The Open Collections website will be unavailable July 27 from 2100-2200 PST ahead of planned usability and performance enhancements on July 28. More information here.

Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The effect of solar radiation on the surface checking of lodgepole pine Urban, Kathrin 2005

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-ubc_2005-0676.pdf [ 11.65MB ]
Metadata
JSON: 831-1.0075075.json
JSON-LD: 831-1.0075075-ld.json
RDF/XML (Pretty): 831-1.0075075-rdf.xml
RDF/JSON: 831-1.0075075-rdf.json
Turtle: 831-1.0075075-turtle.txt
N-Triples: 831-1.0075075-rdf-ntriples.txt
Original Record: 831-1.0075075-source.json
Full Text
831-1.0075075-fulltext.txt
Citation
831-1.0075075.ris

Full Text

THE EFFECT OF SOLAR RADIATION ON THE SURFACE CHECKING OF LODGEPOLE PINE • by KATHRIN URBAN Dipl. Ing. (FH)., Fachhochschule Rosenheim, 2002 THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (FORESTRY) THE UNIVERSITY OF BRITISH COLUMBIA June 2005 © Kathrin Urban, 2005 Chapter One: General Introduction i i ABSTRACT The weathering of wood is caused by a complex combination of chemical and mechanical effects. One of the first signs of weathering is the development of checks at wood surfaces. This study examined the effect of UV-B, UV-A, visible light and infrared light on the development of checks at the surface of lodgepole pine {Pinus contorta Dougl. ex Loud.) deck boards for three periods of twelve weeks, each starting in June 2004. Colour change and change in lignin content were analysed during exposure. A separate experiment examined the loss of weight and tensile strength of microveneers clamped together during exposure. SEM images provided information about structural changes. UV-B had the most significant effect on the quantity, total width, total length and total area of checks that developed at the surface of boards during exterior exposure. UV-B degraded lignin and there was a darkening and reddening of the exposed surface. Defibration of the cell structure at the microscopic level was also observed. UV-A and visible light degraded the surface to a lesser extent. Photodegradation was a surface phenomenon with UV-B only affecting the two upper layers of the veneer blocks (150 um). It can be concluded that selectively protecting wood from UV-B, UV-A and visible light using chemical means might result in less severe checking of lodgepole pine during weathering. Even the blockage of UV-B alone might lead to significantly improved behaviour of surfaces, when the natural look of decks is preferred. Chapter One: General Introduction i i i TABLE OF CONTENTS Abstract .• , ii Table of Contents iii List of Tables vi List of Figures vii Acknowledgements x CHAPTER ONE : 1 1 General Introduction 1 1.1 Introductory Remarks 1 1.2 Hypothesis : 3 1.3 Scope and Importance 7 1.4 Study Outline 8 Chapter Two 9 2 Literature Review 9 2.1 Introduction 9 2.2 Theory 10 2.2.1 Mechanism 10 2.2.2 Quantification of Checking 12 2.3 Factors Affecting the Checking ofWood 16 2.3.1 Water , 16 2.3.2 Light 20 2.3.3 Weathering 25 2.3.4 Wood Species 28 2.4 Structure and Chemical Composition 30 2.4.1 Macroscopic... 30 2.4.2 Microscopic 31 2.4.3 Molecular 32 2.5 Chemical Treatments 35 2.5.1 Preservatives 35 Chapter One: General Introduction iv 2.5.2 Water Repellents 38 2.5.3 Thermal Treatments 41 2.5.4 Mechanical Treatments 42 2.5.5 Kerfing 43 2.5.6 Incising 46 2.5.7 Center Boring 47 2.5.8 Constructive Protection 48 2.6 Summery 49 CHAPTER THREE 51 3 Effect of Solar UV-A, UV-B, Visible Light and IR Light on the Checking and Colour of Flat-Sawn Lodgepole Pine Surfaces 51 3.1 Introduction 51 3.2 Materials and Methods 52 3.2.1 Lumber 52 3.2.2 Decking Samples 55 3.2.3 Exposure 57 3.2.4 Filters 57 3.2.5 Racks ; 59 3.2.6 Measurement of Temperature, Precipitation, Visible Light and UV Light 61 3.2.7 Data Collection 63 3.2.8 Statistical Analysis 64 3.3 Results 65 3.3.1 Checking 65 3.3.2 Weathering Data 77 3.4 Discussion 81 3.5 Conclusion 85 CHAPTER FOUR '. 87 4 Effect of UV-A, UV-B, Visible Light and IR Light on the Integrity of Microveneers formed into Veneer Blocks 87 4.1 Introduction 87 4.2 Materials and Methods 89 Chapter One: General Introduction v 4.2.1 Lumber 89 4.2.2 Veneer Samples 89 4.2.3 Exposure 91 4.2.4 Racks and Filters 92 4.2.5 Measurement of UV Light and Visible Light 94 4.2.6 Measurement of Weight and Tensile Strength Losses of Veneers 95 4.2.7 Experimental Design and Statistical Analysis 95 4.3 Results 96 4.3.1 Effect of Filter Type and Veneer Position 96 4.3.2 Effect of Exposure Time 99 4.4 Discussion 101 4.5 Conclusion 105 CHAPTER FIVE 106 5 General Conclusions and Suggestions for Further Research 106 5.1 General Conclusion 106 5.2 Suggestions for Further Research 108 References ...110 Appendix 3.1 124 Fourier Transform Infra-Red Spectroscopy 124 Materials and Methods 124 Results 124 Appendix 3.2 126 Scanning Electron Microscopy 126 Materials and Methods 126 Results 126 Chapter One: General Introduction v i LIST OF TABLES Table 2.1: Observations of the effects of accelerated weathering on the micro-checking of a range of softwoods and hardwoods (Coupe and Watson 1967) 29 Table 3.1: Length, widths and number of growth rings of the logs 52 Table 3.2: Drying schedule with Ta: temperature (true value), Tn: temperature (command value), EMC-Tr: equilibrium moisture content (true value), EMC-Cr: equilibrium moisture content (command value), DGa: drying gradient (true value), DGn: drying gradient (command value), VNTa: ventilator speed (true value), VNTr (command value) and DCV and FMa,: both drying control values 54 Chapter One: General Introduction v i i LIST OF FIGURES Figure 1.1: Cyclical sequence of events during natural weathering (Groves and Banana 1986) 2 Figure 1.2: Mechanism of check formation (Evans 2004) 5 Figure 2.1: The approximate relative distribution of solar energy (mean noon sea level sunlight) from 200 to 2200 nm (Anonymous 2001) 2 Figure 2.2: Scheme for photodegradation of wood (Kalnins 1966) 3 Figure 2.3: Kerfing and center-boring treatments (Evans et al. 2000) 4 Figure 3.1: The Wood-Mizer LT 15 sawmill used to cut logs into decking boards 5 Figure 3.2: Cutting pattern for the lodgepole pine logs 5 Figure 3.3: Drying diagram showing Ta: temperature (true value), Tn: temperature (command value), EMC-Tr: equilibrium moisture content (true value), EMC-Cr: equilibrium moisture content (command value) and DCV and FMa, both drying control values 5 Figure 3.4: Cutting pattern used to produce decking samples from logs 5 Figure 3.5: Lodgepole pine decking sample 5 Figure 3.6: UV light transmission of colourless Acrylite OP-4, Acrylite GP, AcryliteOP-2 (Anonymous 2001) 5 Figure 3.7: Transmission of Acrylite GP sheet, Colour # 1146-0 and 199-0 (Anonymous 2001) 5 Figure 3.8: Top view and side view of a decking sample rack 6 Figure 3.9: Front view and detail of the decking sample rack 6 Figure 3.10: Rack for decking samples 6 Figure 3.11: Dosimeter of Type III (silver) 6 Figure 3.12: The effect of filter type on check numbers. 6 Figure 3.13: The effect of filter type on total check length. 6 Figure 3.14: The effect of filter type on total check width. 6 Figure 3.15: The effect of filter type on total check area 6 Chapter One: General Introduction V l l l Figure 3.16: The effect of filter type on check shape Figure 3.17: The effect of exposure time on check number Figure 3.18: The effect of exposure time on total check length Figure 3.19: The effect of exposure time on total checks width Figure 3.20: The effect of exposure time on total check area Figure 3.21: The effect of exposure time on check shape Figure 3.22: The effect of the interaction of filter type and exposure time on lightness (L*) Figure 3.23: The effect of the interaction of filter type and exposure time on the green-red colour parameter (a*) Figure 3.24: The effect of the interaction of filter type and exposure time on the yellow-blue colour parameter (b*) Figure 3.25: Average temperature during the exposure trial Figure 3.26: Total precipitation during the exposure trial Figure 3.27: Visible light measured during the exposure trial Figure 3.28: UV light measured during the exposure trial Figure 4.1: Digital micrometer (Lorentz & Wettre HWS 5781) Figure 4.2: Method of constructing blocks from veneers cut from decking samples Figure 4.3: Veneer blocks Figure 4.4: Top view and side view a rack with filter that was used to expose veneer blocks to the weather Figure 4.5: Front view and detail of a rack with filter used to expose veneer blocks to the weather Figure 4.6: Rack with filter for veneer blocks Figure 4.7: Top, side and front view of rack (without filter) used to expose veneer blocks Figure 4.8: Rack without filter for veneer blocks Figure 4.9: Pulmac Troubleshooter HWS 5786 Chapter One: General Introduction Figure 4.10: The interaction of filter type and veneer position on weight losses of veneers 97 Figure 4.11: The interaction of filter type and veneer position on tensile strength losses of veneers 97 Figure 4.12: The interaction of filter type and exposure time on weight losses of veneers 99 Figure 4.13: The interaction between veneer position and exposure time on tensile strength losses of veneers 100 Figure 0.1: Infra-red spectra of lodgepole pine boards. Bands from unexposed control (0), fully exposed control (F) and specimen exposed under filters 1, 2, 3, 4 and 5 125 Figure 0.2: SEM image of the unweathered (left) surface (a) and the surface exposed under filter 1 (right) (b) at a magnification of x40 128 Figure 0.3: SEM image of the surface exposed under filter 2 (left) and filter 3 (right) at a magnification of x40 128 Figure 0.4: SEM image of the surface exposed under filter 4 (left) and filter 5 (right) at a magnification of x40 128 Figure 0.5: SEM image of the unweathered (left) surface and the surface exposed under filter 1 (right) at a magnification of x 100 129 Figure 0.6: SEM image of the surface exposed under filter 2 (left) and filter 3 (right) at a magnification of x 100 129 Figure 0.7: SEM image of the surface exposed under filter 4 (left) and filter 5 (right) at a magnification of x 100 129 Figure 0.8.: SEM image of the unweathered (left) surface and the surface exposed under filter 1 (right) at a magnification of x250 130 Figure 0.9: SEM image of the surface exposed under filter 2 (left) and filter 3 (right) at a magnification of x250 130 Figure 0.10: SEM image of the surface exposed under filter 4 (left) and filter 5 (right) at a magnification of x250 130 Chapter One: General Introduction ACKNOWLEDGEMENTS The research reported in this thesis was made possible through the funding of Chemical Specialties, Inc. (CSI) and the DAAD (Deutscher Akademischer Austauschdienst) Scholarship from September 2003 till May 2005. I am deeply grateful to my supervisor, Dr. Philip D. Evans, Director of the Center for Advanced Wood Processing (CAWP) of the University of British Columbia (UBC). His knowledge about research helped me throughout the project and the preparation of the thesis. I am also deeply to Dr. Colette Breuil (UBC), who helped me in her function as Graduate Student Adviser and formed together with Dr. Gregory Smith my very supportive and helpful committee. Thanks are also due to the following individuals whose assistance contributed to the completion of this research: Dr. Alan Preston from CSI, for his friendly help, his reference for the IRG meeting and his guidance for the future. Dr. Hans Holtschmidt from BioSense, Germany for his expert knowledge of UV dosimeters and their analysis and interpretation. Dr. Kai Morgenstern from the Faculty of Agricultural Sciences (UBC) for assisting me with the weather data. Dr. Bingye Hao (UBC) for his assistance in kiln drying experimental samples. Thanks are also due to other academic and administrative staff in the Center for Advanced Wood Processing (CAWP) and the Wood Science Department, both UBC, for helping with administrative requirements and for their assistance during my study. I especially wish to thank Mr Tom Wray, Mr Robert Myronuk, Mrs Diana Hastings and Mrs Debbie Wong. I would also like to thank Markus Steiniger for his valuable help and assistance before and after I arrived in Canada, the members of my research group Kim-J ana Henze, Chen Huang, Derek Thompson, Jahangir Chowdhury, Steve Ribarits and Johannes Weizenegger. I thank Julia Dordel and Brian Matthews for helping me to look after my exposed samples as well as David Hopkins for his support in taking measurements and encouragement during the writing phase. I would like to extend my deep and sincere gratitude to my parents Jorg and Roswitha Urban, my brother Lutz and my grandparents Alfred and Erna Holzinger for their invaluable understanding and never ending encouragement throughout my study at UBC. Chapter One: General Introduction 1 CHAPTER ONE 1 General Introduction 1.1 Introductory Remarks • Wood is increasingly used outdoors because it has warmth, visual appeal and natural harmony with the surrounding environment. It is also environmentally friendly, easy to recycle or dispose of, naturally renewable and easily processed with low energy consumption (Imamura 2001). Furthermore, it has low thermal expansion, high strength/weight ratio and is inexpensive (Hon and Chang 1984; Kiguchi and Evans 1998; Evans 2001). Wood use exceeds that of all other composites on a volume-basis, and it is possibly the world's most versatile and widely used engineering and structural material (Hon 1981; Hon and Chang 1984). The applications of wood for siding, decking, railing, etc. significantly increased in the 60's and 70's (Hon 1983) and it is of great economic significance in both developing and industrialised countries (Hon 1981; Hon and Chang 1984). However, wood-polymer composites, which compete with wood, will have doubled their market share in the last five years and they have the potential to further displace wood from many traditional end uses. This development is a threat to dozens of building-products, and forest-products companies in the US (Principia Partners 2002). One of the greatest drawbacks of wood used outdoors is its susceptibility to weathering (Sell and Leukens 1969). Freshly prepared wood surfaces undergo biological attacks, and various physico-chemical changes, caused by the combined influence of solar Chapter One: General Introduction 2 radiation, oxygen, atmospheric humidity (dew, rain, humidity, snow) and heat (Chang et al. 1982; Hon 1983; Kleinert 1970). The degradation caused by these factors is collectively termed "weathering", which describes the changes in appearance (gloss and colour), and surface roughness (surface checking) of wood exposed outdoors (Chang et al. 1982; Derbyshire and Miller 1981; Feist 1982 and 1990; Sell 1968). Groves and Banana (1986) summarized the cyclical sequence and interactions of natural weathering in the diagram shown below (Figure 1.1) Exposure of fresh tissue to the weatherine nrocess W o o d exposed out o f doors O W S Erosion, usually by rain, of weathered fragments and whole fibres x a e r u n h g Hygroscopic movement and direct wetting giving rise to warping, checking and splitting Photo-degradation including fragmentation o f l ignin molecule and depolymerisation o f cellulose , Micro-structural changes including removal o f middle lamella, Physical changes such as surface roughening and preferential removal o f lower density tissue destruction o f bordered pits and loss o f adhesion between cell wal l layers Figure 1.1: Cyc l ica l sequence o f events during natural weathering (Groves and Banana 1986) Chapter One: General Introduction 3 Weathering is a surface phenomenon and does not generally affect the structural performance of wood. However, it is generally not aesthetically desirable for consumers, who, as mentioned above, are increasingly inclined to substitute wood used outdoors with alternative materials (Derbyshire and Miller 1981). A survey in the Southern United States showed that aesthetic reasons are an important factor influencing the service life of decking material. CCA-treated decks are replaced after 7 years on average solely because they loose their aesthetic appeal (McQueen and Stevens 1998). The most obvious and displeasing aesthetic changes, caused by weathering, are photo-discolouration (greying) and the development of checks and cracks, which may or may not make a deck structurally unsound. The development of methods to reduce the checking, cracking and splitting of wood used outdoors would help to prolong its appearance and lead to improved customer satisfaction for softwood decking material. 1.2 Hypothesis The causes of checking are manifold. Of all weathering effects mentioned above, stresses imposed by alternating wetting and drying and degradation caused by ultraviolet radiation (from the sun), have the strongest impact on the surface of wood (Chang et al. 1982; Kalnins 1966). Feist (1982) mentioned the formation of "macroscopic to microscopic inter- and intracellular cracks and checks" due to weathering after the initial darkening of the wood. However, the exact process leading to the development of checks is not very well understood due to the complexity of the weathering process (Groves and Banana 1986). Most researchers attribute surface checking of wood to the development of surface stresses and strains generated by shrinking and swelling due to moisture loss and gain Chapter One: General Introduction 4_ within the hygroscopic range that exceeds the elastic limit of wood. This conclusion can be drawn when observing softwood decks exposed to the weather. Since decks are usually exposed horizontally to the weather, only their top surface is directly affected by water and solar radiation, while the rest of the board is protected to varying degrees and restrained by fasteners (Evans 2004). Experiments indicate that the degradation of wood surfaces is highest for horizontal exposure (Evans 1989a, 1996). The upper side is preferentially wetted by rain and is heated by solar radiation to a greater degree (e.g. up to temperatures of 80°C) (Figure 1.2, 1 and 2) (Wengert 1966, Sell and Walchli 1969, Evans 2004). The water eventually penetrates into the core of the board (Figure 1.2, 1). When the top surface dries under the influence of solar radiation, water is first removed from the surface of the board (Figure 1.2, 3). The moisture content drops below the fibre saturation point and the surface starts to shrink. The core of the board, however, stays in a green condition for a longer time (Evans 2004). The board is also restrained by the fasteners, which keep the board from cupping and warping (Figure 1.2, 4). The shell of the board is put under tension, while the core is under compression. If the tension stress at the surface exceeds the tensile strength of the surface cells perpendicular to the grain, the wood checks (Figure 1.2, 5) (Stamm 1964, Sell and Walchli 1969, Evans 2004) Chapter One: General Introduction 5 5 4 Figure 1.2: Mechanism of check formation (Evans 2004) Photochemical degradation of wood plays a significant role in the weathering of unfinished wood. Feist and Hon (1984) described the photon energy in solar radiation as the most damaging component of the outdoor environment, which 'initiates a wide variety of chemical changes at wood surfaces'. Feist (1990) stressed that sunlight and water tend to operate at different times and may accelerate each other's effect on the wood surface. Derbyshire and Miller (1981) found that both UV light and visible light contribute significantly to the degradation of wood. An additional indication of the great effect of solar radiation on the surface of exposed wood, is the necessity to use pigmented finishes to protect the surface from weathering, even if the influence of water is largely excluded (Sell and Leukens 1969). Checking is generally assumed to occur independently of surface photodegradation. For example, Groves and Banana (1986) do not show a link between checking and microstructural changes caused by photodegradation (Figure 1.1). Some researchers, however, have suggested that surface photodegradation of wood may increase the severity Chapter One: General Introduction 6 of checking. Miniutti (1967) stated that the loss of wood substance by photochemical degradation contributes to 'the development of stresses and ultimately checking'. Lignin is degraded more rapidly by UV light than cellulose. Thus, regions of the wood cell wall that have a high lignin content are preferably degraded during irradiation, weakening the surface structure of the wood surface (Miniutti 1973). The reduction of cohesive and adhesive strength due to loss of lignin in the middle lamella causes the failure of various cell wall layers starting at a very early stage during irradiation. Apertures in fibrous wood elements are gradually enlarged leading to weakening of the whole fibre structure (Borgin 1970). These enlarged apertures are considered as microchecks as they cannot be seen with the naked eye. If they occur in individual tracheid walls, their long axis is diagonal to the long axis of the tracheid following the fibril angle of the S2 layer. Microchecks between adjacent walls of neighbouring tracheids develop in or close to the middle lamella and run parallel to the long axis of the cells (Miniutti 1964, 1967). Evans (1989b) found that macroscopic checks in tangential longitudinal surfaces (TLS) were caused by cell wall failure and crack propagation at a microscopic level. He observed that the thin walled parenchyma cells in ray tissue of radiata pine (Pinus radiata D. Don) were degraded first resulting in voids, which progressively enlarged with exposure. Macroscopic checks mainly developed where adjacent cells or tissues differ in cell wall thickness or strength, for example in transverse surfaces at growth ring boundaries and in TLS at the interfaces between rays and tracheids. This observation of a relationship between photodegradation at the microscopic level and the formation of visible macroscopic checks is supported by Yata (2001), who attributed small checks in Japanese Chapter One: General Introduction 7 cypress surfaces (Chamaecyparis obtusa Endl.) to the exposure of wood to rainfall and light irradiation at wavelengths below 500 nm. The importance of solar radiation in the development of checks has not yet been fully determined. Information about the effect of the different components of solar radiation on the degradation of wood, with the exception of the study by Derbyshire and Miller (1981) is limited. There have been no studies that have directly examined the relationship between photo-chemical degradation of wood surfaces and checking. Since there is a link between photochemical degradation of wood and the occurrence of macro and microscopic checks (Evans 1989b) it is a reasonable hypothesis that checking is influenced by exposure of wood surfaces to solar radiation. 1.3 Scope and Importance The scope of this thesis is to examine whether solar radiation has an effect on the development of checks at the surface of lodgepole pine and what wavelengths play a significant role in the degradation process, leading to the development of surface checks. The effect of the different wavelengths on the degradation of thin wood veneers was examined in order to better understand the relationship between surface degradation and the development of checks. Knowledge arising from the study could allow for the development of better photoprotection systems that target the wavelengths that increase the surface checking of wood. Chapter One: General Introduction 8 1.4 Study Outline This chapter (Chapter 1) provides the rationale for the study. Chapter 2 reviews the relevant literature on the factors affecting the checking of wood such as water, light and micro organisms and the role of wood structure and chemical composition on checking. Particular attention is given to the effect of solar radiation on the surface of wood. The physical and chemical treatments that have been used to reduce the checking of wood are also reviewed. In the literature great emphasis was given to the effect of the weather in general and of water in particular on the macrochecking of wood (Schniewind 1963; Stamm 1963, 1964 and 1965; Coupe and Watson 1967; Sell and Leukens 1969; Borgin 1971; Rowell et al. 1981; Feist and Hon 1984; Sandberg 1999; Flaete et al. 2000; Evans 2001 and others). Little if any literature is available that directly addresses the topic of this thesis i.e. the role of solar radiation in the formation of checks at wood surfaces. Chapter 3 is the main experimental chapter that examines the effect of solar UVA, UVB, visible light and IR light on the checking of flat-sawn lodgepole pine surfaces. Derbyshire et al. (1981, 1995, 1996 and 1997) examined the effect of solar radiation on the strength and chemical composition of single microveneers and the effect of solar radiation on single microveneers exposed outdoors under filters for one month. Chapter 4 uses a similar technique to examine the effect of UVA, UVB, visible light and IR light on the integrity of packed microveneers in order to better understand the wavelengths dependent degradation of wood. Finally, Chapter 5 presents a general discussion of the results, conclusions and suggestions for further research. Chapter T w o : Literature Review 9 CHAPTER TWO 2 Literature Review 2.1 Introduction Wood is a natural product that harmonises well with nature. It has a natural feel and has been used as a building material for ages. The continuing popularity of wooden houses, timber decks and exterior furniture are due, in part, to the combination of properties that makes wood suitable for building construction, including its high strength to weight ratio, ease of processing, gluing and painting and its relatively low cost (Coupe and Watson 1967). Wood in favourable environments can be an extremely durable material, but it is prone to weathering when exposed unprotected outdoors (Browne 1960; Feist 1982; Feist and Hon 1984). During exterior exposure checking often occurs at wood surfaces and such checks are one of the first visible signs of the physical deterioration of wood (Feist and Mraz 1978). Flaete et al. (2000) define checks as cracks, which do not penetrate wood deeper than 75 % of the board thickness. Checks are a serious defect that detracts from the appearance of wood and may reduce its durability by promoting fungal colonisation and decay. This occurs because water that is necessary for the growth of fungi and bacteria enters timber through end grain or via deep checks (Sell and Walchli 1969; Borgin 1971; Graham 1979; Morrell and Newbill 1986; Flaete et al. 2000). Preservative treatment of wood produces an Chapter Two: Literature Review 10 external layer of protection. Deep checks can extend beyond the treated shell and allow fungi and insects access to untreated wood (Morrell 1990). Boulton as early as 1884 recognized, "[...] if the crack is deeper than that portion of the wood charged with antiseptic, [the germs] will carry destruction into the center of the log" (Boulton 1884). Checking may also reduce strength (Mackay 1973) and for some applications boards with checks are not useable (Sandberg 1997). Checks can be classified as macrochecks, which are visible for the naked eye, and microchecks, which can only be observed under the microscope. 2.2 Theory 2.2.1 M e c h a n i s m Sandberg (1999) stated that the greater occurrence of checking on tangential surfaces is mainly the result of mechanical degradation induced by stresses, which build up due to "anisotropic moisture movements of the wood material and moisture gradients between the surface" and the core of the softwood member. When the surface has been exposed to moisture and subsequently dries below the fibre saturation point the superficial layer shrinks, but is restrained by underlying wood of higher moisture content (McMillen 1955). While the outer fibres are stressed in tension the inner fibres are stressed in compression. Those stresses are greater for tangential than for radial surfaces because shrinkage is greater in the former direction than in the latter (McMillen 1955; Schniewind 1959; Borgin 1971; Sandberg 1999). Furthermore, rays restrain cells from shrinking freely in the radial direction (Mcintosh 1955). Shrinking and swelling also depends on the orientation of the Chapter Two: Literature Review 11 structural units, the nature and extent of the gross capillary structure, and the stresses, which are generated by the moisture content gradients or forces from the outside (Stamm and Loughborough 1942). If the shrinkage stresses exceed the tensile stress of the exposed surface layer perpendicular to the grain, fibre separation and checking occurs (Stamm 1964). Checking indicates that the cohesive strength of the surface layers is not high enough to resist distortion by swelling and shrinkage of the fibres over a certain length of time (Sell and Leukens 1969, Borgin 1971). The surface begins to roughen and the grain raises, with the springwood rising less than the summerwood. This generates additional stresses and more checks develop on the surface until it becomes very uneven (Browne 1960, Stamm 1963 and 1965, Sell and Walchli 1969). All stresses that are developed within various tissues adjacent to each other are classified as second order stresses (Schniewind and Kersavage 1961). Water also plasticizes the cell walls and debulks them by leaching out soluble degradation products. Thus, the tracheids can be twisted more easily by intermolecular contracting forces and surface tension forces (Miniutti 1973). The middle lamella is a weak layer within the cell structure. It has been suggested that it can be torn apart by thermal or hygroscopic movements even in layers below the exposed surface, where it is protected from solar radiation or in surfaces solely exposed to wetting and drying (Borgin 1971). Another area of weakness is the interface between the primary and the secondary wall layer of the cell wall because at this point the microfibrils change their direction abruptly and the two adjacent layers swell and shrink in almost opposite directions causing large differential movements (Borgin 1971). Thus, the cohesion of the wood substance is reduced and Chapter Two: Literature Review 12 eventually the strength properties decrease or the structural components of wood disintegrate (Raczkowski 1980). Stresses within a single cell due to liquid tension forces or unequal shrinkage of the different cell wall layers are referred to as first order stresses (Schniewind and Kersavage 1961). 2.2.2 Quantification of Checking Manual methods for quantifying checks can be divided into two types of techniques. Either the number of checks per unit length is counted and their size measured using a simple device such as a ruler or checking is estimated visually and subjectively. The easiest way to quantify checks is to assign them visually into different categories or measure them using a simple device such as a ruler. The size of measured checks varies depending on the type of wood product and its end use. The methods used and frequency of assessing checks strongly depend on the purpose of the examinations. For large posts used in-ground, it may be sufficient to only measure checks that are equal to or greater than 1 mm in width (Timber Technology Services 1996), while smaller specimens usually require measurement of larger numbers of checks. Mostly the dimensions of checks are given in mm or 1/10 mm, depending on the aim of the experiment and the preferences of the researcher. In many cases not all of the dimensions of checks are recorded. For example Sandberg (1996) for instance summed up the length of all checks that were recorded for one board, and then calculated the relative length of cracks (defined as the ratio between the total length of the cracks and the length of the board). Chapter Two: Literature Review 13 Crawford et al. (1999) determined the percentage of grain (surface texture) affected by checking. Kabir et al. (1992) assigned checks into three groups according to their widths and calculated the average length of the checks within each group. Rietz (1961) used a Severity Rating Factor and a Numerical Rating Factor to quantify surface checking in red oak during drying. For the former, indices of 0-5 were assigned to the checks depending on the severity of total length of checks. The Numerical Rating Factor used indices of 0-4 depending on the average width and the number of checks. He also defined a surface check rating as the product of the Severity Rating Factor and the Numerical Rating Factor. Assuming that checks can be modelled as two right triangles or two right prisms attached back-to-back, check surface areas and volumes can be calculated, respectively, using the following formulae: Check surface area = (Maximum width x length)/2 Check volume = (Maximum width x maximum depth x length)/6 (Evans et al. 1997) Measurement of checks by hand is tedious and in certain cases is not accurate enough. One step beyond simple visual assessment of checking is to use electronic equipment to obtain dimensions of checks. Coupe and Watson (1967) used a cine camera mounted in a weatherometer to image surface checks. An electronic switch triggered the camera, but the analysis of the image was still done manually. For an optical system to be able to analyse the picture it has to be capable of translating the optical image into Chapter Two: Literature Review 14 electronic signals that can be processed by a computer or displayed on a cathode ray tube screen for an operator to interpret (Szymany and McDonald 1981). Christy et al. (2005) developed a program to automate check measurement at wood surfaces. The surfaces of Southern pine and Douglas fir decking boards were scanned using Adobe Photoshop at 600 d.p.i. resolution and then binarized to obtain a b/w picture. The image was then searched for checks by establishing which black pixels were connected to others. The minimum width of a check that could be reliably captured was about 4 pixels (0.17 mm). The advantage of their automated system was that it could rapidly check sizes. Furthermore, additional information was recorded, for example, vertical coordinates of checks, length, area, width (calculated as (area/length) x shape function) and "shape" (=length/width ratio) of check. The volume of checks, however, could not be calculated. Since light can be absorbed, transmitted, and/or reflected when it strikes an object, optical scanners are available to detect the amount of variation in reflected light. A typical optical scanner consists of a light source, a light sensor (e.g. photo-transistor), an amplifier, and an output device. Optical scanners use either incandescent or fluorescent lamps, light emitting diodes (LED) , or laser beams as the light source (Szymany and MacDonald 1981). Two optical techniques can be used to detect defects including checks. First, a laser beam can be used as a light source. A scanning mirror approximately 1 mm wide directs the beam back and forth across the object under inspection. Thus, line scans are produced at right angles to the motion of the object. To ensure that any change is a trend and not a single event, scans must be recorded at close enough intervals. In the second technique the entire Chapter Two: Literature Review 15 board is illuminated by fluorescent light as it moves past the scanning station. Both board geometry and defects including checks are detected (Szymany and MacDonald 1981). At higher magnification it is possible to view the microchecking that occurs during the early stages of checking. Hence a more detailed picture of checking can be obtained by using a microscope. Miniutti (1964) cut wafers from weathered blocks and examined them using a microscope with a built-in incident illuminator. He used dry objectives from x 3.8 to x 50 and higher to examine microchecking in both pits and cell walls. Miniutt i 's samples, however, could not be reused for further experimentation. Thus, his experiments only showed the extent of micro-checking after certain defined periods of time, but he was not able to show the dynamic process of check formation. Coupe and Watson (1967) took replicas of weathered surfaces using cellulose nitrate film softened with acetone and observed them using a normal transmitted-light microscope. The original specimen could be weathered again to further examine the progressive deterioration of the surface. Mackay (1973) mounted small blocks on a microscope stage and dried them by directing a stream of heated air over the tangential drying face. Thus, surface checks could be observed as they developed in the woody tissues. He also examined microscopic check development in untreated and preservative treated blocks cut from matching material during drying at two temperatures. He only considered the influence of temperature and humidity of air on checks, but not additional aspects of weathering such as exposure of wood to UV-light . Chapter Two: Literature Review 16 Bariska et al. (1988) as well as Chang et al. (1982) closely examined the checking of CCA-treated timber using a scanning electron microscope (SEM). It was possible for them to observe separations in the wood substance between ray and fibre tissue, between single cells in the compound middle lamella region, and between layers of the cell wall. 2.3 Factors Affecting the Checking of Wood 2.3.1 Water Wood is a very hygroscopic material. Thus, water has a great influence on the weathering behaviour of wood (Hosli and Mannion 1991). Water can occur as gas (vapour and atmospheric humidity), fluid (rain, dew) and solid (ice, snow). Depending on its state the impact of water on wood is different. (Ktihne et al. 1968). Liquid water causes leaching of low molecular weight compounds from wood, hydrolysis, swelling and shrinking, distortion, and ultimately cracking and splitting of wood (Coupe and Watson 1967, Feist and Hon 1984). Some researchers maintain that water alone has very little influence on the chemistry of the wood surface, however, its deleterious effect on the physical characteristics of the surface have been clearly demonstrated using scanning electron microscopy (SEM) (Owen et al. 1993). Water has a great effect on the dimensional stability of wood when the moisture content varies within the hygroscopic range (Borgin 1971). Then the whole structure of the wood develops large stresses. A constant moisture level and minimal changes in temperature on the other hand do not promote distortion of the wood structure (Sell and Leukens 1969, Borgin 1971). The degree of change in moisture content also depends on the wood species and grain direction. Pine (Pinus sp.) shows very large Chapter Two: Literature Review 17 large seasonal changes in moisture content. Radial boards generally have a higher moisture content than flatsawn boards (Bottcher 1975). Weathering is generally viewed as a long-term phenomenon, but the effects of water also can be observed on a daily basis. Since temperature decreases at night, water vapour in the wood cells condenses and become a fluid. Hence, moisture and stress gradients are generated from the surface to the core of the wood and cracks may be formed as deep as 10 mm below the surface of untreated timber. In addition the increasing moisture content at the surface, weakens the wood, which leads to additional checking (Bariska et al. 1988). Coupe and Watson (1967) observed microchecking at tangential surfaces in the uniseriate rays of softwoods. Schniewind (1963) explains this phenomenon as being due to the rays having higher tensile strength in the radial direction than the prosenchyma. Much more stress in the radial direction would be required to cause failure perpendicular to the rays (Schniewind 1963). The strains and stresses that lead to the checking of wood occur as a result of shrinkage anisotropy, and in particular the difference between radial and tangential shrinkage of wood (Mcintosh 1955). The ratio of tangential to radial shrinkage (T/R ratio) for most woods is about two (Schniewind 1989). There are several explanations for such transverse anisotropy, which are described in detail by Pentoney (1953), Bosshard (1956), Kelsey (1956), Stamm (1964), Kollmann and Cote (1968) and others. Skaar (1972) outlined the two most accepted theories. The first is based on the assumption that ray tissue restrains radial shrinkage because it shrinks less radially than does longitudinal tissue. The second Chapter Two: Literature Rev iew 18 theory involves the interaction of earlywood and latewood, and states that latewood shrinks more in the tangential direction than earlywood and thus forces the latter to shrink more tangentially than radially. Schniewind and Kersavage (1961) stated that the magnitude of the stresses generated during the drying of wood depend on the difference in shrinkage potential of the tissues, the elasticity of the tissues, the ratio of the volume of ray tissue and prosenchyma, creep and relaxation. After calculating the theoretical stress value arising from the first three factors, it was clear that creep and relaxation effects have a stress reducing influence on the system. The authors further explained that this system is not static as it changes with different moisture contents. As a result of growth stresses, tensile stresses can be found in the ray tissue and compressive stresses in the prosenchyma when the wood is in the green condition. When the fibre saturation point is reached during drying, the stresses reverse and the rays become compressed. As drying continues the compressive stresses in the ray tissue reach a maximum at a moisture content of about 6 to 14 percent. Below this percentage the compressive stresses in the ray tissue decreases again and may even reverse leading to substantial tensile stresses. Thus there is a maximum compressive stress in the ray tissue, which marks the division between compression and tension in the rays. Hale (1957) agreed that rays have a restraining effect on the shrinkage of hardwoods across the grain in radial direction. He measured a tangential/radial shrinkage ratio for springwood of approximately 2/1, and 1/1 for summerwood. He concluded that "the fine ray tissue is responsible for the large tangential to radial shrinkage ratio of the relatively weak springwood." He additionally assumed that the shrinkage of rays had a greater Chapter Two: Literature Review 19 influence on the shrinkage of thin-walled earlywood than on latewood because he viewed the normal shrinkage of ray-free prosenchyma as a function of wood density. Because of its higher density, latewood shrinks and swells more than earlywood. The reason for this can be found in the submicroscopic structure of the secondary walls of prosenchyma cells. Mackay (1972) contradicted the theories of Schniewind and Kersavage (1961) and Hale (1957). He found that checks in messmate (Eucalyptus obliqua L'Herit) originated in vessels and never in rays. Therefore he concluded that the tensile stresses caused by both the normal and collapse shrinkage of the parenchyma were at a maximum in surface vessels, which are relatively large and weak in comparison to the ray tissue. He explained this phenomenon using the weakest link concept, which was developed by Freudenthal (1950, cited by Mackay 1972). The surface layer can be considered as a thin sheet, which is put under uniform tensile stress at right angles to the grain. Tensile stresses concentrate in those weak elements of the system that represent most of the cross-sectional area. Sell and Leukens (1971) concluded that dimensional changes due to sorption have to be minimized in order to reduce checking. Water is also able of generating minute checks, when it freezes and thaws because its volume is increased when it is solid, which causes stresses within the cell lumen and the cell wall depending on the moisture content (Rowell et al. 1981, Furuno 2001). In cold climates the surface is roughened by windblown particles of ice (Mawson 1915) Water also modifies as well as accelerating chemical changes and deterioration of wood initiated by UV light, such as colour, brightness and pH (Webb and Sullivan 1964, Chapter Two: Literature Review 20 Stamm 1965, Jin et al. 1991, Turkulin and Sell 2002). Arnold et al. (1991) concluded that a water-spray system in a weatherometer was essential for removing UV-light-degraded wood material from exposed surfaces and obtaining surfaces similar to naturally weathered wood. 2.3.2 Light The radiation that reaches the earth's surface is generally referred to as solar radiation. The electromagnetic spectrum can be divided into three components: ultraviolet (UV) (200 - 400 nm), visible (400 - 700 nm) and near infrared (700 - 2200 nm) with the UV band representing approximately 3%, the visible band 45% and the infrared band 52% of the total solar energy (Anonymous 2001). Within the ultraviolet component of the spectrum it is possible to distinguish between UV-A (315 - 400 nm), UV-B (280 - 315 nm) and UV-C (200 - 280 nm). The ozone layer absorbs virtually all radiation between 200 and 310 nm before it can reach the earth's surface (Shanklin 2004). Ultra-violet Visible Infrared UVC UVB UVA VIOLET BLUE GREEN YELLOW ORANGE RED P E N E T R A T I N G H E A T R A Y S N O N -P E N E T R A T I N G H E A T R A Y S 2 0 0 -280 nm 2 8 0 -315 nm 315 -400 nm 400 - 700 nm 7 0 0 - 1400 nm 1 4 0 0 - 2 2 0 0 nm Figure 2.1: The approximate relative distribution o f solar energy (mean noon sea level sunlight) from 200 to 2200 nm (Anonymous 2001) Wood absorbs light very well, but at the same time it is also sensitive to light (Hon and Chang 1984). Its absorbing capacity for energetic i.e. photochemically active radiation Chapter T w o : Literature Review 21 is determined by its structure, surface characteristics, colour and chemical composition (Sell and Leukens 1969). Wood is able to absorb all wavelengths in the electromagnetic spectrum such as fluorescent light, terrestrial sunlight and artificial ultraviolet light, which are able to initiate photochemical reactions. The presence of oxygen, moisture and impurities (sensitizers), heat, and the topostereochemistry (spatial arrangement of atoms in cellulose of the wood surface) of cellulose also play important roles in the photodegradation of wood (Hon 1981). Even a short period of exposure to sunlight leads to pronounced losses of surface integrity (Miller and Derbyshire 1981). The chemically and mechanically weakest elements of the wood tissue at the surface fail first (Borgin 1971). The stability of wood to prolonged exposure to sunlight is one of the most important properties affecting its suitability for outdoor applications (Hon et al. 1985). The effect of light on wood occurs over a wide range between the UV and the IR spectrum (Futo 1976, Miller and Derbyshire 1981). In line with energy considerations, UV light contains the most destructive wavelengths and has the strongest effect on the discolouration and degradation of wood (Browne and Simonson 1957, Derbyshire and Miller 1981, Miller and Derbyshire 1981). UV light only accounts for 5% of the total solar energy received at the earth's surface, but it is responsible for the primary photooxidative degradation of wood (Miller and Derbyshire 1981, Hon 1983, Feist and Hon 1984). Wavelength, intensity and duration of exposure to radiation are the most important factors influencing the photodegradation of wood. Solar radiation affects wood's structural constituents, as well as extractives, to different degrees depending on the ability of the chemical constituents to absorb UV light. These chemical reactions are accelerated by heat Chapter Two: Literature Review 22 (Futo 1974). Various studies have focused on the effect of UV light on wood and cellulosic materials. Light is capable of altering wood colour by degrading extractives. It also degrades lignin, cellulose and hemicellulose via free radical mechanisms, which results in deterioration of the ultrastructure of the cell wall (Derbyshire and Miller 1981; Hon et al. 1985). Even though wood as a lignocellulosic material reacts consistently in response to light it is not yet possible to define exactly what wavelengths in the UV spectrum affect the extractives found in most of the common wood species (Sandermann and Schlumbom 1962a). Sandermann and Schlumbom (1962b) concluded that the reactions might differ significantly for each spectral area and wood species. Visible light is also capable of degrading wood. Since a large amount of energy is present in wavelengths longer than 400 nm, timber exposed only to visible and infrared light degrades at about half the rate observed for wood exposed to the full solar spectrum (Derbyshire and Miller 1981). Visible and infrared light causes browning of the wood (Browne and Simonson 1957). Later the surface gets more and more discoloured and visible light is selectively absorbed. The rate of degradation is increased further as its color intensifies, which leads to increases in the absorption of infra-red radiation (Hon 1981) and raises the rate of photochemical and oxidative reactions (Rowell et al. 1981). Exposure of wood to the visible component of the solar spectrum results in changes to wood's ultra-structure, which are associated with tensile strength losses (Derbyshire and Miller 1981). The effect of UV light on the ultra-structure of wood has received some attention. There is evidence that longwave (UV-A) and shortwave (UV-B) UV irradiation cause the Chapter T w o : Literature Review 23 same anatomical changes at exposed wood surfaces, but the rate of change occurs much faster during exposure to shortwave radiation (Miniutti 1967). There are no studies of the effect of solar radiation on the development of checks at wood surfaces. Thus the role of light in the chain of events leading to checking at wood surfaces is not well understood and can only be indirectly inferred from observations described in the literature. Hon (1981) stated that photooxidation of wood surfaces causes extensive surface modification such as roughening and checking after light exposure. Arndt and Willeitner (1969) attributed the stronger roughening of the surface of samples exposed high up in the mountains to the stronger effects of solar radiation. Two types of mechanical degradation can be distinguished at exposed wood surfaces; the erosion and checking of single cell walls and the disintegration of the cell structure, which becomes more pronounced with increased exposure to light (Sell and Leukens 1971a). Erosion is closely linked with the intensity of the UV absorption and the photochemical resistance of the wood constituents (Sell and Leukens 1971a). Raczkowski (1980) found that during the summer months, changes in the structure of Norway spruce (Picea abies L.) microsections were more intense compared to autumn and winter. He concluded that during the summer months solar radiation is the dominating factor contributing to weathering, whereas during winter atmospheric sulphur dioxide (released by the burning of coal) plays the major role. Sandberg (1999) stated that photochemical reactions at wood surfaces accelerated the propagation of checks, which were caused by stresses due to moisture variations. Sell and Leukens (1969) stated that the development of checks parallel to the fibre was caused by shrinkage stresses within the wood, which exceeded the tensile strength perpendicular to Chapter Two: Literature Review 24 the fibre direction. Checking was thought to be facilitated by the photolytic weakening of the surface layers (Sell and Leukens 1969). Stamm and co-workers noted that face checking was more severe in heat-stabilized wood because prolonged exposure to infrared radiation embrittles the wood surface (Stamm et al. 1946, Stamm 1964 and 1965). The link between exposure of wood to solar radiation and the formation of microchecks is stronger than the changes occurring macroscopically. When wood is exposed to UV light, lignin is degraded first (Miller and Derbyshire 1981). Lignin reaches a high concentration in the middle lamella, hence, the cell corners in cross section are preferentially destroyed, followed by the middle lamella adjacent to them (Evans 1989b, Kuo and Hu 1991, Furuno 2001). The progressive breakdown of lignin leads to the separation of tracheids and finally defibration (Jin et al. 1991). Hemicelluloses are also destroyed very easily by irradiation (Miniutti 1973). The bundles of microfibrils found in the different layers of the cell wall are the most stable structural units according to Borgin (1971). Bamber and Summerville (1981) even suggested that cellulose is protected from the solar radiation by lignin since loss of cellulose cannot be detected before the lignin is lost or chemically altered. Therefore both sides of the outer (Si) and inner (S3) layers of the secondary wall are degraded later than the middle lamella (Furuno 2001). Lignin and cellulose differ in the way they are broken down. Lignin is degraded evenly across the entire cell wall, whereas cellulose is decomposed centripetally from the lumen surface outwards toward the middle lamella (Bamber and Summerville 1981). The depolymerisation of cellulose has profound effects on the surface integrity and tensile strength of wood (Miller and Derbyshire 1981). The secondary wall in both earlywood and Chapter T w o : Literature Review 25 latewood starts thinning after prolonged exposure. Subsequently the earlywood cell walls break (Kuo and Hu 1991). Eventually the surface is converted into a layer of weakly bonded cellulose fibres (Sell and Walchli 1969). Radiation - above all UV light - generates intermolecular contracting forces, which make tracheids twist after they have been loosened by the loss of cell wall material, leading to the formation of fissures (Miniutti 1973, Bamber and Summerville 1981). The deterioration of wood due to radiation also depends on the grain direction. Since during the early stages of weathering, the radiation can penetrate deeper through the open pores, transverse surfaces are degraded more rapidly than radial surfaces (Futo 1974). Also, the middle lamella between tangential cell walls photo-degrades at a lower rate than the lamella between radial walls (Kuo and Hu 1991). In radial surfaces, the bordered pits of tracheids are degraded most rapidly. First their membranes are destroyed, followed by checking and the enlargement of the pit apertures, and finally erosion of the pit borders even extending to areas of the cell wall surrounding the pit border (Hon 1981, Furuno 2001). 2.3.3 Weathering Weathering is a complex combination of chemical, mechanical, and light energies, which in most cases results in a loss of the woods' original colour and gloss, defibration, surface checking as well as alteration and degradation of the chemical components of wood (Sell 1968, Chang et al. 1982, Feist 1982 and 1983, Jin et al. 1991, Evans 2001). The susceptibility of wood to weathering is generally seen as one of the greatest drawbacks of using wood outdoors, particularly under severely fluctuating climatic conditions (Coupe Chapter Two: Literature Rev iew 26 and Watson 1967, Sell and Leukens 1969, Kamden and Zhang 2000). The grey colouration and surface roughness of weathered wood are its most obvious features (Sell 1968, Chang et al. 1982, Feist 1982 and 1983, Jin et al. 1991, Evans 2001). Photo-oxidative degradation reduces both the visual appeal of wood and its performance (Hon et al. 1985). The roughened surface of exposed wood eventually gathers dirt and mildew. Surface checks grow into large cracks. Grain may loosen and boards cup and warp and pull away from fasteners, and the wood may become unsightly (Feist and Hon 1984). The interplay of solar radiation (UV, visible, and IR light), moisture (dew, rain, snow, and humidity), temperature (heat and frost), ozone, atmospheric oxygen and the erosive effects of wind, rain, snow hail and other airborne particles on wooden surfaces is not completely understood (Kleinert 1970, Borgin 1971, Feist 1983, Feist and Hon 1984, Evans 2001). The factors involved in weathering act in different ways and the detrimental effect of some of them is proportional to the length and intensity of exposure (Kiihne et al. 1968, Borgin 1971). The weathering agencies are rarely found in a static-stable state as they are subject to constant and often discontinuous alterations (Kiihne et al. 1968). Furthermore, solar radiation and moisture sometimes affect exposed wood at different times. Thus, degradation pathways may vary depending on the interplay of the elements, with solar radiation enhancing the effect of water or the converse (Feist 1990). The degree of weathering degradation also depends upon wood species and the direction of the grain (Stamm 1963). Anderson et al. (1991), Horn et al. (1994) and Owen et al. (1993) showed that artificial weathering is most destructive when wood samples are exposed to both light and Chapter T w o : Literature Review 27 water. Groves and Banana (1986) mention that it is difficult to discriminate between the effects of the three agencies involved in weathering; sunlight, water and oxygen. The exposure of wood to water and light together results in a surface appearance, which differs from that caused by light or water alone (Feist 1982). Upon exposure to both elements the surface of wood darkens, and macroscopic to microscopic intercellular and intracellular checks form because the lignin-rich middle lamella, which bonds adjacent tracheids and fibres together, is eroded and loses its strength. Subsequently, adjacent primary and secondary cell wall layers are progressively thinned. Rainwater washes out degradation products, which leads to erosion of the wood (Feist 1982, 1990, Evans 2001). Yata (2001) describes the development of small checks at wood surfaces after exposure to rainfall and sunlight (<500 nm) caused by the collapse of photodegraded wood cells due to the evaporation of water. Borgin (1971) stated that "the creation of internal stresses between the different anisotropic cell wall layers due to hygroscopic and thermal movements in connection with the loss of cohesion in the cell walls after photodegradation leads to the development of fractures and fissures". Water accelerates photodegradation of wood because it washes away surface cellulose fibres that have been released due to the lignin degradation, which exposes new cell wall material to UV degradation (Rowell et al. 1981). The embrittling effect on wood caused by infrared irradiation is greatly accelerated when the wood is moist (Stamm 1964 and 1965). Yata (2001) suggests blocking either photo irradiation or water uptake to interrupt the linkage between water and light and thus prevent surface checks from occurring during weathering. Chapter Two: Literature Review 28 2.3.4 Wood Species Checks vary significantly not only between softwoods and hardwoods, but also between different species, which seem to develop individual patterns of checks when weathered. For example, aspen {Populus tremula L.) and spruce (Picea spp.) are both relatively low density woods, but aspen develops a high number of relatively short checks when exposed outdoors, whereas spruce develops fewer, longer checks (Flaste et al. 2000). Coupe and Watson (1967) observed that checks in artificially weathered beech (Fagus sylvatica L.), opepe (Sarcocephalus diderrichii De Wild) and oak {Quercus robur L.) differ microscopically as well as macroscopically. Tangential faces of beech developed "large checks within the multiseriate rays often extending from the rays between adjacent vessel and fibre walls", whereas opepe developed "large checks extending from oblique ends of vessels, checks within rays and some longitudinal checks between adjacent fibre walls". Table 2.1 summarises the appearance of checks for selected softwoods and hardwoods (Coupe and Watson 1967). Chapter Two: Literature Review 29 Table 2.1: Observations o f the effects o f accelerated weathering on the micro-checking o f a range of softwoods and hardwoods (Coupe and Watson 1967) Species Location of checks Tangential surface Radial surface Beech Large checks within the multiseriate rays often extending from the rays between adjacent vessel or vessel and fibre walls. Opepe Large checks extending from oblique ends o f vessels. Some checks within rays. Some longitudinal checks between adjacent fibre walls. Oak Large checks within multiseriate rays. Large checks extending from vessels. Longitudinal checks between adjacent fibre walls. Longitudinal checks between adjacent vessel and fibre walls. Some checks between ray parenchyma cells. Large checks originating in vessels. Longitudinal checks between adjacent fibre walls. Large checks within vessels. Some longitudinal checks between fibre walls. Some checks between ray parenchyma. Scots pine Radiata pine Douglas fir Western red cedar Large checks, originating in rays, extending between adjacent tracheid walls especially in latewood. Large checks extending from vertical resin ducts between adjacent tracheid walls. Some checks within rays. Large checks, originating in rays, extending between adjacent tracheid walls especially in latewood. Large checks, originating in rays, extending between adjacent tracheid walls especially in latewood. Longitudinal checks between adjacent tracheid walls especially in latewood. Large checks extending from vertical resin ducts between adjacent tracheid walls. Some longitudinal checks between adjacent tracheid walls mainly in latewood. Checks (some long) between adjacent tracheid walls in or close to latewood. Long checks between adjacent tracheid walls almost exclusively in latewood or late early wood. Some diagonal checking through pits between radial ray parenchyma and longitudinal tracheids. The appearance of checks in softwoods such as Scots pine (Pinus sylvestris L.), radiata pine, Douglas fir (Pseudotsuga menziesii (Mirbel) Franco) and Western red cedar (Thuja plicata D. Don) also varies substantially. For example the tangential surface of Scots pine after artificial accelerated weathering developed "large checks, originating in rays, extending between adjacent tracheid walls especially in latewood" whereas radiata pine had "large checks extending from vertical resin ducts between adjacent tracheid walls. Some checks also occurred within rays" (Coupe and Watson 1967). Chapter Two: Literature Review 30 Sell and Leukens (1971a) found that 20 species of wood showed clear differences in the way they weathered, but the differences became less pronounced with exposure, and after one year all samples had a grey and uniformly weathered surface. Additionally the surfaces became rougher and numerous checks developed, which ran along the fibres. Microscopically, however, differences in weathering behaviour between species were still evident (Sell and Leukens 1971a). 2.4 Structure and Chemical Composition 2.4.1 Macroscopic Macroscopic checks as their name suggests can be seen with the naked eye. Such macrochecks (hereafter referred to as checks) develop where adjacent cells or tissues differ in cell wall thickness or strength, e.g. at growth ring boundaries and in rays and resin canals, where stresses are concentrated, and at the interfaces between rays and tracheids (Borgin 1971, Evans 1989b). In softwoods most checks originate in latewood, while earlywood develops larger but fewer checks (Coupe and Watson 1967). One reason for this might be that drying causes stresses in ray and non-ray tissue that are greater in latewood than in earlywood. Another explanation could be the higher number of resin ducts in latewood because checks often develop in the vertical resin ducts of Douglas fir and pines (Pinus sp.) (Coupe and Watson 1967). Tangential surfaces of softwoods are more prone to checking than the corresponding radial surfaces. Checks on tangential surfaces are greater in number per unit area as well as Chapter Two: Literature Review 31 being wider than those on radial surfaces (Sandberg 1999). In general, wood pieces with growth ring orientation perpendicular to exposed wood surfaces, i.e. quartersawn boards, tend to check less than flatsawn boards, which often have checks extending from the rays radially into the board (Schniewind 1963, Sandberg 1999). 2.4.2 Microscopic Microchecks are caused by stresses imposed by changes in moisture content and temperature, leaching of extractives, degradation of wood substance due to solar radiation and other factors (Miniutti 1964). Coupe and Watson (1967) distinguished between three types of microchecks; longitudinal checks between adjacent walls of neighbouring elements, which occurred in or close to the middle lamella, longitudinal checks in the wall of an element, and diagonal checks through pits, which were thought to, follow the microfibril angles of the S2 layer. The appearance of checks depends on the wood type, wood species and on the type of surface (radial, tangential or cross-grain). Microchecking and thinning of cell walls occurs in both radial and tangential wood cell walls. In radial walls checks often pass diagonally through the orifices of the bordered pits following the microfibril angles of the S2 layer. Since pits are rare in tangential cell walls, checks run parallel to the long axis of the cells between adjacent latewood tracheid walls or close or inside the rays (Coupe and Watson 1967, Sandberg 1999). Tangential surfaces show more and deeper microchecks than radial surfaces. Such checks develop primarily at the annual ring border, but can also be found in earlywood. Furthermore the Chapter Two: Literature Review 32 delamination of the middle lamella is more pronounced in latewood on tangential surfaces (Sandberg 1999). Evans (1989b) found that on transverse and tangential surfaces of radiata pine macrochecks were caused by cell wall failure and crack propagation at a microscopic level. On the tangential surfaces, ray tissue degraded rapidly and the resulting voids enlarged and coalesced to form macrochecks. 2.4.3 Molecular It has been shown that solar radiation causes quite rapid chemical changes at wood surfaces within a few hours of exterior exposure (Evans et al. 1996; Owen et al. 1993). The photon energy of the sun initiates a variety of chemical reactions at wood surfaces, which makes it the most damaging factor involved in weathering (Hon 1983, Feist and Hon 1984; Feist 1983, 1990; Turkulin and Sell 2002). Free radicals are generated by light and play a major role in surface discoloration and deterioration (Coupe and Watson 1967, Hon 1981, Feist and Hon 1984, Jin et al. 1991, Evans 2001). The first obvious sign that wood is undergoing weathering is the change in its colour (Ifebueme 1977). Arndt and Willeitner (1969) observed that wood exposed in mountainous regions showed more pronounced brightening than samples exposed at sea level and concluded that this change was caused by the increasing levels of solar radiation at elevated altitudes. Anderson et al. (1991) as well as Horn et al. (1994) found that light alone had a much more drastic effect on the chemistry of the wood surface than that produced by water alone. The exposure of wood to UV and visible radiation leads to chemical changes, which Chapter Two: Literature Review 33 make the surface of exposed boards soft (Browne 1960, Kiihne et al. 1968). The photochemically degraded and eroded surface layer is quite different from the original wood substance (Sell and Leukens 1971a). Evans (1996) found that radiata pine veneers lost more weight when they were exposed during summer than at any other time during the year. Derbyshire and Miller (1981) showed that the surface of Scots pine veneers exposed to natural sunlight showed extensive degradation of their lignin, cellulose and hemicellulose irrespective of whether they had been exposed to, or sheltered from rain. Kalnins (1966) explained the photodegradation process using the following scheme; Energy gain from U V light . W o o d in an excited state W o o d that contains free radicals Loss o f excitation through fluorescence, phosphorescence, conversion to heat, or energy transfer Oxygen absent 1 Volat i le photolysis products (only at shorter wavelengths) Oxygen present i Volat i le photo-oxidative products Chemical ly altered wood residue Chemical ly altered wood residue i f complete conversion to volatile products is not allowed Figure 2.2: Scheme for photodegradation o f wood (Kalnins 1966) Evans (1989) found that natural weathering causes degradation of the cell wall in areas containing high concentrations of lignin, such as the middle lamella, such a pattern of degradation occurs because lignin is an excellent absorber of UV light with an absorption peak at 280 nm, with a tail extending to over 400 nm (Miniutti 1973, Hon 1981). Lignin is Chapter Two: Literature Review 34 the most photo-labile wood constituent because it interacts strongly with light (Hon 1981, Feist 1982, Feist and Hon 1984). It can also absorb other forms of energy, such as mechanical, chemical and thermal energies, which may also induce free-radical formation (Hon 1981). Unsaturated groups in lignin absorb UV radiation (Horn et al. 1994). First the phenolic hydroxyl groups are destroyed (Sell 1975). The methoxyl content of lignin is reduced, and the acidity and carbonyl concentrations in wood increase. Low-molecular weight products such as carbon monoxide, carbon dioxide, hydrogen, water, methanol, formaldehyde, organic acids, vanillin and syringaldehyde are formed. The presence of these break-down products may explain the loss of gloss of wood when it is subjected to UV irradiation. The polysaccharide portions of wood are also chemically changed by solar radiation, which leads to a reduction in pentosan and cellulose contents and an increase in the ethanol-benzene and 1% sodium hydroxide solubility of wood (Hon 1981). Direct photolysis of cellulose is possible as a result of exposure to UV light 340 nm or shorter in wavelength. Longer wavelengths can have the same effect when sensitizers are present. Such chemical changes are the cause of the discolouration and gradual loss of mechanical properties of the exposed wood surface layer on exposure of light (Hon 1981). Water also modifies and accelerates chemical changes, and deterioration of wood initiated by UV light, such as loss of colour and brightness and pH (Webb and Sullivan 1964, Stamm 1965b, Jin et al. 1991, Turkulin and Sell 2002). Arnold et al. (1991) concluded that a water-spray system in a weatherometer was essential for removing UV-light-degraded wood material from exposed surfaces and obtaining surfaces similar to naturally weathered wood. Chapter Two: Literature Review 35 In the literature, macrochecking has not been related to chemical changes at the molecular level although as described in Section 2.4.2 the photodegradation of the wood cell wall creates voids and microchecks which subsequently enlarge to create macrochecks (Evans 1989b). 2.5 Chemical Treatments 2.5.1 Preservatives Preservatives prevent wood from deterioration by fungi and other organisms, but they are generally not formulated to protect it from physical degradation such as checking (Zahora 1991). Water, oil and solvent-based wood preservatives are widely used by industry to treat timber (Ibach 1999). The following water-based treatments are or have been used by industry; acid copper chromate, ammoniacal copper zinc arsenate, chromated copper arsenate (CCA), ammoniacal copper quat, copper bis(dimethyldithiocarbamate), ammoniacal copper citrate, copper azole - type A, inorganic boron (borax/boric acid) (Ibach 1999) and alkyl ammonium compounds (AAC) (Placket et al. 1984) There is no clear consensus in the literature as to whether these water-borne treatments promote or restrict checking and splitting of wood. The most common preservative for wood used outdoors until recently was CCA. It is generally used to prevent the decay of poles and piles and also wood that is used for commercial building purposes, e.g. for cladding, decking, etc. Consumers have expressed dissatisfaction with the weathering performance of sawn CCA-treated timber due to the severe surface checking that develops when the treated wood is exposed outdoors. Mackay Chapter Two: Literature Review 36 (1973) suggested that CCA caused radiata pine to be less permeable and thus treated wood tended to develop larger drying stresses in comparison to untreated wood when dried at low temperatures. Pits close during initial drying, which slows the rate of moisture leaving the cells. Furthermore, the deposition of CCA salts in cell lumens and walls and on pit margos following treatment contributes to the blocking of flow pathways. During weathering, wood is subjected to wetting and drying and most of the checks, which develop during initial drying, reopen and in some cases increase in length (Mackay 1973). CCA only provides a temporary water repellent effect on Southern yellow pine, which depends directly on the concentration of the treatment (Zahora 1991). In some studies CCA has been found to increase the checking of wood for example of radiata pine shingles exposed in Rotorua, New Zealand (Placket et al. 1984). • Evans et al. (2003) found that after one year of weathering in South-East Australia CCA-treated, sawn, kiln-dried radiata pine decking checked in a similar way to timber that had been pressure treated with water. They concluded that some aspect of the treatment process, possibly re-drying after pressure treatment, may have influenced checking. In comparison to untreated wood, CCA- and water-treated wood developed slightly larger, but less numerous checks. Crawford et al. (1999) treated incised and unincised white pine (Pinus strobus L.), red pine (Pinus resinosa Soland), eastern white spruce (Picea glauca (Moench) Voss), eastern red spruce (Picea. rubens Sarg.), eastern black spruce (Picea mariana P. Mill.), and balsam fir (Pinaceae Abies balsamea (L.) P. Mill.) decks with CCA-Type C for above Chapter Two: Literature Review 37 ground field trials and weathered them for 10 years. Only unincised white pine decks failed due to transverse checks and splits around knots. Bariska et al. (1988) showed that the wood substance between ray and fibre tissue, between single cells in the compound middle lamella region and between layers of the cell wall in radiata pine is hydrolysed and dissolved by CCA during treatment. This was thought to promote checking. CCA fixes or deposits in higher concentrations in the latewood, but the damage to latewood caused by CCA is not significant since latewood is inherently strong. The strength of the earlywood is not markedly affected because earlywood contains a lower CCA concentration due to its lower permeability when it has been dried. However, at the interface between latewood and earlywood high concentrations of CCA in the summerwood come in contact with the weaker springwood and cause hydrolysis of the holocellulose and lignin fractions of the wood substance. Since this interfacial zone is weakened, it is the site for the formation of delamination cracks. The following oil-based treatments have been used commercially to treat wood; creosote, pentachlorophenol solutions (PCP), copper naphthenate, chlorothalonil, chlorpyrifos, oxine copper (copper-8-quinolinlate) and zinc naphthenate (Ibach 1999). Creosote and PCP are heavy oil treatments. They are water-repellent and tend to form a water impermeable oily layer on the wood surface, which reduces checking (Levi et al. 1970). Straight creosote tends to be less effective in reducing checking than creosote solutions (for example creosote-coal-tar solution). However, creosote solutions tend to bleed from treated wood surfaces and penetrate wood with greater difficulty (Ibach 1999). Chapter Two: Literature Rev iew 38 Creosote treated messmate stringybark poles checked less than untreated and salt-treated poles after 5 years of weathering, irrespective of climatic conditions (Gilfedder et al. 1968). Gilfedder et al. (1968) suggested that the ability of creosote to restrict checking was caused by its tendency to inhibit rapid drying. Organic solvent based preservatives form a third group of preservatives. They are generally used to treat jonery and wood used above-ground (Levi et al. 1970). Copper naphthenate-based light organic solvent preservatives (LOSP), however, are considered an alternative to CCA to overcome concerns over the use of rainwater run-off for drinking water, corrosion of fasteners, and checking. LOSP's can provide significantly improved resistance to checking (Placket et al. 1984). 2.5.2 Water Repellents Wood treated with water repellents shows reduced water intake, which is accompanied by an increase in dimensional stability (Borgin 1965) and is less susceptible to checking. Water repellents are usually complex blends of wax, oil, resin and solvent (Borgin and Corbett 1970a). Penetrating water repellents, however, are much more effective than surface treatments (Borgin and Corbett 1969). Borgin and Corbett (1971) found that extractives from the bark of radiata pine combined a remarkable affinity for wood with a high degree of water-repellency even after several wetting-drying cycles. Water-repellent additives have been used to prevent the formation of checks in CCA-treated wood during weathering (Levi et al. 1970, Zahora 1991). The stability of the water-repellent emulsion is critical to the success of the treatment because of the low pH of a Chapter Two: Literature Rev iew 39 solution of electrolytes containing hexavalent chromium. Additives have mainly been used commercially as a self-dispersing solution, which is compatible with CCA and most other salt type preservatives. Treatment is carried out in the same way as for straight CCA. After treatment, the wood dries and the emulsion "breaks" (Belford and Nicholson 1969). No curing is required. The hydrophobic molecules are deposited within the capillary structure of the wood, "over the inner surface of the lumen of the tracheids with a tendency to form discrete deposits within ray parenchyma cells and bordered pits" (Belford and Nicholson 1969). Thus, despite the fact that the wax content in the wood is relatively low, the contact angle is increased over the whole internal capillary system within the treated zone. Wax imparts water repellency to wood and retards the ingress of moisture, but it does not keep the wood from reaching an equilibrium moisture content with the surrounding atmosphere (Belford and Nicholson 1969). Levi et al. (1970) found that rays were the tissues which were most susceptible to weathering degrade when exposed to alternate wetting and drying. They explained the long term effectiveness of water-repellents in dimensionally stabilising wood as being due to their high concentration in the summerwood rays of the sapwood (Levi etal. 1970). The cyclic changes in wood moisture content within the hygroscopic range (< 30%) are lower for wood treated with water-repellent preservatives. Evans et al. (2003) found that checks in CCA-wax treated radiata pine decking, following one year of natural weathering were significantly smaller in width and depth and fewer in number than in untreated, water- and CCA-treated samples (Evans et al. 2003). Chapter T w o : Literature Review 40 Zahora (1992) obtained similar findings for Southern yellow pine boards, which had been treated with a CCA containing a water repellent (WR) emulsion additive. He measured the widths of the checks and found that even the third largest check for CCA-only treated boards was wider than the largest check in the CCA-WR treated boards. The water repellent additive minimized moisture gradients in the treated boards during the first year of weathering and thus stabilized checking characteristics. Surface water beading effects, characteristic of CCA-wax treated boards, disappear after about three years exterior exposure. But, even after nine years of exposure, boards treated with CCA-wax showed smaller variations in moisture content and reduced checking (Zahora 2000). The swelling rates were also lower. Both CCA and CCA-WR treated boards, however, showed damage such as checks caused by extensive UV exposure on the top surfaces, which promoted water-intake into the boards. However, the checks in CCA-WR treated boards were scattered and usually less than 5mm in depth, whereas checks in CCA-only treated boards penetrated much deeper into the wood (Zahora 2000).The water repellency of water-repellent preservatives also depends on the type of preservative. ACQ/WR treated wood showed lower water repellency than CCA/WR treated wood (Cui and Zahora 2000). When used as surface treatment, wax cannot provide lasting protection because of the displacement of the wax phase from the wood surface by water after weathering. The displacement is less pronounced for waxes that contain hydrophilic groups, because the affinity for wood is better, but the hydrophobic properties of such waxes are weaker. An optimum has to be found (Borgin and Corbett 1970b). Chapter Two: Literature Review 41 2.5.3 Thermal Treatments Wood can be treated with heat or frozen to modify its dimensional stability and susceptibility to checking. Heat treatment involves subjecting wood to temperatures close to or above 200°C for several hours in an atmosphere with low oxygen content. Four different major European heat treatments are distinguished; Thermo Wood in Finland, Plato Wood in the Netherlands, Retrification in France and O i l Heat Treatment in Germany (Rapp 2001). In Finland, Scots pine, Norway spruce, birch (Betula verrucosa/pubescens Ehrh.) and European aspen have been heated in a water-steam atmosphere to temperatures of between 150°C - 240°C for 0.5 to 4 hours (Syrjanen and Oy 2001). The Plato-process combines a hydrothermolysis step with a dry curing step. During the hydrothermolysis step green or air dried wood is treated at temperatures typically between 160°C - 190°C under increased pressure., During the curing step the wood is heated again to temperatures of between 170°C - 190°C: Both steps involve either steam or air as the main heating material (Mil i tz and Tjeerdsma 2001). In the French Retrification process, previously dried wood of a moisture content of about 12% is heated slowly to a temperature of 210°C - 240°C in a nitrogen atmosphere containing less than 2% oxygen (Vernois 2001). In Germany, wood has been heated in a hot bath of vegetable oils, for example rape seed, linseed oil or sunflower oil at a temperature of 220°C for maximum durability and minimum oil consumption or at temperatures of between 180°C - 200°C for maximum durability and maximum strength (Rapp and Sailer 2001). Chapter T w o : Literature Review 42 All of the processes described above increase the resistance of the wood to fungi and improve its performance outdoors (Rapp 2001), because they reduce the equilibrium moisture content of wood and its shrinkage and swelling (Syrjanen and Oy 2001). The latter should reduce the tendency of heat-treated wood to check since swelling and shrinkage are the main reasons for the development of checks. On the other hand it has been reported that the high temperatures during treatment also increase the brittleness and the tendency of the wood to check (Rapp and Seiler 2001). Heat treatments also decrease wood strength properties (Rapp 2001). When shrinking and swelling is reduced by 40%, the toughness is less than half that of the original wood. Because of these strength reductions, thermally modified wood should not be used for structural applications (Rowell 1999). Pre-ff eezing has been used experimentally as a pre-treatment to prevent the checking of wood during kiln drying (Cooper 1972), as it reduces drying shrinkage in some woods, e.g. black walnut (Juglans nigra L.). The positive effect of pre-freezing on checking is due to the bulking effect of wood extractive(s) and modification of the rheological properties of wood. Since weathering involves cyclic wetting and drying, pre-freezing might be a way of reducing the shrinkage of wood and thus stresses that lead to checks (Cooper 1972). 2.5.4 Mechanical Treatments Mechanical treatment of poles has two different goals. They are designed to improve the penetration of chemical treatments such as CCA, since the effectiveness of a preservative treatment depends on the thickness of the outer treated shell surrounding the Chapter T w o : Literature Rev iew 43 untreated wood, as well as the amount of the preservative introduced into the wood and the continuity of the treated shell. At the same time the treatments should reduce drying stresses in the wood and prevent checks from developing and from deepening beyond the chemically treated shell (Ruddick and Ross 1979). Mechanical treatments of timber designed to improve preservative penetration include kerfing, center boring, through-boring and incising. These treatments have mainly been applied to softwood poles and the literature indicates that some of the procedures can reduce checking. 2.5.5 Kerfing The term "kerfing" is generally used in connection with poles and involves cutting a slot from the periphery of the pole to its centre and from the butt to the desired point along the axis of the pole (Morrell 1990). Consumers Power Inc., a utility in the Pacific Northwest, used transmission poles with kerfed tops to effectively prevent checking due to bolt holes (Graham 1973). Width, length and number (single or double kerfing) of kerfs differed depending on the chemical treatment, species and use of the pole. Morrell (1990) stated that although the performance of kerfed poles to resist checking was very good, kerfing was only incorporated into pole specifications by one utility in the United States. The literature on kerfing of poles describes the result of two different types of studies; 1, the evaluation of poles in service and specially designed and 2, controlled experiments. Morrell (1990) evaluated 5,000 poles ranging from 0 to 25 years in age that had been in service for various periods of time. He found that if kerfing was performed Chapter Two: Literature Review 44 prior to treatment of poles, the kerf was well treated and shrank and swelled with moisture changes and accordingly relieved internal stresses. In larger poles, kerfing prevented the development of deep checks extending beyond the depth of the treated shell. However, kerfing did not substantially improve the performance of smaller poles. Helsing and Graham (1976) inspected creosote or pentachlorophenol treated Douglas fir transmission poles that contained a single deep saw kerf from the butt to 1.5 m above ground or for two-thirds of the pole length after 5 to 11 years in service. They found that kerfing reduced the width of checks. However, small checks occurred at the base of those kerfs that ceased 41 mm from the center of the pole. This finding emphasises the importance of the kerf going to, or beyond, the center of the pole. Kerfs have a negligible effect on the pole strength and kerfed poles do not show an increased tendency to split (Helsing and Graham 1976). Early experiments by Graham (1973) showed that three kerfs spaced equally around posts widened on initial exposure of the posts to the weather. After a short period of time, however, only one of them continued to widen and deepen, usually to the center of the post. Thus, the author concluded that only one kerf to the center was needed to prevent checking. Ruddick and Ross (1979) examined the effect of kerfing on checking of untreated, unseasoned Douglas-fir pole sections. The pole sections were kerfed full length to the pith with a 4 mm wide saw kerf and exposed outdoors for 44 months. In accord with Morrell (1990) they observed that more than 50 percent of the tangential shrinkage of the weathered poles occurred in the kerf, which rapidly widened during the first 2 years of exterior use. At Chapter Two: Literature Review 45 the same time, the maximum depth and width of the largest checks was found in the unkerfed pole sections. The mean depth of the largest (i.e., most penetrating) checks in the kerfed poles was almost 2.4 times less than that in the unkerfed pole sections. Ruddick (.1981) also examined the effect of weathering on kerfed white spruce poles, which had been commercially pressure treated with either ammoniacal copper arsenate (ACA) or pentachlorophenol (PCP). The check depth and width in the kerfed poles was significantly less than that in the unkerfed poles after ten years of exposure. Ruddick (1981) stated that kerfed ACA treated poles performed better than either unkerfed PCP or ACA treated poles. Ninety percent of all checking occurred during the first three years. After seven additional years of exposure, the depth of the kerf increased only by 10%. Subsequently Evans et al. (1997) found that single and double kerfing reduced the number of checks greater than 1 mm wide, which developed in slash pine posts during air-drying and after they were treated with CCA and exposed to 6 weeks of weathering. Both single and double kerfing also reduced the number of checks and their size in the CCA-treated posts after one year of exterior exposure. However, single kerfs opened very wide (> 10 mm) in dry weather, which reduced the attractiveness of the treatment. Therefore they recommended the use of double kerfs to reduce checking of posts. Kurisaki (2004) confirmed that single and double kerfs prevented deep checking of sugi (Cryptomeria japonica D. Don) posts. Graham (1973) also observed a reduction in check width and depth in CCA-treated red spruce {Picea rubens Sarg) poles that had been kerfed to their centers along their entire length. Chapter Two: Literature Review 46 Evans et al. (2000) examined the effect of kerfing and center-boring on the checking of ACQ treated radiata and slash pine posts (Figure 2.3). They found that posts containing 30 mm and 22 mm deep double kerfs had the lowest check surface area, number of checks and check volume, width, length and depth. Generally, increasing the depth of double and single kerfs resulted in reduced number, width and depth of checks. Single Kerfing K1 (30mm deep) K2 (45mm deep) K3 (60mm deep) Double Kerfing D1 (1 5mm deep) D2 (22mm deep) D3 (30m m deep) Center boring C1 (25mm deep) C2 (35mm deep) C3 (45mm deep) Figure 2.3: Kerfing and center-boring treatments (Evans et al. 2000) Chandler (1968) found that full length saw kerfs to the center of 20-cm-square guardrail posts reduced checking. Splitting was rare in both round and square members that had been kerfed to the center before seasoning (cited by Graham 1973). 2.5.6 Inc is ing Incising originated in Europe and was first used in the USA as a pre-treatment for Douglas fir ties and timbers in Oregon in the 1920's. It should be differentiated from the OOO Chapter T w o : Literature Review 47 perforation of wood with a pattern of drilled holes or slits that run around the outside of poles with a slope of 5° (Graham 1973). The main reason for incising is to improve the penetration of preservatives in refractory (impermeable) species. Ruddick and Ross (1979) hypothesised that incising might encourage the formation of several shallow checks, rather than one major check, which could penetrate beyond the treated shell. However, incising proved to be ineffective at reducing checking in CCA-C treated slash pine poles after air-drying and weathering (Evans et al. 1997). Graham (1973) stated that incising improved the drying process and thus reduced checking indirectly, and it only had a minor strength-reducing effect on preservative-treated poles. 2.5.7 Center Boring As early as 1935 Loughborough concluded from theory that drilling longitudinal holes though the center of round wood might prevent checking (cited by Graham 1973). Mater (1972) described center boring as an ancient art, with some old wooden parts of ships, the stately wood columns in colonial homes, and the circular wood porch supports of homes built until the 1920's being bored through their center to reduce surface checking. A number of studies have examined the effect of center boring on the checking of untreated and treated poles. Graham (1973) found that center boring reduced the width of checks in CCA-treated poles, but it had a strong negative effect on the widening of checks which more than offset its positive effect on checking. The depth of the checks in center bored poles was only reduced in the untreated poles. Chapter Two: Literature Rev iew 48 Mater (1972) found that centre-boring of untreated poles significantly reduced the number and the lengths of checks that developed when poles were exposed outdoors. It was also thought to have had some beneficial effect on the long term strength of poles, even though the initial strength of a centre-bored 30-foot long pole 8 inches over the ground line was about 1 % and 9.8 % weaker in bending and shear, respectively, in comparison to a solid pole. Evans et al. (1997 and 2000) came to the conclusion that center boring of ACQ and CCA-C treated radiata pine posts and slash pine posts prevented checks from developing and from becoming wider or deeper, but had no significant effect on the length of checks. Posts containing center bored holes with a larger diameter (Figure 2.3) developed narrower checks, but there was no significant effect of the diameter of the centre-bored hole on check depth or number (Evans et al. 1997 and 2000). 2.5.8 Constructive Protection Constructive protection is probably the oldest form of wood preservation. As early as 1830 and again after 1880 and around 1910 building construction in Switzerland moved towards wooden cladding and facade elements. Large, extensive wooden facades were protected by roof and wall overhangs. Wooden storefronts were sited above the ground. Walls that were not protected by overhangs were covered with shingles or rough vertical siding. The construction of the walls suited the accepted room climate standards based on thermal isolation and diffusion. Nowadays, constructive wood protection only plays a minor role in architecture. Materials and constructive tools are used in a freer way (Kiihne Chapter T w o : Literature Rev iew 49 et al. 1968). Such practices result in more direct exposure of wood to the weather, which increases the risk of checking and cracking. 2.6 Summery Unprotected wood exposed to the weather even for a short period of time wi l l undergo significant changes in appearance. There is indirect evidence that checking of exposed wood is accelerated by weathering, the complex effects of combination of solar radiation, water, temperature and atmospheric gases on wood (Sell 1968, Chang et al. 1982, Feist 1982 and 1983, Jin et al. 1991, Evans 2001). Although it is difficult to discriminate between the factors involved in weathering (Groves and Banana 1986), researchers generally approach checking by focusing on one aspect of weathering. Emphasis has been placed on the effects of water, which acts in various ways. A t the macroscopic level, absorption of water by wood generates stresses that vary depending on moisture gradients (surface/core), orientation of growth rings (radial/tangential), springwood and summerwood as well as cell structures (parenchyma/rays) (Stamm and Loughborough 1942, Mcintosh 1955, M c M i l l e n 1955, Schniewind 1959, Browne 1960, Schnieweind and Kersavage 1961, Stamm 1963 and 1964, Sell and Walchl i 1969, Borgin 1971, Sandberg 1999). Furthermore water can contribute to checking when it freezes and thaws (Rowell et al. 1981). At the microscopic level, shrinkage stresses cause tracheids to twist (Miniutti 1973) and this results in additional stresses by causing adjacent cell wall layers to swell and shrink in almost opposite directions (Borgin 1971). Schniewind and Kersavage (1961) refer to these intercellular stresses as first order stresses. Water leaches photodegradation Chapter Two: Literature Review 50 products from wood surfaces, swells wood and can hydrolyse hemicelluloses (Furuno 2001). Various studies describe the ability of UV light to alter wood colour and degrade lignin cellulose and hemicellulose, which results in the deterioration of the ultra-structure of the cell wall and finally the conversion of the wood surface into a layer of weakly bonded fibres (Sell and Walchli 1969, Borgin 1971, Miniutti 1973, Bamber and Summerville 1981, Derbyshire and Miller 1981, Miller and Derbyshire 1981, Hon et al. 1985; Evans 1989b, Kuo and Hu 1991, Furuno 2001). At the microscopic level, degradation of bordered pits at radial surfaces, resulting in the formation of microchecks, can be observed (Derbyshire and Miller 1981, Hon 1981, Groves and Banana 1986, Furuno 2001). At tangential surfaces, the degradation of rays results in microchecks (Evans 1989). These microchecks coalesce and develop into large macrochecks that are visible to the naked eye. Hence, there is undoubtedly a connection between the photodegradation of wood and checking. There have, however, been no studies of the effect of solar radiation on the checking of wood. Several authors have reported roughening and checking of the wood surface after the exposure to light (Arndt and Willeitner 1969, Sell and Leukens 1971a, Hon 1981) but neither a quantitative nor a qualitative analysis of the effect of solar radiation on the macrochecking of wood has been undertaken to date. Therefore this thesis focuses on the influence of solar radiation and its spectral components on the checking of wood. Chapter Three: Exposure o f Decking Samples 51 CHAPTER THREE 3 Effect of Solar UV-A, UV-B, Visible Light and IR Light on the Checking and Colour of Flat-Sawn Lodgepole Pine Surfaces 3.1 Introduction Wetting and drying of wood causes changes in its dimensions and thus generates stresses, which lead to checking (Feist 1990). Photodegradation caused by solar radiation plays an important role in the process of weathering (Derbyshire and Miller 1981). While most authors have focused on UV light as the main factor responsible for photodegradation, Derbyshire and Miller (1981) found that the visible part of the spectrum contributes significantly to the degradation of wood. The combined effects of water and sunlight convert wooden surfaces to a layer of weakly bonded cellulose-rich fibres (Sell and Walchli 1969). Yata (2001) observed small checks on the surface of Japanese cypress after exposure of the wood to rainfall and light irradiation at wavelengths below 500nm. Solar radiation degrades rays at exposed tangential surfaces, which creates microscopic voids. These voids coalesce with further exposure resulting in the formation of macroscopically visible surface checks (Evans 1989b). It is hypothesised here that that the exposure of.wood surfaces to solar radiation will enhance the development of checks at wood surfaces in proportion to the ability of the Chapter Three: Exposure of Decking Samples 52 radiation to photodegrade wood. The aim of this Chapter is to test this hypothesis by exposing lodgepole pine wooden decking samples under filters, which transmit selected regions of the solar spectrum while allowing others aspects of weathering (water, temperature and oxygen) to occur, and then quantifying checking in the samples. In addition FTIR spectroscopy and scanning electron microscopy of the surfaces is reported. 3.2 Materials and Methods 3.2.1 Lumber Five lodgpole pine logs were obtained from the Alex Fraser Research Forest close to Williams Lake, B.C., Canada. The logs were felled in February 2004 and were approximately 3.05 m long with an average diameter of 34 cm. The average number of growth rings was 5.3 rings per centimetre (Table 3.1), and the logs were free of blue-stain. Table 3.1: Length, widths and number of growth rings of the logs Log No Length (m) Butt diameter (cm) Top diameter (cm) Number of growth rings 1 3.40 390 345 94 2 2.96 380 310 88 3 3.15 385 310 95 4 2.99 330 290 84 5 3.30 340 310 93 Each log was manually debarked using a bark spud and cut into 2" (40 mm) thick decking boards using a Wood-Mizer LT 15 portable sawmill (Figure 3.1). Chapter Three: Exposure of Decking Samples 53 A 2 " (40 mm) thick board (Board III) was cut from the center of each log parallel to the pith. The adjacent boards (Boards I, II, IV, V) were also 2 " (40 mm) thick. The cutting pattern is shown in Figure 3.2. Board II of each log was selected for the experiments described in Chapters Three and Four. Figure 3.1: The Wood-Mizer L T 15 sawmill used to cut logs into decking boards Figure 3.2: Cutting pattern for the lodgepole pine logs Chapter Three: Exposure of Decking Samples 54 The rough sawn boards were kiln dried for fourteen days to a moisture content of 13% using a moderate drying schedule in order to reduce the formation of drying checks (Table 3.2, Figure 3.3). Then the boards were stored under cover for four days before being converted into decking and veneer samples. Table 3.2: Dry ing schedule with Ta: temperature (true value), Tn : temperature (command value), E M C - T r : equilibrium moisture content (true value), E M C - C r : equilibrium moisture content (command value), D G a : drying gradient (true value), D G n : drying gradient (command value), V N T a : ventilator speed (true value), V N T r (command value) and D C V and F M a , : both drying control values Date Ta Tn EMC-Tr EMC-Cr DGa DGn VNTa TNTr D C V FMa [°C] [°C] [%] [%] [1] [1] [%] [%] [%] [FMa] Feb, 14 th 2004 5.3 65.0 12.7 14.0 2.4 2.1 0 9 71.6 71.6 Feb, 15 th 2004 63.8 65.0 8.1 14.0 3.5 2.1 100 100 73.8 73.8 Feb, 16 th 2004 67.2 65.0 13.8 11.5 2.2 2.6 100 99 65.7 66.7 Feb, 17 , h2004 64.7 65.0 11.8 11.5 2.5 2.6 100 100 66.0 66.0 Feb, 18th 2004 65.7 65.0 11.6 11.5 2.6 2.6 100 100 64.2 64.2 Feb, 19 th 2004 65.3 65.0 11.7 11.5 2.6 2.6 100 100 58.8 58.8 Feb, 20 t h 2004 64.8 65.0 11.5 11.2 2.6 2.7 82 81 48.3 48.3 Feb, 21* 2004 64.2 65.0 11.1 10.8 2.7 2.8 95 94 36.3 36.3 Feb, 22 t h 2004 65.0 65.0 11.3 11.2 2.7 2.7 89 88 487 487 Feb, 23 t h 2004 64.2 65.0 11.5 11.0 2.6 2.7 92 81 39.4 39.4 Feb, 24 t h 2004 64.3 65.0 10.8 10.5 2.8 2.9 87 87 30.2 30.0 Feb, 25 t h 2004 66.4 66.7 9.0 8.9 2.6 2.6 93 92 23.3 23.3 Feb, 26 t h 2004 69.1 69.1 7.8 7.5 2.3 2A 78 77 18.1 18.1 Feb, 27 t h 2004 71.3 71.1 6.4 6.3 2.3 2.3 88 87 14.8 14.8 Feb, 28 t h 2004 71.3 72.1 5.5 5.4 2.2 2.2 72 71 12.3 12.3 Chapter Three: Exposure of Decking Samples 55 TRF Zpom out -so* Zoom Hack -eo •20* Rediaw 'TO g | | i HF || xy 1ST Ta pr tn ST EMC-T 'PT EMC-C rW JO 20 TVNT -T>r VNT - D r p-act r pr DCV pr FMa rMOI-05 rM06-10 -10 0 f" Vents I - Mixere Figure 3.3: Drying diagram showing Ta: temperature (true value), Tn: temperature (command value), EMC-Tr: equilibrium moisture content (true value), E M C - C r : equilibrium moisture content (command value) and D C V and FMa, both drying control values 3.2.2 Decking Samples The five boards were manufactured into decking samples with a final size of 6" (width) x 2" (thickness) (140 mm x 40 mm). All four sides were planed using a Martin T44 jointer and a Martin T54 thickness planer. All longitudinal edges were rounded to a 5 mm radius using a Martin T26 shaper (Figure 3.4). The decking boards were cross-cut into 8 samples each 32 cm in length using an Omga RN 600 radial arm saw (Figure 3.5). Seven decking samples from each log were selected at random. Their end grain was sealed with an "End Sealer For Logs" (LeeValley, Richmond, B.C., Canada). > Figure 3.5: Lodgepole pine decking sample Chapter Three: Exposure o f Decking Samples 57 3.2.3 Exposure Five decking samples from each log were exposed under filters, which cut out certain wavelengths in the solar spectrum (Section 3.2.4). A sixth sample from each board was weathered without being shielded by a filter. The seventh decking sample from each log was kept inside the conditioning room for the duration of the experiment to act as a control. The samples were exposed horizontally to the weather with growth rings orientated convex to the sun. Samples were exposed at Totem Field in Vancouver, Canada (49° 15' 29" N; 123° 14' 58" W; elevation 104.42m), in a shade-free location. The samples were weathered for three periods of 12 weeks each as follows; May, 31st 2004 - August, 22nd 2004; September, 7th 2004 - November, 28th 2004, December, 13th 2004 - March, 6lh 2005. After each exposure period the samples were conditioned for 14 days in a conditioning room maintained at 20°C ± 1°C at 65% ± 5% relative humidity before the checks and the colour of the samples were measured (see Section 3.2.6). 3.2.4 Filters Five different types of Polymethylmethacrylate (PMMA) sheets were obtained from CYRO Industries in Rockaway, USA, which distributes PMMA under the brand name Acrylite. The following types of sheets were used as filters in the experiment; Filter 1: no blocking; Acrylite OP-4 sheet Filter 2: (290-320nm); Acrylite GP sheet; blocks UV-B Filter 3: (290-400nm); Acrylite OP-2 sheet; blocks UV-B and UV-A Filter 4: (290-700nm); Acrylite GP Black 1146-0; blocks UV and visible light Filter 5: (290-3OOOnm); Acrylite GP Black 199-0; blocks UV, visible light and infrared light Chapter Three: Exposure of Decking Samples 58 Filters 1 -3 are acrylic cell cast sheets. Filter 1 (OP-4) transmits most of the radiation in the range from 260 - 370 nm and about 92 % above a wavelength of 370 nm. It will act as control. Filter 2 (GP) transmits very little radation below 345 nm. Most of this range is apportioned to U V - B radiation. Filter 3 (OP-2) absorbs approximately 98% of the incident U V light (Figure 3.6). Filter 4 (GP Black 1146-0) is designed to transmit infrared light but to absorb U V and visible light while Filter 5 (GP Black 199-0) does not transmit any solar radiation (Figure 3.7) (Anonymous 2001). Filters 1 to 4 are 3 mm thick. Filter 5 is 6 mm thick. — - • OP-4 GP GP-2 Wavelength (nm) Figure 3.6: U V light transmission of colourless Acrylite OP-4, Acrylite GP, Acrylite OP-2 (Anonymous 2001) I u u 90 B0 7D B0 90 40 30 20 10 n / < t f / f / I 1 I J J / p / / / Chapter Three: Exposure of Decking Samples 59 100 200 400 600 800 1000 1200 1400 1600 1800 2000 Wavelength (nanometers) Figure 3.7: Transmission of Acrylite GP sheet, Colour # 1146-0 and 199-0 (Anonymous 2001) 3.2.5 Racks Custom-made exposure racks served as supports for the decking samples and the overlying filters (Figure 3.8). They also maintained a distance of 140 mm between the filters and upper surface of the samples. The racks were designed to allow rain water, air and atmospheric pollutants to come in contact with the samples. One rack was assigned to samples from the different logs (5 racks in total). The filters were arranged randomly on top of the racks. To prevent solar radiation from reaching the surface of the samples from the front or back of the racks, the same type of filter as used on the top was also screwed onto the sides of the rack. The decking samples were placed on 2" (40 mm) wide stickers, which were nailed at right angles to the long axis of the rack. Custom made dark brown coated aluminium sheets, supplied by Main Sheet Metal Works Ltd. in Vancouver, Canada, angled at 19° to the filters captured rain water and directed it onto the surface of the decking samples. The material used for the construction Chapter Three: Exposure o f Decking Samples 60 of all racks was CCA-treated SPF. The racks were manually assembled using brackets, nails and screws. To prevent CCA from leaching onto the samples and the solar radiation from being reflected onto the samples the racks were painted dark brown with one coat of semi-gloss acrylic latex Benjamin Moore Style Exterior paint (588 4B). Concrete blocks placed under the main legs of the racks maintained a distance of about 40 cm between the ground and the samples (Figure 3.10). Racks were inspected daily to keep the filters free of dust and prevent weeds from growing under the racks. 4 f mm--^ ^ ^ ^ ^ at:-: it Figure 3.8: Top view and side view o f a decking sample rack Chapter Three: Exposure of Decking Samples 61 Figure 3.10: Rack for decking samples 3.2.6 Measurement of Temperature, Precipitation, Visible Light and U V Light The average temperature and total amount of precipitation at the exposure site were obtained from measurements made by the Faculty of Agriculture, University o f British Columbia, Vancouver, Canada. Chapter Three: Exposure o f Decking Samples 62 The total UV-A and UV-B radiation received by samples was obtained using passive VioSpor Dosimeters from BioSense in Bornheim, Germany. The spore-film filter system is as UV-sensitive device that uses dried DNA-containing bacterial spores. For the film production a DNA-deficient strain of Bacillus subtilis was used. It was shaken in a growth medium and then stained and fixed to allow densitometric protein quantifications at the various film areas to be measured. The dried films were exposed in waterproof castings with a diameter of 32 mm and covered by a filter system that allowed the penetration of UV. The UV radiation penetrates the calibrated detector film of the dosimeter. The stronger the radiation dose the more spores are damaged or killed, which will not be able to develop into bacteria later on. Only bacteria, however, can be photo-metrically detected after the film has been developed. The amount of damaged bacterial spores can thus be directly linked to the quality and intensity of the detected UV radiation (Furusawa et al. 1998). One dosimeter was exposed under each type of filter. One dosimeter was also exposed to the full solar spectrum. For filter 1 and the freely exposed samples, dosimeters of Type 3 dosimeters (silver grey) were used (Figure 3.11). They were replaced every 2 to 3 days during the summer from June until the end of August. For the remainder of the experiment one dosimeter of Type 3 was used per week for samples exposed to the full solar spectrum and beneath filter 1. Dosimeters of Type 2 (black) were used under filter 2. They were replaced every week during the summer and every four weeks during fall and winter. For filters 3, 4 and 5, dosimeters of Type 1 (blue) were chosen. They were changed every 12 weeks for the duration of the experiment. Chapter Three: Exposure of Decking Samples 63 Figure 3.11: Dosimeter of Type III (silver) Visible light received by samples was measured using PMA2132, PAR (visible light) digital detectors. Data from the detectors was fed into a PMA2100 Radiometer/Data Logger from Solar Light Co in Philadelphia, USA. The data was processed on a PC using PMA organizer software supplied by the same company. Two detectors were used at the same time. Measurements were cross-checked under different weather conditions and filters every four weeks. 3.2.7 Data Collection Before weathering and after each exposure period, boards were conditioned (as above) and the tangential surface of each decking sample was photographed with a Sony digital camera (DSC-S75) at a resolution of 3.3 mega pixels. Checks at the surface of deck samples were counted and their length and width were manually measured using a transparent ruler and an optical magnifying glass containing a calibrated graticule (Zoom Lupe 816 from Peak), respectively. The colour of the decking samples was determined according to ISO 7724 Standardising a Minolta spectrophotometer (Model number CM-2600d). A high energy xenon flash light is reflected by the surface of the sample, Chapter Three: Exposure o f Decking Samples 64 diversified by a high resolution monolithic dual beam monochromator and analysed (Minolta 2001). The CIE values L*, a* and b* were recorded. L* axis represent the lightness, a* and b* are the chromaticity coordinates; +a* for red, -a* for green, -b* for yellow and +b* for blue. The L * varies from zero (black) to 100 (white). 3.2.8 Statistical Analysis The experiment used factorial principles to assess the effect of filter type and exposure time on the checking and colour of samples. Random effects arise due to sample variation in wood properties, location of racks and samples and elapsed time between measurements. Analysis of variance (ANOVA) was performed to assess the effects of fixed (exposure time and filter type) and random factors on checking. Statistical computation was performed using Genstat (v. 4.2.1). Before the final analysis, diagnostic checks were performed to see if data conformed to the assumptions of A N O V A , i.e., normality with constant variance, and as a result of such checks data for check sizes was converted to natural logarithms before analysis. Results are presented graphically and a least significant difference bar on each graph can be used to estimate the significance of differences between individual means. Chapter Three: Exposure o f Decking Samples 65 3.3 Results 3.3.1 Checking Effect of Wavelengths on Checking The ANOVA revealed that there were highly significant (p < 0.001) effects of filter type on the quantity and on the total length, total width, total area and shape of checks. 120 l.s.d. = 14.67 100 -\ CD OJ 80H E =3 « 60 o 40 H 20 Filter 1 No blocking Filter 2 Filter 3 Filter 4 Filter 5 290-320nm 290-400nm 290-700nm 290-3000nm blocked blocked blocked blocked Filter Type Figure 3.12: The effect o f filter type on check numbers. Chapter Three: Exposure of Decking Samples 6 6 7.5 7.0-QJ o QJ O o 6.5 6.0-l.s.d. = 0.2328 i n r i r Filter 1 Filter 2 Filter 3 Filter 4 Filter 5 No 290-320nm 290-400nm 290-700nm 290-3000nm blocking blocked blocked blocked blocked Filter Type Figure 3.13: The effect o f filter type on total check length. 1808 1097 665 0) O 403 -»-» O TO 4-* O 2 . 5 -2.0H 1.54 1.0-0 . 5 0 -l.s.d. = 0.1441 1.350 1.284 E E 5 1.161 ? o QJ sz O 75 I- 1 .105 | ° 1.051 Fi l ter 1 No blocking Filter 2 290-320nm blocked Fi l ter 3 290-400nm blocked Fi l ter 4 290-700nm blocked Fi l ter 5 290-3000nm blocked Fi l ter T y p e Figure 3.14: The effect o f filter type on total check width. Chapter Three: Exposure o f Decking Samples 67 E E c 03 < O <D o TO 4—» o 4.5H 4.0-3.5^ 3 .0H 2.5-l.s.d. = 0.3207 \- 1.568 r- 1.492 1.419 h 1.350 1.284 Filter 1 No blocking Filter 2 290-320nm blocked Filter 3 290-400nm blocked Filter 4 290-700nm blocked Filter 5 290-3000nm blocked Filter Type Figure 3.15: The effect o f filter type on total check area E E_ (LP sz O -t—l O 1.733 o. .c CD OJ o 1.648 Filter 1 No blocking Filter 2 290-320nm blocked Filter 3 290-400nm blocked Filter 4 290-700nm blocked Filter 5 290-3000nm blocked Filter Type Figure 3.16: The effect o f filter type on check shape Chapter Three: Exposure of Decking Samples 68 The results from this experiment clearly showed that samples exposed under filter 1 developed significantly more checks than samples exposed under filters 2, 3, 4 and 5. For example, on average 105 checks developed at the surface of samples exposed under filter 1, whereas the comparable value for samples exposed under filter 2 was 67. Samples exposed under filter 2 developed significantly more checks than those exposed under filter 4 and 5 (37 checks), but there was no significant difference between check numbers in samples exposed under filter 2 and 3 (50.5). There was also no significant difference in check number in samples exposed under filters 3, 4 and 5 (Figure 3.12). Total check length for samples exposed under filter 1 (1643 mm) was significantly higher than that of samples exposed under filter 2 (1030 mm). Furthermore, there was a significant difference in check length between samples exposed under filter 3 (808 mm) and samples under filter 4 (463 mm). There was no significant difference in total check length between the samples exposed under filter 2 and those under filter 3, and between samples exposed under filter 4 and filter 5 (Figure 3.13). There was a continuous and significant decline in total check width (Figure 3.14) and total check area (Figure 3.15) for the samples exposed under filter 1, 2, 3 and 4. There was no significant difference in check size parameters for samples exposed under filters 4 and 5. The average total check width was 12 mm for the samples under filter 1, 6.7 mm for samples under filter 2, and 4 mm for samples under filter 3. The greatest difference in check widths occurred for samples exposed under filters 3 and 4 (Figure 3.14). Chapter Three: Exposure o f Decking Samples 69 Total check area decreased significantly from 216 mm2 for samples exposed under filter 1 to 149 mm2 (filter 2) and further to 95 mm2 for samples exposed under filter 3. The most significant difference in total check area was for samples exposed under filter 3 and samples exposed under filter 4 (65 mm , Figure 3.15) The check shape (length/width ratio) parameter increased with increased blockage of wavelengths. The only significant difference for samples exposed under different filters was observed for samples exposed under filters 2 and 3 (Figure 3.16). Effect of Exposure Time on Checking There was a significant (p = 0.002) effect of exposure time on the quantity of checks after weathering, as expected. There were also highly significant (p < 0.001) effects of exposure time on the total length, total width, total area and shape of checks. There was no significant effect of the interaction between filter type and time on checking and hence results for the effect of exposure time on checking area averaged across filter types, in accord with the factorial design of the experiment. Chapter Three: Exposure o f Decking Samples 70 Jun - Aug Sep - Nov Dec - Mar 12 24 Exposure Time (Weeks) Figure 3.17: The effect o f exposure time on check number Jun - Aug Sep - Nov Dec - Mar 1095 898 735 602 493 403 12 24 Exposure Time (Weeks) Figure 3.18: The effect o f exposure time on total check length OJ OJ o TO 4-' O Chapter Three: Exposure o f Decking Samples 71 Jun - Aug Sep - Nov Dec - Mar 12 24 Exposure Time (Weeks) Figure 3.19: The effect o f exposure time on total checks width 4.2-4.0-E £ 3.8-QJ < 3.6-o TO O 3.4-3.2 H 3.0-J u n - A u g , Sep - Nov , D e c - M a r l.s.d. = 0.2484 12 24 36 Exposure Time (Weeks) Figure 3.20: The effect o f exposure time on total check area 66.69 54.60 44.70 £ TO CD 36.60 ^ o 29.96 J2 o 24.53 20.09 Chapter Three: Exposure o f Decking Samples 72 J u n - A u g , S e p - N o v , D e c - M a r 270.43 244.69 h 221.41 h 200.34 QJ Q. TO CO 181.27 o h 164.02 148.41 12 24 Exposure Time (Weeks) Figure 3.21: The effect o f exposure time on check shape The number, total length, total width and total area of checks increased significantly after the first 12 weeks of exposure and then again after 24 weeks (Figure 3.17 - Figure 3.21). After that the values remained almost unchanged for the rest of the exposure time. During the first period of exposure 47.7 checks developed in the desamples. After 24 weeks exposure the number of checks increased significantly to 64.1. It remained almost unchanged until the end the test (Figure 3.17). An average number of 113 checks were observed on the surface of the fully exposed controls after 24 weeks of weathering and 132 checks after 36 weeks. Total check length increased significantly from 626 mm after 24 weeks of exposure to 873 mm after a further 12 weeks of weathering (Figure 3.18). For the fully exposed Chapter Three: Exposure o f Decking Samples 73 samples average total check length was 2285 mm after the first period of exposure and 2823 mm after the second. There was a significant increase in total check width from 2.8 mm to 5.1 mm after 12 and 24 weeks exposure, respectively (Figure 3.19). The total width of checks that developed at the surface of the fully exposed controls after 12 weeks exposure was 14.8 mm and after 24 weeks exposure, 23.7 mm. Total check area increased from 25.1 mm2 12 weeks exposure to 46.9 mm2 after 24 weeks exposure (Figure 3.20). After the same periods of exposure, the check areas in fully exposed samples were 184.2 mm2.and 1423.5 mm2, respectively. The only significant difference in check shape was observed after 12 and 24 weeks exposure and when the check shape parameter decreased from 245 to 187 (Figure 3.21). Colour The ANOVA on colour revealed that there were highly significant (p < 0.001) effects of the interaction of filter type and exposure time on the lightness (L*), green-red parameter (a*) and blue-yellow parameter (b*) of the surface of exposed specimens. Chapter Three: Exposure o f Decking Samples 74 Chapter Three: Exposure o f Decking Samples 75 T — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — r 2 6 10 14 18 22 26 30 34 Time (weeks) Figure 3.24: The effect o f the interaction o f filter type and exposure time on the yellow-blue colour parameter (b*) Analysis of the L* colour parameter clearly showed a significant decrease in lightness for all samples after 36 weeks of exposure. The decrease in lightness was greatest during the first 10 weeks of exposure for those samples exposed under filter 1 and filter 2. For example the L* parameter for samples exposed under filter 2 decreased from an initial value of 82.2 to 59.8. After 14 weeks exposure the lightness of the same samples decreased more modestly from 59.9 to a final value of 52.54. For the first 28 weeks the lightness of samples exposed under filter 3 decreased continuously from an initial value of 81.37 to 51.57. After that lightness did not change significantly. For the first 26 weeks the samples under filters 4 and 5 showed a steady decrease in lightness, which was less rapid than that of samples exposed under filters 1, 2 and 3. The lightness of the samples exposed under Chapter Three: Exposure o f Decking Samples 7 6 filter 4, for example, decreased from 81.70 prior exposure to 56.85 after 26 weeks and then did not change significantly (Figure 3.22). The redness of all samples increased during the first twelve weeks of exposure. The samples exposed under filter 1 showed the greatest increase. For these samples the a* parameter rose from 5.7 prior to exposure to a value of 11.3 in Week 6. Then it decreased slightly to 10 in Week 12. Changes in the green-red-colour of the other samples also occurred, but with increasing blockage of solar radiation the increases in redness were less pronounced. Samples under filter 2, for example reached a maximum a*-value of 9.7 in Week 8 while the redness of samples exposed under filter 3 grew to a maximum of 9.2 in Week 12 and samples exposed under filters 4 and 5 developed a maximum a*-value of 7.9 in Week 8 and 7.5 in Week 12, respectively. In week 14 all a* values decreased rapidly to 6, before increasing again to a peak of 7 for filter 1 and about 8.1 for samples under the remaining filters. During the following weeks the redness of the samples exposed under filters 1, 2 and 3 decreased steadily to values of between 3.1 (filter 1) and 4.2 (filter 2). The samples.under filter 4 and 5 showed slightly lower a*-values of 7.2 and 7.5, respectively in Week 20, which only decreased slightly on further exposure (Figure 3.23). The values for the yellow-blue (b*) parameter for the samples exposed under filters 1, 2 and 3 increased during the first four weeks of exposure from approximately 5.5 prior to exposure to 34.4 (filter 1), 30.3 (filter 2) and 28.2 (filter 3), respectively. They then decreased steadily, with the values for samples exposed under filter 1 showing the greatest decline to a minimum of 20.5 in Week 36. The b* values of the samples under filters 2 and Chapter Three: Exposure o f Decking Samples 77 3 decreased to final b*-values of 3.1 and 4, respectively. During the first eighteen weeks of exposure the b* values of the samples exposed under filters 4 and 5 did not change significantly. Thereafter they decreased slightly between Week 18 and Week 29. They then stayed almost unaltered for the remaining exposure time (Figure 3.24). 3.3.2 Weathering Data Figure 3.25 clearly shows that during the first 12 weeks of exposure, which covered the summer months of June, July and August, the average (monthly) temperature increased continuously from 16.5°C up to a high of 19.3°C. There was a steady decrease in temperature in the following sixteen weeks down to 3.8°C in late December, Then the average temperature increased slightly to 6.6°C until Week 32 before it fell again to an average of 5.0°C during the last four weeks of exposure. Jun Jul Aug Sep Oct Nov Dec Jan Feb 2o _i I I I I I I I I i _ u~i 1 1 1 1 1 1 1 1 r~ 8 16 24 32 Exposure Time (Weeks) Figure 3.25: Average temperature during the exposure trial Chapter Three: Exposure o f Decking Samples 78 25CH | 200 CO CD 150-9z 100-TO -J—I O 50-0-J u n | J u l f A u g f S e p , O c t , N o v , D e c , J a n , F e b T T T T T T T T 4 8 12 16 20 24 28 32 36 Exposure Time (Weeks) Figure 3.26: Total precipitation during the exposure trial During the first 12 weeks of exposure the total precipitation was low with a minimum of 12 mm after eight weeks. A sudden increase occurred after 16 weeks, and a maximum rainfall of 149.6 mm was measured after 24 weeks. During the following four weeks the total precipitation was slightly less, before it reached a maximum of 285 mm in mid January/early February (Week 32). The final four weeks of the exposure trial were once again very dry (Figure 3.26). During the first 12 weeks of exposure the samples experienced about 11.5 hours of strong rainfall (amount/half hour > 0.6 mm). The next period was characterized by 244 hours and the last period by 131.5 hours of strong rainfall. Chapter Three: Exposure of Decking Samples 79 7 10 = _ i a> _Q to 6 105H 5 10 = 4 10°H 3 10" > 2 105 1 10" 0-Jun - Aug Sep - Nov Dec - Mar • Filter 1 El Filter 2 0 Filter 3 0 Filter 4 M Filter 5 12 24 36 Exposure Time (Weeks) Figure 3.27: Visible light measured during the exposure trial Jun - Aug Sep - Nov Dec-Mar 2.5 10" 2 10 -CN E 3 1.510 s-CT) > ZD 1 10° 5 10"H 0-• Filter 1 El Filter 2 El Filter 3 El Filter 4 CD Filter 5 12 24 26 Exposure Time (Weeks) Figure 3.28: U V light measured during the exposure trial Filters 1, 2 and 3 are penetrated by visible light to almost the same degree, while only a negligible amount of visible light reaches the surface of those samples exposed under Chapter Three: Exposure o f Decking Samples 80 filters 4 or 5. During the summertime (first twelve weeks) the amount of energy in visible light, which penetrated filter 1 was highest with 697 MJ/m . It then decreased significantly during the second period to 163 MJ/m and again in the third period to 143 MJ/m . The energy in visible light under filters 4 and 5 ranged between 29 and 5.7 MJ/m for each period of exposure (Figure 3.27). During the first twelve weeks of exposure the amount of UV energy received by samples was highest. The UV irradiation reached a maximum of 241 kJ/m for samples exposed under filter 1. Then the amount of UV energy decreased significantly between the twelfth and the twenty-fourth week when it reached a value of 39.7 kJ/m2. The decrease of energy was less significant between Week 24 and Week 36. During this time the amount of UV energy penetrating filter 1 was 22.9 kJ/m . Since UV-B is the most energetic component of the solar spectrum, blocking this range of wavelengths as was done by filter 2 2 2 resulted in a significant reduction in total energy per m . Only 29.8 kJ/m of UV energy reached the surface of samples exposed under filter 2 during the first twelve weeks of exposure. The UV received by samples under filter 2 decreased to 5.6 kJ/m during the second and 3 kJ/m2 during the third period of exposure, respectively. Samples under filter 3 were exposed to 14.8 kJ/m2 during the first period of exposure. The amount of energy received under this filter decreased then to negligible values of 3.6 kJ/m2 (2nd Period) and 2 2 2 1.5 kJ/m (3rd Period), respectively. Filters 4 and 5 only allowed 9.3 kJ/m and 8.3 kJ/m , respectively, to penetrate during the first twelve weeks. The values then decreased to negligible levels of below 1 kJ/m in the last period of exposure (Figure 3.28). Chapter Three: Exposure of Decking Samples 81 3.4 Discussion A strong relationship was established between the range of wavelengths the lodgepole pine decking samples were exposed to and the development of checks. Blocking UV-B led to the greatest reduction in the quantity of checks. It also reduced the total width, total length and total area of checks. UV-B contains the shortest (290 - 315 nm) and hence the most energetic wavelengths in the solar radiation reaching the earth's surface and is responsible for the most severe photochemical degradation of synthetic materials such as rubber and plastics (Anonymous 2001; Andrady et al. 2003). Andrady and Torikai (1998) also found that mechanical pulp exposed to wavelengths of 290 nm yellowed the most. The inverse correlation between ozone concentration in the atmosphere and transmission of UV light to the earth's surface has been demonstrated in many studies (McKenzie et al. 2003). Hence any reduction in the stratospheric ozone concentration and resulting increase in the UV-B component of terrestrial sunlight will tend to decrease the service life of wood and the other synthetic materials mentioned above (Andrady et al. 2003). The main constituents of wood are cellulose and lignin (Stamm 1964). Lignin strongly absorbs UV light with a distinct maximum at 280 nm and a tail extending beyond 400 nm into the visible part of the spectrum (Kalnins 1966), which makes it highly susceptible to degradation by UV-B. This suggests that lignin degradation is, in part, responsible for increased checking of samples exposed under filter 1 compared to those exposed under the other filters. The findings for colour tend to support this hypothesis. For those samples exposed to UV-B, lightness decreased during the first period of exposure, which was dominated by high amounts of UV-B, high temperatures and little precipitation. Chapter Three: Exposure o f Decking Samples 82 At the same time the samples became redder. Sell and Feist (1986) suggested that the darkening and yellowing of light coloured woods, including most coniferous species, was due to the accumulation of photo-degraded lignin. When the rainy season started during the second exposure period, the degraded lignin was probably leached out from the surface of the samples and hence they became lighter as the concentration of cellulose at the surface increased. Evidence for lignin degradation was obtained using FTIR spectroscopy. Samples exposed to UV-B under filter 1 did not show a peak at wavenumber 1510 cm"1, which indicates delignification. It is likely that the cell structure fails in areas high in lignin such as the middle lamella (Borgin et al. 1975) and this causes the wood structure to split open between the fibres (Borgin 1971; Turkulin and Sell 2002). In support of this suggestion observations made under the SEM revealed that the surfaces of samples exposed to UV-B were the most defibrated of all samples. Rays opened up and caused voids within the cell structures. The latter finding is in accordance with Evans (1989b), who concluded that at tangential longitudinal surfaces macroscopic checking was caused by rapid degradation of ray tissue and subsequent enlargement of the resulting voids. It may become increasingly important to understand the breakdown of wood due to UV-B because of the thinning of the ozone layer, which will lead to increased terrestrial UV-B levels. Borkowski (2000), for example, found an increase in UV-B of 6.1% ± 2.9% per decade over the period 1976 to 1997 in all weather conditions, a trend, which is likely to be observed to various degrees for most parts of the World. Chapter Three: Exposure o f Decking Samples 83 UV-A on the other hand contains longer (315 - 400 run) and less energetic wavelengths than UV-B. It causes tanning and. pigmentation of the human skin (Anonymous 2001) and had the most distinct effect on total check width and check shape. Blockage of this part of the solar spectrum resulted in thinner checks, which were generally distinct from wide, short checks (Figure 3.16). The reason for this observation may be found at the molecular level. Cellulose, which is the main constituent of the longitudinal orientated fibrils, is degraded to a lower degree by UV-A than by the highly energetic UV-B component of the spectrum. Hence the checks could expand in length more easily. Since UV-A is less energetic, it predominantly degrades lignin, which acts as "glue" between the fibrils. The degradation of lignin by UV-A leads to the same colour changes and disappearance of the lignin peak in the FTIR spectrum as shown for samples exposed to UV-B. Lacking lignin, which kept the fibrils together across the grain, the checks grew wider. At the microscopic level UV-A caused the enlargement of rays as noted for samples exposed to UV-B, but changes were less pronounced. When UV-A was blocked the cell structure of exposed samples stayed almost intact. The only difference between the surfaces of samples shielded from the UV radiation and the surface of unexposed samples was the presence of very small raised fibres. It is suggested that visible light and IR light alone are not sufficiently energetic to cause deterioration of cells or parts of cells, and the raised fibres were probably caused by water. Visible light is the only part of the spectrum that can be detected by the human eye and falls between 400 and 700 nm (Anonymous 2001). Blocking this part of the solar spectrum resulted in significantly shorter and slimmer checks and in a smaller total check Chapter Three: Exposure o f Decking Samples 84 area. Visible light is less energetic than UV, but it still degraded the wood resulting in decreased lightness and increasing redness as described for UV. Samples exposed to visible light only, however, showed the presence of a peak at a wavenumber 1510 cm"1 indicating that lignin was present at the surface. Thus the degradation process was slower. This finding is in accord with Derbyshire and Miller (1981), who found that visible light was photochemically active and degraded wood in combination with infrared light at about half the rate of that of the full solar spectrum. However, it appeared that degradation was less severe during the 2nd period of exposure because lightness did not increase to the same extent as samples exposed to UV. Kalnins (1966) proposed that lignin acted as a photo-sensitiser for cellulose oxidation through the range 2554 - 350 nm. This might explain the why blocking visible light resulted in shorter checks at the surface of samples. An unexpected observation was the extent to which all samples, irrespective of filter type, became darker with exposure. Blueness of the samples appeared to be caused by mould, dirt and dust which accumulated at the surface of the samples, particularly during the first ten to twelve weeks of exposure. Microfungi, identified as Aureobasidium pullulans (de Bary) Arnaud, grew on the surface of samples exposed under filters 1, 2 and 3 whereas those exposed under filters 4 and 5 only collected dust and dirt and did not become as blue. Moulds as well as dust only collected at the surface of the samples. Auerobasidium pullulans grows on the lignin breakdown products at weathered wood surfaces (Schoeman and Dickinson 1997). The moulds were more easily washed off by rain than dust and dirt. Thus b* decreased more for the samples under the translucent filters (Figure 3.24). The accumulation of dust also played a role in the decrease of lightness even for those samples Chapter Three: Exposure o f Decking Samples 85 exposed under filters 4 and 5. Additionally the surfaces of those samples may have been darkened by reflected and indirect radiation. Since there was no significant interaction between filter type and exposure time, it is probable that increases in checking with time were mainly caused by swelling and shrinking and the generation of stresses and strains at wood surfaces. Feist (1990) stated that water and solar radiation tend to operate at different times, which appeared to be the case here. In the first twelve weeks of the experiment it rained very little (Figure 3.26), dew did not condense at the surface of the samples and the temperature was almost continuously high (Figure 3.25). Thus, the samples probably dried out during this period. It is likely that most checks developed during this initial period of exposure to relieve surface stresses. The following two periods of exposure were characterised by high rainfall and low temperatures. The checking due to the initial change in the moisture content during the first twelve weeks in combination with the weakening of the surface due to photodegradation was very severe. Dimensional changes during the following 24 weeks may not have been high enough to exceed the flexibility of the check-weakened surface structure. 3.5 Conclusion The hypothesis proposed in the introduction to this Chapter that different wavelengths enhance surface checking of lodgepole pine depending on their ability to degrade wood has been confirmed. From the results it was concluded that UV-A, UV-B and visible light degraded the wood surface. The degradation of wood surfaces, which resulted in darkening and increased redness of the surface of the samples, appeared to be linked to the Chapter Three: Exposure o f Decking Samples 86 development of broader checks. Only the U V radiation, however, caused degradation of cell structures such as rays and cell walls. It can be concluded from observations here that selectively protecting wood from UV -B , U V - A and visible light using chemical means might result in less severe checking of lodgepole pine during weathering. Even the blockage of U V - B alone might lead to significantly improved behaviour of surfaces, when the natural look of decks is preferred. The effect of water on checking may have masked that of light over time. Most checks appeared to develop during summertime, possibly to relieve surface stresses, which had developed due to the high shrinkage of the samples. During the winter the surface was probably already weakened by checks, thus relieving surface stresses. Hence the checking of the samples only increased slightly. Chapter Four: Exposure o f Veneer Blocks 87 CHAPTER FOUR 4 Effect of UV-A, UV-B, Visible Light and IR Light on the Integrity of Micro veneers formed into Veneer Blocks 4.1 Introduction In Chapter 3 it was shown that the surface checking of solid lodgepole pine specimens was greater when blocks were exposed to light containing more energetic wavelengths, particularly UV-B. It was suggested that this was due to weakening of the surface layers of wood specimens due to photodegradation of wood thus reducing the ability of surface layers to resist stresses resulting from dimensional movement of wood. The degree and depths to which wood surfaces were degraded under the different filters was not examined. Reports in the literature indicate that the effects of weathering do not penetrate deeply into wood (Stamm 1963, Evans 2001) because ultraviolet light only penetrates up to 75 um into wood. Hence, chemical changes in wood due to solar radiation are restricted to a very thin surface layer (Browne and Simonson 1957) and the wood a few millimetres under the surface is essentially unchanged and unaffected (Feist 1982). The technique of using thin veneers in order to quantify the degradation of wood by solar radiation and weathering is well-established and has been used by many authors (Kalnins 1966, Raczkowski 1980, Derbyshire and Miller. 1981 and 1995, Evans 1988, Evans and Banks 1988 and 1990, Evans and Schmalzl 1989, Evans et al. 1992, Kataoka et Chapter Four: Exposure o f Veneer Blocks 88 al. 2004, Turkulin and Sell 2002 among others). It has been shown that the measurement of loss of tensile strength in veneers exposed to solar radiation is a consistent and reliable means of determining photodegradation rates for wood (Derbyshire et al. 1995). Losses in tensile strength of wood veneers exposed to the weather can be related to depolymerisation of cellulose (Derbyshire and Miller 1981).Weight loss, on the other hand, can detect small but significant differences in the degradation of materials and is widely used in the assessment of the weathering of polymeric materials (Davis 1981; Qayyum and Davis 1984). Weight losses of wood veneers during weathering can be related to degradation of lignin and hemicelluloses. Hence, measurement of losses of tensile strength and weight of wood veneers during weathering can be used to estimate degradation of cellulose and cellulose respectively (Evans and Schmalzl 1989). Thin wood veneers have been generally used for short-term exposure (about 30 days). In the experiment described in this chapter, veneers were clamped together to form "veneer blocks", which were exposed under filtered light (as described in Chapter 3) for 36 weeks. This method was used to estimate the effect of light at various depths. In this chapter, the hypothesis is assessed that degradation of wood blocks, as assessed by weight and tensile strength losses of veneers under filtered light will increase in relation to the exposure of blocks to progressively shorter wavelengths. Chapter Four: Exposure o f Veneer Blocks 89 4.2 Materials and Methods 4.2.1 Lumber Lodgepole pine was chose as sample material because it is the is the most widespread species in British Columbia and commonly used for decking (Lotan and Critchfield 2005). Five lodgepole pine logs were debarked, milled and kiln dried as described in Chapter 3 (Section 3.2.1). Decking samples were cut out of the dried boards (Section 3.2.2) and one of the eight decking samples obtained from each log was randomly chosen for further processing into thin wood veneers. 4.2.2 Veneer Samples Straight grained wood from each log, free of all visible defects, was cut into smaller blocks measuring 100 mm (longitudinal) x 25 mm (radial) x 18 mm (tangential) using an Altendorf (Model F 45 ELMO) circular saw and an Omega chop saw. Thin wood veneers approximately 70 um thick were microtomed from radial longitudinal surfaces of blocks, after they had been softened by vacuum impregnation with distilled water at room temperature (Evans 1988). The resulting strips were air dried under end restraint in glass jigs for eight hours and then stored in open glass containers in a conditioning room maintained at 20°C ± 1°C at 65% ± 5% r.h. for seven days. The thickness of veneers was measured with a digital micrometer (Lorentz & Wettre HWS 5781, Figure 4.1). Their weight was obtained using an A & D balance Model GR-200 from B.C. Scale Co. Ltd. Figure 4.1: Digital micrometer (Lorentz & Wettre HWS 5781) In order to construct "veneer blocks" forty veneers were batched together in the same order as they had been microtomed from the block. These were clamped together at both ends with 1" wide fold-back clips (Figure 4.3). Twenty-one veneer-blocks were prepared from each log. Chapter Four: Exposure of Veneer Blocks 91 Figure 4.2: Method of constructing blocks from veneers cut from decking samples Figure 4.3: Veneer blocks 4.2.3 Exposure Eighteen veneer blocks from each log were exposed horizontally to the weather. Three blocks were exposed to the weather under each filter. The remaining three blocks for each log were kept in the conditioning room for the duration of the experiment to act as unweathered controls. Samples were exposed for 12, 24 and 36 weeks (as described in Chapter Four: Exposure o f Veneer Blocks 92 Chapter 3, Section 3.2.3). After each period of exposure, the samples were conditioned for 14 days and their weight, thickness and tensile strength were measured. 4.2.4 Racks and Filters The racks provided a means of supporting the veneer blocks beneath the same filters that were described in Chapter 3 (190 mm) (Section 3.2.4). Five racks were built and each contained one of the filters. The racks allowed rain water, air and atmospheric pollutants (dirt and fungi) to come into contact with the samples. The same type of filter used on top of the racks was also screwed onto the sides of the rack to prevent solar radiation from reaching the surface of the samples from the front or back of the racks. Custom made, dark brown coated aluminium sheets, supplied by Main Sheet Metal Works Ltd. in Vancouver, Canada, directed rainwater at an angle of 25° under the filters and onto the surface of the veneer blocks. " Y ; r r t -- T T " t:^r::::::::::::::::r:: j : : ; : ^ : : : : : : : : : : : : : : : — : : Figure 4.4: Top view and side view a rack with filter that was used to expose veneer blocks to the weather Chapter Four: Exposure of Veneer Blocks 93 Figure 4.6: Rack with filter for veneer blocks Veneer blocks were assigned at random to the five racks, each of which contained a different filter. The blocks were held in place by nails on one side and duct tape on the side nearest to the aluminium sheets. The veneer blocks, which were fully exposed to the Chapter Four: Exposure of Veneer Blocks 94 weather without any filter, were placed on a separate horizontal rack 240 mm in height (Figure 4.7). s>-! 1 r'~' Ln A-<•• - H i u If! • fen— i Y T ~ - : — * \ -v L i r-- , i 1 ! - A — 1 i jf" JO H H H H H HH 3 — f . y . * n i H HH Figure 4.7: Top, side and front view of rack (without filter) used to expose veneer blocks Figure 4.8: Rack without filter for veneer blocks 4.2.5 Measurement of UV Light and Visible Light Measurement of the total energy per cm (UV-A, UV-B and visible light) received by samples was described in Chapter 3 (Section 3.2.6). Chapter Four: Exposure o f Veneer Blocks 95 4.2.6 Measurement of Weight and Tensile Strength Losses of Veneers After exposure, all veneer blocks were conditioned (as above) for 14 days. One block exposed under each filter, for the different time periods and the unexposed and freely exposed controls was chosen at random. The weight and thickness of each veneer within the blocks were measured as described in Section 4.2.2. Tensile strength tests were carried out using a Pulmac paper tester (Pulmac Troubleshooter HWS 5786) at zero-span using a clamping pressure of 60 psi (Figure 4.9). Figure 4.9: Pulmac Troubleshooter H W S 5786 4.2.7 Experimental Design and Statistical Analysis Twenty-one veneer blocks each containing forty individual veneers were cut from each of five boards obtained from different lodgepole pine logs. Veneer blocks cut from different logs were exposed to filtered light in separate weathering racks. Veneer blocks exposed under different filters were removed from each of the five racks following 12, 24, and 36 weeks exposure. Therefore for each exposure period five replicate blocks were Chapter Four: Exposure o f Veneer Blocks 96 removed and individual veneers within each wood block were tested for weight and tensile strength losses (as described above). Results from the exposure trials were subjected to separate analyses of variance to determine the effect of exposure (filter type), veneer position (1-40) and exposure time on weight and tensile strength losses of veneers during weathering. A measure of weight loss having appropriate statistical properties (normality and constant variance) was the natural logarithm of the ratio of mass of veneers after weathering to the mass of veneers before weathering. Similarly tensile strength losses are expressed as the natural logarithm of the ratio of the tensile strength of veneers after weathering to the tensile strength of matched unexposed controls. Preliminary analysis of data that included the results for all forty veneers resulted in a distribution of residuals that was not normal, because veneers beyond a depth of 200 microns were not affected by weathering. Hence, the final analysis only included data for the four uppermost veneers exposed to the weather. Weight and tensile strength data are plotted graphically in logarithmic form and the effect of individual factors on weight and tensile strength losses can be compared using a least significant difference bar (p < 0.05) included on each graph. Back transformed (ex) values are included on the Y2 axes of each graph. 4.3 Results 4.3.1 Effect of Filter Type and Veneer Position Analysis of variance showed that there were highly significant (p < 0.001) interactions of filter type and exposure time on the weight losses of veneers. A highly Chapter Four: Exposure of Veneer Blocks 97 significant (p < 0.001) effect was also observed of the interaction between filter type and veneer position on weight losses. 0.1-0.05-o Q: t 0.861 •§ T i — r Position 1 Position 2 Position 3 Position 4 Veneer Position Figure 4.10: The interaction o f filter type and veneer position on weight losses o f veneers 0.2 Si S -0.8-OJ -1--1.2-| l.s.d. = 0.04836 Position 1 • Filter 1 No blocking • Filter 2 290-320nm bid. 290-400nm bid. | Filter 4 290-700nm bid. • Filter 5 290-3000nm bid T 1 1 Position 2 Position 3 Position 4 Veneer Position 1.221 c 0.549 | -0.449 £ h 0.368 0.301 Figure 4.11: The interaction o f filter type and veneer position on tensile strength losses o f veneers Chapter Four: Exposure o f Veneer Blocks 98 There was a highly significant difference in weight loss of the uppermost veneer (position 1) exposed under filter 1 and similar veneers exposed under filters 2, 3, 4 and 5. For example the difference in weight ratio between veneers exposed under filter 1 and filter 2 was 0.23. The weight losses of the uppermost veneers exposed under filter 2 and 3 were significantly higher than those of the veneers exposed under filters 4 and 5. The average weight ratio of the uppermost veneers exposed under filter 2 was 0.95, whereas the weight ratio of the uppermost veneers exposed under filter 4 was 1.09. The weight loss of the veneers below the exposed surface (position 2) and exposed under filter 1 was significantly lower than that of the uppermost veneer. There was a significant difference in weight losses of veneers in position 2 and exposed under filters 1 and 2. There were no significant differences in weight losses of veneers exposed in position 2 under filters 3, 4 and 5. Furthermore, there were no significant differences in weight losses observed between any of the veneers exposed in positions 3 and 4, irrespective of filter type (Figure 4.10). There were significant differences in the losses of tensile strength of the uppermost veneers exposed under filter 1, 2, 3 and 4. No significant difference, however, was observed in the tensile strength losses of the uppermost veneers exposed under filter 4 and filter 5. The largest losses in tensile strength (tensile strength ratio of 0.338) were observed for veneers exposed under filter 1. The greatest differences in tensile strength losses were between veneers exposed under filter 1 and filter 2 (0.610) and between veneers exposed under filter 3 and 4 (0.381). Tensile strength losses were significantly lower for the veneers immediately below the surface and exposed under filter 1, 2 and 3 compared to those of the Chapter Four: Exposure of Veneer Blocks 99 uppermost veneer. There were significant differences in tensile strength losses of the subsurface veneer (position 2) exposed under filter 1 and 2 and 3 and 4. The third veneer showed significant differences in tensile strength losses when exposed under filters 1 to 4. Overall, however, the strength losses of the third and fourth veneers were small irrespective of filter type. In some cases small increases in tensile strength were observed for veneers exposed under filter 5 (Figure 4.11). 4.3.2 Effect of Exposure Time Analysis of variance revealed that the interaction of filter type and exposure time, had highly significant effects (p < 0.001) on weight ratio. It also showed that veneer position and exposure time, had highly significant effects (p < 0.001) on tensile strength of veneers during weathering. 0.1--0.1 i f -0.2H oc £ -0.3H ra I -0.4 -0.5H -0.6 J u n - Aug Sep - Nov . Dec - March l.s.d. = 0.04836 T" • Filter 1 No blocking • Filter 2 290-320nm bid. S Filter 3 290-400nm bid. |- 0.607 El Filter 4 290-700nm bid. • Filter 5 290-3000nm bid. 1 i 1.105 1.000 h 0.905 F- 0.819 OC h 0.741 £ O) 0L> 0.670 ^ 0.549 36 12 24 Exposure Time (Weeks) Figure 4.12: The interaction of filter type and exposure time on weight losses of veneers Chapter Four: Exposure of Veneer Blocks 1 0 0 Jun - Aug Sep - Nov Dec - Mar 12 24 Exposure Time (Weeks) Figure 4.13: The interaction between veneer position and exposure time on tensile strength losses of veneers Veneers exposed under filter 1 lost significantly more weight than veneers exposed under filters 2, 3, 4 and 5 irrespective of exposure time. For example, on average, veneers exposed under filter 1 had a weight ratio of 0.59, whereas the average weight ratio of veneers exposed for 12 weeks under filter 2 was 1.0. The weight ratio of veneers exposed for 12 weeks under filter 1 was 0.82 and the comparable figure for veneers exposed to March was 0.87. There was no significant difference in weight ratio between the veneers exposed for 12 weeks under filters 2, 3, 4 and 5. The weight losses of veneers exposed for 24 weeks under filter 2 were significantly higher than those of veneers exposed under filters 3 to 5. For example, the difference between the veneers exposed under filter 2 and 3 was 0.05. Veneers exposed for 24 weeks under filter 3 lost significantly more weight than Chapter Four: Exposure o f Veneer Blocks 101 veneers exposed under filter 4. Weight losses of veneers exposed for 36 weeks under filter 3 were significantly higher than that of veneers exposed under filter 5 (Figure 4.12). Tensile strength losses of the uppermost veneer increased significantly with exposure time from 0.739 after the first twelve weeks exposure to 0.589 after 36 weeks. There was a significant difference of 0.261 between the tensile strength ratio of the uppermost veneers and the underlying veneer (position 2) after 12 weeks exposure. After 24 weeks exposure significant differences were observed between tensile strength losses of the uppermost and underlying veneer and between veneers in position 2 and 3 (0.140). After 36 weeks exposure the most significant difference in tensile strength losses occurred between veneers on position 1 and position 2 (Figure 4.13). 4.4 Discussion Measurement of weight and tensile strength losses of veneers exposed to the weather is an established means of quantifying the surface photodegradation of wood (Derbyshire and Miller 1981; Evans 1988). Losses in weight of wood veneers during weathering occur due to degradation of lignin and hemicelluloses (Evans 1988). Highly significant weight losses were observed for veneers exposed to all wavelengths (filter 1), which indicates substantial degradation of the lignocellulosic matrix, as expected. The degradation of lignin at wood surfaces exposed under filter 1 was confirmed by FTIR spectroscopy on solid wood samples in Chapter 3. Blocking UV-B (filter 2) alone was sufficient to reduce weight losses of surface veneers during the summer months and reduce weight losses of sub-surface veneers. FTIR spectroscopy on solid wood, however, showed complete Chapter Four: Exposure o f Veneer Blocks 102 delignification of wood exposed under both filter 1 and filter 2. This discrepancy might be caused by the ATR technique used to obtain FTIR spectra which obtains at a depth of 1.2 um. In contrast measurement of veneer weight losses indirectly assessed lignin loss across a much thicker sample. It can be inferred that the UV-B component of the spectrum plays an important role in the surface delignification of wood exposed to the weather, although it is clear that some delignification occurred on exposure to UV-A and visible light. UV-B is the main factor responsible for the photodegradation of both synthetic materials such as rubber and plastics (Andrady et al. 2003). The results clearly showed that UV-B had the most significant effect on the weight and tensile strength of the upper veneer layer, while in comparison the effect of UV-A on the weight and on tensile strength of veneers was small. Hon and Ifju (1978) stated that UV light only penetrated about 75 um into wood. Since the veneers were on average 75 urn thick the effect of all wavelengths (filter 1) and UV-B (filter 2) decreases significantly for subsurface veneers (position 2 and position 3). Visible light only affected the weight of the uppermost veneer and tensile strength of the top two veneers. This suggests that insufficient visible light was present to degrade wood beyond a depth of 150 um despite the fact that Hon and Ifju (1978) found that visible light penetrates about 200 um into wood, and in a recent study Kataoka et al. (2004) detected visible light depth beyond 200 um. These studies of the penetration of light and depth of photodegradation in wood used wood blocks (Browne and Simonson 1957; Hon and Ifju 1978; Park et al. 1996). This study used blocks consisting of assembled veneers with a different structure to that of solid blocks. This could lead to different wetting/drying behaviour because water can penetrate Chapter Four: Exposure of Veneer Blocks 103 and evaporate more easily from the sides of the veneer blocks. Thus, degradation products could be more easily leached out and transported within the block. Hence, the redistribution of photodegradation products by water, which can affect the penetration of light into wood and the growth of micro-organisms, would be different in composite veneer and solid wood blocks. The effect of UV-B is accelerated with higher temperatures during the summer months as it is for the entire solar spectrum (Perbyshire et al. 1997; Andrady et al. 2003). After 24 and 36 weeks exposure there was an unexpected decrease in the weight losses of veneers. In contrast, losses in tensile strength of veneers were positively correlated with exposure. It is very likely that the accumulation of dirt on the surfaces of veneers was responsible for lower weight losses of veneers after prolonged exposure to weathering compared to veneer exposed to the weather for 12 weeks. The contamination of veneer surfaces by dust may also explain weight gains of samples exposed under filter 4 and 5. After the second and third periods of exposure, UV-A had a significant effect on the weight losses of the veneers, as did visible light after the third period. The former finding was in accordance with the lignin degradation observed on the surface of solid lodgepole pine blocks after 36 weeks exposure. Tensile strength results indicated that the rate of degradation was rapid initially and then became slower. The rate of tensile strength loss for the uppermost veneer was greatest during the first 12 weeks exposure. Losses of tensile strength thereafter were relatively small (Figure 4.13). The rapid loss of tensile strength of the upper veneer accords with previous observations of rapid losses in tensile strength of wood veneers during exterior Chapter Four: Exposure o f Veneer B locks 104 exposure and the observations of Raczkowsky (1980) that during the summer months the photolysis of thin veneers was accelerated due to the higher intensity of solar radiation. At the structural level, Turkulin and Sell (2002) found that tensile strength changes of thin wood veneers were consistent with fractographic evidence of the structural changes in wood, as they observed detachment of cells due to cohesive failure of the middle lamella after only 4 days weathering. Marked thinning of cell walls also became apparent at the same time. The authors linked thinning of the cell wall to degradation of lignin in the S2 layer and the steady reduction in strength. Derbyshire and Miller (1981) stressed that the changes in cell structure did not accurately reflect reductions in tensile strength. The loss of tensile strength stemmed from changes at the molecular level and was mainly caused by the depolymerisation of cellulose microfibrils (Derbyshire and Miller 1981). Evans and Schmalzl (1989) supported Derbyshire and Miller's conclusion that the loss of tensile strength of thin wood veneers during weathering, resulted from light induced depolymerisation of cellulose. In a later paper, Derbyshire et al. (1996) linked the initially more rapid degradation of lignin to the loss of tensile strength. Yoshimoto (1972) suggested that lignin transferred UV light energy to cellulose and this transferred energy caused the photodegradation of cellulose. Solid wood subjected to weathering develops a cellulose rich layer on the surface (Borgin 1971). It has been suggested that this layer (0.13 mm thick) protects the underlying wood from photodegradation (Browne and Simonson 1957). Since the veneer blocks in this experiment were not solid, but consisted of separate individual strips of wood the uppermost veneer many have been less able to protect the underlying wood from degradation. The uppermost veneer became increasingly degraded Chapter Four: Exposure o f Veneer B locks 105 with increasing exposure, and this may explain why the subsurface veneer showed significant losses in strengthfrom 12 to 24 weeks of exposure. Losses in tensile strength of the subsurface veneer during the final exposure period from 24 to 36 weeks may have been small because the amount of photochemically active UV and visible light received by the veneers was small (Figure 4.13). 4.5 Conclusion UV-B was the most effective of all wavelengths in reducing weight and tensile strength, of microveneers assembled into composite wood blocks and exposed to the weather for prolonged periods of time. Weight and tensile strength losses indicated loss of lignin, and degradation of the lignocellulosic matrix. Degradation was most pronounced in the upper veneer exposed to light due to the small penetration of solar radiation. The effect of UV-B decreased more rapidly than the effects of UV-A and visible light with increasing depth of penetration. Since the amount of UV-B received by veneers was highest during the summer months, the rate of degradation was greatest during this period. In the following months, the rate of degradation was lower because less UV-B was received and because the upper exposed veneer was already degraded and may have protected the underlying wood to some extent. With prolonged exposure and after high rain fall the uppermost veneers of those samples exposed to UV-B were degraded entirely and this allowed radiation to affect the underlying veneers. Chapter Five: General Conclusion 106 CHAPTER FIVE 5 General Conclusions and Suggestions for Further Research 5.1 General Conclusion Lodgepole pine decking samples and micro-veneers from the same parent wood material clamped together as veneer blocks, were exposed under the same conditions for 36 weeks. It is assumed here that the results gained from these samples are complementary, in accordance with the findings of Derbyshire and Miller (1981) that the weathering behaviour of veneers did not differ from solid wood. Comparisons should, however, be made with caution. For example, the tangential surface of decking samples was exposed to the weather, whereas the microveneers were cut from the radial surface of blocks thereby exposing the radial surface of veneers. Tangential and radial surfaces do not respond to solar radiation in precisely the same way. Kuo and Hu (1991) for example, found that radial middle lamellae were less susceptible to UV light than tangential middle lamellae, which is an indication of the higher resistance of tangential surfaces to solar radiation. Also the veneers blocks exposed here were not solid, which may have influenced movement of water within blocks, altered the leaching of photodegradation products and encouraged the growth of micro-fungi on the top veneers. Chapter F ive : General Conclusion 107 UV-B had a highly significant effect on the surface integrity of both decking samples and veneer blocks. UV-B degraded lignin and the surface of decking samples became darker and redder. Accordingly, veneers exposed to UV-B showed significant weight loss. It is hypothesised that surface degradation and loss of lignin facilitated increases in width of checks at exposed wood surfaces, because lignin acts as "glue", which holds together cells and cell walls. Defibration and extensive surface degradation of wood surfaces exposed under filter 1, which transmitted UV-B was confirmed by scanning electron microscopy. The correlation between findings for decking samples and veneer blocks exposed to UV-A and visible light was less clear. UV-A had an effect on the width and shape of checks, but no significant effect on the weight and tensile strength of veneers. FTIR spectroscopy showed that lignin was completely degraded at the surface of exposed decking boards by UV-A after 36 weeks of exposure. This observation may explain why checks were wider at the surface of samples exposed to UV-A and other components of the solar spectrum (except UV-B) than at the surface of samples exposed to solar radiation in the absence of UV-A. Visible light had a significant effect on check length, width and area of checks, but the checks were smaller and less numerous than those on the surface of samples exposed to low radiation. FTIR spectroscopy showed that about half of the lignin was degraded. There was, however, a less significant effect of visible light on the veneers after 36 weeks of weathering, possibly because dust on the upper, exposed veneer shielded the wood from photodegradation. Chapter Five: General Conclusion 108 Overall, the veneer block method was not sensitive enough to examine the degradation of wood by longer wavelengths. Furthermore, the longer wavelengths may have been adsorbed by dust on the top of the veneer. The veneer blocks are also not suitable for assessing weathering over longer exposure periods, because the upper veneer degraded and was physically damaged during exposure and the blocks did not produce a protective cellulose-rich shell, as has been observed for solid wood (Borgin 1971). Most checks developed during the first period of exposure (summer), when the most significant loss of tensile strength occurred, probably because of high amounts of UV-light that were received. After the second and third period, checking as well as degradation of veneers was less severe. UV-B, UV-A and visible light are the main factors that accelerate the checking of solid lodgepole pine specimens. Blocking these wavelengths, particularly UV-B is recommended to reduce the checking of wood surfaces exposed outdoors. Compounds that absorb visible light are coloured and therefore if a natural look is preferred for decking, the blockage of UV-B should provide benefits. The combination of photoprotection compounds with water repellents should further reduce the incidence of wood checking. 5.2 Suggestions for Further Research The results gained in the present study of the effect of solar radiation on the checking of solid lodgepole pine decking samples were significant and further research involving prolonged exposure of both untreated and treated samples is recommended to analyse the long-term effect of the different components of solar radiation on checking. A parallel Chapter Five: General Conclusion 109 experiment starting time in the fall/winter instead of spring/summer would provide information on whether the sequence of weather events had an effect on checking of samples. For example, after the same total exposure time, it can be anticipated that decks first exposed in summer might check more than decks first exposed in winter because lignin would be degraded earlier and more completely in the former samples. Further research should closely monitor changes in the surface chemistry of exposed samples using FTIR spectroscopy. It is also recommended that other species, namely those used for decking such as Southern yellow pine, as well as ACQ-treated wood with or without water-repellents should be included in future experiments. Additional analysis of the veneers exposed in this study using FTIR spectroscopy would provide a deeper understanding of the chemical changes in various depths and for each length of exposure (12, 24 and 36 weeks). Information from such measurements would complement the weight loss and tensile strength measurements made here. As mentioned above, the veneer block method with forty veneers clamped together, is not suitable for assessing the degradation of wood over long time periods. As an alternative it is suggested that thinner blocks of only five veneers should be exposed for four weeks and then be replaced by new veneer blocks. Based on the findings here it is suggested that UV absorbers, which mainly absorb UV-B would have beneficial effects, on the checking of pine and this should be examined in further experiments. Other experiments should target treatments that stabilize the lignin and include water-repellents to reduce the magnitude of surface stresses that generate checks at wood surfaces. 110 References Anderson E L, Pawlak Z, Owen N L, Feist W C (1991): Infrared studies of wood weathering. Part 1: Softwoods. Applied Spectroscopy 45(4), 641-647. Andrady A L, Hamid H S, Torikai A (2003): Effect of climate change and UV-B on materials. Photochemical and Photobiological Sciences 2(1), 141-149. Andrady A L, Torikai A (1998): Photoyellowing of mechanical pulps III. Polymer Degradation and Stability 66(3), 317-322. Anonymous (2001): Acrylite acrylic sheet light transmission and reflectance technical data. [Online; cited 26th June 2004; 14:00 PST]. Available from: http://www.cyro.com/Internet/SiteContent.nsf/87E49EDDElC7AB5285256B B80070E3A9/$File/1213F+UV+trans+TB.pdf?OpenElement Arndt U, Willeitner H (1969): On the resistance behaviour of wood in natural weathering. Holz als Roh- und Werkstoff21{5), 179-188. Arnold M, Sell J, Feist WC (1991): Wood weathering in fluorescent ultraviolet and xenon arc chambers. Forest Products Journal 41 (2), 40-44. Bamber R K, Summerville R (1981): Microscopic studies of the weathering of radiata pine sapwood. Journal of the Institute ofWood Science 9(2), 84-88. Bariska M, Pizzi A, Conradie W E (1988): Structural weakening of CCA-treated timber. Holzforschung 52(5), 339-345. Belford D S, Nicholson J (1969): Emulsion additives for CCA preservatives to control weathering. Proceedings of American Wood-Preservers' Association 65, 38-51. Borgin K (1965): The effect of water repellents on the dimensional stability of wood. Norsk Skogindustri 15(11), 507-521. Borgin K (1970): The use of scanning electron microscope for the study of weathered wood. Journal of Microscopy 92(1), 47-55. I l l Borgin K (1971): The mechanism of the breakdown of the structure of wood due to environmental factors. Journal of the Institute for Wood Science 5(4), 26-30. Borgin K , Corbett K (1969): Long term stabilizing effect of water repellents on wood. Timber 6(6), 5-19. Borgin K , Corbett K (1970a): The stability and weathering properties of wood treated with various oils. Plastics, Paint and Rubber 14(3), 69-72. Borgin K , Corbett K (1970b): The stability and weathering properties of wood treated with various waxes. Plastics, Paint and Rubber 14 (4), 89-94. Borgin K , Corbett K (1971): The hydrophobic properties of bark extractives. Wood Science and Technology 5(3), 190-199. Borgin K , Parameswaran N , Liese W (1975): The effect of aging on the ultrastructure of wood. Wood Science and Technology 9(2), 87-89. Borkowski J (2000): Homogenisation of the Belsk U V - B series (1976-1997) and trend analysis. Journal of Geophysical Research 105, 4873-4878. Bosshard H H (1956): Ober die Anisotropie der Holzschwindung. Holz als Roh- und Werkstoff14(8), 285-294. Bottcher P (1975): Zum Verhalten unterschiedlich feuchtedurchlassiger Anstriche auf einigen einheimischen Holzarten bei naturlicher Bewitterung. Holz als Roh-und Werkstoff*33,116-120. Boulton S B (1884): On the antiseptic treatment of timber. Proceedings of the Institution of Civil Engineers (London) 78, pp 79-211. Browne F L (1960): Wood siding left to weather naturally. Southern Lumberman 201(2513), 141-143. Browne F L , Simonson H C (1957): The penetration of light into wood. Forest Products Journal 7(10),308-314. 112 Chandler R C (1968): Experimental guardrail posts: final report on a method of reducing post checking. Technical Report 68-4. Materials and Research Division, Oregon State Highway Department, Salem, Oregon. Chang S T, Hon D N-S, Feist W C (1982): Photodegradation and photoprotection of wood surfaces. Wood and Fiber 14(2), 104-117. Christy A G, Senden T, Evans P D (2005): Automated measurement of checks at wood surfaces. Measurement 37109-118. Cooper G A (1972): Prefreezing reduces shrinkage and alters sorption in black walnut. Forest Products Journal 22(5), 54-60. Coupe C, Watson R W (1967): Fundamental aspects of weathering. Proceedings of the Annual Convention of the British Wood Preservers' Association, pp.37-49. Crawford D M, DeGroot R C, Gjovik L R (1999): Ten-year performance of treated north eastern softwoods in aboveground and ground-contact exposures. USDA Forest Service Research Paper FPL-RP-578. Forest Products Laboratory, Madison, Wisconsin. Cui F, Zahora A (2000): Effect of water repellent additive on the performance of ACQ treated decks. The International Research Group on Wood Preservation. Doc. No. IRG/WP/40168. Davis, A (1981): Effect of climate on the weathering of polyacetal. Polymer Degradation and Stability 3(3), 187-198. Derbyshire H, Miller E R (1981): The photodegradation of wood during solar irradiation. Part I: Effects on the structural integrity of thin wood strips. Holz als Roh- und Werkstoff 39, 341-350. Derbyshire H, Miller E R, Turkulin H (1995): Investigations into the photodegradation of wood using microtensile testing. Part 1: The application of microtensile testing to measurement of photodegradation rates. Holz als Roh- und Werkstoff 53(4), 339-345. 113 Derbyshire H, Miller E R, Turkulin H (1996): Investigations into the photodegradation of wood using microtensile testing. Part 2: An investigation of the changes in tensile strength of different softwood species during natural weathering. Holz als Roh- und Werkstojf 54(1), 1-6. Derbyshire H, Miller E R, Turkulin H (1997): Investigations into the photodegradation of wood using microtensile testing. Part 3: The influence of temperature on photodegradation rates. Holz als Roh- und Werks toff 55(5), 287-291. Evans P D (1988): A note on assessing the deterioration of thin wood veneers during weathering. Wood and Fibre Science, 20(4), 487-492. Evans P D (1989a): Effect of angle of exposure on the weathering of wood surfaces. Polymer Degradation and Stability 24(1), 81-87. Evans P D (1989b): Structural changes in Pinus radiata during weathering. Journal of the Institute of Wood Science 11(5), 172-181. Evans P D (1996): The influence of season and angle of exposure on the weathering of wood. Holz als Roh- und Werkstoff'54(3), 200. Evans P D (2001): Wood products: weathering. In: Encyclopaedia of Materials: Science and Technology, pp 9716-9722. Eds. K.H.T. Bushow, R.W. Cahn, M.C. Flemings, B. Iischner, E.J. Kramer, S. Mahagen. Elsevier Science Ltd, Oxford. Evans P D (2004): Engineering the surface of treated wood. Presentation made during the Conference of the American Wood-Preservers' Association in Vancouver, Canada. 17th May, 16.20 PST. Evans P D, Donnelly C, Cunningham R B (2003): Checking of CCA-treated radiata pine decking timber exposed to natural weathering. Forest Products Journal 53(4), 66-71. Evans P D, Michell A J, Schmalzl K J (1992): Studies of the degradation and protection of wood surfaces. Wood Science Technology 26(2), 151-163. 114 Evans P D, Schmalzl K J (1989): A quantitative weathering study of wood surfaces modified by chromium VI and iron III compounds. Part 1. Loss in zero-span tensile strength and weight of thin wood veneers. Holzforschung 43(5), 289-292. Evans P D, Thay P D, Schmalzl K J (1996): Degradation of surfaces during natural weathering. Effects on lignin and cellulose and on the adhesion of acrylic latex primers. Wood Science and Technology 30(6), 411 -422. Evans P D, Wingate-Hill R, Barry S C (2000): The effects of different kerfing and center-boring treatments on the checking of ACQ treated pine posts exposed to the weather. Forest Products Journal 50(2), 59-64. Evans P D, Wingate-Hill R, Cunningham R B (1997): The ability of physical treatments to reduce checking in preservative-treated slash pine posts. Forest Products Journal 47(5), 51-55. Feist W C (1982): Weathering of wood in structural uses. In: Structural Use of Wood in Adverse Environments, pp 156-178, Eds. R.W. Meyer, R.M. Kellog. Van Nostrand Reinhold, New York. Feist W C (1983): Weathering and protection of wood! Proceedings of American Wood Preser\>ers ' Association 79,195-205. Feist W C (1990): Outdoor wood weathering and protection. In; Archaeological Wood: Properties, Chemistry and Preservation, pp 263-298. Eds. R.M. Rowell, R.J. Barbour. American Chemical Society, Washington. Feist W C, Hon D N-S (1984): Chemistry of weathering and protection. In: Chemistry of Solid Wood, pp. 401-451. Ed. R.M. Rowell. American Chemical Society, Washington. Feist W C, Mraz E A (1978): Comparison of outdoor and accelerated weathering of unprotected softwoods. Forest Products Journal 28(3), 38-43. Flaete P O, Hoibo O A, Fjasrtoft F, Nilsen T-N (2000): Crack formation in unfinished siding of aspen (Populus tremula L.) and Norway spruce (Picea abies (L.) Karst.) during accelerated weathering. Holz als Roh- und Werkstoff 58, 135-139. 115 Freudenthal A M (1950): The inelastic Behaviour of Engineering Materials and Structures. John Wiley & Sons Inc., New York. Furuno T (2001): Histochemical study of wood deterioration by UV irradiation and prevention of deterioration using poly PEGMA. In: High-Performance Utilization of Wood for Outdoor Uses. Ed. Y. Imamura. Wood Research Institute, Kyoto University, Kyoto, Japan. Furusawa Y, Quintern L E, Holtschmidt H, Koepke P, Saito M (1998): Determination of erythema-effective solar radiation in Japan and Germany with a spore monolayer film optimized for the detection of UVB and UVA - results of a field campaign. Applied Microbiology and Biotechnology 50(5),597-603. Futo L P (1974): Der photochemische Abbau des Holzes als Preparations- und Analysemethode. Holz als Roh- und Werkstoff '32(8), 303-314. Futo L P (1976): Einfluss der Temperatur auf den photochemischen Holzabbau. 1. Mitteilung: Experimentelle Darstellung. Holz als Roh- und Werkstoff 34(1), 31-36. Gilfedder J, Keating W G, Robertson 1 (1968): Influence of certain preservatives on pole splitting. Forest Products Journal 18(1), 28-30. Graham R D (1973): Preventing and stopping internal decay of Douglas fir poles. Holzforschung 27(5), 168-173. Graham R D (1979): In large timbers fumigants stop rot that good design could have prevented. Forest Products Journal 29(9), 21 -26. Groves K W, Banana A Y (1986): Weathering characteristics of Australian grown radiata pine. Journal of the Institute of Wood Science 10(5), 210-213. Hale J D (1957): The anatomical basis of dimensional changes of wood in response to changes in moisture content. Forest Products Journal 7(4), 140-144. Helsing G, Graham R D (1976): Saw kerfs reduce checking and prevent internal decay in pressure-treated Douglas-fir poles. Holzforschung 30(6), 184-186. 116 Holtschmidt H (1992): Das MeBprinzip. [Online; cited 4th April 2005; 06:00 PST]. Available from: http://www.biosense.de/home.htm. Hon D N-S (1981): Photochemical degradation of lignocellulosic materials. In: Developments in Polymer Degradation, pp 229 - 281. Ed. N. Grassie. Applied Science Publishers, London. Hon D N-S (1983): Weathering reactions and protection of wood surfaces. Journal of Applied Polymer Science, Applied Polymer Symposium 37(1), 845-864. Hon D N-S, Chang S-T (1984): Surface degradation of wood by ultraviolet light. Journal of Polymer Science, Polymer Chemistry Edition, 22(9), 2227-2241. Hon D N-S, Chang S-T, Feist W C (1985): Protection of wood surfaces against photooxidation. Journal of Applied Polymer Science 20(4), 1429-1448. Hon D N-S, Ifju G (1978): Measuring penetration of light into wood by detection of photo-induced free radicals. Wood Science 11,(2) 118-127. Horn B A, Qiu J, Owen N L, Feist W C (1994): FT-IR studies of weathering effects in western red cedar and southern pine. Applied Spectroscopy 48(6), 662-668. Hosli J P, Mannion K (1991): A practical method to evaluate the dimensional stability of wood and wood products. Forest Products Journal 41(3), 40-44. Ibach, RE (1999): Wood preservation. USDA Forest Service Research Paper FPL- GTR-113. Forest Products Laboratory, Madison, Wisconsin. Ifebueme S C (1977): The natural durability of wood with particular reference to West African timbers. In: Proceedings of the International Workshop on Wood Presen>ation. November 7-12 1977, pp 41-96. Forestry Research Institute of Nigeria Ibadan, Nigeria. Imamura Y (2001): Preface. In: High-Performance Utilization of Wood for Outdoor Uses, pp i-iii. Ed. Y. Imamura. Wood Research Institute, Kyoto University, Kyoto, Japan. 117 Jin L, Archer K, Preston A (1991): Surface characteristics of wood treated with various AAC, ACQ and CCA formulations after weathering. The International Research Group of Wood Preservation. Doc. No. IRG/WP/2369. Kabir F R A, Nicholas D, Vasishth R, Barnes M (1992): Laboratory methods to predict the weathering characteristics of wood. Holzforschung 46(5), 395-401. Kalnins M A (1966): Surface characteristics of wood as they affect durability of finishes. Part II: Photochemical degradation of wood. USDA Forest Service Research Paper FPL 57. Forest Products Laboratory, Madison, Wisconsin. Kamden D P, Zhang J (2000): Characterization of checks and cracks on the surface of weathered wood. The International Research Group of Wood Preservation. Doc. No. IRG/WP/40153. Kataoka Y, Kiguchi M, Evans P D (2004): Photodegradation depth profile and penetration of light in Japanese cedar earlywood (Cryptomeria japonica D. Don) exposed to artificial solar radiation. Surface Coatings International Part B: Coatings Transactions 87(B3), 187-193. Kelsey K E (1956): The shrinkage-intersection point - its significance and method of it determination. Forest Products Journal 6(10), 411-16. Kiguchi M, Evans P D (1998): Photostabilisation of wood .surfaces using a grafted benzophenone UV absorber. Polymer Degradation and Stability, 61(1), 33-45. Kleinert T N (1970): Physikalisch-chemische Holzveranderung im Freien. Holzforschung und Holzverwertung 22(2), 21-24. Kollman F, Cote W A (1968): Principles of wood science and technology. I. Solid wood. Springer-Verlag, New York. Kiihne H, Hochweber M, Sell J (1968): Freiland-Bewitterungsversuche an AuBenanstrichen fur Holz. Versuchszeitraum 1962-1967. Report No. 82. Eidgenossische Materialprufungs- und Versuchsanstalt fur Industrie, Bauwesen und Gewerbe, Diibendorf. Kuo M, Hu N (1991): Ultrastructural changes of photodegradation of wood surfaces exposed to UV. Holzforschung 45(5), 347-353. 118 Levi M P, Coupe C, Nicholson J (1970). Distribution and effectiveness in Pinus sp. of a water-repellent additive for water-borne wood preservatives. Forest Products Journal 20(11), 32-37. Lotan J E, Critchfield W B (2005): Lodgepole Pine. [Online; cited 1st June 2005; 14:30 PST]. Available from: http://www.na.fs.fed.us/spfo/pubs/silvics_manual/Volume_l/pinus/contorta.ht m Loughborough W K (1935): Notes on the effect of boring of a hole through the long axis of a timber on its tendency to check. Unpublished Report US Forest Products Laboratory. Mackay J F G (1972): The occurrence, development and control of checking in Tasmanian Eucalyptus obliqua. Holzforschung 26(4), 121-24. Mackay J F G (1973): Surface checking and drying behaviour of Pinus radiata sapwood boards treated with CCA preservative. Forest Products Journal 23(9), 92-97'. Mater M H (1972): Boring 40-foot long utility poles for conduit passage. Forest Products Journal 22(6), 28-31. Mawson D (1915): The Home of the Blizzard. Being the Story of the Australasian Antarctic Expedition. Heinemann, London. Mcintosh D C (1955): Shrinkage of red oak and beech. Forest Products Journal 5(2), 355-359. McKenzie R L, Bjorn L O, Bais A, Ilyasd M (2003): Changes in biologically active ultraviolet radiation reaching the Earth's surface. Photochemical and Photobiological Sciences 2(1), 1-23. McMillen J M (1955): Drying stresses in red oak. Forest Products Journal 5(2), 71-76. McQueen J, Stevens J (1998): Disposal of CCA-treated wood. Forest Products Journal 48(11/12), 86-90. 119 Militz H, Tjeerdsma B (2001): Heat treatment of wood by the Plato-process. [Online; cited 8th March 2005; 14:30 PST]. Available from: http://www.bfafh.de/inst4/43/pdf/heat_net.pdf Miller E R, Derbyshire H (1981): The photodegradation of wood during solar irradiation. In: Proceedings of the second international conference on the durability of building materials and components. September 14-16, 1981, pp 279-287. National Bureau of Standards, Gaithersburg, Maryland. Miniutti V P (1964): Microscale changes in cell structure of softwood surfaces during weathering. Forest Products Journal 14(12), 571-576. Miniutti V P (1967): Microscopic observations of ultraviolet irradiated and weathered softwood surfaces and clear coatings. USDA Forest Service Research Paper FPL 74. Forest Products Laboratory, Madison, Wisconsin. Miniutti V P (1973): Contraction in softwood surfaces during ultraviolet irradiation and weathering. Journal of Paint Technology 45(577), 27-33. Minolta Co., Ltd. (2001): Spectrophotometer CM-2600d. [Online; cited 8th June 2005; 14:00 PST]. Available from: http://bt.konicaminolta.ca/pdf/cm2600d_brochure.pdf Morrell J J (1990): Effect of kerfing on performance of Douglas-fir utility poles in the Pacific Northwest. The International Research Group on Wood Preservation. Document No: IRG/WP/3604. Morrell J J, Newbill M (1986): Kerfing to prevent decay of Douglas fir poles: an update. Forest Products Journal 36(5), 46-48. Owen J A, Owen N L, Feist W C (1993): Scanning electron microscope and infrared studies of weathering in Southern pine. Journal of Molecular Structure 300,105-114. Park B S, Furuno T, Uehara T (1996): Histochemical changes of wood surfaces irradiated with ultraviolet light. Mokuzai Gakkaishi 42(1), 1-9. Pentoney R E (1953): Mechanisms affecting tangential vs. radial shrinkage. Forest Products Journal 3(2), 27-32. 120 Placket D V, Chittenden C M, Preston A F (1984): Exterior weathering trials on Pinus radiata roofing shingles. New Zealand Journal of Forestry Science 14(3), 368-81. Principia Partners (2002): Polymer-wood composites will double share of giant market for boards and railings in residential decking by 2005. [Online; cited 17th August 2004; 15:00 PST]. Available from: http://www.principiaconsulting.com/releases_18growthwpc.cfm Quayyum, M M, Davis A (1984): Ultraviolet radiation for various angles of exposure at Jeddah and its relation to the weathering of polyacetal. Polymer Degradation and Stability 6(4), 201-209. Raczkowski J (1980): Seasonal effects on the atmospheric corrosion of spruce micro-sections. Holz als Roh- und Werkstoff 38, 231-234. Rapp A O (2001): Review on the heat treatment of wood. [Online; cited 8th March 2005; 13:00 PST]. Available from: http://www.bfafh.de/inst4/43/pdf/2review.pdf Rapp A O, Sailer M (2001): Heat treatment of wood in Germany - state of the art. [Online; cited 8th March 2005; 14:30 PST]. Available from: http://www.bfafh.de/inst4/43/pdf/heat_ger.pdf Rietz R C (1961): Accelerated weathering of red oak treated with various preservatives used to treat crossties. Forest Products Journal 11(12), 567-575. Rowell R M (1999): Wood handbook—Wood as an engineering material. USDA General Technical FPL-GTR-113. Forest Products Laboratory, Madison, Wisconsin. Rowell R M, Feist W C, Ellis W D (1981): Weathering of chemically modified southern pine. Wood Science 13(4), 202-208. Ruddick J N R (1981): The effect of kerfing on check formation in treated white spruce (Picea glaucd) poles. The International Research Group on Wood Preservation. Doc. No. IRG/WP/3167 Ruddick J N R, Ross N A (1979): Effect on kerfing on checking of untreated Douglas-fir pole section. Forest Products Journal 29(9), 27-30. 121 Sandberg D (1996): The influence of pith and juvenile wood on proportion of cracks in sawn timber when kiln dried and exposed to wetting cycles. Holz als Roh- und Werstoff54(3),\52. Sandberg D (1997): Radially sawn timber. The influence of annual ring orientation on crack formation and deformation in water soaked pine (Pinus sylvestris L.) and spruce (Picea abies (L.) Kast.) timber. Holz als Roh- und Werkstoff'55(3), 175-182. Sandberg D (1999): Weathering of radial and tangential wood surfaces of pine and spruce. Holzforschung 53(4), 355-354. Sanderman W, Schlumbom F (1962a): On the effect of filtered ultraviolet light on wood - Part I: Photometric and chromatographic investigations on wood powders. Holz als Roh- und Werkstoff'20(7), 245-252. Sanderman W, Schlumbom F (1962b): On the effect of filtered ultraviolet light on wood - Part II: Kind and magnitude of colour difference on wood surfaces. Holz als Roh- und Werkstoff '20(8), 285-291. Schniewind A P (1959): Transverse anisotropy of wood: a function of gross anatomic structure. Forest Products Journal 9(10), 350-59. Schniewind A P (1963): Mechanism of check formation. Forest Products Journal 13(11), 475-480. Schniewind A P, Kersavage P C (1961): A quantitative evaluation of second order drying stresses. Forest Products Journal 11(11), 523-530. Schoeman M, Dickinson, D (1997): Growth of Aureobasidium pullulans on lignin breakdown products at weathered wood surfaces. Mycologist 11(4), 168-172. Sell J (1968): Investigations on the infestation of untreated and surface treated wood by blue-stain fungi. Holz als Roh- und Werkstoff 26(6), 215-222. Sell J (1975): Basic requirements for the surface treatment of exterior wood: A review of pertinent literature. Holz als Roh- und Werkstoff'33(9), 336-340. 122 Sell J, Leukens U (1969): Uber die auBenklimatische und biologische Beanspruchung von unbehandelten und angestrichenen Holzoberflachen. Oberflache - Surface 10(8), 536-539. Sell J, Leukens U (1971a). Untersuchungen an bewitterten Holzoberflachen - Zweite Mitteilung: Verwitterungserscheinungen an ungeschutzten Holzern. Holz als Roh- und Werkstoff "39(1), 23-31. Sell J, Walchli O (1969): Changes in the surface texture of weather-exposed wood. Material und Organismen 4(2), 81-87. Shanklin J (2004): The ozone layer. [Online; cited 24th January 2004; 15:00 PST]. Available from: http://www.theozonehole.com/fact.htm Skaar C (1972): Water in Wood. Syracuse University Press, Syracuse, New York. Stamm A J (1963): Effect of dimensional stabilization on weathering properties of wood. American Paint Journal 48(25), 72-88. Stamm A J (1964): Wood and Cellulose Science. The Ronald Press Co, New York. Stamm A J (1965): Modification of wood for improved finishing. Official Digest. Journal of Paint Technology 37, 707-719. Stamm A J, Burr H, Kline A (1946): Staybwood...heat-stabilized wood. Industrial and Engineering Chemistry 38(6), 630-634. Stamm A J, Loughborough W K (1942): Variations in shrinking and swelling of wood. Transactions of the American Society for Mechanical Engineering 64, pp 379-285. Syrjanen T, Oy K (2001): Heat treatment of wood in Finland - state of the art. [Online; cited 8th March 2005; 14:00 PST]. Available from: http://www.bfafh.de/inst4/43/pdf/heat_fin.pdf Szymany R, McDonald K A (1981): Defect detection in lumber: state of the art. Forest Products Journal 31(11), 34-44. 123 Timber Technology Services (1996): Effectiveness of check reducing techniques in peeled, preservative treated radiata and slash pine poles. Report prepared for CSR Softwoods. ANUTECH Pty. Ltd., ACN No. 008 548 650. Turkulin H, Sell J (2002): Investigations into the photodegradation of wood using microtensile testing. Part 4: Tensile properties and fractography of weathered wood. Holz als Roh- und Werkstoff'60(2), 69-105. Vernois M (2001): Heat treatment of wood in France - state of the art. [Online; cited 8th March 2005; 14:30 PST]. Available . from: http://vvfww.bfafh.de/inst4/43/pdf/heat_fra.pdf Webb D A, Sullivan J D (1964): Surface effect of light and water on wood. Forest Products Journal 14(11), 531-534. Wengert E M (1966): Parameters for predicting maximum surface temperature of wood in exterior exposure. USDA Forest Service Research Paper FPL 62. Forest Products Laboratory, Madison, Wisconsin. Yata S (2001): Occurrence of drying checks in softwood during outdoor exposure. In: High-Performance Utilization of Wood for Outdoor Uses, pp 65-70 Ed. Y. Imamura. Wood Research Institute, Kyoto University, Kyoto, Japan. Yoshimoto T (1972): Photochemical analysis of wood and related substances. Mokuzai Gakkaishi, 18(1), 45-49. Zahora A R (1991): Interactions between water-borne preservatives and emulsion additives that influence the water repellence of wood. The International Research Group of Wood Preservation. Doc. No. IRG/WP/2374. Zahora A R (1992): A water repellent additive's influence on field performance of southern yellow pine lumber. American Wood-Preservers' Association, 148-159. Zahora A R (2000): Long-term performance of a wax type additive for use with water-borne pressure preservative treatments. The International Research Group on Wood Preservation. Doc. No. IRG/WP/40159. Appendices 124 APPENDIX 3.1 Fourier Transform Infra-Red Spectroscopy Materials and Methods A single, conditioned, lodgepole pine decking sample exposed under each of the different filters for 36 weeks, and the fully exposed and unexposed controls were reduced in thickness to 20 mm by planning the unexposed undersides of boards using a Martin T44 jointer. A strip measuring 2 x 6 cm was sawn from the end of each sample and the weathered surface was lightly cleaned with a swab moistened with distilled water. Samples were stored overnight in a vacuum desiccator over silica gel and Fourier transform infra-red (FTIR) spectra of weathered surfaces were obtained using a single bounce attenuated total reflectance accessory (Pike Miracle) attached to a Perkin Elmer Spectrum One spectrometer. The average depth of light penetration was 1.2 urn. Each spectrum represents 16 accumulations between wavenumbers of 650 and 4000 cm"1, although only the finger print region between 1350 and 1800 cm"1 is shown. Results The FTIR spectra of the surface of the unweathered control clearly showed the presence of the peak at 1510 cm"1, corresponding to aromatic C=C bond stretching in lignin (Evans et al. 1996). This peak was absent in the fully exposed samples and those exposed Appendices 125 under filters 1 and 2 indicating delignification of the surface. Samples exposed under filter 3 showed a slight peak at 1510 cm'1 indicating that there was some residual lignin at the surface. Samples exposed under filter 4 and 5 showed no delignification. A 1800.0 1750 1700 1650 1600 1550 1500 1450 1400 1350.0 Figure 0.1: Infra-red spectra of lodgepole pine boards. Bands from unexposed control (0), fully exposed control (F) and specimen exposed under filters 1, 2, 3,4 and 5 Appendices 126 APPENDIX 3.1 Scanning Electron Microscopy Materials and Methods Small specimens 5x5 mm square and 20 mm in length were cut from the surfaces of the deck samples and controls and examined using scanning electron microscopy. Specimens were attached to aluminium stubs using double-sided adhesive tape, with the weathered surface facing up. They were then sputter coated with a layer of gold and weathered surfaces were examined using a field emission scanning electron microscope (SEM) (Hitachi S4700) operating at 10 kV. Selected images of weathered and unweathered wood surfaces were recorded digitally. Results Under the SEM the surface of the unexposed sample shows loose single fibres and slightly damaged cell structures (Figure 3.14). Most cells are, however, still intact. Some debris, most likely slivers from the planing of the surface, can be observed. On the surface of the sample exposed under filter 1 whole bundles of fibres are detached from the surface (Figure 3.14b). The damage to the cell structure is clearly evident in the form of a number of open tracheids and wider rays. More debris has accumulated on the surface. The extent of the degradation of the cell structure decreases with increasing blockage of wavelengths. Appendices 127 The cell bundles detached from the surface are less noticeable in samples exposed under filter 2 than for those exposed under filter 1. Voids due to the degradation of cell walls and rays are also smaller (Figure 3.15). At a magnification of xlOO the surfaces of samples exposed under the filters 3, 4 and 5 can hardly be distinguished from the surfaces of unexposed samples (Figure 3.17-3.19). The photomicrographs taken at a magnification of x 250 showed fewer open tracheids and voids caused by the degradation of rays on the surfaces of the samples exposed under filters 3, 4 and 5 (Figure 3.20 - 3.22). Those surfaces are, however, rougher than the surface of the unexposed samples because they contain more small fibres protruding above the surface and a greater accumulation of debris. Appendices 128 Figure 0.4: SEM image of the surface exposed under filter 4 (left) and filter 5 (right) at a magnification of x40 Appendices 129 SE ll-Apr-05 003165 WD30. 5mm 5.00kV xlOO 500um Figure 0.5: SEM image of the unweathered (left) surface and the surface exposed under filter 1 (right) at a magnification of x 100 Figure 0.6: SEM image of the surface exposed under filter 2 (left) and filter 3 (right) at a magnification of xlOO Figure 0.7: SEM image of the surface exposed under filter 4 (left) and filter 5 (right) at a magnification of xlOO Appendices 130 F igure 0.10: S E M image of the surface exposed under filter 4 (left) and filter 5 (right) at a magnification of x250 

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0075075/manifest

Comment

Related Items