Warsaw University of Life Sciences

Transcription

1 Annals Warsaw University of Life Sciences Forestry and Wood Technology No 73 Warsaw 2011 Contents: SMARDZEWSKI JERZY, KŁOS ROBERT Modeling of joint substitutive rigidity of board elements... 7 ALTUN SUAT, YAPICI FATIH, KORKMAZ ZEHRA. Effects of vacuum drying with infrared heating on some properties of wood BARBOUTIS IOANNIS, MELISSIDES THEODOSIOS Influence of the time between machining and assembly of mortise and tenon joints on tension strength of T-type joints GAWROŃSKI TOMASZ Optimization of furniture technology at design stage HAVIAROVA EVA Approach to furniture design education at Purdue University

2 KAMPERIDOU VASILIKI, BARBOUTIS IOANNIS, VASSILIOU VASSILIOS Correlation between bending and tension strength of corner and middle joints constructed with beech and poplar wood KOŘENÝ ADAM, ŠIMEK MILAN Experimental testing of cam fittings PREKRAT SILVANA, PERVAN STJEPAN, SMARDZEWSKI JERZY Optimization of furniture testing SMARDZEWSKI JERZY, MAJEWSKI ADAM Auxetic spring elements for elastically supporting a sitting or lying TANKUT NURGUL The influence of pilot hole on the moment resistance of screwed T-Type furniture joints TAUBER JIŘÍ, SVOBODA JAROSLAV Ergonomic authentication for dimensions furniture ANDRES BOGUSŁAW Basidiomycetes growing on sleepers reused in small garden architecture ANDRES BOGUSŁAW, MAŃKOWSKI PIOTR Resistance of lime wood (Tilia sp.) impregnated with Paraloid B-72 resin against cellar fungus Coniophora puteana (Schum., Fr.) Karst ANTCZAK ANDRZEJ, MAŃKOWSKI PIOTR, BORUSZEWSKI PIOTR Chemical studies of ozone impact on pinewood (Pinus sylvestris L.) degradation BAJKOWSKI BOGUSŁAW Application of artificial intelligence in the wood industry BAJKOWSKI BOGUSŁAW Computer techniques in automatized wood industrial company BARBU SIMONA-MARIA, BADESCU LOREDANA ANNE-MARIE, JAVOREK LUBOMIR Studies regarding the influence of forces and torques on the quality of surfaces obtained at wood drilling BIERNACKA JUSTYNA The analysis of selected parameters characterizing economic condition of furniture manufacturer - Forte SA BIERNACKA JUSTYNA The analysis of selected parameters characterizing economic condition of Paged SA BIERNACKA JUSTYNA The analysis of selected parameters characterizing economic condition of Grajewo SA

3 BODNÁR FERDINAND, JABŁOŃSKI MAREK Stress distribution along the contour of a circular opening in wooden plate loaded by inplane bending moment BORUSZEWSKI PIOTR, BORYSIUK PIOTR, JASKÓŁOWSKI WALDEMAR, FAJKOWSKA KAROLINA, MAMIŃSKI MARIUSZ, JENCZYK-TOŁŁOCZKO IZABELLA Characteristics of selected fireproof properties of particleboard made from particles impregnated with salt agent BORUSZEWSKI PIOTR, BORYSIUK PIOTR, JASKÓŁOWSKI WALDEMAR, ŚWIĘCKI ANTONI, MAMIŃSKI MARIUSZ, JENCZYK-TOŁŁOCZKO IZABELLA Influence of flakes impregnation with salt flame retardants on selected physical and mechanical properties of OSB BORUSZEWSKI PIOTR, BORYSIUK PIOTR, MAMIŃSKI MARIUSZ Screw holding ability of the lignin-bonded biocomposites BORYSIUK PIOTR, KRAJEWSKI KRZYSZTOF, BORUSZEWSKI PIOTR, JENCZYK-TOŁŁOCZKO IZABELLA, JABŁOŃSKI MAREK Bonding quality of veneers protected with fireproofing preservation based on diammonium hydrogen phosphate, citric acid and sodium benzoate BORYSIUK PIOTR, JABŁOŃSKI MAREK, POLICIŃSKA-SERWA ANNA, RUŽINSKÁ EVA Mechanical properties of glue bonds in black locust wood treated with ammonia BORYSIUK PIOTR, JASKÓŁOWSKI WALDEMAR, BORUSZEWSKI PIOTR, JENCZYK-TOŁŁOCZKO IZABELLA, JABŁOŃSKI MAREK, BYLIŃSKI DAWID Flammability of plywood made from veneers protected with flame retardant based on diammonium hydrogen phosphate, citric acid and sodium benzoate BORYSIUK PIOTR, ZBIEĆ MARCIN, BORUSZEWSKI PIOTR, MAMIŃSKI MARIUSZ, MAZUREK ANDRZEJ Possibilities of single-stage pressing of veneered particleboards BORYSIUK PIOTR, DANIHELOVA ANNA, JABŁOŃSKI MAREK, POLICIŃSKA-SERWA ANNA, RUŽINSKÁ EVA The impact of wood staining with specific synthetic dyes on pine wood gluability BORYSIUK PIOTR, JASKÓŁOWSKI WALDEMAR, BORUSZEWSKI PIOTR, JENCZYK-TOŁŁOCZKO IZABELLA, JABŁOŃSKI MAREK, BYLIŃSKI DAWID Ignitability of wood impregnated with fireproof agent based on diammonium hydrogen phosphate, citric acid and sodium benzoate BURAWSKA IZABELA, TOMUSIAK ANDRZEJ, BEER PIOTR Influence of the length of CFRP tape reinforcement adhered to the bottom part of the bent element on the distribution of normal stresses and on the elastic curve) 186 3

4 BURAWSKA IZABELA, TOMUSIAK ANDRZEJ, TURSKI MICHAŁ, BEER PIOTR Local concentration of stresses as a result of the notch in different positions to the bottom surface of bending solid timber beam based on numerical analysis in Solidworks Simulation environment CHUCHAŁA DANIEL, ORŁOWSKI KAZIMIERZ, KRZOSEK SŁAWOMIR The preparation method of experimental studies of the wood sawing process CYRANKOWSKI MARIUSZ, OSIPIUK JAN, BAREJ PAWEŁ Straw briquette as an energy source CYRANKOWSKI MARIUSZ, KORZEMSKI MICHAŁ Implementation issues in visual systems for wood quality testing CYRANKOWSKI MARIUSZ, OSIPIUK JAN, ADAMCZYK DAWID Plants as an alternative source of energy CZECHOWSKI JACEK, BLUS KAZIMIERZ Eco-friendly method for paper dyeing with reactive dyes ДЕЛИЙСКИ НЕНЧО Система модельно базированного автоматического управления процессом конвективной сушки пиломатериалов ДЕЛИЙСКИ НЕНЧО, ДЗУРЕНДА ЛАДИСЛАВ Вычисление удельной теплоемкости мерзлой древесины во время оттаивания льда в ней от адсорбционно связанной воды DIETZ HANS, KRZOSEK SŁAWOMIR Die Zukunft von Bandsägeanlagen mit Magnetführungen im Sägewerk

5 Board of reviewers: Bogusław Bajkowski Piotr Beer Ewa Dobrowolska Jarosław Górski Adam Krajewski Krzysztof Krajewski Donata Krutul Sławomir Krzosek Hanna Pachelska Jerzy Smardzewski Wacław Szymanowski Piotr Witomski Janusz Zawadzki Scientific council : Arnold Wilczyński (Poland) Kazimierz Orłowski (Poland) Ladislav Dzurenda (Slovakia) Miroslav Rousek (Czech Republic) Nencho Deliiski (Bulgaria) Olena Pinchewska (Ukraine) Włodzimierz Prądzyński (Poland) Dofinansowano ze środków Ministra Nauki i Szkolnictwa Wyższego Polska Akademia Nauk Komitet Technologii Drewna Warsaw University of Life Sciences Press wydawnictwo@sggw.pl SERIES EDITOR Ewa Dobrowolska ISSN Marcin Zbieć PRINT: ZPW POZKAL 5

6 6

7 Annals of Warsaw University of Life Sciences SGGW Forestry and Wood Technology No 73, 2011: 7-15 (Ann. WULS-SGGW, Forestry and Wood Technology 73, 2011) Modeling of joint substitutive rigidity of board elements JERZY SMARDZEWSKI, ROBERT KŁOS Department of Furniture Design, Faculty of Wood Technology, Poznan University of Life Sciences Abstract: Modeling of joint substitutive rigidity of board elements. The study presents alternative methods of numerical modeling of dowel joint rigidity of board elements using for this purpose nodes of substitute linear elasticity modulus. Values of joint deflections obtained by way of laboratory experiments and numerical calculations differed by 3 to 4%. Keywords: dowel joint, substitute linear elasticity modulus, numerical calculations INTRODUCTION Virtual furniture design requires verification of their quality from the point of view of their rigidity and strength of the applied construction solutions. This, in turn, is associated with the development of appropriate mesh models of furniture joints in the environment of programs calculating with the assistance of the finite elements method (FEM). Realistic representation of the examined structure in the FEM environment is very labour-consuming, requires numerous corrections of network geometry and meticulous determination of linear elastic properties of the applied materials (Dzięgielewski, Smardzewski, (1996), Kasal (2008), Kasal et al., (2008a,b), Smardzewski (2004a,b), Smardzewski (2005)). However, it is more practical but equally effective to replace joints with semirigid joints (Kłos, Smardzewski (2004), Nicholls, Crisan (2002), Smardzewski, Kłos (2004), Smardzewski, Ożarska (2005), Smardzewski, Prekrat (2002, 2005)). In this paper, an attempt was made to present alternative methods of numerical rigidity modeling of cabinet furniture dowel joints using nodes of substitute linear elasticity modulus. RESEARCH OBJECTIVE The aim of this research project was to determine values of the substitute linear elasticity modulus of a spatial dowel joint subjected to closing and opening, to ascertain deflections of this joint, to develop alternative mesh models, to compare the obtained results and to select the model most favourable for the virtual prototyping of furniture. MATERIAL AND METHODS Investigations were carried out on angle joints with three 8x32 mm dowel pins of beech wood (Fig.1). Arms of the connections were made of 16 mm thick chipboards and were supported and loaded in accordance with the diagrams presented in Figure 2, measuring values of force P with 0.01 N accuracy and displacements δ ip of point i along the direction of action of force P with 0.01 mm accuracy. For diagrams presented in Figure 2, function equations of the joint arm deflection were determined in the following form: w' ' x for x P l x, it was obtained, w x M y 2 2 M y, where EJ 2P l x ip dx dx Cx D EJ. 2 7

8 Taking into consideration kinematic boundary conditions and rigidity of joint elements, calculating integration constants and converting the above equation in relation to δ ip, the following equations were obtained: a) substitute elasticity model in a closing test: 2 Z Plz 3l1 3l1l z l E z 3 Pl1 3 ip J1 E1 b) substitute elasticity model in an opening test: 2 2 R Plz 3l1 3l1l z lz E z, 3 Pl1 12 ip J1 E1 where: l z length of the near-node section (for the board, it was assumed as l z =2h), l 1 length of the joint arm, b cross-section width of the joint arm, h thickness of the joint arm 16 mm, ip joint total deflection, 3 b h J 1 moment of inertia of joint arm cross-section, J, E 1 Young modulus of the joint arm - chipboard, P load (0.4P max 0.1P max ). 1 2 z 12, Fig. 1. Dowel joint 8

9 a) b) Fig. 2. Diagram of loading of the angle joint in: a) closing, b) opening tests Table 1 collates mechanical properties of beech wood, chipboard and glue bond used for the experimental joints, whereas in Table 2, calculation results of substitute modulus of linear elasticity are presented. Table 1. Mechanical properties of materials used to make joints Material Property Mean value Standard deviation Coefficient of variance [%] Chipboard Linear elasticity modulus [MPa] Bending strength [MPa] Beech wood Linear elasticity modulus [MPa] Bending strength [MPa] Glue bond Linear elasticity modulus [MPa] Table 2. Values of the joint substitute linear elasticity modulus Kind of research Mean value [MPa] Standard deviation [MPa] Coefficient of variance [%] Closing test Opening test When preparing models for numerical calculations, all attempts were made to make sure that geometry and elastic properties attributed to the materials corresponded to real objects as closely as possible. At the same time, issues associated with time necessary for the elaboration of models as well as cost-efficiency of the calculation process were also taken into consideration. For the above reasons, as well as for utilitarian motives, two mesh models were developed for each joint. One of the models constituted a faithful replica of the real joint, while the other, instead of profile and profile-adhesive joints, contained an element of previously determined substitute linear elasticity modulus (Fig.3). 9

10 a) b) Fig. 3. Network of finite elements for a solid body model of angle joint containing: a) dowel pegs, b) element of a substitute modulus of rigidity E z Brick type, eight-node elements were used to model the experimental dowel joints (Fig.3a). Joint arms were attributed earlier determined elastic properties of chipboards, while fasteners elastic properties of beech wood. The glue bond was modeled as an 0.1 mm thick layer of E s =460 MPa linear elasticity modulus. A contact surface was defined between joint elements differentiating the density of the master and slave networks in such a way that the former was at least two times denser at the place of contact than the slave network. The substitute model of the same joint (Fig.3b) was made using an element of determined substitute linear rigidity. In accordance with the diagram in Figure 2, the substitute element included the near-node sections and arms 2h long (two thicknesses of arms) measuring from the inside part of the connection. The remaining sections of the arms of the joint were attributed elastic properties as in the model from Figure 3a. Taking into consideration calculated mean values of the linear elasticity substitute modulus of the examined joints (Tab.2) as well as their non-linear rigidity characteristics, the following parameters were used for numerical calculations (Tab.3): External load P of the values corresponding to 40% of the joint mean breaking load determined for the linear-elastic range, Deflection δ ip corresponding to the determined external load, Linear elasticity substitute modulus: E Z z for the joint closing test and E R z for the opening test calculated for the determined P and δ ip. Table 3. Values of loads, deflections and linear elasticity substitute modulus of the experimental joints Joint Closing test Opening test P [N] δ 1P [mm] E Z z [MPa] P [N] δ 2P [mm] E R z [MPa] Dowel Numerical calculations for closing and opening patterns of the joints were conducted in the environment of Algor software which realises algorithms of the finite element method. Models from Figure 2 and load values from Table 3 were employed as schemes of static support and loading of the examined systems. The results of calculations comprised illustrations of the distribution of reduced stresses (Mises) in elements of joints, values of these stresses, illustrations of deflections of joints in the direction of action of the external force as well as values of these deflections. 10

11 RESEARCH RESULTS When dowel joints were subjected to closing with force constituting 40% of the breaking load value, it was found that the highest reduced stresses concentrated around fasteners and seats (Fig. 4). The value of these stresses did not exceed 3 MPa which means that the only critical condition deciding about the strength of this construction node was the resistance of the chipboard to delamination (k r =<1.0MPa). Fig. 4. Distribution of reduced stresses according to Mises in a dowel joint subjected to closing, σ max 3 MPa a) b) Fig. 5. Distribution of reduced stresses according to Mises in a dowel joint subjected to closing at the point of: a) board mutual pressure, b) pressure of fasteners on boards, σ max 3 MPa The forms of joint deformations presented in Fig.5 indicate strong loading of the front edge of the joint element and a lack of contact (formation of a gap) between elements on external surfaces (Fig.5b). Because the external chipboard surfaces from which elements of joints were made consist of microchips, contact stresses developed on the edge caused by pressures did not exceed 3 MPa and were not dangerous for this joint. 11

12 Fig. 6. Distribution of reduced strains according to Mises on the surface of seats in a dowel joint subjected to closing, σ max 3 MPa The most dangerous stresses for the load bearing capacity of the joint were those concentrating on the surface of seats (Fig.6). Their value was higher than 3 MPa significantly exceeding the strength of chipboards for stratification. Such concentration of stresses first caused damage of seat surfaces made in the loosest and the weakest zone of the chipboard and then resulted in the delamination of the element in which fasteners were mounted longitudinally to wide planes of the board. Fig. 7. Size of gap between boards pressing against each other in the joint subjected to closing, δ max 0,08 mm It is also interesting that within the discussed range of loads, the joint was characterised by distinct rigidity. The size of the gap formed between board elements of the joint did not exceed 0.08 mm in the course of its closing (Fig.7). 12

13 Fig. 8. Deflection value δ 1P of the dowel joint subjected to closing, δ 1P(max) =0.75 mm. The quality of the elaborated model was assessed by comparing the calculation results of deflections determined on the direction of load action with the results of laboratory measurements. In the case of numerical calculations, the obtained deflection value amounted to 0.75 mm (Fig. 8). For the adequate substitute model (Fig. 9a), the value of the deflection was 1.42 mm, whereas in the case of the realisation of the pattern causing opening of the joint, the deflection in the direction of force action amounted to 1.62 mm (Fig. 9b). The comparison of the obtained results with measurements obtained experimentally revealed that the results of numerical calculations conducted on models with substitute modulus of linear elasticity were closer to these values. a) b) Fig. 9. Size of deflection: a) δ 1P of the joint with the element of substitute rigidity (E Z Z) subjected to closing, δ 1P(max) =1.42 mm, b) δ 2P of the joint with the element of substitute rigidity (E R Z) subjected to opening, δ 2P(max) =1.62 mm 13

14 Table 4 collates values of deflections obtained from calculations and laboratory experiments after the comparison of the quality of numerical models elaborated for the examined joints. Values of numerical calculations obtained from models faithfully representing the shape of the examined joints were distinctly lower in relation to the results of laboratory measurements. On the other hand, the comparison of the calculation results obtained from models containing nodes with substitute modulus of linear elasticity with empirical results indicates that, in their majority, they were slightly higher. In the case of the wall angle dowel joint, the difference in deflections during closing amounted to 4,1% and during opening 3,8%. Table 4. Deflections of the examined joints determined experimentally. Joint Closing test δ 1P [mm] Opening test δ 2P [mm] Experiment Brick model Experiment Brick model True Substitute True Substitute Wall CONCLUSIONS Taking into consideration the performed analysis of the obtained research results it should be emphasised that for virtual (numerical) construction modeling of cabinet furniture dowel joints it is recommended to apply cuboid, six- and eight-node networks of finite elements which should be assigned substitute modulus of linear elasticity determined empirically. REFERENCES 1. DZIĘGIELEWSKI ST.; SMARDZEWSKI J. 1996: Moduł sprężystości połączeń a sztywność mebli skrzyniowych - analiza numeryczna. Materiały VIII sesji naukowej Badania dla meblarstwa. Wydawnictwo Akademii Rolniczej im. A.Cieszkowskiego w Poznaniu. Poznań KASAL A. 2008: Effect of the number of screws and screw size on moment capacity of furniture corner joints in case construction. Forest Prod. J. 58(6): KASAL A., ZHANG J., YUKSEL M., ERDIL Y.Z. 2008a: Effects of screw sizes on load bearing capacity and stiffness of five-sided furniture cases constructed of particleboard and medium density fiberboard. ForestProd. J. 58{10}: KASAL A., ERDIL Y.Z., ZHANG J., EFE H., AVCI E. 2008b: Estimation equations for moment resistances of L-type screw corner joints in case goods furniture. Forest Prod. J.58(9): KŁOS R., SMARDZEWSKI J. 2004: Wielokryterialna optymalizacja mebli skrzyniowych metodą algorytmów genetycznych. Roczniki Akademii Rolniczej w Poznaniu 39: NICHOLLS T., CRISAN R. 2002: Study of the stress-strain state in corner joints and box-type furniture using Finite Elements Analysis (FEA). Holz als Roh- und Werkstoff, 60(2002) SMARDZEWSKI J., KŁOS R. 2004: Możliwości polioptymalizacji mebli skrzyniowych z wykorzystaniem metody gradientowej, Przemysł Drzewny 4(LV): SMARDZEWSKI J., OŻARSKA B. 2005: Rigidity of cabinet furniture with semirigid joints of the confirmat type. Electronic Journal of Polish Agricultural Universities Wood Technology 8(2), # SMARDZEWSKI J., PREKRAT S. 2002: Stress distribution in disconnected furniture joints. Electronic Journal of Polish Agricultural Universities, Wood Technology, Volume 5, Issue 2. 14

15 10. SMARDZEWSKI J., PREKRAT S. 2005: Nonlinear strength model of an eccentric joint mandrel. Drvna Industrija 55(2): SMARDZEWSKI J. 2004a: Modeling of semi-rigid joints of the confirmat type. Annals of Warsaw Agricultural university, Forest and Wood Technology 55: SMARDZEWSKI J. 2004b: Modelowanie półsztywnych węzłów konstrukcyjnych mebli. Monografia pod redakcją B.Branowskiego, P.Pohla. Wydawnictwo AR w Poznaniu , SMARDZEWSKI J. 2005: Nieliniowe modele połączeń zaczepowych. Polioptymalizacja i komputerowe wspomaganie projektowania. Wydawnictwo Uczelniane Politechniki Koszalińskiej. Mielno Streszczenie: Modeling of joint substitutive rigidity of board elements. W pracy przedstawiono alternatywne sposoby numerycznego modelownia sztywności połączeń kołkowych elementów płytowych z wykorzystaniem węzłów o zastępczym module sprężystości liniowej. Wartości ugięć połączeń, uzyskane w drodze badań laboratoryjnych oraz obliczeń różniły się o 3-4%. Corresponding authors: Jerzy Smardzewski, Robert Kłos Uniwersytet Przyrodniczy w Poznaniu, Wydział Technologii Drewna Wojska Polskiego 38/ Poznań, JSmardzewski@up.poznan.pl 15

16 Annals of Warsaw University of Life Sciences SGGW Forestry and Wood Technology No 73, 2011: (Ann. WULS-SGGW, Forestry and Wood Technology 73, 2011) Effects of Vacuum Drying with Infrared Heating on Some Properties of Wood SUAT ALTUN, FATIH YAPICI, ZEHRA KORKMAZ Karabuk University, Technical Education Faculty, Department of Furniture and Decoration Education Abstract: Effects of vacuum drying with infrared heating on some properties of wood. Drying of wood is one of the most important industrial processes in wood manufacturing. Mechanical properties of wood are affected from drying process. Heating method of the drying process is the most effective factor on the properties of end product. The aim of this study was to determine the effect of infrared heating in vacuum drying on the drying quality and mechanical properties of wood. For this purpose Oriental beech (Fagus orientalis L.) and Scots pine (Pinus Sylvestris L.) samples were dried by vacuum drying with infrared heating at º C temperatures and at mbar ambient pressure for 48 hours. Final moisture content (MC) distributions, case hardening, modulus of elasticity (MOE), modulus of rupture (MOR) and compression strength (CS) of dried samples were determined. According to the results of the test, although average MC of the pieces met the requirements of the standard drying class, distribution of the individual MC did not meet the requirements in both pine and beech. Vacuum drying with infrared heating increased the MOE, MOR and CS of Scots pine and only CS of beech. MOE and MOR of beech were not affected from drying with infrared heating. Keywords: vacuum drying, infrared heating, drying quality, Scots pine, beech, mechanical properties INTRODUCTION Drying of wood is one of the most important industrial processes in wood manufacturing. There are many methods, which are widespread used to drying lumber such as air drying, shed air drying, forced-air shed drying, warehouse pre-drying, low temperature kiln drying, conventional electric dehumidification kiln drying and conventional steam-heated kiln drying (Denig et.al., 2000). Drying is the most energy-intensive and time-consuming component of the lumber manufacturing process. Currently, over 80% of the produced lumber has been dried in conventional and vent kilns. Although this operation is simple and inexpensive, the product quality is often low and the drying time is very long, especially for hardwoods (Li et al., 2008). An incorrect drying process generates cracking problems, rupture of the cellular structure of the wood, folding in the piece, and in general an inadequate drying stage. These alterations devalue the final product and important losses for the wood industry take place. Drying influences the mechanical properties of wood in three ways, namely through the direct effect of moisture loss, the internal drying stresses and strains. (Taskini 2007) Many studies have been carried out on vacuum drying of wood. Water in wood at the sub-atmospheric pressure can be vaporized and moved at a temperature below C as rapidly as for high temperature drying at the atmospheric pressure. Therefore, vacuum drying has the benefits of high temperature drying without the danger of developing defects with some susceptible species. (Jung et al. 2004) Generally, there are two basic processes (continuous or discontinuous) and three type of heating (by convection, heating plates or microwave) in vacuum drying. In discontinuous vacuum drying the wood undergoes repeated changes, first being heated at atmospheric pressure then demoisturised at low pressure without any heat. There is considerable risk of discoloration owing to the presence of a high oxygen component during heating phase. In continuous vacuum drying there is simultaneous heating and demoisturising most of the time, at reduced pressure; external air is excluded. With convection heating the air speed must be 16

17 suitably high owing to the low density of the drying agent. With direct heating no fans at all required but there are serious problems: heating plates are only used individually and only for small useful volumes owing to their limited size and the very laborious, time consuming job of stacking and removing from the stack. The other direct heating method is the use of microwaves. Apart from the comparatively high capital cost and greater likelihood of malfunctioning it is very difficult to obtain homogenous intensity distribution over the whole stack. (Brunner 1993) Jomaa and Baixeras (1997), Chen (1997), Audebart et al. (1997) and many other researchers studied on vacuum drying with heating by convection or heating plates. Recently, researchers have focused on vacuum drying of wood with radio frequency and microwaves. Koumoutsakos (2001) investigated the energy dissipation coupled with heat and moisture mechanisms in wood during radio frquency/vacuum drying. Li et al. (2008) established a onedimensional mathematical model to describe the process of wood microwave-vacuum drying based on the mechanism of moisture and heat transfer in wood. Hansmann et al. (2008) used the high-frequency energy assisted vacuum drying to improve drying quality of fresh Eucalyptus globulus. Abubakari (2010) studied on the effect of RF heating as preconventional kiln treatment on the drying characteristics and quality of sub-alpine fir lumber. Jung et al. (2004) compared the vacuum drying characteristic of radiate pine using contact heating, radio frequency heating and the combination of both. Möttönen (2006) compared the conventional low temperature drying with vacuum drying in terms of variation in drying behavior and final moisture content. Unlike these studies Perre et al. (2004) used infrared (IR) heating in vacuum drying of wood. They noted that IR heating could be used successfully in the vacuum drying. In most of these studies drying rate and moisture distributions were reported. Very little studies included the effect of drying specifications on the drying quality and mechanical properties of wood. Taskini (2007) made a comparison between microwave, infrared, and convectional drying effects on wood strength and reported that drying time of the microwave heating is significantly reduced, while the strength stays higher than that obtained in convectional and infrared drying. RESEARCH OBJECTIVE The aim of this study was to determine the effect of infrared heating in vacuum drying on the drying quality and mechanical properties of wood. For this purpose Oriental beech (Fagus orientalis L.) and Scots pine (Pinus Sylvestris L.) samples were dried by vacuum drying with infrared heating and final moisture content (MC) distributions, case hardening, modulus of elasticity (MOE), modulus of rupture (MOR) and compression strength (CS) of dried samples were determined. MATERIAL AND METHODS In this study Oriental beech (Fagus orientalis L.) and Scots pine (Pinus Sylvestris L.) were used as a wood material. Samples were sawn with a thickness of 25 mm, a width of 100 mm and a length of approximately 3000 mm. The samples were of best quality without any cracks, cell collapse, warp, or obvious discoloration. Then they were divided into two parts and one part of each sample was dried in a conventional kiln by a commercial lumber supplier according to their professional experience. These samples were used as the control group to determine the mechanical properties. Other parts of samples were dried with vacuum drying with infrared heating in a laboratory scale vacuum chamber. Before drying, 200 mm were cut off on both ends of the boards in order to remove any pre-dried wood and due to manipulation damaged material. A sample with a width of 50 mm was cut of at the middle of each board for determining the initial moisture content (MC) by oven drying at 103 º C and for determining 17

18 the density. Because of smallness of the pilot vacuum drying chamber, samples were cut into a length of 500 mm before drying. Oven dry densities, initial and final MC and drying rate of the samples were given in Table 1. Samples were dried in a sealed vacuum chamber. Heating for drying were supplied by 4 ceramic infrared heaters having 2-10 µm wavelength and 16 kw/m 2 surface ratings and were placed at the two sides of the stack. Ambient temperature of the chamber was measured at the top of the stack, and core temperature of the wood was measured from samples in the middle of the stack. Measuring probe was inserted into half of the thickness of the sample. It was measured 10 º C difference between chamber and core temperature of samples during drying. Vacuum was supplied by an oil vacuum pump. Table 1. Oven dry densities, initial and final MC and drying rate of the samples Wood Oven dry density Initial MC Final MC Drying time Drying rate (g/cm 3 ) % % h %/day Beech 0, Scots pine 0, Stack was placed in the chamber and dried at mbar absolute pressure and º C (core temperature of wood) temperature for 48 hours. After drying, each wood sample was weighed and inspected for occurrence of any drying defects in the form of surface checking. Drying quality of the process was assessed according to TS EN (2006) standard. Samples were marked and cut into wood sections to determine the average final moisture content, the MC distribution, for internal checking and case hardening. Wood sections were quickly weighed with a digital balance having sensitivity 0.01 g to determine the individual MC. Case hardening of the samples was determined according to TS ENV (2005) standard. Effects of vacuum drying with infrared heating on the mechanical properties of wood were investigated. Modulus of elasticity (MOE), modulus of rupture (MOR) and compression strength were determined according to TS 2478 (1976), TS 2474 (1976) and TS 2595 (1977), respectively. RESEARCH RESULTS Beech and Scots pine wood were dried by using vacuum drying with infrared heating for 48 hours. No visual defects were detected on the surface of the dried samples. After sectioning of the samples, each section was inspected and there was no internal check. For beech and pine, the average final MC values were measured as 12% and 10% and drying rate were calculated as 17 %/day and 18 %/day, respectively. The individual MC distribution was between 7.2% and 12.9% in pine samples and between 6.5% and 18.5% in beech samples. The MC distribution across the length and width of the vacuum dried samples are given in Figure 1. According to TS EN standard, allowable range of average MC around the target MC (12%) is ±1.5% and 93.5% of the pieces should have individual moisture content between 15.6% and 8.4%. According to measurements 87.5% of the pine pieces and 75% of the beech pieces had individual MC between this ranges. Although average MC of the pieces met the requirements of the standard drying class, distribution of the individual MC did not meet the requirements in both pine and beech. As can be seen in Figure 1, average MC at the middle of the beech samples was much higher than that of at the end of and outer side of the samples. That is why the distribution of the individual MC exceeded the allowable range for 18

19 the standard drying class. Although the MC difference between inner and outer side of pine samples was relatively low, it was also out of allowable rage Scots pine Beech 15 Average MC (%) end middle outer side middle Across the Length Across the Width Figure 1. MC distribution across the length and width of the vacuum dried samples Case-hardening gap (mm) Beech Scots pine 0 h 24 h 48 h Measurement time (hour) Figure 2. Casehardening gap values of vacuum dried wood samples Casehardening gap values of vacuum dried samples with infrared heating are illustrated in Figure 2. Casehardening gap of beech was higher than that of pine. The gap increased with time. According to Abubakari (2010) casehardening is a condition of stress and set in wood in which the outer fibers are under compressive stress and the inner fibers under tensile stress, after drying. Casehardening is caused by too rapid or uneven drying as a result of too high temperature or too low relative humidity or large fluctuation of both. According to these results it can be said that vacuum drying of Scots pine and beech with infrared heating is not proper enough in terms of drying quality assessment. Mechanical properties of wood are affected from drying process. Because of this, the selection of drying process is very important. The most important mechanical properties of wood are modulus of elasticity (MOE), modulus of rupture (MOR), and compression strength (CS) in the many applications. According to the test results, mean and standard deviation values of these mechanical properties are given in Figure 3. It has been seen that the drying process used in this study affected the MOE, MOR and CS values. It has been found that the MOE, MOR and CS values of the test samples varied between N/mm 2, N/mm 2, and N/mm 2, respectively. The variance analysis was applied on data belonging to these mechanical properties of samples, and the results were shown in Table 2. 19

20 (241.09) ( ) (491.91) 9559 (818.06) 1000 (N/mm 2 ) (2.69) (19.05) (6.08) (7.99) (3.15) (4.82) (4.42) (3.43) 10 1 Control VD Control VD Control VD Control VD Control VD Control VD Scots pine Becch Scots pine Becch Scots pine Becch Modulus of Elasticity Modulus of Rupture Compression Strength Figure 3. Mechanical properties of Scots pine and Beech wood (VD indicates Vacuum Drying and Values in the parenthesis indicate standard deviation) Table 2. The results of variance analysis MOE (N/mm 2 ) MOR (N/mm 2 ) CS (N/mm 2 ) Source F P<0,05 F P<0,05 F P<0,05 Corrected Model , , ,000 Intercept , , ,000 A: Wood species , ,000 B: Drying types , ,000 AxB , ,000 According to the variance analysis, it has been seen that while the effects of drying types, wood species and interaction of them on both modulus of rupture and compressive strength properties of wood were found statically significant at 95% significance level. However, just only the effect of wood species used in this study on the modulus of elasticity value of test sample was not found at 95% significant level. To comparisons of these means were done by employing a Duncan test and the results are given in Table 3. Table 3.The results of Duncan Test MOE (N/mm 2 ) MOR (N/mm 2 ) CS (N/mm 2 ) Experimental Experimental Experimental design Mean HG design Mean HG design Mean HG Scots pinecontrol A Beech-control A Scots pinecontrol A Beech-vacuum drying A Scots pine-control A Beech-vacuum drying B Beech-control A Beech-vacuum drying B Beech-control B Scots pinevacuum drying B Scots pine-vacuum drying C Scots pinevacuum drying C It was shown that vacuum drying with infrared heating did not affect the MOE of beech wood. According to the Duncan test there was no statistical difference between MOE values 20