for trade of wooden furniture for the bedroom group showed well-developed relations between the Czech Republic and Slovakia, as well as between Poland and Slovakia. Generating positive net export values is the driving force for building a competitive advantage of the industry. Slightly lower trade results in this product group were achieved between the Czech Republic and Hungary. The analysis of the Grubel-Lloyd index for the last group of furniture showed a similar level of intra-industry trade between Poland and Slovakia, the Czech Republic and Slovakia and the Czech Republic and Hungary. The calculations performed in this paper, in addition to showing trends in foreign trade between the analysed countries will allow more in-depth analyses of contacts between individual Visegrád Group countries to be carried out in the future, looking at the types of intra-industry exchange and assessing the competitiveness of the furniture industry in the discussed countries. REFERENCES BRODZICKI, T. 2011. Handel zagraniczny państw Grupy Wyszehradzkiej. Zmiany strukturalne i rola handlu wewnątrzgałęziowego, In Analizy i Opracowania KEIE, Uniwersytet Gdański, nr 4/2011 (007), ISSN: 2080-09-40.

186

CVETANOVIĆ, S., NIKOLIĆ, M., CVETANOVIĆ, S. 2019. Wood industry trade competitiveness of selected countries of Southeast Europe. In Drvna Industrija, 70: 209219. DOI: 10.5552/drvind.2019.1803. CZARNY, E. 2002: Teoria i praktyka handlu wewnątrzgałeziowego. Monografie i Opracowania, z. 496. Warszawa : Oficyna Wydawnicza Szkoły Głównej Handlowej, ISBN: 83-7225-137-1. GREENAWAY, D., THARAKAN P. K. M. 1986. Imperfect Competition and lnternational Trade. Policy Aspects of lntra-Industry Trade. Brighton: Wheatsheaf Books. GRUBEL, H.G. 1967. Intra-industry specialization and the pattern of trade. In Canadian Journal of Economics and Political Science, 33: 374–388. GRUBEL, H.G., LLOYD, P.J. 1975. Intra-industry Trade: The Theory and Measurement of International Trade in Differentiated Products. Wiley, New York, ISBN: 0333176243. GRZEGORZEWSKA, E., SEDLIACIKOVA, M., DRABEK, J., BEHUN, M. 2020. Evaluating the international competitiveness of polish furniture manufacturing industry in comparison to the selected EU countries. In Acta Facultatis Xylologiae Zvolen, 62(2): 149164. DOI:10.17423/afx.2020.62.2.14 GRZEGORZEWSKA, E., SEDLIAČIKOVÁ, M., KALAFÚS, J. 2021. A comparison of the importance of the furniture manufacturing in EU countries using cluster analysis and Hellwig's method. In Acta Facultatis Xylologiae Zvolen, 63(1): 151−164. DOI:10.17423/afx.2021.63.1.14 HAJDÚCHOVÁ, I., HLAVÁČKOVÁ, P. 2014. Vplyv globálnej ekonomiky na lesnícko-drevársky sektor v Českej a Slovenskej Republice. In Acta Facultatis Xylologiae Zvolen, 2014, 56(2): 135146, ISSN 1336- 3824. ITC 2021. Trade Map, database. Available online https://www.trademap.org/Index.aspx (accessed on 17 March 2021). KAWECKA-WYRZYKOWSKA, E., AMBROZIAK, Ł., MOLENDOWSKI, E., POLAN, W., 2017. IntraIndustry Trade of the New EU Member States. Theory and Empirical Evidence. PWN, Warsaw. ISBN: 978-83-01-19718-6. KLIMCZAK, Ł. 2016. Handel wewnątrzgałęziowy państw Bałkanów Zachodnich w relacjach wzajemnych oraz z Unią Europejską, Unia Europejska.pl Nr 2 (237), 1018. ŁAPIŃSKA, J. 2014. Determinants of Intra-Industry Trade in Agricultural and Food Products Between Poland and EU Countries. In Danube 5(3): 159172. DOI:10.2478/danb-2014-0009. ŁAPIŃSKA, J. 2017. Ocena możliwości rozwoju polsko-czeskich obrotów towarowych produktami przemysłu chemicznego na podstawie modelu handlu wewnątrzgałęziowego. Instytut Badań Gospodarczych, Polskie Towarzystwo Ekonomiczne Oddział w Toruniu. ISBN 978-83-65605-02-3. MISALA, J. 1985. Wymiana wewnątrzgałęziowa w polskim handlu zagranicznym, Ekonomista, nr 3. MISALA, J., PLUCIŃSKI, E.M. 2000. Handel wewnątrzgałęziowy między Polską a Unią Europejską. Teoria i praktyka, Elipsa, Warszawa. ISBN: 83-7151-364-x. MOLENDOWSKI, E. 2006. Handel wewnątrzgałęziowy krajów CEFTA z państwami Unii Europejskiej. In Gospodarka Narodowa, 56/2006, 7592. MOLENDOWSKI, E., POLAN, W. 2015. Handel wewnątrzgałęziowy - miernikiem międzynarodowej pozycji konkurencyjnej gospodarek. In Studia i Prace Wydziału Nauk Ekonomicznych i Zarządzania Uniwersytetu Szczecińskiego, 41, t.1, 11-23. OECD 2010. Measuring Globalisation. Economic Globalisation Indicators 2010. Available online: https://www.oecd-library.org/deliver?redirecturl=http%3A%2F%2Fwww.keepeek.com%2FDigitalAsset-Management%2Foecd%2Findustry-and-services%2Fmeasuring-globalisation-oecd-economicglobalisation-indicators-2010_9789264084360-en&isPreview=true&itemId=%2Fcontent%2 Fpublication%2F9789264084360-en (accessed on 17 March 2021). PAROBEK, J., PALUŠ, H., LOUČANOVÁ, E., KALAMÁROVÁ, M., GLAVONIĆ, B. 2016. Competitiveness of Central European countries in the EU forest products market with the emphasis on Slovakia. In Acta Facultatis Xylologiae Zvolen, 58(1): 125136, DOI: 10.17423/afx.2016.58.1.14. PLUCIŃSKI E. 1996. Intensywność handlu wewnątrzgałęziowego krajów CEFTA z Unią Europejską i w ramach CEFTA, Instytut Gospodarki Światowej, SGH, Prace i Materiały, nr 150, Warszawa. RATAJCZAK, E., SZOSTAK, A., BIDZIŃSKA G., 2008. Pozycja konkurencyjna polskiego sektora

drzewnego na rynku krajowym i europejskim. In Drewno 51(179), 95-121.

187

SUJOVÁ, A., HLAVACKOVÁ, P., MARCINEKOVÁ, K. 2015. Evaluating the Competitiveness of Wood Processing Industry. In Drvna Industrija, 66: 281288. DOI: 10.5552/drind.2015.1432. TALAR, S. 2012. Wewnątrzgałęziowa wymiana handlowa Polski w warunkach integracji, Studia Ekonomiczne, Uniwersytet Ekonomiczny w Katowicach, nr 123, Międzynarodowe stosunki gospodarcze - integracja regionalna w Europie i na świecie, 199212. THARAKAN, P.K.M., KERSTENS, B. 1995. Does North-South horizontal intra-industry trade really exist? An analysis of the toy industry. In Weltwirtschaftliches Archiv 131, 86–105. https://doi.org/10.1007/BF02709073. TOPOROWSKI, P.E., 2013. The steps of Visegrad countries towards better integration within the EU: a case of intra-industry trade. Paper presented at ETSG 2013, 15th Annual Conference, 12-14 September 2013, Birmingham Business School at the University of Birmingham, UK, https://www.researchgate.net/publication/256375091_The_steps_of_Visegrad_countries_towards_b etter_integration_within_the_EU_a_case_of_intra-industry_trade (accessed on 23 September 2021). WANAT, L., KLUS, S., 1995. Sytuacja konkurencyjna branży i mezoekonomiczne aspekty polityki sektorowej państwa na przykładzie rynku drzewnego w Polsce. In Rynek – Społeczeństwo – Kultura, 1(13), 41-55. WYSOKIŃSKA, Z. 1995. Dynamiczne współzależności wymiany handlowej krajów Europy Środkowej i Wschodniej w świetle teorii integracji i wymiany międzynarodowej, Łódź. 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 23 September 2021).

AUTHOR’S ADDRESS Justyna Biernacka, PhD. Warsaw University of Life Sciences – SGGW Department of Technology and Entrepreneurship in Wood Industry Nowoursynowska 159 02-776 Warsaw Poland justyna_biernacka@sggw.edu.pl ORCID ID 0000-0003-3407-1280

188

ACTA FACULTATIS XYLOLOGIAE ZVOLEN 63 2/2021

Articles inside

INDUSTRY BETWEEN V4 COUNTRIES AND EU-27 USING THE GRUBEL-LLOYD INDEX

TREE IGNITION BY SPARKLERS

ASSESSMENT OF WOOD-BASED PRODUCTS

RADIANT HEAT EXPOSURE

ADJUSTMENT FOR LASER ENGRAVING ON WOOD

INSTALLATION POSITION

PLYWOOD MANUFACTURING

ROZÁLIA VAŇOVÁ – JOZEF ŠTEFKO: ASSESSMENT OF

VALENTIN ATANASOV: EXPERIMENTAL RESEARCH OF THE CUTTING FORCE DURING LONGITUDINAL MILLING OF SOLID

MARIANA SEDLIAČIKOVÁ – PATRIK ALÁČ – MÁRIA MORESOVÁ – IVAN SEDLIAČIK: MAPPING WOOD COLOUR

A.V. CHERAKSHEV  D.E. RUMYANTSEV

ROMAN NÔTA – ZUZANA DANIHELOVÁ: ANALYSIS OF THE

ROZÁLIA VAŇOVÁ: INFLUENCE OF CARBON ACCOUNTING ON

IVETA ČABALOVÁ – MARTIN ZACHAR – MICHAL BÉLIK

MARIA ZYKOVA – VALERIA KASIMIOVA – MIKHAIL