Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Relating pressure response profile during hot pressing to property development in flakeboard Feng, Yan 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1994-0189.pdf [ 4.16MB ]
Metadata
JSON: 831-1.0075242.json
JSON-LD: 831-1.0075242-ld.json
RDF/XML (Pretty): 831-1.0075242-rdf.xml
RDF/JSON: 831-1.0075242-rdf.json
Turtle: 831-1.0075242-turtle.txt
N-Triples: 831-1.0075242-rdf-ntriples.txt
Original Record: 831-1.0075242-source.json
Full Text
831-1.0075242-fulltext.txt
Citation
831-1.0075242.ris

Full Text

RELATING PRESSURE RESPONSE PROFILE DURING HOT PRESSING TO PROPERTY DEVELOPMENT IN FLAKEBOARD by YAN FENG B.Eng., Beijing light industry institute, 1983 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Forestry) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April 1994 © Yan Feng, 1994 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of fO&tS | flyJ The University of British Columbia Vancouver, Canada Date April " , H H DE-6 (2/88) ii ABSTRACT The pressure response profile of a flakeboard during hot pressing was exclusively studied using variables of flake species (aspen and lodgepole pine), flake types (small-size flakes and large-size flakes), resin contents (0 and 8 percent), moisture contents (2 and 6 percent), and press closing times (30, 60 and 90 seconds). This pressure response profile was found to reflect the panel's overall rheological behaviour in relation to each of the variables. Pressure response was related to the development of the panel properties such as vertical density gradient, modulus of rupture, modulus of elasticity, internal bond and dimensional stability, as well as the development of in situ adhesive cure. The development of a panel vertical density gradient was found to depend on the panel overall rheological behaviour during hot pressing and could be well determined either by mat pressure response or by mat displacement response. The development of the panel modulus of rupture, modulus of elasticity, and internal bond, on the other hand, depended not only on the development of the vertical density gradient, but the adhesive cure as well. The in situ adhesive cure could be determined by mat displacement response for a panel with limited rheological response during hot pressing. With large rheological response, the mat displacement was not sensitive to adhesive cure. By monitoring system temperature, mat pressure and displacement response, the developments of panel properties and in situ adhesive cure could be satisfactorily explained. The feasibility of altering panel properties by pressure manipulation was also studied. The introduction of a pressure breathing stage during hot pressing was found to improve properties only in panels with 2 percent moisture content. With 6 percent moisture content, panel properties were adversely affected. iii TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iii LIST OF TABLES vi LIST OF FIGURES vii ACKNOWLEDGEMENTS x ABBREVIATIONS USED xi DEDICATION xii 1.0 INTRODUCTION 1 2.0 BACKGROUND 3 2.1 Hot pressing 3 2.1.1 Hot pressing process 3 2.1.2 Pressure response during hot pressing process 4 2.2 Viscoelastic behaviour 7 2.2.1 Viscoelastic behaviour of wood 7 2.2.2 Viscoelastic behaviour of wood components 12 2.2.3 Viscoelastic behaviour of adhesive 14 2.3 Pressure relaxation of flakeboard 16 2.4 Physical and Mechanical properties 18 2.4.1 Vertical density gradient 18 2.4.2 Bending strength and stiffness 20 2.4.3 Internal bond strength 20 2.4.4 Dimensional stability 20 3.0 METHODOLOGY 22 3.1 Experimental design 22 3.2 Materials and preparation 25 3.3 Reproducibility of pressure response 27 3.3.1 Equipment 27 3.3.2 Mat preparation 27 IV 3.3.3 Pressing procedure 27 3.4 The development of in situ adhesive cure 29 3.4.1 Differential Scanning Calorimetry (DSC) 29 3.4.2 Instrumentation 30 3.4.3 Sample preparation 30 3.4.4 Mat preparation 30 3.4.5 Pressing procedure 33 3.4.6 DSC examination 33 3.5 The development of physical and mechanical properties 33 3.5.1 Mat preparation 34 3.5.2 Pressing procedure 34 3.5.3 Property evaluation 34 3.5.3.1 Specimen preparation 34 3.5.3.2 Vertical density gradient 36 3.5.3.3 Bending strength 36 3.5.3.4 Internal bond strength 36 3.5.3.5 Dimensional stability 36 3.6 Exploration of pressure manipulation 38 3.6.1 Mat preparation and pressing procedure 38 3.6.2 Property evaluation 38 3.7 Supplementary experiment 38 3.7.1 Image analysis 39 3.7.2 Instrumentation 39 3.7.3 Experimental procedure 39 3.7.4 Results 41 RESULTS AND DISCUSSIONS 44 4.1 Pressure response curve study 44 4.1.1 Species and flake size effects 48 4.1.2 Moisture content effects 50 4.1.3 Press closing time effects 52 4.1.4 Resin content effects 52 V 4.2 The development of in situ adhesive cure 55 4.2.1 Adhesive cure 55 4.2.2 The development of in situ adhesive cure 57 4.3 The development of panel properties 63 4.3.1 Panel vertical density gradient 63 4.3.2 Panel mechanical properties 70 4.3.3 Panel dimensional stability 78 4.4 The effect of pressure manipulation on panel properties 86 4.4.1 Effect on panel vertical density gradient 86 4.4.2 Effect on panel mechanical properties 90 4.4.3 Effect on panel dimensional stability 90 4.5 Statistical analysis 90 5.0 CONCLUSIONS 96 LITERATURE CITED 98 APPENDIX I 108 APPENDIX H 117 LIST OF TABLES vi Table 3.1 Experimental design 23 Table 3.2 Balanced design 24 Table 3.3 Species and dimensions of flakes used in this experiment 26 Table 4.1 Degree of in situ adhesive cure of phenol-formaldehyde powder resin in three vertical locations of a panel at press times of 1, 2, 4 and 7 minutes 60 Table 4.2 Mean specific MOR, MOE and IB for four pressure drops and different levels of MC, PCT, species and flake sizes 71 (a) Specific MOR 71 (b) Specific MOE 72 (c) Specific IB 73 Table 4.3 Mean percentage of thickness swell for four pressure drops and different levels of MC, PCT, species and flake sizes 79 (a) Thickness swell after conditioned in conditioning chamber reaching 22% EMC (TS22%) 79 (b) Thickness swell after treated in VPS treatment (TSvps) 80 (c) Irrecoverable thickness swell (TSJ 81 Table 4.4 Mean specific MOR, MOE, IB and percentage of thickness swell with variables of pressure breathing and MC 88 Table 4.5 Summary of ANOVA results comparing experimental variables for the properties of specific MOR, MOE, IB and dimensional stability 94 Table 4.6 Summary of ANOVA results comparing pressure breathing and MC for the properties of specific MOR, MOE, IB and dimensional stability 95 LIST OF FIGURES vii Figure 2.1 Pressure changes within the board as a function of time 5 Figure 2.2 Variation of pressure (total resistance to compression of the mat) with time after press closure to stops 8 Figure 2.3 Variation of pressure with time after press closure to stops for boards made with and without resin 9 Figure 2.4 The moisture content dependence of Tg for in situ lignin and hemicelluloses as predicted by the Kwei model 13 Figure 2.5 A typical platen pressure curve denoting specific regions 17 Figure 3.1 A typical pressure response curve 28 Figure 3.2 A DSC capsule fixed in between of two veneers 31 Figure 3.3 The positions of DSC samples in a mat 32 Figure 3.4 Pressure response curves showing each of the four pressure drops at 1, 2, 4, or 7 minutes after reaching maximum pressure 35 Figure 3.5 The size and position of each test specimen in a panel 37 Figure 3.6 Image analysis system 40 Figure 3.7 Intensity histograms for flake images with different resin contents 42 Figure 3.8 Relationship between resin coverage and resin content 43 Figure 4.1 Pressure curve reproducibility study 45 (1) Pressure curves of flakeboards with aspen small flakes, without resin 45 (2) Pressure curves of flakeboards with aspen small flakes, 8% RC 46 (3) Pressure curves of flakeboards with aspen larger flakes, lodgepole pine small flakes and lodgepole pine larger flakes, 8% RC, 30 sec. PCT 47 viii Figure 4.2 Species and flake size effects on pressure curves, 8% RC, 30 sec. PCT 49 Figure 4.3 Moisture content effects on pressure curves, flakeboards were manufactured with aspen small flakes 51 Figure 4.4 Press closing time effects on pressure curve, flakeboards were manufactured with aspen small flakes 53 Figure 4.5 Resin cure and resin content effects on pressure curves, flakeboards were manufactured with aspen small flakes 54 Figure 4.6 A DSC thermogram of phenol-formaldehyde powder resin sample 56 Figure 4.7 Comparison of DSC thermograms of the adhesive sample taken at 1, 2, 4, and 7 minutes hot pressing 58 (1) 6% MC 58 (2) 2% MC 59 Figure 4.8 Degree of in situ adhesive cure of resin sample in three layers of a panel after 1,2,4 and 7 minutes in hot press following application of full pressure 61 Figure 4.9 The monitored temperature changes during hot pressing in three layers in a panel 62 Figure 4.10 Pressure response and mat displacement of flakeboards in relation to sequential pressure drops 64 (1) Flakeboards made with aspen small flakes, 6% MC 64 (2) Flakeboards made with aspen small flakes, 2% MC 65 (3) Flakeboards made with aspen larger flake,lodgepole pine small flakes and lodgepole pine larger flakes, 6% MC, 30 sec. PCT 66 Figure 4.11 Vertical density profiles of flakeboards corresponding to sequential pressure drops 67 (1) Flakeboards made with aspen small flakes, 6% MC 67 (2) Flakeboards made with aspen small flakes, 2% MC 68 ix (3) Flakeboards made with aspen larger flakes, lodgepole pine small flakes and lodgepole pine larger flakes, 6% MC, 30 sec. PCT 69 Figure 4.12 Development of panel mechanical properties with different MC levels 74 Figure 4.13 Development of panel mechanical properties with different PCT levels 75 Figure 4.14 Development of panel mechanical properties with different species 76 Figure 4.15 Development of panel mechanical properties with different flake sizes 77 Figure 4.16 Development of panel thickness swell with different MC levels . . . . 82 Figure 4.17 Development of panel thickness swell with different PCT levels . . . . 83 Figure 4.18 Development of panel thickness swell with different species 84 Figure 4.19 Development of panel thickness swell with different flake sizes . . . . 85 Figure 4.20 Pressure and mat displacement response curves of flakeboards showing each of the pressure breathing cycle 87 Figure 4.21 Vertical density profiles of flakeboards made with and without pressure breathing 89 Figure 4.22 Pressure breathing effects on panel mechanical properties 91 Figure 4.23 Pressure breathing effects on panel thickness swell 92 X ACKNOWLEDGEMENTS First of all I would like to thank my supervisor Dr. P.R. Steiner for his invaluable guidance, attention, encouragement and help during conducting this research. A special thanks must go to Dr. S.C. Ellis and Dr. S. Avramidis for their support and help as my committee members. I extend my thanks to Mr. A. Sidhu, and Mr. R. Johnson for their helpful technical assistance with the experiment. I also wish to thank CAE Industries (Vancouver, B.C.) for assistance with flake cutting and Alberta Research Council (Edmonton, Alberta) for assistance with the testing of panel vertical density gradient. A grateful acknowledgement is made to Dr. A.H. Buchanan for his support and help while I was working in New Zealand. Finally, a special acknowledgement is given to my mother for her love, care, encouragement and understanding during my graduate career. xi ABBREVIATIONS USED ANOVA AOI CTH DSC EMC IB MC MOE MOR OSB PCT PF RC T TS VDG VPS analysis of variance area of interest constant temperature and humidity differential scanning calorimetry equilibrium moisture content internal bond moisture content modulus of elasticity modulus of rupture oriented strand board press closing time phenol formaldehyde resin content glass transition temperature thickness swell vertical density gradient vacuum pressure soak xii This is dedicated to my dear husband XiXian Deng. 1 1.0 INTRODUCTION In the manufacture of flakeboard, the hot pressing process is the key step controlling the development of many physical and mechanical properties. Studies on this process have been extensively carried out to examine the effects of pressing variables such as temperature, moisture content, press closing speed, adhesive cure, etc. on the ultimate panel properties (Strickler, 1959; Suchsland 1962; Heebink et al., 1972; Plath et al., 1974; Geimer et al., 1975; Kelly, 1977; Smith, 1982; etc.). However, the interactions of these variables and their effects on panel properties are complicated and still not fully understood. During hot pressing, a complex set of conditions is known to exist within a mat. The process of simultaneously consolidating and heating an initially cold and loose-packed mat results in a transient temperature, gas pressure, and moisture content gradient in vertical and horizontal directions within the mat. These gradients, in turn, cause a non-uniform change in both compression properties of wood and resin polymerization. The rheological behaviour of solid wood has been well described by creep and stress relaxation studies (Bodig, et al., 1985). This rheological behaviour is observed to increase with either increasing moisture content, increasing temperature, or a combination of both (Youngs, 1957; Kunesh, 1961). During hot pressing, creep and stress relaxation of individual flakes within a mat occur as functions of time, mat structure, temperature, moisture content, stress level, and degree of adhesive cure. The overall rheological behaviour of flakes within the mat can be represented by the apparent creep, pressure relaxation, and vertical density gradient distribution. Thus, it was postulated that the development of vertical density gradient and the effects of adhesive cure can be detected by monitoring the pressure response curve obtained during hot pressing. The objective of this study was to establish how the in situ pressure response of a mat related to panel vertical density gradient formation, mechanical property 2 development, and dimensional stability properties. In order to accomplish this major objective, the following sub-objectives were considered: (1) The reproducibility of the pressure response curve. (2) The relationship between the pressure response curve and the degree of adhesive cure. (3) The relationships between the pressure response curve and sequential development of vertical density gradient, bending strength, internal bond strength and dimensional stability. (4) The feasibility of altering bending, internal bond properties and dimensional stability by manipulating the pressure response curve. 3 2.0 BACKGROUND 2.1 Hot pressing Many wood composites, such as fibreboard, particleboard, flakeboard, waferboard, OSB, etc., are manufactured by similar production processes, with hot pressing being one of the most critical steps. The basic function of hot pressing is twofold: to compact the loose mat into the desired panel thickness and density, and to polymerize the adhesive in order to bond the wood elements and maintain the mat in this consolidated state when removed from the press. To facilitate the mechanical compaction as well as the chemical reaction, all industrial production presses adopt certain methods of simultaneously compressing and heating the mat. Most commonly, this is accomplished by heated platens which contact the mat surface (Moslemi, 1974; Kollmann, et al., 1975). There are several types of hot presses used in pressing systems and press heating systems. This study will only focus on the single opening flat platen press and contact heating system. While flakeboard will be the main product discussed in this presentation, the pressing principle can be applied to other wood composites. 2.1.1 Hot pressing process Most platen pressing processes for flakeboard consist of three major steps: press closing, thickness control, and press opening. The process appears to be rather simple, but in reality the total interactions of many variables involved in hot pressing are rather complicated and still neither theoretically nor experimentally fully understood. Temperature, moisture content, pressure, press closing speed, and the cure of adhesive are the most important factors which affect panel property development. During press closure, a loose mat of wood flakes is compacted under pressure. As the heat from the hot platens, which have been set to temperatures ranging from 150°C to 225°C, starts to penetrate the cold mat, two phenomena will occur. Heat will 4 be transferred from the mat faces to the core by means of conduction and convention. At the same time moisture in the wood near the heated platens is converted to vapour causing total gas pressure to increase in the outer portions of the mat. This heat laden vapour is driven vertically to the core and horizontally to the edges of the mat. This simultaneous heat and mass transfer creates transient temperature, gas pressure, and moisture content gradients within the mat which in turn result in non-uniform changes in both compression properties of wood flakes and resin polymerization. Press pressures of 3 to 5 MPa are gradually applied which cause good contact among flakes and bring the mat to the desired thickness. After the target mat thickness is achieved, the press pressure required to maintain this thickness declines. Mechanically the mat is undergoing an apparent stress relaxation under constant deformation. Within the mat, the varying transient environmental conditions which exist throughout the thickness cause differential creep and stress relaxation of individual flakes. The flakes in some regions of the mat will continue to densify with time while others recover stored elastic deformation. At the same time, increased resin polymerization levels develop from the outer mat surfaces into the core. On completion of the press cycle, the press slowly opens, and pressure applied to the mat decreases, resulting in a release of any internal gas pressure. Elastic recovery can occur during this period. Additional viscous recovery will occur after the panel has been removed from the press. The elastic and viscous recovery depends on the viscoelastic behaviour of both wood and adhesive during hot pressing (Maku et al., 1959; Suchsland, 1962; Kelly, 1977; Kamke et al., 1988 a, b; Wolcott et al., 1990). 2.1.2 Pressure response during the hot pressing process In 1969, Liiri measured total pressure applied to a particleboard as a function of press time. He found a sharp maximum pressure occurred at the point at which target thickness was attained, followed by an immediate drop to a low pressure at the end of the press cycle: this pattern was obtained at all board densities and all press closing times. The maximum pressure was higher for the higher density boards as shown in Figure 2.1 5 SO | MO "So E Jl density=0.80 g/cm3 1 t l //\o.65 / /A 0.50 ^ j —^___ ; r J * 5 Pressing time (min.) (a) with varying board densities (press closing time: 57 sec.). s 8 •0 JO 20 10 *2 6 s1 53s i I 93s i I i M i l / \ A press closing time=129s i i l ' v / X / i 1 \ \ ' \ / w \ \ \ i 2 3 4/ Pressing time (min.) (b) with varying press closing times (board density: 0.65 g/cm3). Figure 2.1 Pressure changes within the board as a function of time (liiri, 1969). 6 (a). This maximum pressure level decreased and shifted to a later time as the press closing time increased (Figure 2.1 (b)). He therefore concluded that the pressure on the board increased quickly during mat compaction and decreased once the thickness was obtained. Lehmann et al. (1973) obtained similar results for board pressure versus press time using catalyzed urea formaldehyde Douglas-fir flakeboards. The pressure diagrams for boards of three thicknesses at density 0.64 g/cm3 were nearly identical; a maximum pressure of approximately 4 MPa was obtained at a press time of 1 minute. The 0.72 and 0.8 g/cm3 boards attained a pressure of approximately 6 and 6.5 MPa, respectively, at a press time of 1 minute. Heebink et al. (1972) found similar sharp decreases in board pressure and suggested that the minimum press time could be determined as the point at which the required pressure fell below the internal bond strength of the board. Maloney (1977) explained that the tendency for lower pressure at the end of press time was due to two reasons. Firstly, much of the resin was cured and the particles were being held in their set position. Secondly, plastic deformation of wood occurs and densifies the particles, which can be demonstrated by hot pressing a mat containing no resin. Bolton et al. (1989) studied stress development within a mat during hot pressing and pointed out that stress relaxation was one of the major reasons which caused mat resistance (pressure) to decline. During compaction of a flake mat, considerable stress and strain are introduced into the flakes. The stress and strain of individual flakes will be primarily in the compressive and bending modes with some shear mode action. Movement associated with wood fracture may also be involved. The strain results in an apparent net compressive stress in the mat. The maximum stress occurs at the moment when the press closes to stop positions. Because wood is a viscoelastic material, this stress relaxes with time. However this may only be apparent after reaching press stops. 7 During press closure, this relaxation phenomenon is obscured by new strains and stresses developed with increased mat compression. Stress development and relaxation are known to be strongly influenced by moisture content and temperature (Youngs, 1957; Kunesh, 1961). The variation of pressure-time diagrams with changes in moisture content and temperature are shown in Figure 2.2 (a), (b). Shrinkage and swelling of mat flakes associated with in situ moisture change and water vapour pressure also affect stress relaxation. As mentioned earlier, resin cure develops bond strength between flakes and therefore enhances the relaxation of stresses by locking in elastic and delayed elastic strains. A graph showing the pressure-time relation of boards made with and without resin is presented in Figure 2.3. To better understand pressure relaxation and in situ stress relaxation of flakes, an understanding of the viscoelastic behaviour of wood and cure behaviour of adhesive is necessary. 2.2 Viscoelastic behaviour 2.2.1 Viscoelastic behaviour of wood Pressure relaxation during flakeboard hot pressing can be explained by the viscoelastic nature of wood. Like many other viscoelastic material, wood exhibits both elastic and viscous behaviour when subjected to load. The predominant approach to Theological studies of wood has been from the phenomenological point of view. More recently, the viscoelastic behaviour of wood components, specifically the amorphous lignin and hemicellulose matrix in the wood cell wall, were investigated (Kelly, et al., 1987; Wolcott, et al. 1990). Rheology is the study of time dependent deformation and flow of material resulting from the application of deforming forces (Pentoney, et al. 1962). From the mechanical point of view, the influence of time on the stress-strain characteristics of a viscoelastic material can be reflected by the three dimensional stress-strain-time diagram (Bodig, et al. 1985, Chapter 5). 8 p> s 0 0 - M 0 100 200 500 400 500 Time after press closure (s) (a) with varying moisture content (Temp.=160oC, s.g.=0.65). ool -60 0 » 0 100 KO 400 WO Time after press closure (s) (b) with varying temperature (MC=11%, s.g.=0.65). Figure 2.2 Variation of pressure (total resistance to compression of the mat) with time after press closure to stops (Bolton et al. 1989). 9 rt 3 CO P Time after press closure (s) Figure 2.3 Variation of pressure with time after press closure to stops for boards made with and without resin (Bolton et al. 1989). 10 The rheology of wood may be characterized by linear viscoelastic behaviour in which the elastic effects obey Hooke's Law and viscous effects are Newtonian. During creep, the total deformation (or total creep) 5t can be expressed by the sum of three components, based on linear viscoelastic theory: an elastic part 5e, a delayed elastic part 5^, and a viscous component 5^ 8t = 6e + 5de + 5V The elastic deformation component 5e is an instantaneous deformation upon loading and is recoverable. The delayed elastic deformation 5^ is time-dependent and is recoverable with time, while the viscous deformation is permanent and irrecoverable. Stress relaxation is governed by the same time-dependent characteristics of the material as creep. While the total deformation during stress relaxation is kept constant, the magnitude of the elastic, delayed elastic, and viscous components alters as time progresses. Initially all the deformation is elastic. With elapsing time, the elastic deformation subsides as the delayed elastic and viscous components respond. At any particular time, however, the sum of the three deformation components is constant. Removal of the residual force, reduces instantaneously the elastic deformation to zero. Kitazawa (1947) studied the relaxation of clear wood specimens under compression perpendicular to the grain. Based on linear viscoelastic theory, he developed an empirical equation for the stress relaxation given by: F = Fl ( 1 - m log(t)). Where F = stress at time t minutes, F! = stress at time 1 minute, m = relaxation coefficient, and t = relaxation time (t> 1). He found an increase in the relaxation coefficient m indicated an increase in plasticity. He also found a negative correlation existing between specific gravity and the relaxation coefficient m. Youngs (1957), Kunesh (1961), Pentoney et al. (1962), Grossman (1963), and Ranta-Maunus (1975) reported the effects of temperature and moisture content on the creep and stress relaxation of wood. In a study of creep and stress relaxation of red oak in compression perpendicular to the grain, Youngs discovered that both creep and stress relaxation were increased by either increasing moisture content, temperature or both. He 11 found the effect of raising the temperature from 27 °C to 82 °C at constant moisture content was considerably greater than the effect of raising moisture content from 12 percent to the green condition at constant temperature. Moreover, the combined influence of high temperature and high moisture content provided the greatest effect. Kunesh (1961), working with relatively large strains in compression perpendicular to grain, found that the rate of relaxation was greater at 49°C than at 27°C, but there was little difference between the rates at 49°C, 71 °C, and 93°C. The rate increased as the moisture content was raised from 6 to 18 percent, but in green material the rate declined. Bariska (1985) studied the creep of wood specimens at low moisture contents of 2 to 3 percent in tension, compression, and shear. He proposed that in compression perpendicular to grain, creep was affected by various causes including micro failures due to buckling of cells, widening of wrinkles or severing of collapsed cells. Schniewind (1968) presented a critical review on wood rheology studies. The effects of temperature and moisture content on creep and relaxation in wood can be distinguished between direct effects and interaction effects. In the former case, the temperature and moisture content are constant during the stress (or strain) history of the experiment, as carried out by Youngs (1957) and Kunesh (1961). In the later case, there is an interaction between stress, temperature, and moisture content histories (Schniewind, 1968). Changes in wood temperature which occurred simultaneously with moisture content variations resulted in complex changes in creep and stress relaxation. Arima (1974 a) studied the effect of temperature and moisture contents on Theological behaviour of wood during hot pressing. Tang et al. (1990) studied the perpendicular-to-grain Theological behaviour of loblolly pine during press drying1. These two studies showed that 1 "Press drying" - is a wood drying technique and can be defined as the application of heat to opposite faces of a board by heated platens to remove moisture from the board. Press drying temperatures generally range from 120°C to 230°C and platen pressures are from 0.2 to 0.5 MPa. During drying, heat is transferred mainly by conduction, from the platens to the wood causing air in the wood to expand and water to vaporize. The vapour and liquid then escape through the surfaces, if the ventilated cauls are used, and through the edges. 12 rheological properties are a function of pressure, temperature, moisture content, and time. This process is dynamic with wood temperature gradually increasing and the moisture content gradually decreases and these changes take place simultaneously during loading. The rheological behaviour is complex and the linear viscoelastic theory can no longer be applied. Tang et al. (1990) developed a viscoelastic model to describe the perpendicular-to-grain rheological behaviour as a function of pressure, temperature, and press drying time. Youngs (1957) observed that increasing stress levels resulted in increased relaxation rates while recoverable deformation appeared to increase linearly with increasing stress under the same conditions of temperature and moisture content. 2.2.2 Viscoelastic behaviour of wood components From the molecular structure point of view, wood is a unique composite material which consists of three structural polymers: cellulose, hemicelluloses, and lignin. The overall viscoelastic responses of wood to external stresses or strains can be rationalized from the interactions between the individual polymers. Kelley et al. (1987) investigated the viscoelastic properties of wood using dynamic mechanical thermal analysis (DMTA) and differential scanning calorimetry (DSC). Two separate glass transitions (Tg) were identified with both techniques and were attributed to amorphous lignin and hemicelluloses. Like all other amorphous polymers, lignin and hemicelluloses can exhibit a range of properties from viscous fluids to linear elastic solids depending on the temperature, diluent concentration, and time (Ward, 1983). Several researchers have studied the effect of moisture content on the Tg of extracted hemicelluloses and lignin (Goring, 1971; Irvine, 1984). Kelley et al. (1987) also studied the variations in Tg of hemicelluloses and lignin in situ with moisture content using the Kwei model. The Tg of lignin decreases with increasing moisture content and begins to plateau at 70°C at 10 to 15 percent moisture content. For hemicelluloses, Tg will continue to decrease with increasing moisture content until it reaches a value of approximately -20°C near 30 percent moisture content. Figure 2.4 shows such variations of Tg with moisture content calculated from the Kwei equation. In solid wood , the decrease in mechanical properties associated with the phase changes will be different from hemicelluloses and lignin. Because of the spiral winding and reinforcing nature of the 250 225 200 175 o 0,150 CD 3 1 2 5 LU § 1 0 0 H 2 7S LU ^ 50 LU H 25 0 -25 - 5 0 0 --ft \ \ ^ ^ Lignin ^ Hemicellulose i i i i i i 5 10 15 20 25 30 35 Moisture Content (%) Figure 2.4 The moisture content dependence of Tg for in situ lignin and hemicelluloses predicted by the Kwei model (Kelley et al. 1987). 14 cellulose microfibrils, the decrease in modulus will be less than that individual polymer undergoing the phase change. 2.2.3 Viscoelastic behaviour of adhesive It is believed that the rheological behaviour and cure development of wood adhesives during hot pressing play another major role in the pressure relaxation. During hot pressing, thermosetting adhesives cure by transforming from relative low molecular weight prepolymers to a highly cross-linked, three-dimensional system. Physically, the adhesive changes from a liquid or powder form to a rubbery state, and finally to a rigid solid form. The cured adhesive bonds the wood elements together which is believed to enhance the relaxation of stresses by locking in the elastic and delayed elastic strains. This will in turn influence the apparent pressure relaxation response. Many efforts have been made to understand the curing behaviour of wood adhesives in a mat during hot pressing. Most of this research has been done by an indirect approach based on spectroscopic analysis, thermal analysis, dynamic mechanical analysis of the degree of cure of the adhesive (Pillar 1966; Chow et al., 1969, 1972; Bolton et al., 1977; Steiner etal., 1981, 1987; Young, 1986; Geimeretal., 1990, 1991). In general, the degree of cure increased with higher temperature and longer heating time. In a study of the cure of neat phenol formaldehyde resin (PF) in a three-ply plywood panel under several pressing temperatures and times, Chow et al. (1969) found that there was a critical temperature (110°C) above which the degree of cure increased rapidly and linearly with time. Below this temperature, the degree of cure developed in a different pattern. It initially increased to about a 27 to 35 percent level, then apparently decreased before increasing again. The lower the temperature, the more significant the apparent decrease was. This phenomenon was attributed to the water which was present in the resin or was produced from the condensation polymerization reaction. Since bonding pressure of 1.4 MPa was used and the temperature was below 100°C, water remained in the glue line. This excessive water would dilute the reactants and hence a sharp drop in cure rate could occur. In a further study in 1972, Chow et al. found the degree of cure of the adhesive was linearly related to moisture content at 15 temperatures ranging from 100 to 160°C. The higher the adhesive moisture content, the lower is the degree of cure. The rheological behaviour of wood adhesives was first extensively examined by Steiner et al. (1981). A Torsional Braid Analysis (TBA) technique was used in their study. The rheological behaviour of the adhesives was represented by the relative storage modulus (rigidity) and relative loss modulus (damping) values. These two values for a plywood PF resin were found to develop nonlinearly with an increase of temperature of 3°C/min.. A constant but low rigidity was observed in the temperature range of 25 to 80 °C indicating very low levels of cure had developed at this stage. Between 80 and 120°C, a minor increase in rigidity was found and above 120°C the rigidity increased dramatically as cure took place. At the same time, damping began to decrease. A rapid decrease in damping took place when the temperature was above 140°C indicating a high degree of cure had been reached. The study of Theological behaviour of the wood adhesives was further developed by Young (1986) and Geimer et al. (1990, 1991) using Dynamic Mechanical Analysis (DMA) techniques. The damping level achieved as the adhesive cured was defined as the percentage of mechanical cure. A study which correlated mechanical cure (from DMA), chemical cure (from DSC), and bonding strength development (IB) by Geimer et al. (1990, 1991) showed that mechanical cure developed at a much higher rate than chemical cure for most phenolic resin, while their bonding property (IB) correlated more to mechanical cure. This result indicated that bonding properties could be well developed prior to the fully chemical cure of the adhesive. Through part of this literature review, some information about chemical, physical, and mechanical aspects of the curing process in wood adhesives has been compiled. Pressure, as an important pressing parameter in wood composite manufacture, seemed to have no direct effect on the cure of adhesives. As mentioned earlier, these studies were based on indirect approaches and the experimental works were conducted either at isothermal or controlled conditions. In reality, cure development of the adhesive in flakeboard will be 16 more complex due to the dynamic and rapid changes in temperature, moisture content, and vapour pressure. The rheological behaviour and the curing process of the adhesives could differ from the ones observed above. Therefore, it appears that an in situ examination of cure development is needed. This was undertaken as part of this research study as described in section 3.4. 2.3 Pressure relaxation of flakeboard Knowledge of the viscoelastic behaviour of wood and wood adhesives provides valuable insight into the fundamental behaviour of wood composites during the hot pressing process. Based on the viscoelastic behaviour of solid wood, Kunesh (1961) first introduced these concepts into the wood-composites-hot-pressing-process, and pointed out that hot pressed wood materials behaved inelasticly. Since then, several viscoelastic theories have been applied to wood composite manufacture (Laufenburg, 1986; Kelley et al., 1987; Wolcott et al., 1990). These rheological behaviour of wood and wood adhesives could be helpful in explaining the creep and pressure relaxation phenomena in flakeboard which occurred during hot pressing. In the early stage of the hot pressing, the rheological behaviour of the wood mat was dominated by the behaviour of wood elements and mat structure. As described earlier, with the application of heat to the mat during the pressing operation, the stress-strain relationship in compression will change as functions of time, temperature, and moisture content. Flake layers will be softened and deformed successively from the mat face towards the core resulting in an apparent creep of the mat under constant pressure until the mat reaches its target thickness. At this thickness, the rheological behaviour of the mat will be governed by stress relaxation at constant deformation and result in an apparent pressure relaxation (Suchsland, 1962). The total pressure relaxation period can be divided into two regions according to Wolcott et al. (1990): the transient relaxation region, which starts when the pressure on the mat decreased rapidly with time; and the asymptotic relaxation region which begins when changes of pressure with time decrease to a relatively lower rate (Figure 2.5). Wolcott et 17 u at z> vt </i u tr o. z A B J C 0 rJA i ~1V PRESS TIME Figure 2 .5 A typical platen pressure curve denoting specific regions. A: press closure; B: transient relaxation; C: asymptotic relaxation; D: venting (Wolcott, et al . , 1990). 18 al. (1990) related the pressure relaxation of a flake mat during hot pressing to the formation of vertical density gradient in the board and explained that the density gradient might still continue to actively form after press closure in the transient relaxation region. In this region, the density gradient is formed from differential stress relaxation of individual flakes throughout the thickness of the panel. However, the role that the stress relaxation plays during the asymptotic relaxation region has not yet been addressed clearly. As both temperature and moisture content gradients exist in the mat and they change with press time, the non-uniform cure of adhesive must be taken into account when relating the pressure relaxation to the density gradient formation. 2.4 Physical and Mechanical properties Traditionally the physical and mechanical properties of wood composites have been determined after they are manufactured. In order to understand the relationship between the pressure relaxation response and panel property development, a literature study was undertaken on panel physical and mechanical properties. 2.4.1 Vertical density gradient The transient environmental conditions within the mat during hot pressing strongly influence the physical and mechanical properties of flakeboard. Vertical density gradient is a typical example. If moisture and heat are equally distributed throughout the mat and remain equally distributed throughout the entire press cycle, the stress-strain-time relationship for all flakes in a homogeneous flakeboard will be equal and no vertical density gradient will result (Suchsland, 1962). However this is not the case in flat panel hot pressing. The transient temperature, vapour pressure and moisture content gradient plasticize the wood in a non-uniform manner through the thickness of the mat by influencing the wood components glass transition Tg. This plasticization takes place first in the face layers. With a fast press closing time, the mat reaches the desired thickness before sufficient heat is transferred from the platens to the mat interior. The plasticized faces compress to a higher density under pressure. The core, still having low to moderate temperature, resists the pressure. When sufficient heat reaches the core layers, the face layers are already at higher density and the 19 cured resin has set these layers at the higher density. As a result, there is insufficient pressure exerted on the core to permit compression to a higher density at a given thickness (Strickler, 1959; Suchsland, 1962; Maloney, 1977). The relationship between press closing time and vertical density gradient was studied extensively by Strickler (1959), Suchsland (1962), and Smith (1982). Very fast press closing speed, associated with high initial pressure resulted in a sharp vertical density gradient with high density board surfaces and lower density core. Slower press closing speeds developed lower initial pressure which allowed the interior of the panel to attain a high temperature and reduced compressive strength before final thickness was reached. The surface and interior portions of the panel would be more equal in density. It was concluded that the slower the press closing speed, the lower the vertical density gradient (Suchsland, 1962). The press closing speed must not be too slow since adhesives used in flakeboard production are temperature sensitive. The polymerization of adhesives takes place rapidly at elevated temperature. Longer press closing times will cause the adhesive on the flakes in contact with the hot platens to polymerize before sufficient interflake contact has occurred. This phenomenon is referred to as precure. Heebink et al. (1972) examined the effect of various processing factors on average density and vertical density gradients in particleboard. Variables involved were species, particle geometry, board average density, board thickness, moisture content, moisture content distribution, press time, temperature, and press closing speed. It was found that moisture content and distribution within the mat were the most important variables affecting the vertical density gradient. Press closing speed and press temperature also contributed to the vertical density gradient; all other variables, however, had secondary effects. His results were in general agreement with those of Strickler (1959): fast press closing results in high density board faces and low density core. Increasing moisture content, results in a very severe density gradient, while higher press temperature increases the core density because of fast heat transfer into the core. 20 2.4.2 Bending strength and stiffness Vertical density gradient in flakeboard substantially influences strength properties, as demonstrated by many researchers (Strickler, 1959; Rice et al., 1967; Heebink et al., 1972; Plath et al., 1974; Geimer et al., 1975; Kelly, 1977). Certain properties, most notably bending strength are significantly enhanced by the presence of this vertical density gradient because bending stress is the highest at the panel surface. Many observed flakeboard properties attributed to various processing parameters may be the result of differences in vertical density gradient. Rice (1960) found by increasing the mat moisture content into the press from 9 to 15 percent for sweet gum flakeboard, MOR increased 18 percent. Heebink et al. (1972) found surface densification increased as the press closing speed increased and a direct relationship between MOR, MOE and surface density. Bismarck (1974) reported an increase in MOE as speed of press closure was increased from 15 to 30 mm/min.. These results indicated that different strength values can be achieved from similar average board density simply by changing the processing variables affecting the vertical density distribution. 2.4.3 Internal bond strength Vertical density gradient adversely affects internal bond (IB) strength. Flakeboard with a well-cured adhesive bond will normally fail near the mid point of the panel thickness when stressed in tension perpendicular to the board surface. This is usually the lowest density region in a hot pressed panel. Therefore depending upon which properties are most critical in the ultimate application of the panel, modifications of the pressing parameters may be justified either to enhance or to restrict the formation of this gradient (Kelly, 1977). Most researchers have found higher IB values can be achieved by increasing board average density, resin content, press time and temperature. 2.4.4 Dimensional stability Like solid wood, flakeboard shrinks and swells when subjected to environmental exposure. This dimensional change is more severe than with solid wood especially in the thickness direction (Kollman et al. 1975). Severe thickness swell is found to be caused by 21 two factors: hygroscopic swelling and compressive stress relief, often referred to as springback. The total thickness swelling for flakeboard exposed to moisture, thus, is the sum of these two components. During hot pressing, at least a portion of the wood flakes experience compressive failure. Moisture content reductions in the mat reduces the plasticity of the wood and results in a semipermanent set of these compressive stresses. If panel moisture content increases, the wood will plasticize allowing these stresses to be relieved, producing an expansion in the thickness direction. Redrying of the flakeboard will cause shrinkage in wood flakes due to desorption, however, the thickness swell due to the compressive stress release will be irrecoverable (Neusser et al. 1965; Beech 1975). Higher density boards will possess more compressive set than lower density boards when both are made with the same wood furnish. Therefore, it is not surprising that thickness swell is normally increased with increasing board density (Childs 1956; Halligan et al. 1972). Childs (1956) studied the effects of density, resin content, and width of particles on the springback of particleboards made of sweet gum particles bonded with urea-formaldehyde resin. The springback of the board was independent of particle width, but it increased with increasing board density and decreased with an increase in resin content. Increased board density results in more compressive failure of particles which tends to springback upon rewetting. Increased resin content results in improvement to interparticle bonding and, therefore, will stabilize the component particles, reducing springback as well as thickness swelling. Saito (1972) found boards bond with phenolic resin were more stable than boards bonded with urea formaldehyde resin. Thickness swelling results from the flat pressing process. However it can be reduced by adjustments in some pressing variables. Attempts have been made to reduce springback by reducing either the hygroscopicity of the flakes or reducing the compressive set in the final board. So far it has been found that steam and/or heat treatment of the cured particleboard seems to effectively reduce springback in phenolic bonded boards (Heebink et al. 1969; Halligan et al. 1972; and Beech, 1975). 22 3.0 METHODOLOGY This study was undertaken to gain a better understanding of mat pressure relaxation during hot pressing and its relation to panel property development. It was divided into four parts: (1) the reproducibility of pressure response and its general evaluation, (2) adhesive cure development, (3) board property development, (4) pressure manipulation experiment. The properties examined included vertical density gradient (VDG), adhesive cure, bending strength (MOR) and stiffness (MOE), internal bond (IB) and dimensional stability. 3.1 Experimental design Throughout the experiment, five experimental variables were involved, two species (aspen and lodgepole pine), two flake types (small-size flakes and larger-size flakes), two resin content (RC) levels (0 and 8 percent), two moisture content (MC) levels (2 and 6 percent), and three press closing times (PCT) (30, 60 and 90 seconds). The experiment was arranged in a partially factorial design shown in Table 3.1 with six overlapping complete factorial portions, which are referred as six sub-designs (SD) shown in Table 3.2. In the sub-designs, SD, and SD2 were aimed at studying the reproducibility of pressure response and to evaluate the general pressure response with respect to various experimental variables. SDj provided comparisons of pressure response for two levels of RC, MC and three levels of PCT. SD2 provided comparisons for different species, and flake types and MC. SD3 was included to examine the development of in situ adhesive cure. SD4 and SD5 were designed to study the development of board properties. SD6 was proposed to empirically test the feasibility of altering panel properties by manipulating pressure response. The pressure manipulation was undertaken by means of pressure reduction which will be described in detail in section 3.6. A total of a hundred and forty-seven 250 x 250 x 12 mm boards were manufactured which included three replicates for each of the treatments. Target specific gravity for all Table 3.1 Experimental design. 23 Code ASL-30 ASL-60 ASL-90 ASH-30 ASH-60 ASH-90 ASRL-30* ASRL-608 ASRL-90 ASRH-30* ASRH-60* ASRH-90 ALRL-30 ALRH-30 LSRL-30 LSRH-30 LLRL-30 LLRH-30 Pressure drops D, 3" 3 3 3" 3 3 (3) (3) (3) D2 3R 3 3 3R 3 3 3 3 3 D3 3" 3 3 3R 3 3 3 3 3 DF 3 3 3 3 3 3 3R 3 B 3 3" 3B 3 3 3 3 3 3 3 Notes: In the code, each character represents each of the factor levels: (i) the first character represents species, A = aspen, and L = lodgepole pine; (ii) the second character represents flake types, S = small-size flakes, L = larger-size flakes; (iii) the third character represents RC levels, it is omitted in the first six codes to mean 0% RC, in the other codes, R = 8% RC; (iv) the fourth character represents MC levels, L = 2% MC, and H = 6% MC; (v) the number following the dash represents each of the PCT levels, 30, 60 and 90 second PCT. (vi) Number 3 represents three replicates. Those within brackets could not be practically manufactured in the experiment, (vii) The superscript R indicates where degree of adhesive cure was examined, (viii) The superscript B indicates where pressure manipulation was carried out. Table 3.2 Balanced design Species aspen aspen lodgepole pine aspen aspen aspen lodgepole pine aspen Flake type small small larger small larger small small small larger small larger small RC 0 8 8 8 8 8 8 MC 2 6 2 6 2 6 2 6 2 6 2 6 2 6 2 6 6 2 6 PCT 30 60 90 30 60 90 30 60 90 30 60 90 30 30 30 30 30 30 30 30 30 30 30 60 90 30 60 90 30 30 30 30 60 60 Pressure drops D P DF DF DF Dp Dp DP Dp D P DF Dp Dp Dp Dp Dp DF Dp DF DF Dp D, D2 D3 Dp D, D 2 D3 Dp D, D2 D3 Dp D, D2 D3 Dp D, D2 D3 Dp D, D 2 D 3 Dp D, D 2 D3 DF Dj D 2 D3 Dp D, D 2 D3 Dp (D.) D 2 D 3 Dp (D.) D2 D3 Dp (D,) D2 D3 Dp B, B2 Dp B, B2 Dp 1 Sub-design code SD, SD2 SD3 SD4 SD5 SD6 Note: Pressure drop D, within brackets could not be practically carried out in the experiment. 25 boards was 0.65 based on dry wood weight and resin content. The general pressing parameters were: Pressing temperature 200°C Maximum pressure 3.4 MPa Pressing time 7 minutes2 Target thickness 12 mm. 3.2 Materials and preparation Aspen small-size flakes were selected as the primary test material in the study. In addition, aspen larger-size flakes, lodgepole pine small-size flakes, and lodgepole pine larger-size flakes were selected as secondary test material. The geometric size of these designated materials are shown in Table 3.3. Flake sizes smaller than typical commercial flakes were chosen in order to eliminate edge effects and problems of flake bridging when using a small forming box. Aspen and lodgepole pine commercial flakes of average size 100 x 100 x 0.7 mm and 6 to 8 percent MC were used to produce the small flakes. These flakes were processed through a laboratory grinder which had a 10 mm diameter mesh screen. The ground flakes were then passed over a 5 x 5 mm mesh to eliminate fines. The aspen and lodgepole pine larger-size flakes were cut from logs to desired flake length and thickness using a laboratory flaker at CAE Industries in Vancouver B.C. The cut flakes were then hand split to the desired width. Flakes were either dried in an over at 80°C to give 2% MC or conditioned at 25°C, 50% relative humidity (RH) to give 6% MC. They were stored in polypropylene bags to maintain these MC conditions until panel manufacture. A phenol formaldehyde (PF) powder resin, W3154N (BORDEN Chemicals) applied at resin level of 8 percent to dry wood weight was used throughout the study. This resin content level was selected to ensure full resin coverage on the flakes. A study of the relationship between resin content and resin coverage was carried out in a supplementary experiment. 2 The pressing time is measured from the moment when maximum pressure is reached. The sum of pressing time and press closing time is defined as press time. Table 3.3 Species and dimensions of flakes used in this experiment. Species aspen lodgepole pine Specific gravity 0.44 0.47 Flake type small larger small larger Length (mm) mean 15.21 36.99 13.52 37.82 s.d. 13.40 11.56 17.25 6.37 Width (mm) mean 2.95 12.15 3.08 13.44 s.d. 0.92 17.44 1.16 16.37 Thickness (mm) mean 0.65 0.84 0.63 0.79 s.d. 0.02 0.02 0.01 0.02 Note: Specific gravity values are taken from Wood handbook 1989, United states, Forest Products Lab. Wis. 27 3.3 Reproducibility of pressure response The experiments were undertaken based on SDj and SD2 in Table 3.2 in order to empirically examine the reproducibility of pressure response curves during hot pressing. Mat rheology, pressure relaxation, and mat deformation with experimental variables were also studied. 3.3.1 Equipment The equipment used to manufacture the flakeboards was a computer controlled 300 x 300 mm hot press (Wabash 27,000 kg load), programmed with pressure and thickness control and containing a data acquisition system. The press could simultaneously monitor mat pressure response, mat deformation, and interior temperature changes during the pressing process. With this press, pressure manipulation to produce a breathing stage could be applied and the responses of the pressure (mat resistance), mat deformation, and temperature changes upon breathing were observed. 3.3.2 Mat preparation Wood flakes and enough resin for six panels were weighed separately and 10 percent extra of each material was included to compensate for any losses through the process. Flakes were placed in a blender first and the resin sprinkled over the flakes. The mixture was tumbled for 10 minutes. The blended flakes for one panel were weighed out and randomly applied into a 250 x 250 mm forming box, which was placed on an aluminum caul, ensuring the flake were evenly spread to give a uniform mat height. The forming box was carefully removed, and the mat weighed again. It is important that mats have identical weight when examining the pressure response curve during hot pressing. An aluminum caul was laid carefully on the top of the mat prior to pressing. Three mats were assembled at one time to reduce variation in the three replicates. 3.3.3 Pressing procedure The press was programmed to apply maximum pressure of 3.4 MPa to the mat at 30 (or 60, 90) seconds press closing time and keep this pressure until a thickness of 12 mm was reached. The program then was switched to press position control and pressure relaxation 28 3 4 5 6 Press time (min.) Figure 3.1 A typical pressure response curve. 29 occurred as the mat cured. A typical pressure response curve is shown in Figure 3.1. After pressing for 7 minutes, the remaining pressure was released and the press was opened. Upon removal from the press, the panel was laid on the floor and cooled to room temperature. Three replicates were pressed in sequence. Pressure response data were collected automatically by the computer. Panels made in this experimental session were stored in the constant temperature humidity (CTH) conditioning room at 25 °C temperature and 50 percent relative humidity for future property evaluation. 3.4 The development of in situ adhesive cure It was assumed that the pressure relaxation during hot pressing was governed by two major effects, rheology of wood flakes and rheology of adhesive. Since the rheology of adhesive could likely be illustrated by the cure process, a study of sequential in situ degree of cure was undertaken. The purposes of this part of study were to empirically understand the development of in situ adhesive cure, its relationship to the pressure response, and its influence on the vertical density gradient formation. The focus was on the pressure relaxation period when the press had closed to the target thickness. To facilitate this part of the study, adhesive samples were placed in three vertical positions in an active mat which was then pressed to one of four stages of pressure relaxation. These samples were then taken out and the degree of cure was examined. The experiment was carried out based on SD3 in Table 3.2. 3.4.1 Differential Scanning Calorimetry (DSC) Differential scanning calorimetry was used to examine degree of cure in this study. In this technique, the thermal response of a sample is compared to a reference capsule which is usually empty. As the system temperature is increased at a constant rate, heat will be input to either the sample capsule or the empty capsule in order to keep the difference in temperature between them at zero. By measuring this heat input, the energy involved in the reaction of the sample can be determined. During adhesive cure, an initial heat input is necessary for polymerization. As the reaction takes place, the system either absorbs or emits 30 heat. The more molecules involved in the reaction, the higher degree of polymerization will be achieved and the more heat be generated. As a result, the degree of adhesive cure can be defined by the amount of heat generated which is directly related to the degree of polymerization or cross-linking of the molecules. To determine the in situ degree of cure of an adhesive sample which has been hot pressed and partially cured for time t, it is necessary to know both the total heat generated for an uncured adhesive sample H0, and the heat generated for the partially cured adhesive sample Ht. The in situ degree of cure (a) can then be calculated by a(%) =—5—^xlOO Ho 3.4.2 Instrumentation A Du Pont DSC 2910 calorimeter with a pressure cell was used. The pressure cell is designed to suppress the interference of water during the cure of the adhesive. 3.4.3 Sample preparation Approximately 6 mg of W3154N PF powder resin was carefully weighed into a DSC pan, which was then sealed with a DSC lid; two approximately 0.2 mm holes were punched into the lid. The weight of the resin and the code of the resulting sample capsule was recorded. Nine sample capsules were prepared for each panel. Enough aspen veneer of size 30 x 25 x 3 mm was cut and a 5 mm diameter hole was drilled into each veneer at the central position. The veneers were conditioned or dried to give MC of 2 or 6 percent which were similar to the MC of the flake furnish material. Each of sample capsules was fixed in between two veneers as shown in Figure 3.2. 3.4.4 Mat preparation An aluminum caul was tared on a balance, and placed underneath a 250 x 250 mm forming box. Flakes for one panel were weighed out and half amount of these flakes were spread randomly into the forming box to evenly cover the mat area. Three sample 31 <t)5 ^ K m o " CO 12.5 25 veneers Figure 3.2 A DSC capsule fixed in between of two veneers. 32 •$—0—0 0—0—0 o CO o ID CNi —5 — 2 t^ 8 - • 6 i —9-1 * - S - Q - M Figure 3.3 The positions of DSC samples in a mat. 33 capsules #1, #2 and #3 were placed in three positions as shown in Figure 3.3. Another quarter of the amount of flakes were spread evenly into the forming box, on the top of the sample capsules. Sample capsules #4, #5 and #6 were then placed into the mat. After the remaining quarter of the amount of flakes were evenly spread into the forming box, #7, #8 and #9 sample capsules were positioned on the top surface of the mat. A top aluminum caul was placed carefully on the top of the mat. Three thermocouples were inserted into the mat at three positions, the top surface, the quarter position from the top by weight, and the middle layer, corresponding to the regions where the sample capsules were placed. 3.4.5 Pressing procedure Panels were pressed at a 30 second press closing time to a maximum pressure of 3.4 MPa and the pressure maintained until the target thickness was reached. The press was then switched to position control and the pressure started to relax. At press times of 1, 2, 4 or 7 minutes after reaching maximum pressure, the platen was programmed to open and the panel removed immediately from the press. Designations of D^ D2, D3 and DF were assigned to the cases where the platen opened at 1, 2, 4 and 7 minutes press time respectively. The sample capsules were removed by a core drill bit immediately after the panel was removed from the press, cooled at room temperature then stored in a cold room. 3.4.6 DSC examination The DSC capsules with partially cured resin were run at a pressure of 4.2 MPa under Nitrogen gas and heated at a rate of 10°C/min. from 20 to 250°C. The thermograms of each sample were automatically acquired using the available software package. 3.5 The development of physical and mechanical properties The purpose of this experiment was to understand the relationship between pressure relaxation and panel property development such as the sequential development of VDG, MOR, MOE, IB strength and dimensional stability, in relation to the in situ degree of adhesive cure. The strategy was to reduce panel pressure to a lower level at the same pressing stages as selected in 3.4.5, i.e., at pressing time of 1, 2, 4 and 7 minutes after reaching maximum pressure. It was assumed that the development of VDG as well as those 34 mechanical properties would be affected by this pressure reduction and the significance would depend on flake mat rheological properties and the degree of adhesive cure. The level to which the pressure was dropped was set to be below the pressure remaining at the end of the pressing cycle. This pressure level was selected because it was believed that it would alter the development of panel properties but maintain mat integrity and temperature, allowing the adhesive to be fully cured. In this way, the influence of insufficient adhesive cure on properties could be neglected. The experiments were carried out based on sub-design SD4 and SD5 shown in Table 3.2. 3.5.1 Mat preparation The mats for all panels were prepared following the procedure described in section 3.3.2. 3.5.2 Pressing procedure Based on the strategy explained above, the press was programmed to reach the maximum pressure of 3.4 MPa at a specific closing time and keep this pressure until the target thickness was reached. The press was then switched to position control. After closing to maximum pressure for either 1 (2, 4 or 7) minutes, pressure was dropped to 0.4 MPa and kept there until the end of pressing cycle. Pressure response curves showing each of the pressure drops are present in Figure 3.4. Similarly, D,, D2, D3 and DF were assigned to the pressure drops which occurred after press times of 1, 2, 4 and 7 minutes respectively. After removal from the press, panels were cooled at room temperature and then stored in CTH conditioning room for future property evaluation. 3.5.3 Property evaluation The properties of all panels manufactured in section 3.3.3 and 3.5.2 were evaluated for VDG, MOR, MOE, IB, and dimensional stability. 3.5.3.1 Sample preparation Panels which had been stored in the CTH room for at least two weeks were trimmed to 240 x 240 x 12 mm. Two 240 x 50 x 12 mm bending specimens, four 50 x 50 x 12 mm 35 A s (A 1 • 3 4 5 6 Press time (min.) D, D2 D3 7 8 D« Figure 3.4 Pressure response curves showing each of the four pressure drops at 1, 2, 4, or 7 minutes after reaching maximum pressure. 36 IB specimens, and one 100 x 100 x 12 m dimensional stability specimen were cut from each panel. One VDG specimen was chosen from the IB specimens for each panel pressing condition. The size and position of each test specimen in the panel are given in Figure 3.5. Due to the small size of the panels, edge effect on testing specimens is expected especially with the two ends of bending specimens. With other test specimens, this effect will not be significant. 3.5.3.2 Vertical density gradient The vertical density gradient examination was carried out by Alberta Research Council at Edmonton Alberta using a gamma radiation instrument. 3.5.3.3 Bending strength The bending strength was tested according to ASTM standard method ASTM D 1037-87. However a span of 180 mm had to be used due to the small size of the panels. Therefore, results could only be compared within this study. Samples were tested to failure on a MTS testing machine (MTS 810 Material Test System). The breaking load and deformation were recorded automatically by data acquisition program. 3.5.3.4 Internal bond strength Samples were bonded to aluminum blocks using a hot-melt adhesive. They were tested according to ASTM standard method ASTM D 1037-87 on an Instron machine. Breaking strength and the failure position were recorded for each sample. 3.5.3.5 Dimensional stability Dimensional stability was exclusively carried out through the examination of thickness swell which was believed to relate pressure response more strongly than other dimensional properties. Specimens were initially measured thickness after conditioned in the CTH conditioning room at 25 °C and 50 percent relative humidity for 2 weeks. For each of specimen, four measurements were taken using a dial gauge. The point of the measurement is 25 mm from the middle point of each edge. The study of thickness swell was based on following treatments: 37 IB IB Bending DS Bending IB VDG o s o S 50 100 50 240 Figure 3.5 The size and position of each test specimen in a panel. 38 (1) Conditioned in a humidity chamber at 26°C and relative humidity of 90 percent for 20 days, which gave each specimen 22 percent equilibrium moisture content (EMC). (2) Treated to vacuum-pressure-soak (VPS) condition in water with vacuum for one hour and pressure for one hour. (3) Oven-dried and conditioned back in the CTH room for 20 days. After each treatment, the specimen thickness was measured again. The thickness swell was calculated as a percentage of difference based on the initial thickness. 3.6 Exploration of pressure manipulation The purpose of this experiment was to study the feasibility of altering panel properties by manipulating the pressure during the pressure relaxation period. This was done by introducing a breathing cycle in which the pressure was dropped to a lower level of 0.4 MPa for 10 seconds and then immediately brought back to the previous pressure level. 3.6.1 Mat preparation and pressing procedure Only aspen small flakes were selected to run this experiment. Flake mats were prepared as before. The breathing cycle was programmed to occur after the press closed to the maximum pressure for either 1 or 2 minutes. Bj and B2 were assigned to breathing cycle at 1 and 2 minutes pressing after reaching maximum pressure. After completing the press cycle, panels were cooled at room temperature and stored in the CTH conditioning room. 3.6.2 Property evaluation Panels stored in the CTH room for at least 2 weeks were cut for each of the testing specimens as described in section 3.5.3. The properties of VDG, MOR, MOE, IB and dimensional stability were tested according to the procedure described in section 3.5.3.2, 3.5.3.3, 3.5.3.4, and 3.5.3.5 respectively. 3.7 Supplementary experiment This experiment was conducted to examine the relationship between the resin content and the resin coverage of flakes using W3154N PF powder resin. Resin coverage was determined by image analysis. 39 3.7.1 Image analysis An image appears to the human eye as a continuum of varying brightness and colour. In image analysis, the image is rendered as a set of discrete cells called pixels, individual points of varying brightness whose composite forms the image displayed on a raster screen. Since each pixel is given a numeric value corresponding to the brightness of the image at that location, the image can be subjected to mathematical analysis and stored on disk. The brightness or intensity is measured by the grey level of the pixel ranging numerically from 0 (for black) to 255 (for white). An intensity histogram shows the number of occurrences of each grey level in an image or part of an image. The histogram is a bar chart in which the horizontal dimension represents the range of all possible grey levels (0 to 255), and the vertical dimension shows the number of occurrences of each grey level. Light coloured flakes mixed with PF resin and fully cured in an oven will appear dark in areas covered by the dark coloured resin. The total darkness of the flakes depends on the areas covered with resin, the amount of the resin, or the number of pixels which yield lower grey levels than that of uncovered flakes. If assuming that the lowest limit of the grey level for an uncovered flakes is Lm, the resin coverage (6) on a flake can be determined by dividing the number of pixels (N) which have grey level lower than Lm by the total number of pixels (N,) in an image or a portion of the image (area of interest) of the flake. jg(%) =JLX100 3.7.2 Instrumentation The image analysis system consisted of a Zeiss/Jena Technival 2 stereomicroscope, a Javelin JE3462HR colour video camera, a Targa+ 16/32 frame grabber (Truevision, Inc.), a Javelin CVM13A colour monitor, an IBM 486- compatible microcomputer and Java image analysis software (Jandel Scientific), shown in Figure 3.6. 3.7.3 Experimental procedure The light coloured aspen larger-size flakes were used in this experiment to give a better image resolution. Ten replicate flakes were selected for the first sample group at zero Camera ay ii Test Instrumen C3CZJ VCR ' • » • " ' Computer Monitor Video Monitor Figure 3.6 Image analysis system. 41 percent RC. Flakes were blended with 2 percent W3154N PF powder resin, based on the weight of dry flakes. Ten flakes were picked out for the second sample group at 2 percent RC. In total, six sample groups were prepared with either 0, 2, 4, 6 or 8 percent resin and an ideal resin coated group (FC), which was fully covered with resin. The flakes were heated in an oven at 150°C for 10 minutes to develop the colour. An image of each flake was viewed in a stereomicroscope and was displayed on a monitor. The image was then frozen by the frame grabber and stored in a disk. After all the flake images were collected, the image analysis was conducted. The image was retrieved from the disk, and displayed on the monitor. A similar sized representative area of interest (AOI) was selected from all images. The histogram of the AOI was displayed and analyzed for the resin coverage calculation. Figure 3.7 shows the intensity histograms for flake images with different resin contents. 3.7.4 Results and discussions A relation between resin coverage and RC was displayed in Figure 3.8. Resin coverage increased dramatically with RC below 4 percent. Its increasing rate decreased with RC between 4 to 8 percent and the resin coverage reached 90 percent at 8 percent RC. Further increasing RC resulted in little increase in the resin coverage due to the limited ability of the flakes to pick up more resin. 100 150 Gray level Figure 3.7 Intensity histograms for flake images with different resin content. 43 100 2 g O c "w o DC 6 8 Resin content (%) FC Resin content (%) Resin coverage (%) 0 0 2 17 4 59 6 73 8 89 FC 100 Figure 3.8 Relationship between resin coverage and resin content. 44 4.0 RESULTS AND DISCUSSIONS 4.1 Pressure response curve study Forty-five pressure response curves were examined for the study of panel pressure response during hot pressing. The reproducibility of the pressure curve for all treatments as determined from three replicates are demonstrated in Figure 4.1. These graphs confirm that the pressure responses of hot pressed panels are reproducible at constant pressing parameters and conditions. All pressure response curves appeared to follow the same trend, ie. at first, the pressure increased linearly with time, after reaching the maximum value, it remained constant for a while and then started to decline curvilinearly until the end of the pressing cycle. Exceptions were observed in Figure 4.1 (3), (d) and (f) where the panels were manufactured with larger-size flakes and at 2 percent initial MC. Due to the use of relatively lower maximum pressure, the pressure response curves consisted of five regions instead of the four denoted by Wolcott (1990): I: The press closure region, where pressure increased linearly with time as determined by the PCT. II: The mat creep region, where the pressure remained constant at a maximum value while the mat densifies with time. This creep region is completed when the mat reached its target thickness and average density value. Ill: The pressure transient relaxation region, where the pressure declined rapidly with time. It began after the mat reached the target thickness. Mechanically the mat is undergoing stress relaxation at constant deformation. IV: The pressure asymptotic relaxation region which began when the pressure gradient decreased to a relatively lower level. V: The press opening region, where the pressure decreased rapidly to zero during which time the press opened to allow the pressed panel to be unloaded. The pressure response curve was very sensitive to many factors as demonstrated in Figure 4.1. It represented a combination of all the factors affecting the rheological 45 ASH30 ASL30 0 1 2 3 4 5 6 7 8 9 Press time (min.) (a) 30s PCT, 6% MC. ASH60 0 1 2 3 4 5 6 7 8 9 Press time (min.) (b) 60s PCT, 6% MC. ASH90 0 1 2 3 4 5 6 7 8 9 Press time (min.) (C) 90s PCT, 6% MC. 0 1 2 3 4 5 6 7 8 9 Press time (min.) (d) 30s PCT, 2% MC. ASL60 0 1 2 3 4 5 6 7 8 9 Press time (min.) (e) 60s PCT, 2% MC. ASL90 0 1 2 3 4 5 6 7 8 9 Press time (min.) (f) 90s PCT, 2% MC. Figure 4.1 Pressure curve reproducibility study. (1) Pressure curves of flakeboards manufactured with aspen small flakes, without resin. 46 ASRH30 ASRL30 • 1 ' I I1 • ' " I 1 " T " 0 1 2 3 4 5 6 7 8 9 Press time (min.) (a) 30s PCT, 6% MC. A . *t- * I 3 Pressure -» ro n • ASRH60 A / ^ — . / I 0 1 2 3 4 5 6 7 8 9 Press time (min.) (b) 60s PCT, 6% MC. ASRH90 0 1 2 3 4 5 6 7 8 9 Press time (min.) (c) 90s PCT, 6% MC. 0 1 2 3 4 5 6 7 8 9 Press time (min.) (d) 30s PCT, 2% MC. ASRL60 0 1 2 3 4 5 6 7 8 9 Press time (min.) (e) 60s PCT, 2% MC. ASRL90 0 1 2 3 4 5 6 7 8 9 Press time (min.) (f) 90s PCT, 2% MC. Figure 4.1 (continued). (2) Pressure curves of flakeboards manufactured with aspen small flakes and 8% resin content. 47 ALRH30 0 1 2 3 4 5 6 7 8 9 Press time (min.) (a) Aspen larger flake, 6% MC. LSRH30 0 1 2 3 4 5 6 7 8 9 Press time (min.) (b) Lodgepole pine small flake, 6% MC. LLRH30 0 1 2 3 4 5 6 7 8 9 Press time (min.) ALRL30 0 1 2 3 4 5 6 7 8 9 Press time (min.) (d) Aspen larger flake, 2% MC. LSRL30 t I 3 Pressure 5 -» to / I I I V • 0 1 ~ ~ ^ 1 2 3 4 5 6 7 8 9 Press time (min.) (e) Lodgepole pine small flake, 2% MC. LLRL30 0 1 2 3 4 5 6 7 8 9 Press time (min.) (c) Lodgepole pine larger flake, 6% MC. (f) Lodgepole pine larger flake, 2% MC. Figure 4.1 (continued). (3) Pressure curves of flakeboards manufactured with aspen larger flakes, lodgepole pine small flakes, and lodgepole pine larger flakes. Resin content is 8%. Press closing time is 30 seconds. 48 characteristics of the mat. In order to separate the independent effects of each factor, each factor was examined individually by holding other factors constant. 4.1.1 Species and flake size effects The influences of species and flake size on pressure response of a mat are presented in Figure 4.2. In all cases, mats with 6 percent MC exhibited a shorter mat creep region, a faster pressure relaxation rate in the transient relaxation region, and a lower pressure value in the asymptotic relaxation region than that of 2 percent MC mats. This phenomenon indicated that mats with 6 percent MC had larger rheological responses and therefore they crept and relaxed more than the mats with 2 percent MC during hot pressing. The influences of the MC on the rheological behaviour of the mat will be discussed exclusively in the next section. Significant differences in pressure responses were observed for the two species and the two flake types (Figure 4.2). In the 6 percent MC case, the mats furnished with aspen small-size flakes took approximately 1 minute to creep to the target thickness under given maximum pressure while the lodgepole pine small-size flake mats took more than 2 minutes to reach the target thickness. Moreover, in the pressure relaxation regions, the former relaxed quickly to a pressure of 1 MPa over a half minute period and continued to relax at a slower rate until 0.5 MPa before the press was opened. The later however initially relaxed at the same rate to 2.4 MPa pressure, then the rate slowed until the end of the pressing cycle where the remaining pressure was 1.4 MPa. The higher pressure remaining at the end of the cycle indicated a higher interior stress within the mat. The same trend was observed for the larger-size flake cases. This result is in accord with the work reported by Kitazawa (1947) in which he did a study of wood relaxation with respect to specific gravities. Aspen is categorized as a hardwood and has relatively lower specific gravity (0.44) than lodgepole pine (0.47) which is a softwood. Kitazawa found that a negative correlation existed between specific gravity and relaxation coefficient, i.e., the higher the specific gravity, the lower the relaxation coefficient or the lower the plasticity. Specific gravity varies within species from heartwood to sapwood and from early-wood to late-wood. Such variations are different for different species. In lodgepole pine for example, larger variation is know to exist between 49 S2F2H (a) 6% MC. S2F2L Press time (min.) (b) 2% MC. aspen small flake 'aspen larger flake lodgepole pine small flake lodgepole pine larger flake Figure 4.2 Species and flake size effects on pressure curves. Resin content is 8%. Press closing time is 30 seconds. 50 early-wood and late-wood compared with aspen. Since a mat is formed with flakes from a random source and were cut randomly from a log, when studying the rheological behaviour of the mat during hot pressing, such variation must also be taken into account. Flake size influenced pressure response through its effect on mat structure. In the case of a random flake mat, flake geometry determines to a large extent within mat voids. The voids increase as the flake aspect ratio (flake length / width) and slenderness ratio (flake length / thickness) increase. Mats with higher amounts of voids will result in an increase in the number of flake layers if the final overall density at a given final thickness is to remain unchanged. Therefore more severe interior compression stress and horizontal density variation will result which will reduce the mat's overall creep and relaxation (Suchsland 1959, 1967). This is illustrated in Figure 4.2 (a) and (b). With mats containing larger-size flakes, a shift of the pressure response curve in the transient relaxation region to the right hand side (Figure 4.2 (a)) was evident; this arose from the reduced rate of mat creep due to high densification in some areas of the mat. The rate of pressure relaxation of mats made with small-size flakes appeared to be similar to the one with larger-size flakes in the transient relaxation region but slower in the asymptotic relaxation region. The final mat pressures with small- and larger-size flakes of the same species are quite similar at the end of asymptotic relaxation region. The most severe case of compression stress was observed in mats with larger-size flake (both aspen and lodgepole pine) and 2 percent MC where no pressure relaxation was apparent throughout the pressing process (Figure 4.2 (b)). This phenomenon indicated that a mat with such a structure and low MC creeps so slowly under the given pressure that it never reaches its target thickness by the end of the pressing cycle. 4.1.2 Moisture content effects Based on the plasticization effect of MC on wood there is little doubt that MC present within the mat will enhance creep and relaxation during hot pressing. Figure 4.3 shows comparisons of pressure response curves of mats with 2 and 6 percent MC. These comparisons involved only one furnishing material, i.e., aspen small flakes, but two RC levels (0 and 8 percent) and three PCT (30, 60 and 90 seconds). The trend was clear; compared with the 2 percent MC condition, the mat with 6 percent MC always exhibited ASH&L30 ASRH&L30 0 1 2 3 4 5 6 7 Press time (min.) (a) 30s PCT, no resin. 8 9 <£ I 4 • 3 • 2 • 1 n • / / / f • « \ « \ . *•••- ,_ ASH&L60 """""*"•———... H—H 0 1 2 3 4 5 6 7 8 9 Press time (min.) (b) 60s PCT, no resin. ASH&L90 2 3 4 5 6 7 Press time (min.) (c) 90s PCT, no resin. 0 1 2 3 4 5 6 7 8 9 Press time (min.) (d) 30s PCT, 8% RC. ASRH&L60 (e) 60s PCT, 8% RC. ASRH&L90 0 1 2 3 4 5 6 7 8 9 Press time (min.) (0 90s PCT, 8% RC. 6%MC 2%MC figure 4.3 Moisture content effects on pressure curves. Flakeboards were manufiactured with aspen small flakes. 52 faster creep in the mat creep region, greater pressure relaxation rate in the transient relaxation region and lower pressure level in the asymptotic relaxation region. Xhis MC effect appeared more significant when the mat was pressed with a slower press closing speed (Figure 4.3 (c) and (f)) than at higher press closing speed (Figure 4.3 (a) and (d)), and when there was no RC within the mat (Figure 4.3 (a), (b), (c)) compared with 8 percent RC (Figure 4.3 (d), (e), (f)). Xhis indicated that an interaction existed among MC, RC, and PCX. 4.1.3 Press closing time effect From Figure 4.4, a shift of the pressure response curve to the right was observed in all cases. PCX influenced the mat pressure response curve more in the mat creep region than in the pressure relaxation regions. In general, short PCX resulted in a longer mat creep region but the same pressure relaxation rate in the transient relaxation region shown by parallel curves as the longer PCX. Xhe pressure relaxation rate in the asymptotic relaxation region differs slightly for different PCX, however, they all reached the same pressure value at the end of this region. Xhe longer mat creep region at shorter PCX might be due to the delay of heat and moisture transfer into the interior of the mat. At shorter PCX, the flakes on the mat surfaces were heated and plasticized as the pressure was applied whereas the flakes in the mat interior were compressed before the moisture and heat moved towards the centre of the mat. Xherefore a longer creep region is needed for the mat to reach the target thickness under the given maximum pressure. In contrast, longer PCX allowed many of the flakes to be plasticized by water vapour before the compressive force was fully applied. Xherefore the mat was deformed more easily and required less time to creep to its target thickness. Similar pressure relaxation in the transient relaxation region and the same pressure level at the end of the asymptotic relaxation region for the different PCX indicated that the rheological properties of the flake mats were identical after reaching target thickness although the heat and mass transfer for different PCX might be slightly different. 4.1.4 Resin content effects Bolton et al. (1989) have indicated that the cure of resin enhanced the relaxation of stresses by developing bond strength between flakes and therefore locking in elastic and 3 2 1 :•// ••\V V ASH369 """.—1 '. 1 L ASL369 0 1 2 3 4 5 6 7 Press time (min.) (a) 6% MC, no resin. 3 2 1 fffi\ :,'/ \ \ \ •'' / '• \ \ ASRH369 : i r-S-A— 0 1 I 3 4 5 6 7 Press time (min.) (c) 6% MC, 8% RC. 2 3 4 5 6 7 Press time (min.) (b) 2% MC, no resin. 4 g 3 £ 2 £ 1 ASRL369 0 . • " / — .* ' / • ' / • / / • ' / • - / / ' / 1 / <£ r -• 7 n — s •v \ '•\ \ ^ • r Z ^ - -""""—•i"™* ~ \ : \ 0 1 2 3 4 5 6 7 Press time (min.) (d) 2% MC, 8% RC. 30s PCT 60sPCT •90sPCT Figure 4.4 Press closing time effects on pressure curves. Flakeboards were manufactured with aspen small flakes. 54 R&NR-H30 3 2 1 / \ ^aa«=s=ssa X-**» • „ _ V OftRC 4«RC 0 1 2 3 4 5 6 7 8 9 Press time (min.) (a) 30s PCT, 6% MC. R&NR-H60 3 2 1 / V j \ / V / ^ 0 1 2 3 4 5 6 7 8 9 Press time (min.) (b) 60s PCT, 6% MC. R&NR-H90 3 2 1 / Y / \ / V .^ / ^"**5r —•>*•• • ! ,mWmtmm^ s^—sa 0 1 2 3 4 5 6 7 8 9 Press time (min.) (c) 90s PCT, 6% MC. R&NR-L30 3 2 1 . f \ x — — —• — 4ft BC / \ o * « - """""*—•• 0 1 2 3 4 5 6 7 8 9 Press time (min.) (d) 30s PCT, 2% MC. R&NR-L60 0 1 2 3 4 5 6 7 8 9 Press time (min.) (e) 60s PCT, 2% MC. R&NR-L90 0 1 2 3 4 5 6 7 8 9 Press time (min.) (f) 90s PCT, 2% MC. Figure 4.5 Resin cure and resin content effects on pressure curves. Flakeboards were manufactured with aspen small flakes. 55 delayed elastic strains. In this study, however, the influence of resin cure was distinguishable only in the 2 percent MC cases (Figure 4.5 (d), (e) and (f)) but not in the 6 percent MC cases (Figure 4.5 (a), (b) and (c)). At 2 percent MC, different RC showed a different degree of influence as illustrated in Figure 4.5 (d). Here the mat with 8 percent resin content showed the highest rate in the mat creep and the transient relaxation region, reducing to the lowest pressure level at the end of the asymptotic relaxation region. The effect of resin content on this mat rheological behaviour might be also due to the plasticization effect of resin on the wood elements. In a study of plasticizing wood, Kollmann et al. (1975) stated that at elevated temperatures, uncured phenol-formaldehyde resin plasticizes the wood more than water. In the 6 percent MC cases, this plasticization effect of resin was not evident since no difference was observed for the mats made with different resin contents. It seems reasonable to conclude that at 6 percent MC, the contribution of the MC alone to the rheological properties of the mat was identical to that of interaction effect of MC and RC. 4.2 The development of in situ adhesive cure As stated earlier, the development of in situ adhesive cure involves a sequential examination of the degree of cure of adhesive samples during hot pressing. Such information is important for studying the formation of panel vertical density gradient (VDG) and the development of other properties. 4.2.1 Adhesive cure The cure of the PF resin was initially examined by DSC calorimetry. The resulting thermogram is displayed in Figure 4.6. This thermogram has a large exothermic peak starting at a temperature of approximately 90 °C and finishing at 229 °C indicating that the polymerization reaction of adhesive occurred in this temperature range at a heating rate of 10°C/min.. The total area under the peak represents the heat generated during adhesive polymerization which is 211 J/g for this resin. This large exothermic peak appeared to consist of three peaks ranging in temperature from 90 to 140°C, 140 to 195°C, and 195 to 229°C. These three peaks might indicate that three reactions took place during adhesive polymerization. These reactions might be the three important steps for this adhesive to be 56 eo 1.0 O.B 0.6-0.4 0.0--0.2--0.4' 151.63»C I40.92*C Sll.OJ/B — i — 50 100 ISO 200 2S0 Temperature (°C) Figure 4.6 A DSC thermogram of phenol-formaldehyde powder resin sample. It was scanned at 10°C/min. heating rate from 25 to 250°C under pressure of 4.2 MPa. 57 fully cured. However, what these three peaks really represent is not yet clear. Further study on this adhesive is necessary. 4.2.2 The development of in situ adhesive cure DSC degree of in situ adhesive cure at three vertical locations of a mat (top surface layer (S-layer), one-quarter to the top surface layer (1/4-layer), and middle layer (M-layer)) were examined in relation to four sequential pressing times designated by D1? D2, D3 and DF. The DSC thermograms of these precured adhesive are shown in Appendix I. Comparisons between thermograms are given in Figure 4.7 (1) and (2). In general, the longer the adhesive remained in the hot press, the more cure resulted and the observed exothermic peak became smaller. This is seen by a graduate decrease of exothermic peaks from Dx to DF in (a), (b), (c) of Figure 4.7 (1) and (2). For the same pressing period, the adhesive in the S-layer cured faster than that in the 1/4-layer and much faster than the one in the M-layer. This is well illustrated by Figure 4.7 (1) and (2) in which the thermograms in (c) show the smallest exothermic peaks compare to those in (a) and (b) while the thermograms in (a) show the largest peaks. The differences in the exothermic peaks are more significant for Dx and D2 which occurred during the earlier period of hot pressing. The percentage degree of cure calculated for each condition is presented in Table 4.1 and graphically shown in Figure 4.8. This table and graph clearly showed how adhesive cure developed vertically from mat surfaces to its core and with time during hot pressing. Initially, a large gradient in the degree of cure existed within the mat in the transient relaxation region but this became smaller in the asymptotic region where a satisfactorily sufficient degree of cure was gradually achieved. An exception was observed with the 6 percent MC case in which a large gradient still existed between the 1/4-layer and M-layer in the asymptotic relaxation region. It is believed that higher MC hindered the development of adhesive cure as illustrated by a consistently lower degree of cure exhibited in the 6 percent MC case in both the 1/4-layer and M-layer (Figure 4.8). A graph showing monitored temperature changes in these three layers of a mat during hot pressing is given in Figure 4.9. Similar temperatures were observed in S-layer for both 2 and 6 percent MC panels. But a delay in temperature increase was evident in the 1/4- and M-layers for the 6 Temperature (°C) Temperature (°C) I O I Temperature (°C) Figure 4.7 Comparison of DSC thermograms of adhesive samples taken at 1, 2, 4, and 7 minutes hot pressing. (1) 6% moisture content. 00 lio irio Temperature (°C) rfo Temperature (°C) Temperature (°Q Figure 4.7 (continued). (2) 2% moisture content. 60 Table 4.1 Degree of in situ adhesive cure of phenol-formaldehyde powder resin in three vertical locations of a panel at press times of 1, 2, 4 and 7 minutes. Moisture Condition 2% 6% Pressure Drops » 1 D2 D3 DF D, D2 D3 DF Pressing Time (min.) 1 2 4 7 1 2 4 7 Cure (%) M-layer 0.00 11.51 67.84 81.17 0.00 11.76 39.20 55.10 1/4-layer 11.11 54.56 70.18 81.14 25.12 32.93 64.65 67.18 S-layer 54.13 74.99 75.09 79.82 69.34 80.38 69.29 68.19 ^*% & © Z5 O *s © © 1 _ CD o 100-1 90-80-70-60-50-40-30-20-10-o-Pressure drops • y-bjer, 6% A l/4-kjer. 6% + S-hjar. 6% O Ifr-hyer, 2% 2 l/4-feyer, 2% * S-hjar, 2% figure 4.8 Degree of in situ adhesive cure of resin samples in three vertical locations of a panel after 1, 2, 4 and 7 minutes in the hot press following application of full pressure. D, D2 D3 DF 250 200 u o * w r V 1 -2 C3 I M ^ 6 u H 150 100 50 S - layer 0 4 3 4 5 6 Press time (min.) (a) 6% MC. 250 200 S - layer i i i i 0 1 2 3 4 Press time (min.) 8 • i i i 5 6 7 8 9 (b) 2% MC. Figure 4.9 The monitered temperature changes during hot pressing in three lay 63 percent MC after reaching 100°C; this delay appeared to be longer in the M-layer due to the presence of higher MC. This overall slower rate of temperature increase, in turn, affected the development of adhesive cure. 4.3 The development of panel properties Panels manufactured with either one of four sequential pressure stages (drops) were evaluated for VDG, MOR, MOE, IB and dimensional stability in order to reveal the development of these properties during hot pressing. In the manufacture of ALRH-30, LSRH-30, and LLRH-30 panels, pressure drop Dj was omitted since mats did not reach the target thickness at 1 minute pressing. A new pressure drop D4 which occurred at 5.5 minutes hot pressing was introduced instead. Figure 4.10 shows the pressure and mat displacement response curves in relation to each of the pressure drops. 4.3.1 Panel VDG The development of panel VDG with different furnish materials, MC, and PCT are displayed in Figure 4.11. Each VDG curve is an average of three specimens from each of three replicate panels. From the graphs, it appeared that panels made with Dx had the lowest density values (Figure 4.11 (1) and (2)). For the ALRH-30, LSRH-30 and LLRH-30 panels where Di was omitted, panels made with D2 showed the lowest density values (Figure 4.11 (3)). These panels were all manufactured when pressure dropped in the transient relaxation region where high interior compression stresses existed, represented by the high pressure response in Figure 4.10. This high interior stresses or pressure response implied a tendency of immediate springback upon pressure reduction. It should be also remembered that the degree of adhesive cure in this early stage of hot pressing is 0 percent in the M-layers for both 2 and 6 percent MC, 11 and 25 percent in the 1/4-layer, 54 and 69 percent in S-layer respectively for 2 and 6 percent MC. Upon pressure reduction, the density gradient would redistribute by increasing panel thickness or by adjustments in the inner portion of the panel thickness if surface flakes had already densified and cured. Panels manufactured with pressure drops after the transient relaxation region exhibited quite similar VDG curves for ASRH panels indicating that the panel VDG might have been satisfactorily developed. But for ASRL, ALRH, LSRH, and LLRH panels slight differences in VDG curves were ASRH30-P 2 3 4 5 6 7 8 9 Press time (min.) (a) 30s PCT. ASRH60-P 2 3 4 5 6 7 Press time (min.) 8 9 (b) 60s PCT. ASRH90-P 2 3 4 5 6 7 Press time (min.) (c) 90s PCT. t 6 ° £ 50 ~ 40 ' o £ 30' M 20 I 1 0 0 ASRH30-O ^ • l a f c M M U A t M . M • e<rfMfa«^*-*^«^fcrf^e^J • " " • " • ff ' " T " I n f i l l I I B f — l « l f • N ^ H ^ H f M 0 1 2 3 4 5 6 7 8 9 Press time (min.) (d) 30s PCT. ASRH60-0 1 2 3 4 5 6 7 Press time (min.) (e) 60s PCT. ASRH90-D 1 2 3 4 5 6 7 Press time (min.) (f) 90s PCT. D, D2 D, figure 4.10 Pressure response and mat displacement of flakeboards in relation to sequential pressure drops. (1) The flakeboards were made with aspen small flakes and 6% MC. Pressure (MPa) O -» KJ O) A 1 1 1 1 D •B ? 1 s s to 9 3 • 3 12 3 4 5 6 Press time(min.) oo CD "i II s to 9 1— ? •o Displacement (mm) c ) 1 2 3 Press r 1 ' o> 03 - • IO U ^ Ol c 0 o o o o o c c j ft 3 ? • > 33 r-«o o 6 Pressure (MPa) Pressure (MPa) Displacement (mm) - * KJ G) 4* Ol O) O O O O O O o fo 1" r 5 6 (min.) v j • 00 (O L Displacement (mm) -* ro w *> Ol 0) o o o o o o o l\ 0\ in 66 ALRH30-P £ 3 • £ 2 0 1 2 3 4 5 6 7 8 9 Press time (min.) (a) Aspen larger flakes. H ' I 3 ' | 2 • n • / k ^ > ' : L LSRH30-P 0 1 2 3 4 5 6 7 8 9 Press time (min.) (b) Lodgepole pine small flakes. LLRH30-P 0 1 2 3 4 5 6 7 8 9 Press time (min.) (c) Lodgepole pine larger flakes. ALRH30-D I 5 0 1 2 3 4 5 6 7 8 9 Press time (min.) (d) Aspen larger flakes. LSRH30-D E E § I 60 0 1 2 3 4 5 6 7 8 9 Press time (min.) (e) Lodgepole pine small flakes. URH30-D 50 • 40 • 30 20 10 i"""'" , l"l ^ • ^ - • •• - r - " 1 f 0 1 2 3 4 5 6 7 8 9 Press time (min.) (f) Lodgepole pine larger flakes. D2 D3 D4 DP Figure 4.10 (continued). (3) The flakeboards were made with aspen larger flakes, lodgepole pine small flakes, and lodgepole pine larger flakes, all at 6% UC and 30s PCT. Density (g/cm3) o o o o o o o o o o ^ r o u ^ u i o ^ b o c o - t vO o CO*a q 3 n 7? 3 Zs s 3 -* N» co 01 0 ) •>i 00 CO Ni CO s u> o 3 7T 3 ? 3 Density (g/cm5) o o o o o o o o o ^ foco ikcno i^boco S 2J 1 0 . o\ ^ • 4 0 a 1* 8-0 0 . va * 0 O 3 CO ft. 0 * •9 a 1 CM C 3 0 . 1 VI cr 0 CO 8 0 ^ ^ 9 Q. 5" Oq O 1 n 3 EX. BL • a a. s •id •3 0 a» 8 0 c •a 3 JC 3 l 1 l 3 Density (g/cm3) o o o p p p o p p o^-'Mco'-P*cjibj>jboco-' ex s CO •tf q H sr 0 ?r s 8 </) ^^ 3 3 -* NJ CO -£k cn 0 ) •vl 00 CO M CO L9 SO O 9 Density (g/cm3) o o o o o o o o o O ^ N ) C J ^ u i b > ^ b o c o - > Density (g/cm3) 3 o OS a PPPPPPPPP 0 : - » N J W ' A C J l b 5 > J C » C O - » O I S I a* o a 3. Q . CO # Q S 3 O | 1 Density (g/cm3) o o p o o p p p p o ^ f o c o ^ u i a i ^ i o i o - ' 3 3 i £ ^ £ Z, ~ » £ * I et. 3 e 8. 8 3 rt S f* C/> f 3, o -NJ • U • A • ui • OJ • >4 • 00 • CO • o " ISJ ' OJ ' t ^ ' _J/— ^ *** ^ * • * . . - - ' .•V> ? 4* > • ^ f T w 3 ''*/ > -'1 f •As K *m it ^^ \£ / 1 ;» <P i . 8 $ 89 Density Cg/cm3) Density (g/cm3) ppppppppp O i • • • • i , - , • ppppppppp o ' - > N ) l > > A u i b > ^ j b o ( 0 - t TO re >o 3 s* (W re •1 Izb 1* i** 3 o 7? 3 fll V) ? 1 > 3 TO a a 3 3 CB re CA CO > • *• B. § o\ 8. Si °- 3 ° s M TO 3 3 s* 3 .* CTQ rt • 1 a re TO ^ 1 3 re 3 re o 3 Q» 3 gi i i a. 3 e s. Density (g/cm3) p p o p o p o o p o ' -»roc j^b ib i ; »jboco p o •d 1 1 1 1 o b» TO « re •o 3 re 0) 3 2> § • £ n 7? 3 is CM 3 E to u •t» O l CD CO CO o CO 69 70 observed between panels manufactured with pressure drops near the transition point and in the asymptotic region indicating that the panel VDG were continuously developing. From the pressure and mat displacement response curves in Figure 4.10, for ASRH panels, the pressure near the transition point at D2 were as low as 1 MPa and no immediate springback was observed upon pressure reduction although it was known that the adhesive within the mat was not satisfactorily cured at this stage. While for ASRL, ALRH, LSRH and LLRH panels, the pressure near the transition point was approximately 1.7 MPa and an immediate springback was observed. Therefore it could be reasonably concluded that panel VDG would continue to develop until the pressure relaxed to a value below 1 MPa or there would be no immediate springback upon pressure reduction caused either by plasticization or adhesive bonding. The rate of the VDG development would depend on the panel rheological behaviour exhibited during hot pressing, which was faster with panels possessing larger rheological response but slower with panels possessing limited rheological response. Some differences in thickness values were apparent, but the order of such differences was not consistent, making it difficult to draw any satisfactory conclusions. These differences could reflect errors in panel thickness measurements arising from surface roughness. 4.3.2 Panel mechanical properties In this study, panel mechanical properties, MOR, MOE, and IB were expressed by their specific values, being based on strength-specific gravity ratios to eliminate the effect of varying densities among specimens. Table 4.2 shows respectively the mean values of specific MOR, MOE, and IB with each of the experimental variables. The development of these properties is shown in Figures 4.12 to 4.15. A trend of increasing properties as pressing progress was observed in all cases. The increases were significant for panels with 6 percent MC, registering an almost constant rate. With 2 percent MC, the increase was significant in the transient relaxation region but became slight thereafter. Referring to the development of VDG that panels with 6 percent MC yield similar VDG after the transient relaxation region, it could be seen that a well established 71 Table 4.2 Mean Specific MOR, MOE and IB for four pressure drops and different levels of MC, PCT, species and flake sizes. (a) Specific MOR Specific MOR (MPa) Species aspen aspen lodgepole pine Average2 Flake size small small larger All1 small larger AH1 small larger All1 MC (%) 2 6 Average2 6 6 6 PCT (s) 30 60 90 All1 30 60 90 All1 30 60 90 All1 30 30 30 Pressure drops D, 22 17 9 16 21 25 26 24 21 21 17 20 — — ~ D2 23 19 19 20 25 26 27 26 24 23 23 23 25 50 37 25 42 33 25 46 35 D3 25 21 19 22 28 31 31 30 26 26 25 26 28 52 40 26 44 35 27 48 38 Dp 29 21 21 23 33 34 32 33 31 27 26 28 33 56 44 26 44 35 30 51 41 Table 4.2 (continued). (b) Specific MOE Specific MOE (MPa) Species aspen aspen lodgepole pine Average2 Flake size small small larger AH1 small larger All1 small larger AH1 MC (%) 2 6 Average2 6 6 6 PCT (s) 30 60 90 AH1 30 60 90 All1 30 60 90 All1 30 30 30 Pressure drops Di 3587 3164 1952 2901 3996 3838 3556 3796 3791 3501 2754 3349 — — — D2 3694 3219 3246 3387 4262 3916 3634 3937 3978 3567 3440 3662 4262 6511 5386 4146 5377 4761 4204 5944 5074 D3 3680 3431 3309 3473 4702 4333 3617 4218 4191 3882 3463 3845 4702 6585 5644 4137 5465 4801 4420 6025 5222 DF 4113 3673 3423 3736 5591 4648 4073 4771 4852 4161 3748 4254 5591 6550 6070 4226 5665 4946 4909 6107 5508 Table 4.2 (continued). (c) Specific IB Specific IB (kPa) Species aspen aspen lodgepole pine Average2 Flake size small small larger AH1 small larger All1 small larger AH1 MC (%) 2 6 Average2 6 6 6 PCT (s) 30 60 90 All1 30 60 90 All1 30 60 90 All1 30 30 30 Pressure drops Dt 675 543 282 500 639 1090 1173 967 657 816 728 734 — — — D2 861 758 970 863 876 1201 1166 1081 868 980 1068 972 876 1029 952 1318 1175 1247 1097 1102 1100 D3 898 786 930 871 955 1383 1199 1179 926 1085 1065 1025 955 1182 1069 1361 1186 1274 1158 1184 1171 DF 957 842 894 898 1312 1470 1300 1361 1135 1156 1097 1129 1312 1268 1290 1465 1280 1372 1388 1274 1331 Notes: (1) All1 represents the average value for different PCT levels and different flake sizes, respectively. (2) Average2 represents the average value for different MC levels and different species, respectively. 74 • 2%UC * €%MC • 2%MC * 6%MC 200- — — • 0-1 1— 1 1 1 1 1 t~ D, D2 D, DF Pressure drops • 2%MC * 6%MC Figure 4.12 Development of panel mechanical properties with different MC levels based on PCT All1 values in Table 4.2. All panels were made with aspen small flakes, (a) Specific MOR. (b) Specific MOE. (c) Specific IB. D, D, D, Pressure drops (a) D, • 30 MC. o 60 MC. « 90 MC. (b) Pressure drops • 30 MC. a 60 MC. * 90 MC. I CD o 0> D. CO Pressure drops (c) • 30 M C . o 00 MC. « 80 MC. Figure 4.13 Development of panel mechanical properties with different PCT levels based on MC Average2 values in Table 4.2. All panels were made with aspen small flakes, (a) Specific MOR. (b) Specific MOE. (c) Specific IB. D, (a) Pressure drops DF uptn * todg*pol* pin* 8000 7000 e o. LU O 2 _o "o 0) a w 6000 9000 4000 3000 2000 1000 D, (b) Pressure drops DF •span * lodg*pol* pin* D, 1600-1200-"S i . 1000-m o 800-CO 0-- • " • • " • —• 1 1 1 1 1 1 1 ' (c) Pressure drops DF • aspen * lodgtpol* pin* Figure 4.14 Development of panel mechanical properties with different species based on flake size All1 values in Table 4.2. All panels were made with 6% MC and at 30 sec. PCT. (a) Specific MOR. (b) Specific MOE. (c) Specific IB. 77 S. 5 o 8 CL CO 90-40-30-20-10-T - m (a) D, D, Pressure drops tnwll laigtr DF 8000 7000 <0 Q_ 6000 "*•' 9000-2 4000 S= 3000-O Q. 2000-co 1000 D, D2 Pressure drops •mall iatgar (b) DF I 00 o 0) a w 1600 1400 1200 1000 800-600 400 200 -1 r-D, 0, Pressure drops • tnwll * tergtr (c) DF Figure 4.15 Development of panel mechanical properties with different flake sizes based on species Average2 values in Table 4.2. All panels were made with 6% MC and at 30 sec. PCT. (a) Specific MOR. (b) Specific MOE. (c) Specific IB. 78 VDG after the transient relaxation region did not necessarily mean well developed mechanical properties. A mat might have crept and relaxed sufficiently to have its VDG developed well after the transient relaxation region, but with insufficient adhesive cure it could yield poor bonding since the pressure drop would suddenly release the interflake contact. This poor bonding would not only affect IB but MOR and MOE as well through interlayer shearing. In this study range, 6 percent MC provided overall higher MOR, MOE and IB than 2 percent MC. Fast press closing speed (30 second PCT) provided slightly superior MOR and MOE but inferior IB. Such slight influence might be due to the lower specified maximum pressure employed in this study. Aspen material provided better MOR and MOE than lodgepole pine but poorer IB. Flake size positively influenced the MOR and MOE but exhibited little effect on IB. 4.3.3 Panel dimensional stability Based on the experiment, three types of thickness swell were analyzed, the thickness swell after conditioning in a humidity chamber to 22 percent EMC (TS22%), the thickness swell after VPS treatment (TSvps), and the irrecoverable thickness swell, ie. the thickness swell after oven-drying and conditioning in the CTH conditioning room back to initial condition (TS^). The mean percentage swell values for each of the pressing conditions are shown in Table 4.3. The development of these thickness swells are displayed in Figures 4.16 to 4.19. In general, the development of dimensional stability was quite unstable in the transient relaxation region but became more stable thereafter, especially in the asymptotic region. Panels with 6 percent MC, which exhibited larger rheological responses, provided better dimensional stability. Panels with 2 percent MC, which had limited rheological responses during hot pressing, exhibited poor dimensional stability because of more compressive stresses. The irrecoverable thickness demonstrated the extent of the set compressive stresses. The thickness swell was extremely high for the 2 percent MC panels in the transient relaxation region Dj likely arising from poor bonding. An improved dimensional stability 79 Table 4.3 Mean percentage of thickness swell for four pressure drops and different levels of MC, PCT, species and flake sizes. (a) Thickness swell after conditioned in conditioning chamber giving 22% EMC (TS22%) TS22* (%) Species aspen aspen lodgepole pine Average2 Flake size small small larger All1 small larger All1 small larger AH1 MC (%) 2 6 Average2 6 6 6 PCT (s) 30 60 90 All1 30 60 90 All1 30 60 90 All1 30 30 30 Pressure drops D, 15.0 18.7 28.7 20.8 17.0 16.0 15.3 16.1 16.0 17.3 22.0 18.4 — — — D2 15.7 17.7 18.7 17.3 17.3 15.7 16.7 16.6 16.5 16.7 17.7 16.9 17.3 15.0 16.2 19.7 20.3 20.0 18.5 17.7 18.1 D3 17.3 19.3 17.3 18.0 16.0 15.7 17.3 16.3 16.7 17.5 17.3 17.2 16.0 14.0 15.0 25.7 22.0 23.8 20.8 18.0 19.4 DF 16.0 18.7 20.3 18.3 16.3 16.3 16.0 16.2 16.2 17.5 18.2 17.3 16.3 13.3 14.8 24.7 19.3 22.0 20.5 16.3 18.4 Table 4.3 (Continued). (b) Thickness swell after treated in VPS treatment (TS J. TSvp8 (%) Species aspen aspen lodgepole pine Average2 Flake size small small larger All1 small larger All1 small larger All1 MC (%) 2 6 Average2 6 6 6 PCT (s) 30 60 90 AH1 30 60 90 All1 30 60 90 AH1 30 30 30 Pressure drops D, 30.7 40.7 75.0 48.8 37.0 30.0 32.7 33.2 33.8 35.3 53.8 41.0 — — --D2 32.7 38.3 39.0 36.7 39.0 32.7 32.7 34.8 35.8 35.5 35.8 35.7 39.0 38.7 38.8 41.7 48.0 44.8 40.3 43.3 41.8 D3 36.3 41.0 38.7 38.7 33.0 30.0 35.7 32.9 34.7 35.5 37.2 35.8 33.0 40.3 36.7 52.3 54.7 53.5 42.7 47.5 45.1 DF 32.7 40.7 42.7 38.7 33.3 32.0 30.7 32.0 33.0 36.3 36.7 35.3 33.3 35.3 34.3 49.3 48.7 49.0 41.3 42.0 41.7 Table 4.3 (continued). (c) Irrecoverable thickness swell (TStt). TSi, (%) Species aspen aspen lodgepole pine Average2 Flake size small small larger AH1 small larger All1 small larger All1 MC (%) 2 6 Average2 6 6 6 PCT (s) 30 60 90 All1 30 60 90 All1 30 60 90 AH1 30 30 30 Pressure drops D j 22.3 33.7 73.7 43.2 30.0 22.3 23.0 25.1 26.2 28.0 48.3 34.2 — — ~ D2 22.7 30.0 30.0 27.6 33.0 24.7 24.0 27.2 27.8 27.3 27.0 27.4 33.0 31.7 32.3 32.7 43.7 38.2 32.8 37.7 35.3 D3 28.0 32.3 29.3 29.9 25.7 22.3 26.3 24.8 26.8 27.3 27.8 27.3 25.7 34.3 30.0 45.7 51.7 48.7 35.7 43.0 39.3 DF 23.3 31.3 34.3 29.7 25.0 23.3 23.0 23.8 24.2 27.3 28.7 26.7 25.0 29.3 27.2 41.7 44.3 43.0 33.3 36.8 35.1 Notes: (1) All1 represents the average value for different PCT levels and different flake sizes, respectively. (2) Average2 represents the average value of different MC levels different species, respectively. 82 Pressure drops • 2%MC « 6 \MC (a) 9S & (A a. > V) 90 49 40 30 25 20-1 r-D, D2 D, Pressure drops • 2%MC * 6SMC (b) DF 60 99 90 ^^ * k_ V) H 49 40 35 30 29 20 D, D, Pressure drops 2%MC «%MC (c) DF Figure 4.16 Development of panel thickness swell with different MC levels based on PCT All1 values in Table 4.3. All panels were made with aspen small flakes, (a) TS22% (b) TS^ (c) TS*. (a) Pressure drops • 3d o « i • n , 99 30 30 29 20 0, (b) D3 Pressure drops DF so. o GO. * so. D, (C) Pressure drops D f 30. D 60. * 90. Figure 4.17 Development of panel thickness swell with different PCT levels based on MC Average2 values in Table 4.3. All panels were made with aspen small flakes, (a) T S ^ (b) TS™ (c) TS*. 26 ^ M eo «• *~ 16 14-D, (a) Pressure drops D, •pan lodgapol* pin* 55 90 # « > CO 40 30 25 20 01 (b) Pressure drops DF • up*n iodgtpol* pin* 50 --> 45 £ _ 40' CO I- 35 25 20' D, (C) Pressure drops of • nptn - iodg*pol* pin* Figure 4.18 Development of panel thickness swell with different species based on flake size All1 values in Table 4.3. All panels were made with 6% MC and at 30 sec. PCT. (a) TS^* (b) TS^ (c) TS^. 85 28-26-£ n # 20 CO " • " 16-10-D, (a) D, Pressure drops D, •mall kvgar •0 55 SO 0) a CO 40 X ' 85 20 D, D2 Pressure drops • imall * larger (b) DF D, Pressure drops •rmll Mirgtr (C) Figure 4.19 Development of panel thickness swell with different flake sizes based on species Average2 values in Table 4.3. All panels were made with 6% MC and at 30 sec. PCT. (a) TS^* (b) TS^ (c) TS*,. 86 was observed with pressure drop D2, indicating that D2 probably provided an optimum opportunity to allow some set stresses to be released. The 30 second PCT provided slightly better dimensional stability than the 60 and 90 second PCT, which provided similar dimensional stability after the transient relaxation region. Aspen provided better dimensional stability than lodgepole pine. Increasing flake size increased the thickness swell since large-size flakes would result in larger variations in panel horizontal density. The high-density regions therefore would exhibit more swell upon wet conditions. An exception was observed in Figure 4.19 where panels with large-size flakes provided lower TS2296. The reason for this exception was not established. 4.4 The effect of pressure manipulation on panel properties Based on the study of the development of panel properties, an empirical study of the feasibility of altering panel properties by pressure manipulation, ie. by pressure breathing, was carried out. Pressure and mat displacement response curves associated with each of the pressure breathing stages, Bx and B2, are shown in Figure 4.20. Properties of the manufactured panels were evaluated and were compared with panels manufactured without pressure breathing B0. Table 4.4 shows the mean values of specific MOR, MOE, IB and thickness swells TS2296, TSvps and Tsir with variables of pressure breathing and MC. 4.4.1 Effect on panel vertical density gradient The comparisons in VDG of panels manufactured with B0, Bi and B2 are shown in Figure 4.21. An examination of this graph indicated that pressure breathing appeared to have little effect on panel VDG. A slight difference was observed in the panel manufactured with 6 percent MC and B0 in the higher thickness range comparing to those manufactured with Bt and B2. This difference is more likely attributed to the precure of adhesive on the mat bottom surface arising from the variations of initial upper platen position rather than the effect of pressure breathing (referring to Figure 4.11). This suggests that pressure breathing could influence the panel properties by altering mat interior compression stresses rather than by altering panel VDG. 87 ASRH60-P ASRH60-D 0 1 2 3 4 5 6 7 8 9 Press time (min.) ( a ) 6% MC. 0 1 2 3 4 5 6 7 8 9 Press time (min.) ( C ) 6% MC. 4 i 2 Pressure n ASRL60-P J • »• i • I-/ • r / • V / • "V / ' * V-f * * \**A. f »• U 1 ASRL60-D 0 1 2 3 4 5 6 7 8 9 Press time (min.) ( b ) 2% MC. 0 1 2 3 4 5 6 7 8 Press time (min.) ( d ) 2% MC B, &2 Figure 4.20 Pressure and mat displacement response curves of flakeboards showing each of the pressure breathing cycle. All flakeboards were made with aspen small flakes and at 60 sec. PCT. 88 Table 4.4 Mean specific MOR, MOE, IB and thickness swell with variables of pressure breathing and MC. MC (%) 2 6 Pressure breath Bo B, B2 Bo B, B2 MOR (MPa) 21 28 27 34 26 23 MOE (MPa) 3673 3629 3504 4648 3489 3590 IB (KPa) 842 1558 1376 1470 1040 892 TS22% (%) 18.7 8.3 8.7 16.3 8.7 9.3 TSyps ( % ) 40.7 27.3 28.7 32.0 36.3 37.0 TSir (%) 31.3 17.0 18.3 23.3 26.7 28.0 89 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Thickness (mm) ( a ) 6% MC. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Thickness (mm) ( b ) 2% MC B0 - ~ Bj B, Figure 4.21 Vertical density profiles of flakeboards made with and without pressure breathing. All panels were made with aspen small flakes and at 60 sec. PCT. 90 4.4.2 Effect on panel mechanical properties Comparisons in specific MOR, MOE and IB for panels pressed with and without pressure breathing are shown respectively in Figure 4.22 (a), (b) and (c). Pressure breathing appeared to improve these properties at 2 percent MC level only. At 6 percent MC, pressure breathing adversely affect these properties. Bj provided slightly higher values than 4.4.3 Effect on panel dimensional stability Pressure breathing dramatically improved dimensional stability especially at 2 percent MC as shown in Figure 4.23 (a), (b) and (c). At 2 percent MC, panels manufactured with pressure breathing provided approximately 40 to 50 percent lower values in all TS22%, TSvps and TS;, comparing to those manufactured traditionally. At 6 percent MC, the influence of pressure breathing was not consistent. This positive influence was observed only in T S ^ , but not in TSvps and TSjf. Again, Bj provided a slightly superior effect than B2. From the foregoing, it could be reasonably concluded that pressure breathing positively affects panel properties with the 2 percent MC condition but adversely affects panel properties with the 6 percent MC condition. The design of the pressure breathing appeared to be in favour of the situations where limited rheological response were exhibited by a mat during hot pressing. Pressure breathing provided an opportunity to allow some compressive stresses to be released upon pressure reduction; restoring it after a few seconds of reduction ensured the panel VDG and other properties would develop continuously. For panels possessing large rheological responses during hot pressing, pressure breathing would interfere with the development of bonding. 4.5 Statistical analysis To test the significant differences in each of the properties of specific MOR, MOE, IB and dimensional stability exhibited by the experimental variables and by the pressure breathing, analysis of variance (ANOVA) was performed based on the complete factorial design in SD4, SD5 and SD6 in Table 3.2, respectively. All two-way interactions and three-way interactions were also examined for statistical significance. The ANOVA tables 2%MC 91 (a) 6% MC re n 5 Hi O 5 a 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0 (b) 2% MC 6% MC 1800 1600 5 1400 6 1200 (C) 2% MC 6% MC B, Figure 4.22 Pressure breathing effects on panel mechanical properties. All panels were made with aspen small flakes and at 60 sec. PCT. (a) Specific MOR. (b) Specific MOE. (c) Specific IB. 92 (a) 2%MC 6% MC (b) 2% MC 6% MC (C) 2% MC 6% MC M B0 I B , • B2 Figure 4.23 Pressure breathing effects on panel thickness swell. All panels were made with aspen small flakes and at 60 sec. PCT. (a) T S ^ (b) TS^ (c) TS,. 93 comparing all the experimental variables are presented respectively in Appendix II. The four levels of pressure drops were considered as one of the variables in the ANOVA analysis in order to test the significant differences exhibited by these pressure drops. Tables 4.5 and 4.6 summarize the F-values from the ANOVA table for each of the properties with all the experimental variables and pressure breathing. After comparing these F-values with the values from a F-table, statistical significance is symbolized by *, **, and n.s.; these symbols represent significant at 99 percent confidence level, significant at 95 percent confidence level, and non-significant respectively. s *~* TO > b u § C/3 / • • s . *& s—^  ^ H /"\ fcS ' * • ' CO H CO H ^ ^ y TO c-CQ *-« £ w o 2 /-•s TO 2 w a: o 2 * * * * * * * * * * * * * * t * r « . - < ^ N eo Os -* c>J O) rn so O r-^  oo f i f i wi ci so If ) *M M ^ H P 4 H * * * * * * * * * J * * * * r-; oo ^ p •* rn — os o\ os *c' © -* m' vo -* *" ro — M -H * * « * « « * * * * * * * * t * « t es o 1 O O «•' SC in vc N J * * * * * * * * * * * * so f- S! £2 <^ r- <*> ~ tr> ON 00 ^ vO SO S5 o^ J; SJ 00 — 00 »—1 v ^ * * * CO « _ M * * * • I * Jt C * * * J> = $ £ o S S 3 $ S - 2 » 2 2 SO N - H J . * ! * * * * * * * * ; ; ; so K 2 !Q r- r- t -. t*; . ""J <—i ^ SO iCl •""* )£ rr^ SO l/"i Os PJ <s 2 ^ in & s MC PCT Drops MC*PCT MC*Drops PCT*Drops MC*PCT*D * * * * * * * * * * * * * * *i *""* ts so *•* **• t s *». •"": <•<•> r - « o «/•» » o j ^ ^ c i 6 * d « » in « K en t/5 * d e c * e c 5 os t-» —< m r-» so « se « o o so t_ r m N ' 6 wi d « * « to to _ ce * * c c * * e so r« *-' *-' «*> ^ *^ r so in rt ^ * t^ {5 S <N> © oo ro c4 ^^  * * * * * # * « * * * ; « ^ S S £ S $ o ^ § • § 2 * ^ I 2 * * * * * * * * * * * * ~ ° ! co <i (g | s a S Jf ^ N K.' ft ^ * J * * "» "9 * * * * c c © S r? jo TJ- — «t ^ oo • • ^ ^ ^ *-« to O. £ - P . CL * Species Size Drops Species*Size Species*Dro Size*Drops Species*Size Table 4.6 Summary of ANOVA results comparing pressure breathing and MC for the properties of specific MOR, MOE, IB and dimensional stability. Source Breath MC Breath*MC F-values MOR (MPa) 7.74 ** 12.26 ** 84.60 ** MOE (MPa) 25.06 ** 14.42 ** 17.69 ** IB(KPa) 14.07 ** 20.40 ** 187.10 ** TS22% (%) 184.20 ** 1.07 n.s. 4.87 * TS^ (%) 2.86 n.s. 3.20 n.s. 12.80 ** TSir (%) 4.99 * 6.49 * 15.78 ** Notes: *, significant at 99% confidence level; **, significant at 95% confidence level; n.s., non-significant. in 96 5.0 CONCLUSIONS Based on the study of pressure responses of panels during hot pressing and their relationship to the panel properties as well as in situ adhesive cure, it can be concluded: 1 The pressure responses of hot pressed panels were reproducible at constant pressing parameters and conditions. 2 In this study range, aspen flake mats exhibited more extensive Theological behaviour, i.e., crept faster, relaxed to a higher degree in the transient relaxation region, and yielded a lower pressure level at the end of the asymptotic relaxation region than lodgepole pine flake mat. 3 Increasing flake size extended the mat creep region but the mat relaxed at the same rate in the transient relaxation region and yielded the same pressure level at the end of the asymptotic region. 4 Press closing time influenced mainly the rate of mat creep. Faster press closing speed resulted in an extended creep region. The rates of pressure relaxation in the transient relaxation region were the same, and pressure values at the end of the asymptotic relaxation region were identical regardless of press closing time. 5 Based on a given furnish material, increased moisture content further enhanced mat Theological response and resulted in an even faster mat creep, faster pressure relaxation in the transient relaxation region and lower pressure value, at the end of asymptotic region. 6 Increasing resin content increased the mat Theological response at 2 percent moisture content condition, but had little effect at 6 percent moisture content condition. 7 Moisture content, press closing time and resin content interactively influenced mat Theological properties during hot pressing at a given temperature. 8 The in situ adhesive cure developed non-uniformly from mat face to core. This cure gradient was large in the transient relaxation region but smaller in the asymptotic region. The development of in situ adhesive cure could be determined by the mat displacement response for panels with limited Theological response but not for those with large Theological response (aspen small-size flakes with 6 percent moisture 97 content, for example) since the mat displacement response became less sensitive with adhesive cure. 9 The development of panel vertical density gradient also depended on panel overall rheological behaviour during hot pressing. This development would continue until the pressure relaxed to a value below 1 MPa or when no immediate mat springback was observed upon pressure reduction. 10 A panel exhibiting large rheological response during hot pressing would yield better dimensional stability in thickness. 11 The development of panel mechanical properties depended on both the development of vertical density gradient and adhesive cure. 12 The pressure breathing manipulation method influenced panel properties by altering interior compression stresses rather than panel vertical density gradient. At 2 percent moisture content level, panel modulus of rupture, modulus of elasticity and internal bond were improved, while panel dimensional stability were greatly improved. However, at 6 percent moisture content, panel properties were adversely affected. This study provided possibilities of detecting the development of panel properties during hot pressing and altering these properties by pressure manipulation. The methodology used could be applied in industry. However, fundamental experiment with commercial flakes on a larger scale hot press is necessary. LITERATURE CITED Arima, T. 1974 a. Studies on rheological behaviour of wood under hot pressing. II. Effects of temperature and moisture contents in wood on deformation under hot pressing. Mokuza Gakkaishi. 20(8): 355-361. Arima, T. 1974 b. Studies on rheological behaviour of wood under hot pressing. III. A consideration on the mechanism of deformation. Mokuza Gakkaishi. 20(8): 362-367. Bariska, C. 1974. Optimizing the pressing of particleboards. The manufacture of particleboards with urea-formaldehyde binders using special automated regulation systems for the pressing process. Holz-Zentralbl. 100(80): 1247-1249. Bariska, M. 1985. Creep and fracture phenomena in wood tissues. Symposium on forest products research international-achievements and the future:22-26. Beech, J.C. 1975. The thickness swelling of wood particleboard. Holzforschung. 29(1):11-18. Bodig, J., and B.A.Jayne. 1985. Mechanics of wood and wood composites. Van Nostrand Reinhold. New York. NY. Bolton, A J . and P.E.Humphrey. 1977. Measurement of the tensile strength development of urea formaldehyde resin-wood bonds during pressing at elevated temperatures. J. Inst. Wood. Sci. 7(5): 11-14. Bolton, A.J., P.E.Humphrey, and P.K.Kawouras. 1989 a. The hot pressing of dry-formed wood-based composites. Part III. Predicted vapour pressure and temperature variation with time, compared with experimental data for laboratory boards. Holzforschung. 43(4):265-274. Bolton, A.J., P.E.Humphrey, and P.K.Kawouras. 1989 b. The hot pressing of dry-formed wood-based composites. Part IV. Predicted variation of mattress moisture content with time. Holzforschung. 43(5):345-349. Bolton, A.J., P.E.Humphrey, and P.K.Kawouras. 1989 c. The hot pressing of dry-formed wood-based composites. Part VI. The importance of stresses in the pressed mattress and their relevance of the minimisation of pressing time, and the variability of board properties. Holzforschung. 43(6):406-410. Childs, M.R. 1956. The effect of density, resin content, and chip width on springback and certain other properties of dry formed flat pressed particleboard. M.S. Thesis, Dept. of Wood and Pap. Sci., North Carolina State Univ., Raleigh. Chow, S.-Z. 1969. A kinetic study of the polymerization of phenol-formaldehyde resin in the presence of cellulosic materials. Wood Science. 1(4):215-221. Chow, S.-Z., and W.V.Hancock. 1969. Method for determining degree of cure of phenolic resin. For. Prod. J. 19(4):21-29. Chow, S.-Z. 1972. Thermal analysis of liquid phenol-formaldehyde resin curing. Holzforschung. 26(6):229-232. Chow, S.-Z., and H. N. Mukai. 1972. Polymerization of phenolic resin at high vapour pressure. Wood Science. 5(1):65-72. Chow, S.-Z. 1973. Residual reactivity of partially cured phenolic resin: A bromination, X-ray analysis method. Wood Science. 6(2): 143-145. Chow, S.-Z. 1977. A curing study of phenol-resorcinol-formaldehyde resins using infrared spectrometer and thermal analysis. Holzforschung. 31(6):200-205. 100 Christiansen, A.W. 1985. Differential scanning calorimetry of phenol-formaldehyde resols. J. Appl. Polymer Sci. 30:2279-2289. Deppe, H.J., and K.Ernst. 1964. Technologie der spanplatten. Holzzentralblatt Verlag GmbH, Stuttgart, 156-177. Ellis, S.C. 1989. Some factors affecting the flow and penetration of powdered phenolic resins into wood. Ph.D. thesis. University of British Columbia. CA. Geimer, R.L., H.M.Montrey, and W.F.Lehmann. 1975. Effects of layer characteristics on the properties of three-layer particleboards. For. Prod. J. 25(3): 19-29. Geimer, R.L., R.A. Follensbee, A.W. Christiansen, J.A. Koutsky, and G.E. Myers. 1990. Resin characterization. Prepared for publication in 24th International Washington State University Particleboard/composite Materials Symposium. Geimer, R.L., A.W. Christiansen. 1991. Adhesive curing and bonding: response to real time conditions. In Proceeding of the Adhesive and Bonded Wood Products Symposium, at Seattle, Washington. Gerard, J.C. 1966. Dimensional behaviour of particles in simulated particleboard constructions. For. Prod. J. 16(6): 40-48. Goring, D.A.I. 1971. Polymer properties of lignin and lignin derivatives. Wiley-Interscience, New York, NY. Gross, B. 1947. On creep and relaxation. I. Jour. Appl. Phys. 18:212-221. Gross, B. 1948. On creep and relaxation. II. Jour. Appl. Phys. 19:257-264. Gross, B., and H.Pelzer. 1951. On creep and relaxation. III. Jour. Appl. Phys. 22:1035-1039. Grossman, P.U.A., and R.S.T.Kingston. 1954. Creep and stress relaxation in wood during bending. Australian. J. Appl. Sci. 5:403. Grossman, P.U.A. 1963. Research on the rheology of wood. Holzforschung. Berlin. 17(5): 146-149. Halligan, A.F., and A.P.Schniewind. 1972. Effect of moisture on physical and creep properties of particleboard. For. Prod. J. 22(4):41-48. Harless, T.E.G., F.G.Wagner, P.H.Short, R.D.Seale, P.H.Mitchell, and D.S.Ladd. 1987. Amodel to predict the density profile of particleboard. Wood and Fibre Science. 19(l):81-92. Heebink, B.G., and F.V.Hefty. 1969. Treatments to reduce thickness swelling of phenolic-bonded particleboard. For. Prod. J. 19(11): 17-26. Heebink, B.G., W.F.Lehmann, and F.V.Hefty. 1972. Reducing particleboard pressing time: exploratory study. U.S.D.A.Forest Service Research Paper FPL 180. U.S. Department of Agriculture/ Forest Service Forest Products Laboratory-Madison, Wis. Howard, E.T. 1974. Slash pine root wood in flakeboard. For. Prod. J. 24 (6): 29-35. Humphrey, P.E., and A.J.Bolton. 1979. Urea formaldehyde resin bond strength development with reference to wood particleboard manufacture. Holzforschung. 33:129-133. Irvine, G.M. 1984. The glass transitions of ligmn and hemicellulose and measurement by differential thermal analysis. Tappi. 67(5): 118-121. Kamke, F.A., and L.J.Casey. 1988 a. Gas pressure and temperature in the mat during flakeboard manufacture. For. Prod. J. 38(3):41-43. Kamke, F.A., and L.J.Casey. 1988 b. Fundamentals of flakeboard manufacture: internal-mat conditions. For. Prod. J. 38(6):38-44. Katovic, Z. 1967. Curing of resole-type phenol-formaldehyde resin. J. Appl. Polymer Sci. 11:85-93. Kawai, S., and H.Sasaki. 1986. Mokuzai Gakkaishi. 32(5):324-330. Kelly, M.W. 1977. Critical literature review of relationships between processing parameters and physical properties of particleboards. USDA Forest Service. Forest Products Laboratory. General Technical Report. FPL-10. U.S. Department of Agriculture. Forest Service. Forest Products Laboratory. Madison. WIS. Kelly, S.S., T.G. Rials, and W.G. Glasser. 1987. Relaxation behaviour of the amorphous components of wood. Journal of Materials Science. 22:617-624. Kingston, R.S., and B.Budgen. 1972. Some aspects of the Theological behaviour of wood. Part IV: Non-linear behaviour at high stresses in bending and compression. Wood Sci. Technol. 6: 230-238. Kitahara, K. 1957. Stress relaxation of chip-board in hot press. Bulletin of the Tokyo University Forests. No. 53. Kitazawa, G. 1947. Relaxation of wood under constant strain. (A study of the viscoelastic property of wood). New York State College of Forestry. Tech. Publ. No. 67. KoUmann, F. 1960. Verfermung und Fliefien bei Querdruckbelastung von Holzwurfeln. Materialprufung. 2(8):289-294. KoUmann, F. 1961. Rheologie und strukturfestigkeit von Holz. Holz als Roh- Werkst. 19(3):73-80. KoUmann, F.F.P., and A.C.Jr.Wilfred. 1968. Principles of wood science and technology. Vol. I. Solid wood. Springer-verlag, New York. KoUmann, F.F.P., E.W.Kuenzi, and A.J.Stamm. 1975. Principles of wood science and technology. II. wood based materials. Spriner-Verlag. New York. Heidelberg. Berlin. Kuhne, H. 1961. Beitrag zur theorie des mechanischen formanderungsverhaltens von Holz. Holz als Roh- Werkst. 19(3):81-82. Kunesh, R.H. 1961. The inelastic behaviour of wood: a new concept for improved panel forming processes. For. Prod. J. 11(9): 395-406. Laufenberg, T.L. 1986. Using gamma radiation to measure density gradients in reconstituted wood products. For. Prod. J. 36(2):59-62. Lehmann, W.F., R.L.Geimer, and F.V.Hefty. 1973. Factors affecting particleboard pressing time: interaction with catalyst systems. U.S.D.A.Forest Service Research Paper FPL 208. U.S. Department of Agriculture/Forest Service Forest Products Laboratory-Madison, Wis. 104 Liiri, V.O. 1969. The pressure in the particleboard production. Holz Roh- Werkst. 27(10):371-378. Mahoney, R.J. 1980. Physical changes in wood particles induced by the particleboard hot-pressing operation. Proc-Wash-State-Univ-Int-Symp-Particleboard. 213-223. Maku, T., R.Hamada, and H.Sasaki. 1959. Studies on the particleboard. Rept. 4. Temperature and moisture distribution in particleboard during hot-pressing. Wood Research, Kyoto Univ. 21:34-46. Maloney, T.M. 1977. Modern particleboard & Dry-process fibreboard manufacturing. Minami, Y. 1939. Tensile creep tests on wood. J. Aero. Res. Inst., Tokyo Imperial Univ., No. 174 Moslemi, A.A. 1974. Particleboard. Vol. II. Neusser, H., V.Krames, and K.Haidinger. 1965. The effect of moisture on particleboards with special consideration of swelling. Holzforschung. Holzverwert. Pt. I. 17(3):43-53; Pt. II. 17(4):57-69. Northcott, P.L. 1955. Bond strength as indicated by wood failure or mechanical test. For. Prod. J. 5:118-123 Pentoney, R.E. and R.W.Davidson. 1962. Rheology and the study of wood. For. Prod. J. 12(5): 243-248. Pillar, W.O. 1966. Determining curing properties of an adhesive in contact with wood. For. Prod. J. 16(6):29-37. Pizzi, A. 1983. Wood adhesives. Plath, L., and E.Schnitzler. 1974. The density profile, a criterion for evaluating particleboard. Holz Roh-Werkst. 32(11):443-449. Raczkowski, J. 1969. The effect of moisture content changes on the creep behaviour of wood. Holz Roh-Werkst. 27(6):232-237. Ranta-Maunus, A. 1975. The viscoelasticity of wood at varying moisture content. Wood Sci. Technol. 9:189-205. Rice, J.T. 1960. The effects of selected variables on the properties of flat pressed flakeboard. M.S. Thesis, Dept. of Wood and Pap. Sci. North Carolina State Univ., Raleigh. Rice, J.T., J.L.Snyder, and C.A.Hart. 1967. Influence of selected resin and bonding factors on flakeboard properties. For. Prod. J. 17(8):49-57. Saito, F. 1972. Spring back of particleboards. Wood Ind. (Jap). 27(1): 14-18. Schaffer, E.L. 1972. Modelling the creep of wood in a changing moisture content environment. Wood Fibre. 3(4):232-235. Schniewind, A.P. 1968. Recent progress in the study of the rheology of wood. Wood Sci. Technol. 2:188-206. Schniewind, A.P., and J.D.Barrett. 1972. Wood as a linear orthotropic viscoelastic material. Wood Sci. Technol. 6: 43-57. Seborg, R.M., and A.J.Stamm. 1941. The compression of wood. U.S. For. Prod. Lab. Re. No. R1258. Selbo, M.L. 1958. Curing rates of resorcinol and phenol-resorcinol glues in laminated oak members. For. Prod. J. 8(5): 145-149. Siegmann, A., and M. Narkis. 1977. Thermal analysis of thermosetting phenolic compounds for injection moulding. J. Appl. Polymer Sci. 21:2311-2318. Smith, D. 1980. Considerations in press design for structural boards. Proc-Wash-State-Univ-Int-Symp-Particleboard. 95-104. Smith, D.C. 1982. Waferboard press closing strategies. For. Prod. J. 32(3):40-45. Steiner, P.R., and S.R.Warren. 1981. Rheology of wood-adhesive cure by torsional braid analysis. Holzforschung. 35:273-278. Steiner, P.R., and S.R.Warren. 1987. Behaviour of urea-formaldehyde wood adhesives during early stages of cure. For. Prod. J. 37(l):20-22. Strickler, M.D. 1959. Effect of press cycles and moisture content on properties of douglas-fir flakeboard. For. Prod. J. 9:203-215. Suchsland, O. 1959. An analysis of the particleboard process. Michigan Agricultural Experiment Station, Quarterly Bulletin 42(2): 350-372. Suchsland, O. 1962. The density distribution in flakeboard. Quart. Bull., Michigan. Agri. Exp. Sta., Michigan State Univ. 45(1):104-121. Suchsland, O. 1967. Behaviour of a particleboard mat during the press cycle. For. Prod. J. 17(2): 51-57. Tang, -Y., and -W.T.Simpson. 1990. Perpendicular-to-grain Theological behaviour of loblolly pine in press drying. Wood and Fibre Science. 22(3):326-342. Troughton, G.E., and S.-Z.Chow. 1974. Cross-linking in phenol-formaldehyde resins. Holzforschung. 28(2):55-57. Vital, B.R., W.F. Lehmann, and R.S. Boone. 1974. How species and board densities affect properties of exotic hardwood particleboards. For. Prod. J. 24 (12): 37-45. Ward, I.M. 1983. Mechanical properties of solid polymers. 2nd ed. Wiley-Interscience. New York, NY. Wolcott, M.P., F.A.Kamke, and D.A.Dillard. 1990. Fundamentals of flakeboard manufacture: viscoelastic behaviour of the wood component. Wood and Fibre Science. 22(4):345-361. Wood handbood. 1989. United States. Forest Products Laboratory, Madison, Wis. Young, R.H. 1986. Cure and durability of adhesives for wood-based composites. Composite systems from natural and synthetic polymers, elsevier science publishers B.V., Amsterdam. 225-231. Youngs, R.L. 1957. The perpendicular to grain mechanical properties of red oak as related to temperature, moisture content, and time. U.S. Forest Products Laboratory Report. No.2079. Zhang, Q.L., and H.H.Wen. 1990. Research on processing of CBPF. International Timber Engineering Conference. Tokyo. 40-48. iOb APPENDIX I. DSC thermograms of phenol-formaldehyde powder resin samples pressed in hot press for 1, 2, 4 and 7 minutes respectively. Heat flow (W/g) Heat flow (W/g) Heat flow (W/g) "aT 2 J-SJ 3 a 3 T* ^^  • 03 1 2 & 9 e 601 Heat flow (W/g) s Heat flow (W/g) Heat flow (W/g) 5 o 3 I i t 0X1 • . » • t . t -w JH i : -«.f-- • .4 TC -Temperature (°C) Temperature (°C) r r a. . . (b) i « . « t H . H J / I Temperature (°C) Figure 1-3. Pressed for 4 minutes, 6% MC. (a) M-layer. (b) 1/4-layer. (c) S-layer. (a) -3 1 1 1 0 . 7 -0 .B-0 . 8 -0 . 4 -0 . 3 -o . j J 0 . 1 -0 . 0 -DF-H30M IBB 103 197.39'C / ^ oj /o Temperature (°C) 60 (c) £ 1 I «7.0i*C 84 .87J/g Temperature (°C) O 0.3 I" (b) 16a.54"C 72.32J/g 75T Temperature (°C) Figure 1-4. Pressed for 7 minutes, 6% MC. (a) M-layer. (b) 1/4-layer. (c) S-layer. ts» (a) I" Temperature (°C) » o (c) ,2 ••" Temperature (°C) 60 I-S ••'-(b) T3T Temperature (°C) Figure 1-5. Pressed for 1 minute, 2% MC. (a) M-layer. (b) 1/4-layer. (c) S-layer. U ) (a) | .... I"" I» I .»"C ?03.IJ/f 13T Temperature (°C) (c) §. .<H Temperature (°C) (b) Temperature (°C) Figure 1-6. Pressed for 2 minutes, 2% MC. (a) M-Iayer. (b) 1/4-layer. (c) S-layer. *. (a) g ] W7. *>*C 75.S3J/fl (C) | ...J 2 too t4a Temperature (°C) 40.0ZJ/I 7. .KMJ/t Temperature (°C) 00 «5 U I M " C M.?3J/f 13T Temperature (°C) Figure 1-7. Pressed for 4 minutes, 2% MC. (a) M-layer. (b) 1/4-layer. (c) S-layer. (b) u» (a) -3 I Temperature (°C) Temperature (°C) M 1 1 (b) Temperature (°C) Figure 1-8. Pressed for 7 minutes, 2% MC. (a) M-laycr. (b) 1/4-layer. (c) S-Iayer. o Ill APPENDIX II. ANOVA tables comparing all the experimental variables for specific MOR, MOE, IB and dimensional stabilities. 118 Table II-l. Analysis of Variance for MOR. (a) with variables of MC, PCT and pressure drops. Source MC PCT Drops MC*PCT MC*Drops PCT*Drops MC*PCT*Drops Error Total DF 1 2 3 2 3 6 6 120 143 SS 2304.00 176.39 1392.39 591.50 58.72 104.28 184.28 381.00 5192.56 MS 2304.00 88.19 464.13 295.75 19.57 17.38 30.71 3.17 F 725.67 27.78 146.18 93.15 6.17 5.47 9.67 P 0.000 0.000 0.000 0.000 0.001 0.000 0.000 (b) with variables of species, flake sizes and pressure drops, Source species sizes drops species*sizes species*drops sizes*drops species*sizes*drops Error Total DF 1 1 2 1 2 2 2 60 71 SS 523.85 8008.93 350.23 120.67 42.49 2.69 37.39 381.89 9468.15 MS 523.85 8008.93 175.11 120.67 21.25 1.35 18.69 6.36 F 82.30 1258.31 27.51 18.96 3.34 0.21 2.94 P 0.000 0.000 0.000 0.000 0.042 0.810 0.061 Table II-2. Analysis of Variance for MOE. (a) with variables of MC, PCT and pressure drops. 119 Source DF SS MS MC PCT Drops MC*PCT MC*Drops PCT*Drops MC*PCT*Drops Error Total 1 2 3 2 3 6 6 120 143 23393350 17415580 15428486 105251 1161707 2029969 4324297 4641983 68500624 23393350 8707790 5142829 52625 387236 338328 720716 38683 604.74 225.11 132.95 1.36 10.01 8.75 18.63 0.000 0.000 0.000 0.260 0.000 0.000 0.000 (b) with variables of species, flake sizes and pressure drops. Source species sizes drops species* species* 'sizes 'drops sizes*drops species* Error Total 'sizes*drops DF 1 1 2 1 2 2 2 60 71 SS 13438637 41301812 2336717 595810 754333 954073 1764114 2942110 64087604 MS 13438637 41301812 1168359 595810 377166 477037 882057 49035 274, 842, 23, 12, 7, 9, 17, F .06 .29 .83 .15 .69 .73 .99 P 0.000 0.000 0.000 0.001 0.001 0.000 0.000 120 Table II-3. Analysis of Variance for IB. (a) with variables of MC, PCT and pressure drops. Source MC PCT Drops MC*PCT MC*Drops PCT*Drops MC*PCT*Drops Error Total DF 1 2 3 2 3 6 6 264 287 SS 9541532 692945 6061427 2707098 810983 491153 1626653 1852605 23784396 MS 9541532 346472 2020476 1353549 270328 81859 271109 7017 1359, 49. 287. 192. 38. 11. 38. F .69 .37 .92 .88 .52 ,67 ,63 P 0.000 0.000 0.000 0.000 0.000 0.000 0.000 (b) with variables of species, flake sizes and pressure drops. Source species sizes drops species* species* 'sizes •drops sizes*drops species'* Error Total <sizes*drops DF 1 1 2 1 2 2 2 132 143 SS 1354896 27945 1348443 704480 271753 137874 104827 960749 4910968 MS 1354896 27945 674222 704480 135876 68937 52414 7278 F 186.15 3.84 92.63 96.79 18.67 9.47 7.20 P 0.000 0.052 0.000 0.000 0.000 0.000 0.001 121 Table II-4. Analysis of Variance for TS 2296 ' (a) with variables of MC, PCT and pressure drops. Source MC PCT Drops MC*PCT MC*Drops PCT*Drops MC*PCT*Drops Error Total DF 1 2 3 2 3 6 6 48 71 SS 95.681 74.083 24.375 94.694 37.597 64.583 130.861 86.000 607.875 MS 95.681 37.042 8.125 47.347 12.532 10.764 21.810 1.792 F 53.40 20.67 4.53 26.43 6.99 6.01 12.17 P 0.000 0.000 0.007 0.000 0.001 0.000 0.000 (b) with variables of species, flake sizes and pressure drops. Source Species Sizes Drops Species*Sizes Species*Drops Sizes*Drops Species*Sizes*Drops Error Total DF 1 1 2 1 2 2 2 24 35 SS 393.361 61.361 11.556 0.250 38.889 16.889 12.667 55.333 590.306 MS 393.361 61.361 5.778 0.250 19.444 8.444 6.333 2.306 F 170.61 26.61 2.51 0.11 8.43 3.66 2.75 P 0.000 0.000 0.103 0.745 0.002 0.041 0.084 122 Table II-5. Analysis of Variance for TS^,. (a) with variables of MC, PCT and pressure drops. Source MC PCT drops MC*PCT MC*drops PCT*drops MC*PCT*drops Error Total DF 1 2 3 2 3 6 6 48 71 SS 1005.01 573.58 394.15 1038.53 450.15 984.64 1134.81 692.00 6272.88 MS 1005.01 286.79 131.38 519.26 150.05 164.11 189.13 14.42 F 69.71 19.89 9.11 36.02 10.41 11.38 13.12 P 0.000 0.000 0.000 0.000 0.000 0.000 0.000 (b) with variables of species, flake sizes and pressure drops. Source species size drops species*size species*drops size*drops species*size*drops Error Total DF 1 1 2 1 2 2 2 24 35 SS 1406.25 72.25 89.06 0.25 197.17 26.17 57.17 470.00 2318.31 MS 1406.25 72.25 44.53 0.25 98.58 13.08 28.58 19.58 F 71.81 3.69 2.27 0.01 5.03 0.67 1.46 P 0.000 0.067 0.125 0.911 0.015 0.522 0.252 123 Table II-6 Analysis of Variance for TSj,.. (a) with variables of MC, PCT and pressure drops. Source MC PCT drops MC*PCT MC*drops PCT*drops MC*PCT*drops Error Total DF 1 2 3 2 3 6 6 48 71 SS 975.35 610.86 669.93 1478.36 774.82 1274.69 1624.31 808.00 8216.32 MS 975.35 305.43 223.31 739.18 258.27 212.45 270.72 16.83 F 57.94 18.14 13.27 43.91 15.34 12.62 16.08 P 0.000 0.000 0.000 0.000 0.000 0.000 0.000 (b) with variables of species, flake sizes and pressure drops. Source species size drops species*size species*drops size*drops species*size*drops Error Total DF 1 1 2 1 2 2 2 24 35 SS 1626.78 245.44 139.06 16.00 272.72 22.72 105.50 502.67 2930.89 MS 1626.78 245.44 69.53 16.00 136.36 11.36 52.75 20.94 F 77.67 11.72 3.32 0.76 6.51 0.54 2.52 P 0.000 0.002 0.053 0.391 0.006 0.588 0.102 124 Table II-7. Analysis of Variance for MOR with variables of pressure breathing and MC. Source Breath MC Breath*MC Error Total DF 2 1 2 30 35 SS 48.74 38.61 532.91 94.48 714.74 MS 24.37 38.61 266.45 3.15 F 7.74 12.26 84.60 P 0.002 0.001 0.000 Table II-8. Analysis of Variance for MOE with valuables of pressure breathing and MC. Source Breath MC Breath*MC Error Total DF 2 1 2 30 35 SS 2953060 849771 2084805 1767728 7655364 MS 1476530 849771 1042402 58924 F 25.06 14.42 17.69 P 0.000 0.001 0.000 Table II-9. Analysis of Variance for IB with variables of pressure breathing and MC. Source Breath MC Breath*MC Error Total DF 2 1 2 66 71 SS 383676 278134 5101880 899838 6663528 MS 191838 278134 2550940 13634 F 14.07 20.40 187.10 P 0.000 0.000 0.000 125 Table 11-10. Analysis of Variance for TW22% with variables of pressure breathing and MC. Source Breath MC Breath*MC Error Total DF 2 1 2 12 17 SS 307.000 0.889 8.111 10.000 326.000 MS 153.500 0.889 4.056 0.833 F 184.20 1.07 4.87 P 0.000 0.322 0.028 Table 11-11. Analysis of Variance for TW with variables of pressure breathing and MC. Source Breath MC Breath*MC Error Total DF 2 1 2 12 17 SS 67.00 37.56 300.78 140.67 546.00 MS 33.50 37.56 150.39 11.72 F 2.86 3.20 12.83 P 0.097 0.099 0.001 Table 11-12. Analysis of Variance for TW^ with variables of pressure breathing and MC. Source Breath MC Breath*MC Error Total DF 2 1 2 12 17 SS 98.778 64.222 312.111 118.667 593.778 MS 49.389 64.222 156.056 9.889 F 4.99 6.49 15.78 P 0.026 0.026 0.000 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0075242/manifest

Comment

Related Items