Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Thermal modification of Western hemlock (Tsuga heterophylla) Nourian, Sepideh 2018

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
24-ubc_2018_september_nourian_sepideh.pdf [ 1.41MB ]
Metadata
JSON: 24-1.0368548.json
JSON-LD: 24-1.0368548-ld.json
RDF/XML (Pretty): 24-1.0368548-rdf.xml
RDF/JSON: 24-1.0368548-rdf.json
Turtle: 24-1.0368548-turtle.txt
N-Triples: 24-1.0368548-rdf-ntriples.txt
Original Record: 24-1.0368548-source.json
Full Text
24-1.0368548-fulltext.txt
Citation
24-1.0368548.ris

Full Text

 THERMAL MODIFICATION OF WESTERN HEMLOCK (Tsuga heterophylla)   by Sepideh Nourian M.Sc., The International University of Imam Khomeini, 2013   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF SCIENCE in The Faculty of Graduate and Postdoctoral Studies (Forestry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   May 2018  © Sepideh Nourian, 2018ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis/dissertation entitled:  Thermal Modification of Western Hemlock (Tsuga heterophylla)  submitted by Sepideh Nourian  in partial fulfillment of the requirements for the degree of Master of Science  in The Faculty of Graduate and Postdoctoral Studies (Forestry)  Examining Committee: Dr. Stavros Avramidis Supervisor  Dr. Julie Cool Supervisory Committee Member  Dr. Philip D. Evans Supervisory Committee Member   Dr. Luiz Oliveira Supervisory Committee Member Dr. Savvas Hatzikiriakos Additional Examiner  Additional Supervisory Committee Members: Dr. Taraneh Sowlati Neutral Chair     iii  Abstract  Thermal modification is a controlled degradation process of wood cell wall material at high temperatures that results in improving some of its properties. This process, albeit quite developed in Europe, is still in its infancy in Canada. So, use of such method might add value to local wood species and thus, allow the development of new products and markets. Western Hemlock (Tsuga heterophylla) is a locally abundant species used mostly in construction, so the development of other uses could bring extra revenues to the industry.   This research focuses on exploring thermal treatment levels and their effect on some material properties in order to achieve the optimum combination for hemlock. In this regard, kiln-dried western hemlock boards with two cuts (flatsawn and quartersawn) were subjected to the Thermowood® process at three different maximum treatment temperatures (170, 212, and 230°C) for 2hrs. Samples were cut from both the treated and untreated boards for property evaluation tests including basic density, equilibrium moisture content, water absorption, anti-swelling efficiency, color change, Janka hardness, and dynamic modulus of elasticity.  Data analysis revealed that there was no significant difference between the two cuts and temperature was the only factor that affected the wood properties. Basic density, equilibrium moisture content, and water absorption were lower at higher treatment temperatures, while dimensional stability considerably increased. The impact of higher treatment temperatures on hardness and stiffness of samples was hardly noticeable, but it visually influenced the color of samples and made them darker. Within the scope and limitations of this study, the optimum treatment temperature, namely, the one that enables improved dimensional stability and provided a darker color without significantly affecting the wood strength suggested establishing at 212°C. Further research is required to fully determine the performance of thermally modified wood for interior and exterior applications.          iv  Lay Summary Thermowood® process is an environmentally friendly industrial treatment method, which uses high temperatures and steam to modify the properties of wood. The extent of altered wood properties immensely depends on the treatment process parameters and the wood species that are being modified. Even though western hemlock is an important commercial tree species of B.C., not much effort has ever been done to thermally modify this species and evaluate its altered properties. Therefore, using Thermowood® process, I thermally modified western hemlock at different process conditions and determined the extent of some of its altered physical and mechanical properties. My main goal was to find the effects of the combination of treatment temperature and wood’s sawing pattern (flatsawn and quartersawn) on final quality and appearance of western hemlock.                              v  Preface This dissertation is an original intellectual product of the author, Sepideh Nourian. The proposed methodology in this manuscript is original, unpublished, and independent work by the author at the UBC Department of Wood Science.                     vi  Table of Contents Abstract……… ............................................................................................................................. iii Lay Summary……. ...................................................................................................................... iv Table of Contents ......................................................................................................................... vi Preface……………………………………………………………………………………………VList of Tables….………………………………………………………………………………..viii List of Figures…. .......................................................................................................................... ix Acknowledgments ........................................................................................................................ xi Dedication……… ........................................................................................................................ xii 1. Introduction ............................................................................................................................ 1 1.1     Outline of thesis ................................................................................................................... 2 2. Literature Review and Objective ......................................................................................... 4  Thermal Modification ............................................................................................................... 4  Industrial Thermal Modification Processes .............................................................................. 5  Proving Lasting Advanced Timber Option (Plato)................................................................... 5  Retification ............................................................................................................................... 6  Le Bois Perdure ........................................................................................................................ 6  Oil-Heat Treatment (OHT) ....................................................................................................... 6  ThermoWood ............................................................................................................................ 6 2.2.5.1. Process .......................................................................................................................... 8 2.2.5.2. Equipment ..................................................................................................................... 9  Alterations in Chemical Structure .......................................................................................... 10 2.3.1. Hemicelluloses .................................................................................................................. 10 2.3.2. Cellulose ............................................................................................................................ 11 2.3.3. Lignin ................................................................................................................................ 11 2.3.4. Extractives ......................................................................................................................... 12  Alterations in Physical Properties .......................................................................................... 12 2.4.1. Hygroscopicity and Dimensional Stability........................................................................ 12 2.4.2. Density .............................................................................................................................. 14 2.4.3. Color…… .......................................................................................................................... 14  Alterations in Mechanical Properties ..................................................................................... 16 2.5.1.     Strength ............................................................................................................................... 16 2.5.1.     Hardness ............................................................................................................................. 17 vii   Western Hemlock ................................................................................................................... 19  Objective ................................................................................................................................ 20 3. Materials and Methods ........................................................................................................ 22 3.1. Materials ................................................................................................................................. 22 3.2. Heat Treatment ....................................................................................................................... 23 3.3. Sample Cutting Protocol ........................................................................................................ 23 3.4. Dimensional Stability and Water Absorption ......................................................................... 24 3.5. Equilibrium Moisture Content ................................................................................................ 26 3.6. Basic Density .......................................................................................................................... 26 3.7. Color ....................................................................................................................................... 26 3.8. Stiffness .................................................................................................................................. 28 3.9. Hardness ................................................................................................................................. 29 3.10. Statistical Analysis ................................................................................................................. 30 4. Results and Discussion ........................................................................................................ 31 4.1. Basic Density .......................................................................................................................... 31 4.2. Water Absorption and Equilibrium Moisture Content ........................................................... 33 4.3. Anti-Swelling Efficiency ........................................................................................................ 36 4.4. Color ....................................................................................................................................... 41 4.5. Hardness ................................................................................................................................. 46 4.6. Stiffness .................................................................................................................................. 50 5. General Discussion ............................................................................................................... 54 6. Conclusion ............................................................................................................................ 58 7. Further Research ................................................................................................................. 59 References………… .................................................................................................................... 62        viii  List of Tables Table 2.1 Standard process temperatures for Thermo-S and Thermo-D ...................................................... 9 Table 3.1 Heat treatment protocol…………………………………………………………………………22 Table 3.2 Experimental design.................................................................................................................... 30 Table 4.1 Descriptive statistics of basic density of samples….…………………………………………..31 Table 4.2 Likelihood ratio tests for basic density. ...................................................................................... 32 Table 4.3 Tukey’s pairwise comparisons for basic density at different levels of temperature. .................. 32 Table 4.4 The descriptive statistics of water absorption (WA) and equilibrium moisture content (EMC) of samples. ....................................................................................................................................................... 33 Table 4.5 Likelihood ratio tests for water absorption (WA). ....................................................................... 36 Table 4.6 Likelihood ratio tests for equilibrium moisture content (EMC).................................................. 36 Table 4.7 Tukey’s pairwise comparisons for water absorption (WA) and equilibrium moisture content (EMC) at different levels of temperature. ................................................................................................... 36 Table 4.8 The descriptive statistics of radial and tangential swelling of samples. ..................................... 37 Table 4.9 Descriptive statistic of anti-swelling efficiency (ASE) for heat-treated samples. ....................... 38 Table 4.10 Likelihood ratio tests for anti-swelling efficiency (ASE). ......................................................... 40 Table 4.11 Tukey’s pairwise comparisons for anti-swelling efficiency (ASE) at different levels of temperature ................................................................................................................................................. 40 Table 4.12 Descriptive statistics of color coordinates and color change. ................................................... 42 Table 4.13 Likelihood ratio tests for color change. .................................................................................... 45 Table 4.14 Tukey’s pairwise comparisons for color change at different levels of temperature. ................ 45 Table 4.15 Descriptive statistics of side hardness in samples ..................................................................... 46 Table 4.16 Likelihood ratio tests for side hardness..................................................................................... 47 Table 4.17 Tukey’s pairwise comparisons for side hardness at different levels of temperature. ............... 47 Table 4.18 Descriptive statistics of end hardness in samples. .................................................................... 48 Table 4.19 Likelihood ratio tests for side hardness..................................................................................... 49 Table 4.20 Tukey’s pairwise comparisons for end hardness at different levels of temperature. ................ 49 Table 4.21 Descriptive statistics of stress-wave velocity (V) in samples at different temperatures. .......... 51 Table 4.22 Descriptive statistics of stiffness of samples. ............................................................................ 52 Table 4.23 Likelihood ratio tests for stiffness. ............................................................................................ 53 Table 4.24 Tukey’s pairwise comparisons for dynamic modulus of elasticity (ED) at different levels of temperature. ................................................................................................................................................ 53  ix  List of Figures Figure 2.1 The share of ThermoWood® market areas (ThermoWood® Production Statistics 2016). .......... 7 Figure 2.2 ThermoWood® production in 2001-2016 ................................................................................... 7 Figure 2.3 Schematic of the Thermowood® process. Phase 1: High-Temperature Drying, Phase 2: Heat Treatment, Phase 3: Cooling and Conditioning. ........................................................................................... 8 Figure 2.4 Changes in the wood components during heat treatment (From Navi and Sandberg 2012). .... 10 Figure 2.5 The distribution map of western hemlock (Natural Resources Canada, 2015) ......................... 19 Figure 3.1 The Jartek industrial thermal modification kiln at Scottywood Corporation (Abbotsford, BC)……………………….………………………………………………………………………………. 22 Figure 3.2 Scheme for thermal modification process in accordance with suggested schedules. ................ 23 Figure 3.3 Cutting protocol of samples. ...................................................................................................... 24 Figure 3.4 Measuring: a) the thickness at four positions, b) the length and the width at three positions. .. 24 Figure 3.5 Submerged samples in the water bath. ...................................................................................... 25 Figure 3.6 The three-dimensional CIEL*a*b* color space (Johansson 2008). .......................................... 27 Figure 3.7 Minolta Colorimeter. ................................................................................................................. 27 Figure 3.8 Metriguard 239A Stress Wave Timer set up to measure the stiffness. ...................................... 28 Figure 3.9 Hardness test using a Universal testing machine. ...................................................................... 29 Figure 4.1 The average basic density of treated and untreated samples. Error bars indicate standard error……………………………..……………………………..……………………………..……………32 Figure 4.2 The average: a. water absorption (WA), and b. equilibrium moisture content (EMC) of samples at different temperatures. Error bars indicate standard error. ..................................................................... 34 Figure 4.3 The correlation between water absorption (WA) and equilibrium moisture content (EMC) of samples. ....................................................................................................................................................... 35 Figure 4.4 The tangential and radial swelling of samples. Error bars indicate standard error. ................... 38 Figure 4.5 The average dimensional stability of heat-treated samples at three temperatures. Error bars indicate standard error. ................................................................................................................................ 39 Figure 4.6 Scatter plot of anti-swelling efficiency (ASE) versus water absorption (WA) of samples with fitted line. .................................................................................................................................................... 40 Figure 4.7 Thermally modified western hemlock at (a unmodified, (b) 170°C, (c) 212°C, and (d) ) 230°C. Top row: quartersawn, bottom row: flatsawn. ............................................................................................ 41 Figure 4.8 The lightness (L*) of heat-treated samples. Error bars indicate standard error. ........................ 43 Figure 4.9 The average of: a. red/green coordinates (a*), b. blue/yellow coordinates (b*), and c. color change (ΔE) in samples. Error bars indicate standard error. ....................................................................... 44 x  Figure 4.10 Side hardness of samples treated at different temperatures. Error bars indicate standard error…... ...................................................................................................................................................... 47 Figure 4.11 End hardness of samples at different temperatures. Error bars indicate standard error. ......... 48 Figure 4.12 The scatter plot of a. side hardness, and b. end hardness versus density. ................................ 50 Figure 4.13 The correlation between density and dynamic modulus of elasticity (ED) of samples. ........... 51 Figure 4.14 Dynamic modulus of elasticity (ED) of samples at different temperatures. Error bars indicate standard error. ............................................................................................................................................. 52                   xi  Acknowledgments  I would like to express my sincere appreciation to my supervisor Dr. Stavros Avramidis for mentoring me through the field of wood science. Your support and scientific advice during my time as your student truly helped me to conquer the challenges of studying this new field. I am much honored to have received the opportunity to be a part of your research group. I would like to extend my gratitude to my supervisory committee members Dr. Julie Cool, Dr. Philip D. Evans and Dr. Luiz Oliveira, for your time and helpful suggestions. I would like to thank Scottywood Corporation for allowing me the use of their thermal modification kiln without which this project would have never been started in the first place. A big thank you goes to Pooria Kuchebaghi, Farbod Moheb and Rui Song for all their help when it came to the hard work of sample preparation. Also, thank you Sina Heshmati, Vahid Nasir, and Mahdi Shahverdi for your friendship, support, and for always finding the time to help me. Special thanks to Dr. Valeri Lemay for your priceless help when it came to the statistical analysis of the data. My sincere gratitude to the donors of the Robert and Averil Kennedy Wood Science Graduate Scholarship, Dr. J. David Barrett Memorial Scholarship in Wood Science, and VanDusen Graduate Fellowship in Forestry for their financial support.           xii  Dedication To my beloved husband, Arash, who has been a constant source of support and encouragement for me during the challenges of graduate school and life. I am truly thankful for having you in my life.                       1  1. Introduction Wood is one of the oldest construction materials because of its advantageous characteristics such as strength-to-weight ratio, renewability, and versatility. Nevertheless, wood has some undesirable properties such as high hygroscopicity, dimensional instability, and biodegradability. These properties could be improved by means of various chemical treatments and modifications. However, environmental concerns coupled with the increased need for durable wood products have led to the development of an environmentally safe method called thermal modification (Johansson 2008, Rowell et al. 2012).  Thermal modification changes the cell wall ultrastructure and chemistry of wood by means of heating it at elevated temperatures. These wood modifications result in uniform color changes, decreased hygroscopicity, increased dimensional stability, decay resistance, weather resistance, and heat insulation. While, strength properties may decrease pending on treatment conditions (Hill 2006, Esteves and Pereira 2009, Unsal et al. 2009). With specific properties improved, thermally modified wood can now be used in higher end-value applications such as floors, windows and doors, claddings, saunas, musical instruments, and many other outdoor and indoor applications (Gunduz et al. 2009, Albrektas and Navickas 2017).  The extent of chemical changes and subsequently improved properties depend on process parameters such as time and temperature of treatment, using an open or closed system, treatment atmosphere, and intrinsic features of wood species (Tjeerdsma et al. 1998b, Hill 2006, Gérardin 2016). In recent decades, several thermal modification processes based on different parameters have been introduced to European markets including Thermowood from Finland Plato from the Netherlands, Retification and Le Bois Perdure from France, and OHT (oil-heat treatment) from Germany (Vernois 2001, Militz 2002).  In retrospect, for more coherent data on the influence of modification process parameters, studies with one specific wood species and a systematic schedule for the variation of modification process parameters for that species are required. Accordingly, western hemlock (Tsuga heterophylla) as a locally abundant softwood which plays a key role in British Columbia’s annual timber production is chosen to be investigated in this thesis (Middleton and Munro 2001, Lazarescu and Avramidis 2012). Thermal modification of western hemlock would be an economical and environmentally friendly way to acquire aesthetically valuable wood products from a domestic wood species mimicking the appearance of exotic tropical hardwoods such as Ipe 2  (Handroanthus spp.), or valuable western red cedar (Thuja plicata Donn ex D.Don) without using any chemicals or stains.   Beside the treatment parameters, according to Syrjänen (2001), another influential factor on final properties of thermally modified wood is the sawing pattern. He mentions a lot of peeling in the flatsawn treated pieces (annual rings with less than 30° from the surface), while the quartersawn treated pieces (annual rings with 60-90° angle from the surface) were less susceptible to deformation, had much better final look, and better surface hardness.   For aforementioned reasons, the current study examines the impact of temperature and sawing pattern on the properties of western hemlock after thermal modification process using Thermowood® method. I hypothesized that thermally modified western hemlock would acquire darker shades of color and enhanced physical properties such as increased dimensional stability, lower equilibrium moisture content, and lower water absorption, while it's basic density, stiffness and hardness do not significantly decrease. As forest industry embraces more value-adding philosophy, the mixture of improved properties and aesthetics through the application of Thermowood® process on western hemlock will definitely provide beneficial market opportunities in future. Nevertheless, further research is required to fully determine the true capabilities of thermally modified western hemlock, and to figure out how to refine its properties by applying different treatment parameters, to adapt thermally modified western hemlock to certain markets. 1.1     Outline of thesis Chapter 1 of this thesis briefly explains what the research was about and why it was carried out. Chapter 2 provides background information on several thermal modifications and specifically explains the ThermoWood process. Moreover, the alterations in chemical structure, physical and mechanical properties due to thermal modification are explained along with a description of western hemlock properties and applications. Chapter 2 ends with indicating the objective of this thesis. Chapter 3 illustrates the methodology used for thermal modification, preparing samples and then testing the properties of thermally modified samples. The tested properties include basic density, equilibrium moisture content, dimensional stability, water absorption, color change, side and end hardness, and stiffness. In chapter 3, the results of experiments are given and the analysis of results are discussed. Finally, Chapter 4 discusses the general results from chapter 3 and how 3  they support the objective of this thesis. Then, makes recommendations for future research, and draws overall conclusions.                          4  2. Literature Review and Objective  Thermal Modification    Wood modification is an industrial procedure where the properties of wood are altered so that after disposal at the end of the product lifecycle, environmental hazards are eliminated in comparison with an unmodified wood (Hill 2006). Conventional wood treatment methods use active biocides to increase the longevity of wood products under service conditions. However, these chemicals will inevitably appear as waste at the end of the life cycle of that wood and could have adverse effects on both humans and the environment (Hill 2006, Chen 2013). Thermal modification is an industrial process that utilizes heat as a means to improve specific wood properties in an environmentally friendly way thus substituting traditional chemical treatment methods (Awoyemi and Jones 2010). The concept of heating wood for performance enhancement dates back to thousands of years ago when Egyptians bent wood for bows and chairs using hot water (Ostergard 1987), Vikings bent wood by heat for building ship parts (Navi and Sandberg 2012), and cask makers used heat-bent wood 2000 years ago (Twede 2005).    In the early 1900’s, it was realized that “heat treatment” or “thermal modification” of wood is a potentially useful method to improve dimensional stability and decay resistance (Hill 2006). Tiemann (1915) investigated the effect of high-temperature drying at 150°C for 4 hours on air-dried wood with superheated steam and concluded a 10-25% reduction in moisture sorption and a decrement in strength. Later, Stamm and Hansen (1937) mentioned an unpublished report dated 1916, on high-temperature drying of black gum (Nyssa sylvatica N. biflora) at 205°C for 6 hours, which resulted in hygroscopicity reduction to almost one-half of its original value. Further studies of this method resulted in the development of several commercial processes such as “Lignstone” and “Lignifol” in Germany, which used high-temperature drying and densification by hot-press (Kollmann 1936). In 1945, the same method used by Seborg et al. (1945) in the United States called “Staypack”. A year later, Stamm et al. (1946) reported a thermal modification method at temperatures between 140 and 320°C, which improved dimensional stability and increased resistance against microbial attack in sitka spruce (Picea sitchensis (Bong.) Carr.), and called it “Staybwood”. Many more studies presented later by Seborg et al. (1953), Kollmann and Fengel (1965), Burmester (1973), Tjeerdsma et al. (1998a) in thermal modification. 5    Most of these studies in the early 20th century were small scale and had limited success in the market mostly due to the availability of naturally durable species (Esteves and Pereira 2009). It took about 50 years for thermal modification to be commercialized on a larger industrial scale. The significant increase in the interest in thermal modification after 50 years was mainly due to increasing concerns about the detrimental effects of chemicals on health and environment, which led to the issuance of governmental restrictions and regulations on the use of chemicals for treatment of wood (Chen 2013). The other reason is the rising demand for timber, which has led to concerns of deforestation and timber shortage by the end of the 21st century (Hill 2006).    In summary, the demand for environmentally friendly and of high-durability wood exploded in recent decades, and thus, it led to significant developments in the area of thermal modification (Navi and Sandberg 2012).  Industrial Thermal Modification Processes   Several historic names have been given to various methods of thermal modification over the last decades. All these methods are generally based on degradation of cell wall material at high temperatures (160-260°C). However, the process variables such as duration, temperature level, number of steps, pressure, heating medium, inert gas, and the velocity of heating or cooling down differs from one process to another one. The most known industrial thermal modification processes are mentioned below.  Proving Lasting Advanced Timber Option (Plato)    The Plato wood process is a hydrothermal modification of pre-dried timber (moisture content of 14-18%) in four stages. First comes the hydrothermolysis stage where wood is heated up to 150–180°C with high-pressure saturated steam (about 6-8 atmospheres) for 4-5 hours. Second comes the drying stage that is carried out in a conventional kiln to reach a moisture content of 8-9% over a period of 5-21 days. Third comes the curing stage where the wood is heated again to 150–190°C under dry atmospheric conditions in order to reach a moisture content of less than 1% for 12-16 hours depending on wood species and size. Finally, conditioning by using saturated steam, wood moisture content is raised to a level which is necessary for manufacturing (4-6%). The Plato® Wood plant is located in Arnhem, Netherlands with capacity of 35,000 m3 per year, and they mainly treat spruce (Picea spp.), Scots pine (Pinus sylvestris L.), Douglas-fir (Pseudotsuga 6  menziesii (Mirb.) Franco), poplar (Populus spp.), and birch (Betula spp.)  (Militz and Tjeerdsma 2001, Rowell et al. 2012).  Retification    Retification has been developed in France at Ecole des Mines de Saint-Etienne, in collaboration with the license owner NOW S.A. (NewOption Wood) and the kiln manufacturer Fours & Brûleurs Rey. The process is industrialized since 1997 under the brand name Retiwood. In this, pre-dried wood with moisture content of approximately 12% is heated in an oven at temperatures between 180°C and 250°C under a nitrogen atmosphere where oxygen is less than 2%. Once the maximum temperature is reached, wood is cooled by sprinkling of water (Navi and Sandberg 2012, Rowell et al, 2012).  Le Bois Perdure     This is a Canadian process developed in Québec, Canada in which the first step involves drying of green wood and then, heating to 200°C to 230°C in a steam atmosphere generated out of the moisture from green wood (Navi and Sandberg 2012, Sandberg and Kuntar 2016). The process only produces very limited amounts of liquid residues and no atmospheric emissions (Bois Perdure 2016).  Oil-Heat Treatment (OHT)    In 2000, the Menz Holz Company in Germany started to use a hot oil bath instead of inert gas atmosphere to thermally modify wood. The process takes place in a closed vessel at temperatures 180–220°C for up to 18 hours. Using hot vegetable oils result in fast and even heat transfer to wood while separating oxygen from wood. Oil absorbance by wood and unpleasant smell of oils, that diminish after some time are two common process drawbacks (Rapp and Sailer 2001, Esteves and Pereira 2009).  ThermoWood     In early 1990’s, to improve the wood properties, an industrial scale process called ThermoWood® was established by the Finnish Research Center VTT together with the Finnish industry. Their first commercial plant was built in Mänttä Finland. The process is carried out in the presence of steam, without pressure, and with an air speed of at least 10 m/s (Syrjänen and Kangas 7  2000).  Presence of steam minimizes the amount of oxygen in contact with wood to 3-5% which results in reduced oxidative degradation of wood (Syrjänen 2001). Today, this process is licensed to the members of the International ThermoWood Association (established in December 2000), who is by far the largest producer of thermally modified timber in the world and as it is depicted in Figure 2.1, ThermoWood® is dominating the European markets (Sandberg and Kutnar 2016).   Figure 2.1 The share of ThermoWood® market areas  (ThermoWood® Production Statistics 2016).   According to the latest ThermoWood® Production Statistics (Figure 2.2), the production volume has been progressively increasing since 2001, and in 2016, the production of ThermoWood reached 179507 m3.   Figure 2.2 ThermoWood® production in 2001-2016  (ThermoWood® Production Statistics 2016). 8  2.2.5.1.  Process The ThermoWood process is feasible with both green as well as pre-dried wood and is suitable for both hardwoods and softwoods. However, it must be optimized separately for each wood species (Johansson 2008). The process consists of three phases (Figure 2.3). 1) High-Temperature Drying – Using heat and steam, the kiln temperature is quickly raised to around 100ºC. Afterward, the temperature increases steadily to 130ºC, when the high-temperature drying takes place and the moisture content in the wood decreases to nearly zero. Therefore, the initial moisture content of wood does not play any role in the success of heat treatment since the wood is heated until it is totally dry; however, it can affect the duration of this phase. The duration of high-temperature drying also depends on both wood species and thickness, and it is usually the longest phase amongst all. 2) Heat Treatment- Using heat and steam, the temperature is raised up to 185-230°C and kept there for 2-3 hours depending on the desired end use. 3) Cooling and Conditioning- Using water spray systems, the temperature is cooled down to 80-90°C and the wood is conditioned to moisture content appropriate for the end use (normally between 4-7%). (Syrjänen 2001, Finnish ThermoWood Association 2003, Navi and Sandberg 2012).  Figure 2.3 Schematic of the Thermowood® process. Phase 1: High-Temperature Drying, Phase 2: Heat Treatment, Phase 3: Cooling and Conditioning. 9  As seen in Figure 2.3, there are two standard treatment classes: Thermo-S and Thermo-D, where “S” stands for stability and “D” for durability. These two processes are carried at different temperature levels (Table 2.1), which results in different properties and consequently different end uses of the modified wood (Finnish ThermoWood Association 2003). Table 2.1 Standard process temperatures for Thermo-S and Thermo-D. Class Softwood Hardwood Thermo-S 190°C 185°C Thermo-D 212°C 200°C 2.2.5.2. Equipment Since process conditions are corrosive, thermal modification equipment is made of stainless steel. To generate the required heat, either electric heating or hot-oil heating systems is used. Eighty percent of the heat energy is used for drying of the wood thus, the total energy demand is only 25% higher than that of the normal kiln drying process.  A steam generator is used to produce the required steam, and a reversible fan is used to make even and sufficient air circulation (10 m/s) in the chamber. In addition, there are several sensors around the chamber to measure the heat and moisture content of both air and the wood and are connected to a programmable controller. The controller automatically adjusts the treatment conditions based on the initial and collected data. More specifically, to prevent the surface and internal checking of wood due to temperature alterations, the collected wood’s internal temperature and moisture content during the process regulates the temperature rise in the chamber. Because of automation, the process can be recorded and rechecked again.  The treatment chamber is also equipped with water spray systems for equalizing and cooling, an incinerator to burn the volatile compounds evaporated from wood to avoid the odor nuisance imposed to environment, and a special clarification basin for separating the solid components from the waste water (Syrjänen 2001, Militz 2002, Finnish ThermoWood Association 2003, Navi and Sandberg 2012).  10   Alterations in Chemical Structure  As mentioned earlier, thermal modification is based on degradation of wood cell wall at high temperatures, which consequently results in wood chemical composition changes. According to Bourgois et al. (1991), these changes depend upon treatment conditions especially temperature level. Drying begins at low temperatures up to 150°C by losing free water and bound water (Fengel and Wegener 1984). As the temperature rises up to 180-250°C (a range in which thermal modification occurs), wood undergoes significant chemical transformations. And, above 250°C, pyrolysis takes place which results in carbonization of wood and formation of CO, CO2, and other pyrolysates (Bourgois and Guyonnet 1988, Esteves and Pereira 2009). In temperatures higher than 300°C, wood undergoes a severe degradation (Kim et al. 2001). However, since simultaneous endothermic and exothermic processes occur during heating, it is difficult to determine the exact temperatures at which the different reactions become dominant (Hill 2006). A summary of alterations of wood constituents due to heat treatment are listed in Figure 2.4.  Figure 2.4 Changes in the wood components during heat treatment  (From Navi and Sandberg 2012). 2.3.1. Hemicelluloses    According to Fengel and Wegener (1984), as hemicelluloses have low molecular weight and a branching structure, they are the first cell wall polymers that start to degrade at low temperatures. As wood is heated, hemicelluloses side chains (hydroxyl groups) are split and acetic acid forms due to hydrolysis (Kollmann and Fengel 1965, Tjeerdsma et al. 1998a). The released acetic acid catalyzes depolymerization of hemicellulose and increases the decomposition of polysaccharides (Klauditz and Stegmann 1955, Tjeerdsma et al. 1998a). Subsequently, decomposed carbohydrates are dehydrated to aldehydes, resulting in the formation of furfural from pentoses and 11  hydroxymethylfurfural from hexoses. It should be noted that pentosans are more vulnerable to thermal degradation than hexosans, which is why hardwoods with higher amounts of pentosans are less thermally stable than softwoods (Burtscher et al. 1987, Fengel and Wegner 1984, Gérardin 2016).  2.3.2. Cellulose Cellulose is considered more resistant to thermal degradation than hemicellulose due to its high degree of polymerization and crystallinity and therefore degrades at much higher temperatures (Bourgois and Guyonnet 1988, Esteves and Pereira 2009, Gérardin 2016). As the treatment temperature rises, the amorphous regions of cellulose, which are more susceptible to degradation, begin to disintegrate progressively and lead to an increase in cellulose crystallinity that consequently decreases the accessibility of hydroxyl groups to water molecules (Bhuiyan et al. 2000, Boonstra and Tjeerdsma 2006). These amorphous regions presumably show comparative thermal properties to the hexose components of hemicellulose and form furans such as hydroxymethylfurfural and furfural as they degrade (Hill 2006, Navi and Sandberg 2012). While crystalline regions of cellulose consist of highly ordered cellulose chains which limit the acid diffusion throughout the cellulosic structure during hydrolysis and provide a great stability up to 300°C (Fengel and Wegener 1984, Kim et al. 2001). Extended heating results in chain scission of the cellulose and breakdown products, which are not an issue here, since it happens over the range of heat treatment temperatures (Johansson 2008, Navi and Sandberg 2012). 2.3.3. Lignin According to Sanderman and Augustin (1964), lignin is considered the most stable compound of wood during heat treatment; yet, it will experience some structural changes. These structural modifications include cleavage of methoxyl groups in C3 position of the aromatic nuclei of lignin, which results in its increased reactivity. As the temperature rises, propane side chains are cleaved, resulting in lignin weight loss (Boonstra 2008, Kim et al. 2014). More weight loss occurs due to depolymerization of lignin when ether linkages such as β-O-4 and α-O-4 are cleaved and phenolic compounds are released. Simultaneously, CO homolysis results in the formation of free radicals which generate condensation reactions and cross-linking within the lignin (Tjerdsmaa et al. 1998b, Westermark et al. 1995, Kim et al. 2014). Extended heating increases the free radical 12  content and consequently increases the portion of methylene bridges connecting two phenolic nuclei leading to more extensive cross-linking of the lignin network (Hill 2006, Boonstra 2008). These reactions could explain the origins of the improved dimensional stability in heat-treated wood (Johansson 2008, Gérardin 2016). 2.3.4. Extractives Extractives are not the structural constituents of the wood and most of them can easily evaporate while heat-treated (Finnish ThermoWood Association 2003). During the first stage of heat treatment, most of the extractives especially the volatiles like terpenes that naturally exist in wood might evaporate (Mohareb et al. 2010, Navi and Sandberg 2012). While, at temperatures from 160°C, new extractives can appear due to degradation of cell wall material and as the treatment time and temperature increases, these new compounds could undergo more degradation. The by-products of these reactions can either remain in wood or evaporate (Gérardin 2016). Moreover, migration of low-molecular-weight extractives such as fatty acids, fats, and waxes to the surface of wood during heat-treatment affects the surface properties, causing difficulties in adhesion of coatings (Nuopponen et al. 2003, Inari et al. 2011).  Alterations in Physical Properties  2.4.1. Hygroscopicity and Dimensional Stability As mentioned in section 2.3, chemical changes in the cell wall at high treatment temperatures lower the accessibility of hydroxyl groups to water, which consequently results in lower hygroscopicity. Depending on how the treatment is performed, the reduction of the equilibrium moisture content (EMC) is about 0–50% (Chirkova et al. 2005). This variation in EMC highly depends on treatment intensity and the characteristics of each species (Ferrari et al. 2012, Yang et al. 2016). Esteves et al. (2007) found a decrease of 46% to 61% in maritime pine (Pinus pinaster Aiton) and blue gum (Eucalyptus globulus Labill.) wood samples when thermally modified at 190, 200, and 210°C for 2, 6, and 12 hrs (Unsal et al. 2009). Akyildiz and Ates (2008) examined the effect of treatment temperature (130, 180, and 230°C) and holding time (2 and 8 hrs) on sessile oak (Quercus petraea Lieb.), chestnut (Castanea sativa Mill.), calabrian pine (Pinus brutia Ten.) and black pine (Pinus nigra Arnold.) and concluded that the highest decrease in EMC (47.9%) 13  occurred in the most intense treatment condition (at 230°C for 8 hrs) while lowest decrease (2.9%) happened at the least intense conditions (at 130°C for 2 hrs).  Decreased EMC in the heat treated wood results in enhanced dimensional stability, which is described by anti-swelling efficiency (ASE). Similar to EMC, many authors found ASE to be highly depended on both temperature and time of the treatment (Yildiz 2002, Bekhta and Niemz 2003, Welzbacher et al. 2007, Korkut et al. 2008, Poncsak et al. 2011).  Guller (2012) studied the heat-treatment of black pine and reported the highest ASE of 56-66% was attributed to the most severe treatment condition (at 225°C for 3 hrs), while the lowest ASE (7-13%) belonged to the lowest temperature and time (at 190°C for 1 hr). He also concluded that the treatment temperature is more effective on ASE change than the duration of treatment, especially for temperatures over 200°C. Yang et al. (2016) investigated the effect of three temperatures (170, 190, and 210°C) and treatment times (1, 2, and 4 hrs) on ASE of Japanese red cedar (Cryptomeria japonica (L.f.) D. Don) and reported an increase from 20.3% up to 54%, as treatment conditions got more severe.  According to Chang and Keith (1978), the improved dimensional stability of heat-treated wood relies upon the unique features of species used, and it is more apparent in the radial direction due to anisotropic properties of wood. However, Tjeerdsma et al. (1998a) determined ASE of 10%, 13%, 11% and 35% in radial, and ASE of 13%, 23%, 40% and 40% in the tangential direction of heat treated beech (Fagus spp.), birch, spruce, and radiata pine (Pinus radiata  D. Don.), respectively. Esteves et al. (2007) reported an ASE of 57% and 90% in the radial direction of maritime pine and blue gum heat-treated at 190, 200, and 210°C for 2, 6, and 12 hrs. They also mentioned that the improvement in dimensional stability was slightly higher in the tangential direction. Nevertheless, Mitani and Barboutis (2014) reported much higher swelling (less dimensional stability) in tangential than radial direction in heat-treated European beech (Fagus sylvatica L.) at 180°C, for 2, 4, 6, 8, and 10 hrs. However, as the duration of treatment extended, the swelling in both directions (especially in tangential) decreased compared to the untreated specimens. In addition, other parameters such as treatment atmosphere, whether the system is open or closed (Seborg et al. 1953), and the presence of a catalyst while heating (Stamm and Baechler 1960) are determined to have a high impact on the dimensional stability of heat treated wood.  14  2.4.2. Density Density is one of the physical properties of wood, which prominently affects almost all the material properties during heat-treatment. Density is mass over volume and decreases after heat-treatment. This reduction could be the result of both mass loss due to degradation of wood components into volatiles, and dimensional changes after heat-treatment (Boonstra 2008, Källander 2016).  Many researchers showed that there is a general decreasing tendency in density of wood by increasing the treatment time and temperature (Yildiz 2002, Finnish ThermoWood Association 2003, Korkut et al. 2008, Gunduz et al. 2008). Chotikhun and Hiziroglu (2016) investigated the effect of longer treatment times on the density of red oak (Quercus falcata Michx.). They found that the specimen treated at the temperature of 190°C for 3 hrs had density value of 642 kg/m3 which further reduced to 628 kg/m3 when exposure treatment duration was increased to 8 hrs. Moliński et al. (2016) investigated the effect of higher temperature on the density of Ash (Fraxinus excelsior L.). They found a 3.5% reduction in density after treating at 190°C for 2 hrs while increasing the temperature to 200°C resulted in 9.3% reduction. By heat-treatment of black pine at a combination of maximum temperatures (190, 200, 212, and 225°C) and holding times (1, 2, and 3 hrs), Guller (2012) found the highest decrease in density (12.6%) occurred at 225°C and 3hrs. At the same temperature, decreasing the treatment duration to 2 hrs resulted in 10.7% decrease; while decreasing the temperature for the same treatment duration (2 hrs) reduced the density only by 4.5%. He then concluded that treatment temperature has more impact on the density compared to treatment duration. For the same treatment conditions, Ghalehno and Nazerian (2011) also reported a close result of 4.9% reduction in density of European hornbeam (Carpinus betulus L.). However, only 3% decrease reported by Metsä-Kortelainen et al. (2006) for heat-treated Scots pine at the same conditions. The effect of species could also be observed in the work of Chaouch et al. (2010) where they determined a decrease in density of 5 different species namely beech, poplar, ash, pine, and fir (Abies spp.), after heat-treatment at 230°C and the greater decrease in wood density was obtained for samples of greater starting density. 2.4.3. Color Color plays a significant role in the final consumer since it is the key factor for choosing specific wood when the decorative and visual perspective of the end application is important. In 15  this regard, thermal modification could be beneficial by turning a low-value wood into a darker wood, which is more valuable (Hoang 2009, Candelier et al. 2016). According to Gunduz et al. 2009, the new color becomes uniform throughout the thickness of the wood. Patzelt et al. (2003) suggested that color could be used as a classification method of modified wood due to its correlation with treatment intensity, mass loss, and the thermal process used. Poncsak et al. (2011) demonstrated that for Jack pine (Pinus banksiana Lamb.) the color turns out to be significantly darker and slightly reddish with increasing the maximum heat treatment temperature. However, raising the heating rate did not have any considerable effect on the color. Similarly, Guller (2012) concluded that increasing the temperature (from 190°C to 225°C) and/or duration of heat treatment (from 1 to 3 hrs) gives the black pine wood a uniformly darker color. Therefore, the color could be used as an indicator of the severity of process conditions. Pleschberger et al. (2014) also heat treated Norway spruce (Picea abies L. Karst.) and ash at maximum temperatures of 200, 210 and 220 °C for 2.5 hrs and concluded that increasing the intensity of heat treatment results in the decrease of the lightness. They also found that there is a good linear correlation between the maximum breaking load and the lightness on all the surfaces of both species, which could be beneficial in the non-destructive assessment of strength for quality control of heat-treated wood.  To provide a quality control indicator, many other attempts have been made to find a relationship between color change, treatment conditions, and mechanical properties of thermally modified wood. Bekhta and Niemz (2003) reported that heat treatment at 200°C induced significant darkening and reddening of spruce wood. They also found a strong correlation between total color differences and both bending strength and modulus of elasticity, which suggests that color could be used for predicting the wood strength properties. However, Johansson and Moren (2006) concluded that color is not a proper predictor of wood quality since the distribution of color through the modified boards is not homogeneous. Regardless, color was introduced as the best predictor of most of the classical mechanical strength properties by González-Peña and Hale (2009) only within a temperature range of 190-245°C and a heating duration of 0.3-16 hrs.  It should be mentioned that in addition to the severity of the treatment conditions (higher temperatures and longer heating times), the use of air instead of steam during heating, would also result in darker shades of modified wood (Hill 2006, Aydemir et al. 2012). Wood moisture content is also another effective parameter on color change. Brauner and Conway (1964) showed that green walnut wood specimens experience greater and more rapid darkening than air-dried specimens. 16   Alterations in Mechanical Properties 2.5.1.     Strength  The main reason that limits the use of thermally modified wood for most construction applications is the increased brittleness and reduced mechanical strength (Hill 2006, Johansson 2008, Esteves and Pereira 2009). Degradation of hemicellulose, and to a lesser extent lignin degradation at elevated temperatures result in lower bending and tension strengths of heat-treated wood. Therefore, the damaged cell wall matrix results in reduced modulus of rupture (MOR) of wood (Niemz et. al 2010, Lekounougou et al. 2011, Navi and Sandberg 2012). While in many cases, an increase in modulus of elasticity (MOE), has been reported and attributed to crystallization of cellulose and cross-linking of lignin (Gunduz et al. 2009, Esteves and Pereira 2009).  Millett and Gerhards (1972) heated several dried species at temperatures between 115°C to 175°C and found an initial increase in MOE whereas higher treatment temperatures led to diminished MOE. These contradictory results in MOE of woods have been reported many times and could be explained by various thermal modification conditions such as time and temperature of treatment, initial moisture of samples, open or closed heating system, heating in air or anaerobic conditions, and intrinsic characteristics of each wood species. In addition, hardwoods exhibit higher strength losses than softwoods (Stamm et al. 1946, Militz 2002, Hill 2006).  Bekhta and Niemz (2003) thermally modified conditioned spruce at 200°C for 2, 4, 8, and 10 hrs and at 100°C, 150°C, and 200 for 24 hrs. They showed that both MOR and MOE decreased when the treatment temperature was over 100°C. The bending strength drastically decreased by 44-50%, while the modulus of elasticity was hardly affected (only 4-9%). Using the ThermoWood process, the same 50% decrease in MOR and insignificant change in MOE of structural timber of Scots pine and Norway spruce was reported by Bengtsson et al. (2002).  Shi et al. (2007) thermally modified five Quebec wood species, namely, spruce,  pine (Pinus spp.), aspen (Populus spp.), and birch at 200°C, 202°C, and 212°C for 3 hrs and found that higher treatment temperatures have a smaller impact on reduction of MOE than MOR. Intensified treatments resulted in the loss of MOR from 0% to 49% in all species except for birch where it increased slightly after treatment; however, they affected MOE to a lesser extent. A 4% to 28% decrease was found in MOE of thermally modified spruce and pine, while for fir, aspen and birch, it increased from 15% to 30% at higher temperatures.  17  Similarly, Yang et al. (2016) found that increased treatment intensity resulted in lower MOR and less significantly MOE values in thermally modified Japanese cedar (Cryptomeria japonica) treated at 170°C, 190°C, and 210°C for 1, 2, and 4 hrs. In addition to the maximum treatment temperature and holding time, Lekounougou et al. (2011) investigated the effect of heating rate on strength properties of thermally modified white birch (Betula papyrifera Marshall). The results show a marked decrease in MOR with increasing temperature while MOE did not seem to be affected. However, increasing the holding time increased MOE while it did not seem to affect MOR. Increasing heating rate also resulted in lower MOE, but insignificant difference in MOR. As the only investigation on thermal modification of western hemlock, Williams (2011) reported a slight increase in both MOR and MOE of samples when treated at 180°C, but only an increase in MOE when treated at 212°C.  Dynamic modulus of elasticity (Ed) of thermally modified wood was also investigated by the means of non-destructive testing by several researchers. Del Menezzi et al. (2014) showed that thermal modification of marupa (Simarouba amara Aubl.) at 160 ºC for 180 min and 200 ºC for 70 min improved Ed at a higher temperature even though the density was decreased. Garcia et al. (2012) thermally modified eucalypt wood (Eucalyptus grandis Hill ex Maiden) at 180°C, 200°C, 215°C, and 230°C for 15min, 2 and 4 hrs. Using the stress wave nondestructive method, they revealed that Ed decreased by approximately 13% at the most severe treatment conditions (230°C for 4 hrs), while the other treatment conditions did not significantly make a change. Similarly, Borůvka et al. (2015) concluded that higher treatment temperatures correspond to lower stiffness and strength properties when Douglas-fir and common alder (Alnus glutinosa Gaertn.) were thermally modified at 165°C and 210°C for 3 hrs. This decrease was more evident for common alder with lower density and lower hemicellulose content (because of being a hardwood), compared to Douglas-fir. 2.5.1.     Hardness Hardness is defined as the resistance of a material to indentation (Garratt 1931) and is a commonly tested mechanical property for woods used in paneling, furniture, flooring, and decking, which can reflect their durability (Bowyer et al. 2003, Leitch 2009).  According to previous studies, the hardness of thermally treated woods may increase or decrease based on several factors during thermal modification. 18  Yildiz (2002) observed a 25.9%, 45.1%, and 41.8% decrease in longitudinal, radial, and tangential hardness of oriental beech (Fagus orientalis Lipsky) treated at 180°C for 10 hrs, while for spruce (Picea spp.), the observed decrease was 19.7%, 43%, and 42.5% longitudinally, radially, and tangentially. For the same treatment conditions, Unsal and Ayrilmis (2005) reported a 23.9%, 44.2%, and 33.6% decrease in the longitudinal, radial, and tangential hardness of river red gum (Eucalyptus camaldulensis Dehnh.). Sundqvist et al. (2006) also investigated European white birch (Betula pubescens Ehrh.) at the same treatment temperature 180°C, and two more temperatures 160°C, and 200°C for 1, 2.5, and 4 hrs. Treatment for 2.5 and 4 hrs, decreased the hardness drastically as the temperature raised; however, samples treated for 2.5 hrs showed slightly higher hardness at 200°C compared to 180°C. For treatment duration of 1 hr, they found a slight decrease in the hardness, when treated at 160°C and then a continuous increase as temperature increased. Similarly, Percin et al. (2016) thermally modified oriental beech at 150°C, 175°C, and 200°C for 1, 3 and 5 hrs and reported an increase in hardness at the initial stage of treatment and a decrease afterward. Chotikhun and Hiziroglu (2016) reported a 12% and 24% reduction in tangential hardness of southern red oak due to heat-treatment at 190°C for 3 and 8 hrs, respectively.  Poncsak et al. (2006) studied maximum treatment temperature, holding time at this temperature, heating rate, and gas humidity on the mechanical properties of white birch and found that the hardness increases slightly with temperatures above 200°C. They also concluded that the longer heat-treatment times and consequently slower heating rates result in increased longitudinal hardness in birch. At the same temperature, a 36%, 22%, and 37% rise in the radial, tangential, and longitudinal hardness obtained by Shi et al. (2007) for birch; while, aspen exhibited a 26%, 39%, and 15% decrease in radial, tangential, and longitudinal hardness, respectively. The hardness of heat-treated spruce increased in all directions at 212°C; however, by increasing the duration of heat treatment, a slight 6% decrease was observed in tangential hardness. The same happened to fir by increasing the temperature except in this case there was a marked decrease of 27% in the tangential hardness. They concluded that the hardness of the thermally modified wood increases or decreases depending on the species, directions of the test (radial, tangential, and longitudinal), and treatment intensity.  According to Rowell et al. (2012), degradation of hemicelluloses and consequently the damaged lignin-hemicellulose matrix reduces the load sharing capacity and the effective stress transfer in wood which might have a negative impact on the hardness. Lekounougou et al. (2011) 19  suggested that the increase in the hardness might be as a result of the increase in the amount of the highly ordered crystalline cellulose due to the crystallization of amorphous cellulose supporting the wood structure which happens at high temperatures above 180°C (Boonstra and Tjeerdsma 2006). Boonstra (2008) also referred the slightly improved hardness to an increased cross-linking of lignin because of heat treatment.   Western Hemlock Western hemlock is an important commercial softwood species of the Pacific coast and the northern Rocky Mountains. It grows from Alaska to northern California mostly on the west side of Coast Mountains and covers the entire coast of British Columbia. It is also scattered over the humid interior of British Columbia from Prince George to northern Idaho and Montana on the west side of Rocky Mountains (Figure 2.5).The best climate for the growth of this species is mild, humid weather where there are frequent fog and precipitation during the growth season (Packee 1990, Pojar et al. 1991). At these culminating growth conditions, western hemlock trees can reach a diameter of 3-4 ft and a height of 175-225 ft (Johnson and Gibbons 1929).     Figure 2.5 The distribution map of western hemlock (Natural Resources Canada, 2015) Western Hemlock is quite light in weight and soft, with fairly straight and even grain. It has a medium to fine, coarse, and somewhat lustrous texture. The color of both sapwood and heartwood are whitish to light yellowish brown, which makes it hard to distinguish between them. However, sapwood is a little lighter than heartwood and is roughly 25 mm wide (Alden 1995). The growth rings are distinct and fairly narrow but uniform in width, delineated by a band of darker latewood with purplish to reddish brown tinge that is distinct to the naked eyes. It could be easily seen that 20  approximately two-thirds or more of the growth ring is occupied with early wood and the transition from earlywood to latewood is almost gradual (Panshin and deZeeuw 1980, Hoadley 1990).  Western hemlock is tasteless, non-resinous, and odorless when dry; but it has a sour odor when green or containing wetwood (Panshin and deZeeuw 1980, Hoadley 1990). Wetwood is an undesirable phenomenon frequently occurs in the heartwood of western hemlock and results in a higher density and moisture content in comparison with adjacent areas (Hartley et al. 1961); consequently resulting in uneven drying of western hemlock, chemical stains, distortions, lower dimensional stability and lower strength (Ward and Pong 1980).   The basic density of this wood in green form ranges between 470 to 490 kg/m3 which is much higher compared to many other conifers (Rohrbach 2008). In addition, its average moisture content in green form reported to be 85% in heartwood and 170% in sapwood (Ross 2010). When drying, this high initial moisture content results in a relatively high 12.4% volumetric, 4.2% radial, and 7.8% tangential shrinkage of the wood (Ross 2010). With its moderately high hardness, stiffness, and strength, along with its satisfactory machinability, western hemlock is considered one of the popular species for general construction in North America and overseas (Mullins and McKnight 1981, Alden 1995). Western hemlock is mostly being exported to Japan (Lazarescu and Avramidis 2012), to be used for posts and beams in traditional housing and in smaller sizes for roof rafters. The straight grain and medium to fine texture of western hemlock that allows it to sand to a smooth silky finish with nearly no tendency to split appeals to remanufacturers and wood workers. The moderate nail and screw-holding ability without splitting, satisfactory glueability and adhesive-holding, easy sawing without splintering, and high acceptability of a variety of finishes make western hemlock a good choice for builders. Accordingly, it is widely used for building material such as flooring, siding, roof decking, studding, planking, joists, rafter, doors, and windows. Other applications include cabinetry, plywood manufacturing, boxes, crates, and pallets. Smaller quantities are also used for furniture and ladders. It is also the most important pulpwood species in British Columbia (Mullins and McKnight 1981, Wallace et al. 2003, Softwood Export Council 2004, Rohrbach 2008).   Objective The current study aims to develop effective thermal modification techniques for treatment of Western hemlock by optimizing the ThermoWood process in terms of temperature and sawing 21  pattern of samples, in order to improve dimensional stability, minimize equilibrium moisture content and water uptake and achieve uniform color changes without significantly affecting the wood’s basic density, stiffness and hardness.  The results are expected to provide the local industry with preliminary information and an option for thermal modification of western hemlock at reduced times and cost together with enhanced productivity and quality of produced lumber. In turn, such environmentally friendly treatment is expected to add value to this important BC wood species and open up new opportunities for its application.    22  3. Materials and Methods 3.1. Materials Fifty-six flatsawn (FS) and fifty-six quartersawn (QS) kiln-dried and defect free local Western hemlock boards with dimensions of 35. 6 × 101.6 × 1432.6 mm3 were supplied from Tenryu Canada Corp. Eighty-four boards were randomly picked and subjected to ThermoWood® process in a Jartek industrial heat treatment kiln at Scottywood Corporation located in Abbotsford, BC (Figure 3.1) using the schedules listed in Table 3.1.  In order to investigate the impact of both treatment temperature and the cut (flatsawn and quartersawn), each run of treatment included seven boards per cut. In addition, fourteen boards per cut were kept untreated as controls. Table 3.1 Heat treatment protocol. Treatment Run Max. Temperature (°C) Holding  Time (h) quartersawn boards flatsawn boards Total boards per treatment 1 230 2 7 7 14 2 230 2 7 7 14 3 212 2 7 7 14 4 212 2 7 7 14 5 170 2 7 7 14 6 170 2 7 7 14 Control — — 14 14 28 Total 56 56 112  Figure 3.1 The Jartek industrial thermal modification kiln at Scottywood Corporation (Abbotsford, BC). 23  3.2. Heat Treatment  ThermoWood® process was carried out at three different maximum temperatures of 230°C, 212°C, and 170°C, and 2 hours of holding time at these maximum temperatures (Figure 3.2) which resulted in a total number of three treatments. Moreover, to meet the proper experimental design requirements, a replicate of each run of treatment was also implemented (Table 3.1).  Upon completion of the treatments, all boards were shipped to UBC for further processing and assessment. To minimize the moisture loss, all boards were wrapped in plastic.  Figure 3.2 Scheme for thermal modification process in accordance with suggested schedules. 3.3. Sample Cutting Protocol As shown in Figure 3.3, both thermally modified (TM) and unmodified boards were planed and cut into shorter lengths according to the corresponding standards of each test. Then, they were all kept in a conditioning room at relative humidity of 65±7% and temperature 20±3°C for at least 4 weeks to reach equilibrium moisture content. In total, the number of samples per test was 336 (three samples from each board per test).   0204060801001201401601802002202402600 4 8 12 16 20 24 28 32 36 40Temperature (°C)Time (hours)Run 1 and 2Run 3 and 4Run 5 and 624   Figure 3.3 Cutting protocol of samples. 3.4. Dimensional Stability and Water Absorption To evaluate the improvement of dimensional stability as a function of treatment, ASE (anti-swelling/shrinkage efficiency) is commonly used. ASE expresses the difference between swelling of treated and untreated wood at relative humidity between 30-90%, similar to the end-use conditions of wood (Militz 2002, Esteves and Pereira 2009). According to ASTM 1037-12 Standard (ASTM Intl. 2012), treated and untreated samples with dimensions of 33 × 101.6 × 101.6 mm3 were submerged in 20 ± 3°C water, 25.4 mm under the water surface, at room temperature of 25°C for 24 hrs. To achieve this water temperature, a mixture of cold and hot water was used while the temperature was monitored with a thermocouple. In order to maintain the 20°C temperature, the water was replaced after each measurement.  Before and after submerging, the samples were weighed with a digital balance (precision ±0.001 g) and their dimensions were measured with a digital caliper with ±0.01 mm accuracy at marked positions as shown in Figure 3.4, for calculating their volume.       Figure 3.4 Measuring: a) the thickness at four positions, b) the length and the width at three positions. a b Reject Reject 25           Figure 3.5 Submerged samples in the water bath.  Upon completion of the test, all samples were oven-dried at 103±2°C for 24h, and their dimensions and weight were re-measured. Then, using an average value of measured dimensions, the volumetric swelling coefficient (S) was determined as below (Hill 2006, Guller 2012): S (%) = 𝑉𝑉𝑠𝑠−𝑉𝑉𝑜𝑜𝑉𝑉𝑜𝑜 ×100                                                                                                           (1)                                                                        where Vs is the volume after wetting with water (mm3), and Vo is the oven-dried volume (mm3) of the samples. Anti-swelling efficiency (ASE) as a term to describe the degree of dimensional stability was then determined as:  ASE (%) = 𝑆𝑆𝑡𝑡−𝑆𝑆𝑐𝑐𝑆𝑆𝑐𝑐 ×100                                                                                                       (2) where St is the volumetric swelling coefficient of treated samples, and Sc is the volumetric swelling coefficient of control samples. Water absorption (WA) by the samples was also calculated as below:  WA (%) = 𝑊𝑊𝑠𝑠−𝑊𝑊𝑜𝑜𝑊𝑊𝑜𝑜 ×100                                                                                                                   (3) where Ws is the wet weight of samples after soaking in water, and Wo is the oven-dry weight of samples. 26  3.5. Equilibrium Moisture Content The test for evaluating the equilibrium moisture content was carried out according to ASTM D 4442-16 standard (ASTM Intl. 2016) on the same samples used in 3.4. As mentioned earlier, the samples were kept in a conditioning room (relative humidity 65±7%, temperature 20± 3°C) for 4 weeks to reach equilibrium. Therefore, the EMC of samples could be determined as below: EMC (%) = 𝑚𝑚𝐶𝐶−𝑚𝑚0𝑚𝑚0 ×100                                                                                                        (4) where mC is the weight of conditioned samples (g), and mo is the oven-dried weight of the samples (g). 3.6. Basic Density According to the ASTM D2395-17 Standard Test Method (ASTM Intl. 2017), the same treated and untreated samples used for determining EMC were used to calculate the basic density:  ρbasic = 𝑚𝑚𝑜𝑜𝑉𝑉𝑀𝑀 ×100                                                                                                                                        (5)         where ρbasic is the basic density (g/cm3), mo is the oven-dried mass (g), and VM is the volume (cm3) of conditioned samples.  3.7. Color The most common method used in wood color studies is called CIE L*a*b*.  There, the color is introduced as three coordinates in 3-D (Figure 3.6): lightness (L*) that ranges from black (0) to white (100), and chromatic coordinates a* ranges from green (-100) to red (100), and b* ranges from blue (-100) to yellow (100). Color changes (ΔE) after any treatment has also a parabolic correlation with these coordinates and could be calculated afterwards. The similarity between this system and human vision makes it advantageous for camera or scanner imaging editing (Johansson 2008, Esteves and Pereira 2009). 27   Figure 3.6 The three-dimensional CIEL*a*b* color space (Johansson 2008). The color test was carried out on three marked positions (in the middle, and 25mm from both ends) on the surfaces of 33 × 48 × 254 mm3 samples, using a Minolta spectrophotometer (Model #CM-2600d) with a 6-mm sensor head and using a D65 illuminant and a 10-degree Standard observer based on the ASTM D2244-16 (ASTM Intl. 2016) test standards. The spectrophotometer records two sets of parameters namely SCI (specular component included) and SCE (specular component excluded) (Konica Minolta 2007). Here, the SCI parameters were selected based on the previous studies, and the fact that SCI values measure the true color of an object without considering the surface conditions. After measuring the color coordinates, the mean values were used to calculate the difference in the lightness (∆L*) and chroma coordinates (∆a* and ∆b*). Finally, the total color change (∆E*) was calculated as below (Bekhta and Niemz 2003, Johansson 2008): ∆Eab* = [(∆L*) 2+ (∆a*) 2+ (∆b*) 2] 1/2                                                                                       (6)  Figure 3.7 Minolta Colorimeter. 28  3.8. Stiffness  Stiffness was determined by calculating the modulus of elasticity with a non-destructive evaluation method using a Metriguard 239A stress-wave timer. The tests were performed on the same samples subjected to color evaluation. The stress wave timer consists of an impact device (pendulum ball) to initiate the stress wave and two accelerometers (start and stop) to detect the stress wave signal (Figure 3.8).   Figure 3.8 Metriguard 239A Stress Wave Timer set up to measure the stiffness.  The pendulum was used to create 10 consecutive stress waves. The time taken for the waves to travel over a distance of 214 mm was averaged and used to calculate the stress wave velocity as below (Dunn 1992, Gray et al. 2008): V= 𝐿𝐿𝑡𝑡                                                                                                  (7) where V is the velocity (m/s), L is the propagation distance (m), and t is the time of propagation after subtracting the zero offset time (s). Finally, the dynamic modulus of elasticity (ED) calculated from the stress wave velocity using the following formula: Timer Impact ball Sample holder Sample holder Start accelerometer Stop accelerometer Sample 29   ED= V2 ρ                                                                                                                           (8) where ED is the dynamic modulus of elasticity (Pa), V is the velocity (m/s), ρ is the wood density at equilibrium moisture content (kg/m3).  3.9. Hardness The general term for the hardness test used is the Janka one. The test performed in accordance with the ASTM D143-14 standard (ASTM Intl. 2014) on samples with dimensions of 33 × 48 × 152.4 mm3 using a Universal testing machine equipped with a Janka ball hardness testing tool with a diameter of 11.3 mm and the penetration rate of 6 mm/min. The maximum load required to embed the ball to one-half of its diameter is then measured. Although ASTM prescribes 50.8 mm thickness for hardness testing, the thickness of the available boards is limited to 33 mm, and according to Green et al. (2006), hardness values are independent of thickness over 25–75 mm.  Two penetrations were made on the tangential surface, two on the radial surface, and one on each end. The radial and tangential hardness values were averaged and tabulated as “side hardness” (Green et al. 2006). In addition, the average of values from both ends corresponds to longitudinal hardness.  Figure 3.9 Hardness test using a Universal testing machine. 30  3.10. Statistical Analysis  Data assessment was performed by means of general descriptive statistics, and statistical analysis using R software. A linear mixed-effects model and likelihood ratio tests were carried out to find out if the fixed factors (temperature, cut, and their interaction) have a significant impact on the treatment’s means. Moreover, Tukey’s test for multiple comparisons among means was carried out to rank the levels of each significant factor. Prior to statistical data analysis, tests for data normality and variance homogeneity were done in order to verify whether data respected the assumptions of the test. In case of abnormality, log-transformations were applied. The experimental design including the fixed and random factors and sources of errors are listed in Table 3.2. Table 3. 2 Experimental design. Source                      df Type Whole Plot Temperature  (4 levels)                          3 Fixed-Effect Error 1: Temperature Reps.        4 Random-Effect Subtotal  7 - Split-Plot Cut (2 levels) 1 Fixed-Effect Cut *Temperature 3 Fixed-Effect Error 2: Cut *Temperature Reps. 4 Random-Effect Subtotal 15 - Error 3 Boards (Split-plot Reps.) 12 Random-Effect Boards*Temperature 36 Random-Effect Boards*Temperature Reps. 48 Random-Effect Subtotal 111 Random-Effect Subsamples Error 4: samples from boards 224 Random-Effect Total= no. samples-1 335 -        31  4. Results and Discussion 4.1. Basic Density  The amount of cell wall material in wood could be measured by basic density, which is a key indicator of wood quality (Panshin and DeZeeuw 1980). In this regard, the average basic density of treated and untreated samples was measured and listed in Table 4.1. The highest basic density is for control samples with an average of 455.67 kg/m3 (CV=4.94%), followed by an average basic density of 448.57 kg/m3 (CV=4.83%), 441.01 kg/m3 (CV=4.10%), and 436.74 kg/m3 (CV=4.84%) for treatment at 170°C, 212°C, and 230°C, respectively. Table 4.1 Descriptive statistics of the basic density of samples. Temperature (°C) Cut Mean (kg/m3) StDev (kg/m3) CoefVar (%) Min (kg/m3) Max (kg/m3) Control FS 455.84 19.8 4.34 424.14 493.21 QS 455.51 25.18 5.53 424.41 517.09 170 FS 448.50 25.97 5.79 401.98 500.72 QS 448.60 16.66 3.71 418.1 478.11 212 FS 443.85 16.51 3.72 396.75 469.14 QS 438.15 19.30 4.41 406.49 478.86 230 FS 437.48 27.94 6.39 381.57 474.65 QS 436 11.05 2.53 415.61 454.45 Figure 4.1 shows that basic density has a downward trend as temperature rises. The basic density of samples decreased by 1.6% when treated at 170°C, while it decreased for 3.2% and 4.1% when treated at 212°C, and 230°C. These results are in good agreement with the only density values reported for thermally modified western hemlock by Williams (2011). However, the treatment conditions of his study were different. Williams (2011) reported a 1.5% decrease in density of samples treated at 180°C when controls had a density of 466 kg/m3, and 6.3% decrease for samples treated at 212°C while the controls had a density of 475 kg/m3. As stated earlier in section 2.4.3, the reason for this reduction in basic density is the degradation of wood components and mass loss during thermal modification. 32   Figure 4.1 The average basic density of treated and untreated samples. Error bars indicate standard error.  Statistical analysis using likelihood ratio (LR) test was performed to find the effect of fixed-factors on basic density (Table 4.2). At a significance level of 0.05, results show that temperature is the only factor that has a significant effect on basic density with test statistics of LR= 8.64, and P-value= 0.034. In addition, a pairwise Tukey’s test with 95% confidence level was carried out to find if the basic density in each level of temperature is significantly different from one another (Table 4.3). The temperatures that do not share a letter are significantly different. The grouping indicates that except for 170°C, the mean of basic density of samples treated at the other two temperatures significantly differs from controls. Table 4.2 Likelihood ratio tests for basic density. No. Model LogLik L.Ratio P-Value 1 Ln(ρbasic)~ Intercept+ Temperature+ Cut +Temperature* Cut 564.6429 0.8562993 0.836 2 Ln(ρbasic)~ Intercept+ Temperature+ Cut 564.2148 2 Ln(ρbasic)~ Intercept+ Temperature+ Cut 564.2148 8.635844 0.0345 3 Ln(ρbasic)~ Intercept+ Cut 559.8968 2 Ln(ρbasic)~ Intercept+ Temperature+ Cut 564.2148 0.3582635 0.5495 4 Ln(ρbasic)~ Intercept+ Temperature 564.0356 Table 4. 3 Tukey’s pairwise comparisons for basic density at different levels of temperature. Temperature (°C) Mean (kg/m3) Grouping Control 455.67 A   170 448.57 A B  212 441.01  B C 230 436.74   C 410420430440450460470Control 170 212 230Basic density (Kg/m3 )Temperature (°C)33  4.2. Water Absorption and Equilibrium Moisture Content The descriptive analysis of WA and EMC are given in Table 4.4. Comparing WA and EMC values at different treatment temperatures indicates that they both followed a downward trend as temperature raised (Figure 4.2). The highest water absorption was attributed to the control samples with an average WA of 30.98% (CV= 12.04%) and average EMC of 12.41% (CV= 5.85%). Then as temperature raised, both WA and EMC decreased gradually to their minimum values of 16.06% (CV=10.11%) and 5.07% (CV=6.95%) at 230°C, respectively. Table 4.4 The descriptive statistics of water absorption (WA) and equilibrium moisture content (EMC) of samples. Property Temperature (°C) Cut Mean (%) StDev (%) CoefVar  (%) Min  (%) Max  (%) Water Absorption Control FS 32.36 3.981 12.30 26.06 41.54 QS 29.59 2.89 9.78 23.55 36.29 170 FS 24.67 3.29 13.34 20.10 33.10 QS 25.05 2.41 9.61 19.63 30.43 212 FS 19.09 2.04 10.70 14.40 23.53 QS 17.13 1.52 8.87 14.44 21.84 230 FS 15.87 1.93 12.18 12.03 19.29 QS 16.24 1.24 7.61 14.21 18.95 Equilibrium Moisture Content Control FS 12.47 0.76 6.09 11.26 13.94 QS 12.35 0.70 5.64 11.44 13.85 170 FS 8.53 0.49 5.74 7.66 9.25 QS 8.61 0.47 5.46 7.62 9.39 212 FS 5.99 0.43 7.26 5.16 6.91 QS 5.72 0.29 5.04 4.98 6.18 230 FS 5.04 0.36 7.19 4.35 5.82 QS 5.10 0.34 6.73 4.37 5.83 34    Figure 4.2 The average: a. water absorption (WA), and b. equilibrium moisture content (EMC) of samples at different temperatures. Error bars indicate standard error. To be more specific, treatment at 170°C caused 19.75% decrease in WA compared to controls, while treatment at 212°C and 230°C resulted in 41.53% and 48.17% decrease, respectively. As for EMC, treatments at 170°C, 212°C, and 230°C resulted in 30.93%, 52.85%, and 59.16% decrease, respectively. This is in good accordance with the only EMC values reported for heat-treated 0510152025303540Control 170 212 230Average Water Absorption (%)Temperature (oC)a.0246810121416Control 170 212 230Average Equilibrium Moisture Content (%)Temperature (°C)35  western hemlock by Williams (2011). For treatment at 180°C, Williams reported a 44% reduction in EMC of western hemlock, while for treatment at 212°C he reported a 52% reduction. The most likely reason for this reductions, as stated in Section 2.3, is the degradation of hemicellulose (the most hydrophilic polymer of the wood) and the loss of hydroxyl groups at intensified treatment temperatures. Since there is a reduction in the number of hydroxyl groups, less water is absorbed by the cell walls, so water affinity and the equilibrium moisture content of the heat-treated wood decreases (Esteves and Pereira 2009, Navi and Sandberg 2012). In addition, higher crystallinity in the cellulose of heat-treated wood and cross-linking of lignin also result in less accessibility of hydroxyl groups to water molecules and therefore a reduction in hygroscopicity (Esteves and Pereira 2009). This correlation between water absorption and equilibrium moisture content is illustrated in Figure 4.3 for treated and untreated samples.  Figure 4.3 The correlation between water absorption (WA) and equilibrium moisture content (EMC) of samples.  Statistical analysis using likelihood ratio test carried out in order to find the main effects of fixed-factors on WA and EMC, separately (Table 4.5 and 4.6). Based on the test statistics for WA (LR= 32.56, P-value<0.0001) and for EMC (LR=44.931, P-value<0.0001) at significance level of 0.05, only the effect of temperature is evident among treatments. A paired t-test using Tukey’s method with α=0.05 was also used to find the difference between WA means at different levels of temperatures and EMC means at different levels of temperatures (Table 4.7). The means that do 15.012.510.07.55.04540353025201510R-Sq 83.5%EMC (%) WA = 6.352 + 2.025 EMCWater Absorption (%) Equilibrium Moisture Content (%)  WA =6.325 + 2.025 (EMC) R2=83.5% 36  not share a letter are significantly different. The grouping indicates a significant difference between WA and MC means at all levels of temperature.  Table 4.5 Likelihood ratio tests for water absorption (WA). No. Model LogLik L.Ratio P-Value 1 Ln(WA)~ Intercept+ Temperature+ Cut+ Temperature* Cut 295.861 6.249935 0.1001 2 Ln(WA)~ Intercept+ Temperature+ Cut 292.736 2 Ln(WA)~ Intercept+ Temperature+ Cut 292.736 32.55921 <0.0001 3 Ln(WA)~ Intercept+ Cut 276.4564 2 Ln(WA)~ Intercept+ Temperature+ Cut 292.736 1.73872 0.1873 4 Ln(WA)~ Intercept+ Temperature 291.8666 Table 4.6 Likelihood ratio tests for equilibrium moisture content (EMC). No. Model LogLik L.Ratio P-Value 1 Ln(EMC)~ Intercept+ Temperature+ Cut +Temperature* Cut 482.5528 4.316568 0.2292 2 Ln(EMC)~ Intercept+ Temperature+ Cut 480.3945 2 Ln(EMC)~ Intercept+ Temperature+ Cut 480.3945 44.931 <0.0001 3 Ln(EMC)~ Intercept+ Cut 457.9290 2 Ln(EMC)~ Intercept+ Temperature+ Cut 480.3945 0.4099865 0.522 4 Ln(EMC)~ Intercept+ Temperature 480.1895 Table 4.7 Tukey’s pairwise comparisons for WA and EMC at different levels of temperature. Property Temperature (°C) Mean Grouping Water Absorption Control 30.98 A    170 24.86  B   212 18.11   C  230 16.06    D Equilibrium Moisture content Control 12.41 A    170 8.57  B   212 5.85   C  230 5.07    D 4.3. Anti-Swelling Efficiency The average thickness swelling and width swelling of samples are shown in Figure 4.4 and Table 4.8. The data shows that the average swelling decreases in both directions as the treatment temperature increases. The average thickness swelling of flatsawn samples and width swelling of quartersawn samples that occur in radial direction ranges from 1.74% to 3.87%. While the width 37  swelling of flatsawn samples and thickness swelling of quartersawn samples occur in tangential direction and range from 2.14% to 5.17 %. These values show that the tangential swelling is higher than radial and they are in a good agreement with the findings of Williams (2011). In terms of radial swelling, Williams reported 1.8% swelling at 180°C while the control samples had 2.8% swelling; and for control samples with swelling of 2.6%, he reported 1.9% swelling at 212°C. For tangential swelling, he found 4% swelling at 180°C while the control samples had 4.5% swelling; and for control samples with swelling of 5.3%, he reported 3.6% swelling at 212°C. As mentioned in section 2.4.2, this substantial swelling in tangential direction compared to radial is due to anisotropic properties of wood.  Shrinkage/swelling in the tangential direction of wood is approximately two times greater than radial direction. Panshin and de Zeeuw (1980) associated this lower radial swelling/shrinkage to wood being restricted from swelling by ray cells and to the interaction between bands of high-density latewood and low-density earlywood, which is a controlling factor in the tangential direction. However, looking at the data from table 4.8, the T/R ratio of swelled samples are 1.33 for controls, 1.36 for treatment at 170°C, 1.40 for treatment at 212°C, and 1.23 for treatment at 230°C which is less than the aforementioned value (~2). Lower T/R swelling ratio, less than two, was also achieved in other studies such as Mitani and Barboutis (2014), Aytin and Korkut (2016), Chotikhun and Hiziroglu (2016). Table 4.8 The descriptive statistics of radial and tangential swelling of samples. Direction Temperature (°C) Mean (%) StDev (%) CoefVar (%) Min (%) Max (%) Tangential Control 5.17 0.95 18.32 2.71 7.38 170 3.90 0.49 12.69 2.46 5.14 212 2.79 0.40 14.26 1.30 3.49 230 2.14 0.39 18.47 1.19 2.95 Radial Control 3.87 0.83 21.52 1.51 5.76 170 2.86 0.54 18.75 1.42 3.78 212 1.98 0.28 14.10 1.00 2.62 230 1.74 0.36 20.72 0.98 2.78  38   Figure 4.4 The tangential and radial swelling of samples. Error bars indicate standard error. The descriptive statistics of anti-swelling efficiency of samples are listed in Table 4.9.  In quartersawn, the highest ASE of 57.15% (CV=7.12%) was found at 230°C. The second highest ASE was 47.81% (CV=8.5%) at 212°C, and the lowest ASE was 25.38% (CV=33.05%) happened at 170°C. For flatsawn samples, ASE was found to be 57.09% (CV=7.15%), 47.06% (CV=13.47%), and 24.55% (CV=15.87%) at 230°C, 212°C and 170°C, respectively. It is concluded that by increasing the treatment temperature, an increase in ASE could be seen which is much evident between 170-212°C (89.99%) than between 212-230°C (20.42%) (Figure 4.5). Table 4.9 Descriptive statistic of anti-swelling efficiency (ASE) for heat-treated samples. Temperature (°C) Cut Mean (%) StDev (%) CoefVar (%) Min (%) Max (%) 170 FS 24.55 3.90 15.87 17.80 33.19 QS 25.38 8.39 33.05 13.09 37.69 212 FS 47.06 6.34 13.47 36.07 60.81 QS 47.81 4.06 8.5 41.10 55.74 230 FS 57.09 4.08 7.15 48.24 64.68 QS 57.15 4.07 7.12 50.35 65.39  01234567Tangential RadialSwelling (%)Control170°C212°C230°C39   Figure 4.5 The average dimensional stability of heat-treated samples at three temperatures. Error bars indicate standard error. At first, this increased dimensional stability at higher temperatures was attributed to cross-linking of cellulose chains by Stamm and Hansen (1937). Later, Tjeerdsma et al. (1998b) concluded that the cross-linking of aromatic rings in the lignin is responsible for the improved dimensional stability. However, Burmester (1973) introduced the degradation of hemicellulose as the primary reason for this phenomenon that was later corroborated by Weiland and Guyonnet (2003). They believed the loss of hemicellulose at high temperatures and destruction of hydroxyl groups result in diminished water affinity and therefore higher dimensional stability. In this regard, a scatter plot of ASE over WA is shown in Figure 4.6. The graph shows that increased water absorption which happens at lower treatment temperatures (part 4.2), results in less dimensional stability in treated samples (R2=68.1%). 010203040506070170 212 230Average Anti-swelling Efficiency (%)Temperature (°C)40   Figure 4.6 Scatter plot of anti-swelling efficiency (ASE) versus water absorption (WA) of samples with fitted line. Statistical analysis using likelihood ratio tests at a significance level of 0.05 was also done to find the effect of temperature and cut on ASE of samples (Table 4.10). As concluded, the temperature is the only factor that has a significant effect on ASE with LR=31 and P-value <0.0001. Also, the results from paired t-test using Tukey’s method with 95% confidence level are presented in Table 4.11, which confirm that the ASE means at all the levels of temperature significantly differ from one another. Table 4.10 Likelihood ratio tests for anti-swelling efficiency (ASE). No. Model LogLik L.Ratio P-Value 1 ASE~ Intercept+ Temperature+ Cut +Temperature* Cut -771.1201 0.116  0.9436 2 ASE~ Intercept+ Temperature+ Cut -771.1781 2 ASE~ Intercept+ Temperature+ Cut -771.1781  30.999  <0.0001 3 ASE~ Intercept+ Cut -786.6776 2 ASE~ Intercept+ Temperature+ Cut -771.1781  0.285  0.5933 4 ASE~ Intercept+ Temperature -771.3207 Table 4.11 Tukey’s pairwise comparisons for anti-swelling efficiency (ASE) at different temperature levels. Temperature (°C) Mean (%) Grouping 170 24.97 A   212 47.44  B  230 57.12   C 70605040302010353025201510R-Sq 68.1%ASE(%)WA(%)WA = 30.42 - 0.2488 ASEWater Absorption (%) WA =30.42 + 0.2488 (ASE)             R2= 68.1% Water Absorption (%) Anti-swelling Efficiency (%)  41  4.4. Color The variation in surface color of heat-treated samples is seen in Figure 4.7. Visual inspection reveals that the wood becomes darker as the temperature rises. To be more specific, measured L*, a*, and b* are presented in Table 4.12. Heat-treated samples showed the color parameters were influenced to a different degree after each heat treatment, yet they almost looked similar within both cuts (Figures 4.8 and 4.9). Since the color parameters were tested on the surfaces of each sample, flatsawn samples had their tangential surface tested while quartersawn samples were tested on their radial surface. In this case, visually speaking, it can be concluded that there is not much difference between the color parameters in tangential and radial directions. Same results were achieved for radial and tangential directions of European beech heat-treated at 180°C for five different durations ranging from 2 to 10 hrs by Mitani and Barboutis (2014). However, it is seen that quartersawn color parameters are slightly higher than flatsawn’s which could be attributed to anatomical differences between radial and tangential surfaces such as order of cells, rays and spiral grain (Nishino et al. 2000).                                                                    (a)  (b)                                   (c)                                (d) Figure 4.7 Thermally modified western hemlock at (a) unmodified, (b) 170°C, (c) 212°C, and (d)  230°C. Top row: quartersawn, bottom row: flatsawn.   42  Table 4.12 Descriptive statistics of color coordinates and color change. Color Parameter Temperature (°C) Cut N Mean StDev CoefVar (%) Min Max L* Control FS 42 73.51 1.95 2.65 69.83 77.01 QS 42 74.24 2.24 3.02 68.07 77.96 170 FS 42 64.51 1.85 2.86 60.42 67.20 QS 42 65.06 2.09 3.21 60.77 67.96 212 FS 42 47.80 1.64 3.43 43.93 51.79 QS 42 48.49 2.58 5.32 44.55 52.48 230 FS 42 41.23 1.46 3.54 38.26 43.47 QS 42 41.47 1.26 3.05 39.00 43.23 a* Control FS 42 8.21 0.54 6.60 7.028 9.41 QS 42 8.22 0.63 7.62 7.38 10.36 170 FS 42 8.99 0.51 5.71 8.10 10.43 QS 42 9.00 0.62 6.89 8.06 10.77 212 FS 42 10.43 0.26 2.46 9.98 11.06 QS 42 10.58 0.33 3.13 9.42 11.12 230 FS 42 9.45 0.68 7.19 8.22 10.46 QS 42 9.48 0.31 3.25 9.02 10.17 b* Control FS 42 20.56 0.98 4.76 19.01 23.31 QS 42 20.85 1.03 4.94 19.14 23.68 170 FS 42 23.84 0.94 3.93 21.39 25.20 QS 42 24.77 1.25 5.06 22.54 27.34 212 FS 42 21.49 1.23 5.73 18.79 24.07 QS 42 22.64 1.86 8.22 18.37 25.29 230 FS 42 17.80 1.79 10.05 12.92 20.81 QS 42 18.08 1.09 6.05 16.02 20.17 ΔE 170 FS 42 9.68 1.73 17.91 7.23 13.71 QS 42 10.13 2.00 19.78 7.15 14.51 212 FS 42 25.85 1.59 6.14 22.11 29.64 QS 42 25.99 2.48 9.53 22.31 29.85 230 FS 42 32.47 1.51 4.65 30.08 35.45 QS 42 32.92 1.31 3.97 31.09 35.46 43  Over both cuts, the maximum L* mean value of 73.87 (CV=2.87%) was measured on control samples. The samples treated at 170°C became 12.3% darker than controls reaching to L* mean value of 64.78 (CV=3.05%). A considerable decrease in the L* values (mean= 48.14, CV=4.52%) of samples occurred at 212°C and samples became 34.97% darker than controls. Finally, at 230°C, the L* reached its minimum value of 41.35 (CV=3.30%) where it became 43.91% darker than controls. So, it is clearly seen that as the temperature rises, the samples become darker.   Figure 4.8 The lightness (L*) of heat-treated samples. Error bars indicate standard error. Nevertheless, the change in chromatic coordinates follow another pattern: The red/green (a*) mean values increased from 8.22 (CV=7.09%) in controls up to 10.50 (CV=2.91%) in samples treated at 212°C, then it decreased to 9.47 (CV=5.54%) for samples treated at 230°C (Figure 4.9.a). The increased redness could be ascribed to formation of condensation, degradation, and/or oxidation products (Chen et al. 2012); while, the decreased a* at 230°C could be due to volatilization of the phenolic compounds from lignin (Gonzalez de Cademartori et al. 2013). The blue/yellow (b*) mean values increased from 20.70 (CV=4.88%) in controls up to 24.30 (CV=4.92%) in samples treated at 212°C and afterwards, it had a falling trend reaching the value of 17.94 (CV=8.26%) for samples treated at 230°C (Figure 4.9.b). It is seen that heating the wood resulted in more yellowness compared to control samples except at 230°C.  This could be justified by the fact that low-molecular-weight phenolic substances which mainly come from lignin are pale yellow (Hiltunen et al. 2006) and the reason for less yellowness at 230°C compared to controls might be, as stated before, volatilization of the phenolic compounds at that high temperature.  020406080100Control 170 212 230L*Temperature(°C)FlatsawnQuartersawn44      Figure 4.9 The average of: a. red/green coordinates (a*), b. blue/yellow coordinates (b*), and c. color change (ΔE) in samples. Error bars indicate standard error. 051015Control 170 212 230a*Temperature(°C)a. Red/Green CoordinateFlatsawnQuartersawn051015202530Control 170 212 230b*Temperature(°C)b. Blue/Yellow CoordinateFlatsawnQuartersawn051015202530Control 170 212 230b*Temperature(°C)c. Color ChangeFlatsawnQuartersawn Control 45  Moreover, the color change of heat-treated samples compared to controls is depicted in Figure 4.9c. Averaging over both cuts, the lowest color change is attributed to samples treated at 170°C with an average of 9.90 (CV=18.94%), then there is an abrupt increase in ΔE where it reaches to 25.92 (CV=7.98%) at 212°C. Afterwards, the color change increased slowly until reached its maximum value of 32.70 (CV=4.35%) at 230°C. The wood discoloration after heat treatment has been reported by many researchers and has been attributed to the formation of colored degradation products from hemicellulose, especially pentosans (Bourgios et al. 1991, Sehlstedt-Persson 2003), oxidation products (Tjeerdsma et al. 1998b, Bekhta and Niemz 2003), and extractives (Sundqvist and Morén 2002).  The data clearly indicates that as the temperature rises, the color change (ΔE) intensifies which is in accordance with the literature (Section 2.4.4.).  To make sure which factor has a significant effect on ΔE means, statistical analysis using likelihood ratio tests at significance level of 0.05 was carried out (Table 4.13). Test results acknowledge that temperature is the only factor that immensely affects color change (LR= 41.437 and P-value <0.0001). Also, the results from paired t-test using Tukey’s method with 95% confidence level (Table 4.14) showed that every three levels of temperature result in different color changes. Table 4.13 Likelihood ratio tests for color change. No. Model LogLik L.Ratio P-Value 1 ΔE0.8 ~  Intercept+ Temperature+ Cut+ Temperature* Cut -256.2880 0.230 0.8912 2 ΔE0.8 ~  Intercept+ Temperature+ Cut -256.4031 2 ΔE0.8 ~  Intercept+ Temperature+ Cut -256.4031 41.437 <0.0001 3 ΔE0.8 ~  Intercept+ Cut -277.1216 2 ΔE0.8 ~  Intercept+ Temperature+ Cut -256.4031 1.022 0.3121 4 ΔE0.8 ~  Intercept+ Temperature -256.9139  Table 4.14 Tukey’s pairwise comparisons for color change at different levels of temperature. Temperature (°C) Mean (%) Grouping 170 9.90 A   212 25.92  B  230 32.70   C 46  4.5. Hardness According to Table 4.15 the highest side hardness is for control samples with an average of 2810.21 N (CV=16.16%) and the second highest is for samples treated at 170°C with an average of 2766.87 N (CV=15.27%). These values are closely similar to the ones reported by Williams (2011). He calculated the side hardness of 2766.79 N for samples treated at 180°C while their control samples had side hardness of 2815.72 N. By comparing his results with mine, the 10°C decrease in temperature resulted in a slight change in the side hardness. However, for samples treated at 212°C, a sudden decrease to 2400.58 N (CV=13.03%) is seen. Williams (2011) also reported a lower side hardness of 2379.8 N for samples treated at 212°C while their control had a side hardness of 3024.79 N. Finally, the lowest side hardness belonged to samples treated at 230°C with an average of 2206.65 N (CV=12.59%). Table 4.15 Descriptive statistics of side hardness in samples Temperature (°C) Cut Mean (N) StDev (N) CoefVar (%) Min (N) Max (N) Control FS 2809.6 458.2 16.31 2105.9 3923.1 QS 2810.8 455.7 16.21 2051.8 4097.8 170 FS 2763.1 390.7 14.14 1546.7 3481.3 QS 2770.1 456.6 16.48 2110.4 3622.1 212 FS 2458.7 320.1 13.02 1858.1 3190.7 QS 2342.5 297.9 12.72 1831.3 3116.2 230 FS 2189.2 291.9 13.33 1706.8 2792.4 QS 2224.1 265.3 11.93 1792.3 2779 It seems that as temperature increases, the side hardness decreases (Figure 4.10). There is only a 1.54% decrease in the side hardness of samples treated at 170°C compared to controls. While, a more distinct decrease of 14.58% for samples treated at 212°C is noticed. The most noticeable change is for samples treated at 230°C where their side hardness decreased by 21.48% compared to controls. 47        Figure 4.10 Side hardness of samples treated at different temperatures. Error bars indicate standard error. Statistical analysis using likelihood ratio tests also gives the same results indicating that the temperature is the only factor that has a significant effect on side hardness with LR= 25.28 and P-value<0.0001, where significance level is 0.05 (Table 4.16). The results from paired t-test using Tukey’s method with 95% confidence level are shown in Table 4.17. The grouping indicates that the side hardness of all the temperature levels differs significantly from one another, except for the controls, which are not significantly different from samples treated at 170°C. Table 4.16 Likelihood ratio tests for side hardness. No. Model LogLik L.Ratio P-Value 1 Ln(Side Hardness)2=Intercept+ Temperature+ Cut + Temperature* Cut -745.4931 1.7723 0.621 2 Ln(Side Hardness)2=Intercept+ Temperature+ Cut -746.3792 2 Ln(Side Hardness)2=Intercept+ Temperature+ Cut -746.3792 25.27904 <.0001 3 Ln(Side Hardness)2=Intercept+ Cut -759.0187 2 Ln(Side Hardness)2=Intercept+ Temperature+ Cut -746.3792 0.1542028 0.6946 4 Ln(Side Hardness)2=Intercept+ Temperature -746.4563 Table 4.17 Tukey’s pairwise comparisons for side hardness at different levels of temperature. Temperature (°C) Mean (N) Grouping Control 2810.2 A   170 2766.9 A   212 2400.6  B  230 2206.6   C 0500100015002000250030003500C 170 212 230Average Side Hardness (N)Temperature (°C)Control 48  The same falling trend for end hardness of samples is noted as the temperature rises (Figure 4.11). However, end hardness presents much higher values than side hardness due to the orientation of polymers in the wood and the presence of strong bonds parallel to the grain (Boonstra 2008). For end hardness, according to Table 4.18, the average values for controls (mean=4804.10 N, CV=11.62%) and samples treated at 170°C (mean=4792.90 N, CV=9.28%) are approximately the same while a slight fall to 4567.90 N (CV=9.07%) is seen for samples treated at 212°C. Similarly, the lowest end hardness is attributed to samples treated at 230°C with values of 4334 N (CV=10.63%). Table 4.18 Descriptive statistics of end hardness in samples. Temperature (°C) Cut Mean (N) StDev (N) CoefVar (N) Min (N) Max (N) Control FS 4786.7 576.8 12.05 4091 6351.1 QS 4821.4 545.8 11.32 3519.7 6018.9 170 FS 4774.3 483.7 10.13 3586.3 5918.6 QS 4811.6 406.8 8.45 3824 5897.6 212 FS 4579.3 453.3 9.9 3542.7 5516.1 QS 4556.6 377 8.27 3795.9 5488.5 230 FS 4269.5 438.5 10.27 3098.6 5237.6 QS 4398.5 478.1 10.87 3282.6 5549.5        Figure 4.11 End hardness of samples at different temperatures. Error bars indicate standard error. 3800400042004400460048005000C 170 212 230Average Side Hardness (N)Temperature (°C)Control 49   Based on the likelihood ratio tests results, at a significance level of α=0.05 (Table 4.19), temperature is the only factor that has a significant effect on the end hardness (LR= 16.94, P-value= 7e-04). Paired t-test results using Tukey’s method with 95% confidence level also revealed the same results as for side hardness. It showed that the end hardness at all the levels of temperature is significantly different from one another except for controls and treatment temperature of 170°C (Table 4.20). Table 4.19 Likelihood ratio tests for side hardness. No. Model LogLik L.Ratio P-Value 1 Ln(End Hardness)=Intercept+ Temperature+ Cut +Temperature* Cut 293.3771 0.8855378 0.8289 2 Ln(End Hardness)=Intercept+ Temperature+ Cut 292.9343 2 Ln(End Hardness)=Intercept+ Temperature+ Cut 292.9343 16.93788 7e-04 3 Ln(End Hardness)=Intercept+ Cut 284.4654 2 Ln(End Hardness)=Intercept+ Temperature+ Cut 292.9343 0.69507 0.4044 4 Ln(End Hardness)=Intercept+ Temperature 292.5868 Table 4.20 Tukey’s pairwise comparisons for end hardness at different levels of temperature. Temperature (°C) Mean (N) Grouping Control 4804.1 A   170 4792.9 A   212 4567.9  B  230 4334   C As discussed in 2.5.1, the diminution of hardness at higher treatment temperatures is primarily due to degradation of cell wall material specially hemicellulose which results in mass loss and consequently reduced density (Hillis, 1984). To show the effect of density on the hardness, the correlation between both side and end hardness with sample density is plotted in Figure 4.12. It indicates that as density decreases at higher treatment temperatures, the hardness decreases, however the correlations found to be poor. For density and side hardness R2=17% while for density and end hardness R2=5.1%.  50         Figure 4.12 The scatter plot of (a) side hardness, and (b) end hardness versus density. 4.6. Stiffness In non-destructive evaluation of stiffness, one of the parameters used to calculate the modulus of elasticity is the velocity of the applied stress wave on the wood. The stress-wave velocity in the longitudinal direction of wood at moisture contents between 9-15% has been reported to range between 3000 to 6000 m/s (Han et al. 2006) which is in accordance with my findings depicted in Table 4.21. It is seen that increasing the treatment temperature slightly increases the stress-wave velocity from 4786.1 m/s in controls (CV= 3.37%) up to 4876.7 m/s (CV=3.48%) at 230°C. 600550500450400400035003000250020001500R-Sq 17.0%i  k  Side Hardness = - 361.6 + 6.042 ρ60055050045040065006000550050004500400035003000R-Sq 5.1%Density (kg/m3)  End Hardness = 2826 + 3.738 ρ(a) (b) R2= 17% Density (kg/m3) Density (kg/m3) End Hardness= 2826 + 3.738 (ρ) R2= 5.1% Side Hardness= -361.6+ 6.042 (ρ) Side Hardness (N) End Hardness (N) 51  According to Gray et al. (2008) and Del Menezzi et al. (2014) the decreased stress-wave velocity at higher treatment temperatures is due to lower moisture content at those intensified temperatures. Table 4.21 Descriptive statistics of stress-wave velocity in samples at different temperatures. Temperature (°C) Mean (m/s) StDev (m/s) CoefVar  (%) Min (m/s) Max (m/s) Control 4786.1 161.2 3.37 4328.8 5062.8 170 4820.7 154.0 3.19 4331.5 5037.9 215 4855.4 199.1 4.10 4215.7 5156.6 230 4876.7 169.6 3.48 4354.5 5205.2   The other parameter used to calculate the dynamic modulus of elasticity is density. As stated earlier (2.4.3), thermal modification generally reduces the density due to degradation of wood cell components. Since the internal stresses will be spread over less wood material, the strength properties reduce (Boonstra 2008). In this regard, the correlation between the density and stiffness of samples is depicted in Figure 4.13. As it could be seen from the graph, lower densities that attribute to higher treatment temperatures, exhibit lower stiffness (R2=35.7%).      Figure 4.13 The correlation between density and dynamic modulus of elasticity (ED) of samples. According to Table 4.22, ED of flatsawn samples are slightly higher than that of quartersawn samples; however, this difference was not considered significant after carrying out the statistical 600550500450400140001300012000110001000090008000R-Sq 35.7%Density (kg/m3)Ed (MPa)Ed = 2123 + 19.01 DensityDensity (kg/m3) ED = 2123 + 19.01 (ρ) R2= 35.7% Dynamic Modulus of Elasticity (MPa) 52  analysis. In fact, this difference directly attributes to the higher calculated density of flatsawn samples compared to quartersawn samples. The highest stiffness was for control samples with an average of 11620.54 MPa (CV=7.85%) over both cuts. At 170°C, there was a marginal decrease of 0.62% in the ED, reaching the average of 11549 MPa (CV=8.33%). A more evident decrease of 3.16% and 5.49% occurred for samples treated at 212°C and 230°C, respectively, where the ED of samples reached 11253.45 MPa (CV=9.72%) and 10983.08 MPa (CV=8.29%). The downward trend of ED by increasing the treatment temperature is shown in Figure 4.14.  Table 4.22 Descriptive statistics of stiffness of samples. Temperature (°C) Cut Mean (MPa) StDev (MPa) CoefVar  (%) Min (MPa) Max (MPa) Control FS 11737 816 6.95 10091 13729 QS 11504 995 8.65 10086 13789 170 FS 11682 744 6.37 10195 12917 QS 11416 1134 9.93 8871 13424 212 FS 11510 1123 9.75 9122 13644 QS 10997 1015 9.23 8477 12889 230 FS 11073 952 8.6 8875 13035 QS 10893 870 7.98 8567 12313  Figure 4.14 Dynamic modulus of elasticity (ED) of samples at different temperatures. Error bars indicate standard error. 1000010500110001150012000Control 170 212 230Average Ed(MPa)Temperature (°C)53  Statistical analysis using likelihood ratio tests also revealed that the only factor that significantly affects the stiffness is temperature with test statistics LR=9.61, and P-value= 0.0222 (Table 4.23). Moreover, the results from pairwise t-test using Tukey’s method at 95% confidence level (Table 4.24) showed that the mean stiffness at all the treatment temperatures significantly differ from one another except for temperature 212°C  and 230°C. Table 4.23 Likelihood ratio tests for stiffness. No. Model LogLik L.Ratio P-Value 1 ED=Intercept+ Temperature+ Cut +Temperature* Cut -2718.388 0.6689223 0.8805 2 ED =Intercept+ Temperature+ Cut -2718.723 2 ED =Intercept+ Temperature+ Cut -2718.723 9.610556 0.0222 3 ED =Intercept+ Cut -2723.528 2 ED =Intercept+ Temperature+ Cut -2718.723 3.5838 0.0583 4 ED =Intercept+ Temperature -2720.515 Table 4. 24 Tukey’s pairwise comparisons for dynamic modulus of elasticity (ED) at different levels of temperature. Temperature (°C) Mean (MPa) Grouping Control 11621.54  A  170 11549.00  A  212 11253.45  A B 230 10983.08   B             54  5. General Discussion There has been plentiful research on altered properties of wood species using ThermoWood modification process. However, the main focus of all the published research has only been on a few wood species such as pine, spruce, birch, ash, and eucalyptus. Yet, there are a lot more wood species whose properties could be enhanced using this modification process. With western hemlock being a predominant species in British Columbia’s forests and the fact that there has been only one study of thermal modification of western hemlock in the past, I chose to study this species. In this thesis, I hypothesized that thermal modification of Western hemlock using the ThermoWood process would enhance some of the wood’s physical properties such as increasing dimensional stability, decreasing equilibrium moisture content and water uptake without significantly affecting its basic density and  mechanical properties like stiffness and hardness.  Also, thermally modified Western hemlock would gain darker color.  To test this hypothesis, I performed experiments and computer analysis, which are explained in Chapter 3. More specifically, I investigated the effect of three modified ThermoWood schedules to find the schedule that can significantly change the quality of Western hemlock in a cost, energy, and time–effective manner. The results were explained and discussed in chapter 4.  Wood is a hygroscopic material due to the presence of hydroxyl groups in its cell wall polymers. Therefore, when surrounded by a setting containing moisture, it can absorb or desorb water until it reaches EMC (Akyildiz and Ates 2008). The ThermoWood process can lead to a 40-50% reduction in EMC of modified wood (ThermoWood® brochure) due to chemical changes at high treatment temperatures and less accessibility of hydroxyl groups to water molecules (Hill 2006). My experimental results showed an even higher reduction in EMC at higher temperatures than the reported percentage. A moderate reduction of 30.9% in EMC of western hemlock treated at 170°C was seen. Nevertheless, when treated at higher temperatures of 212°C and 230°C, much larger reduction of 52.6% and 59.2% was achieved, respectively. Clearly, a higher treatment temperature of 230°C resulted in lower equilibrium moisture content. Yang et al. (2016) found a 3% reduction in EMC of Japanese red cedar treated at 170°C for 2 hrs. Akyildiz and Ates (2008) heat treated several species at 230°C for 2 hrs and reported EMC reductions of 51% in sessile oak, 47% in chestnut, 47.3% in Calabrian pine and 41.7% in  black pine. Majano-Majano (2012) heat treated European beech and ash at the same temperature and reported 43.5% and 46.7% reduction 55  in their EMCs, respectively. It is seen that high temperature had more impact on EMC of treated western hemlock than aforementioned species and reduced it to more extent. A good correlation (R2=83.5%) between EMC and WA of treated western hemlock samples was found, which confirmed the loss of water affinity as the temperature rose. After a moderate reduction (19.7%) in water absorption at 170°C, a large decrease (52.6%) appeared when samples treated at 212°C. The most decrease in WA (59.2%) occurred when treated at 230°C. Respectively, higher treatment temperatures led to more diminished water absorption. The decreased EMC and WA also resulted in higher dimensional stability of thermally modified wood. A fair correlation (R2=68.1%) was found between WA and ASE of samples, which confirmed the fact that the loss of water affinity at higher temperatures resulted in escalated dimensional stabilities. ThermoWood products have been reported to have 40-50% higher dimensional stability than untreated wood (Finnish ThermoWood Association 2003). Using ASE as an indicator of dimensional stability, my experimental results showed that thermally modified western hemlock samples treated at 212°C were 89.9% more stable than samples treated at 170°C; they were also 20.4% less stable than the ones treated at 230°C. The results indicated that the dimensional stability of samples increased prominently at higher treatment temperatures. This high dimensional stability of thermally modified western hemlock makes it a good choice for door frames and window sashes where 10% loss of strength is tolerable (Giebeler 1983). The high dimensional stability can prevent doors and windows from sticking to the frame when the moisture content of their surroundings changes (Christmas et al. 2007). High dimensional stability also reduces the cracking and peeling of the surface coating when in contact with high humidity environment (Finnish ThermoWood Association 2003), which could be further investigated (Section5.2). Basic density is a key indicator of sample’s quality that affects almost all the properties of wood. Thermally modified wood has been reported to have a lower density compared to untreated wood due to degradation of wood components at high temperature and consequently mass loss. (Finnish ThermoWood Association 2003, Moliński et al. 2016). This decrease in density is inevitable for all the species when heated (Chaouch et al. 2010). My experimental results showed that basic density decreased slightly as the treatment temperature increased. The results were in accordance with the results from previously thermally modified western hemlock by Williams (2011).  The minimum reduction in basic density (1.6%) occurred at 170°C, and it was doubled 56  (3.2%) when treatment temperature raised to 212°C. Treatment temperature of 230°C induced a reduction of 4.1% in the basic density of western hemlock. When comparing this value with the findings of Chaouch et al. (2010) for the density of poplar, beech, ash, pine, and fir treated at 230°C, it is seen that except for pine, all the other species had lost their density 2-3 times more than western hemlock.  Mechanical properties are highly dependent on the density of wood (Boonstra 2008). Due to limitations of sample sizes in this study, I was only able to conduct two mechanical tests including Janka hardness and stiffness. Based on the literature (section 2.5.2), the hardness of thermally modified wood may increase or decrease depending on the treatment conditions and species. Side hardness of thermally modified western hemlock was measured to slightly decrease (1.5%) at 170°C. A more significant decrease (14.6%) happened when samples treated at 212°C, and the biggest decrease (21.5%) occurred when treated at 230°C. Poor correlation (R2=17%) was found between density and side hardness of samples. For end hardness of samples, an insignificant reduction (0.2%) was observed when treated at 170°C. The change was almost twenty-four times bigger (4.9%) but still only slightly different from controls, when treated at 212°C. The most reduction (9.8%) in end hardness occurred when treated at 230°C. Again, poor correlation (R2=5.1%) was found between density and end hardness of samples. In conclusion, higher treatment temperature led to decreased hardness with the effect being more adverse on side hardness compared to end hardness. Among the three treatment temperatures, 212°C seemed to affect hardness less, while noticeably improving the other aforementioned properties. On the contrary, some other wood species such as spruce, pine, and fir with initial lower basic density (360, 410, and 430 kg/m3) and lower side hardness (2300, 2100, and 1900 N)  compared to western hemlock (Ross 2010), were reported to have increased hardness (3-52%) when treated at 212°C. However, their treatment duration was 3hrs (Shi et al. 2007). Therefore, it can be argued that the combination of treatment temperatures and durations in addition to the initial basic density of samples might have different outcomes on hardness and needed to be investigated.  Modulus of elasticity is the key property when choosing a wood for where supporting the load without breakage (MOR value) is not important such as in stairways or decks (Esteves et al. 2007). Non-destructive stress-wave timer test was done to determine the ED of samples. The stress wave velocity propagated along the samples marginally increased at higher treatment temperatures where there was less moisture content. A reasonable correlation between ED and density 57  (R2=35.7%) of samples was found which agreed with the aforementioned fact regarding degradation of chemical components at higher treatment temperatures and consequently decreased density and mechanical properties. Higher treatment temperature resulted in a slight reduction               (0.6%) in ED of samples when treated at 170°C. The decrease was almost five times bigger    (3.2%) when treated at 212°C, but still only hardly different from controls. The most decrease in ED (5.5%) occurred when treated at 230°C. Non-destructive testing of thermally modified wood has not been in the center of attention. Nevertheless, among those who have studied this subject, Garcia et al. (2012) reported a 6.77% reduction in ED of thermally modified eucalypt wood treated at the same treatment temperature of 230°C and duration of 2hrs as in my study. Comparing the higher density and the close ED value of treated eucalyptus (542 kg/m3, 10404 MPa) to that of treated western hemlock (461.16 kg/m3, 10983 MPa) at this intense temperature shows that the congenital characteristics of each species and the stability of their chemical constituents under treatment conditions, play a significant role in achieving different ED values .  Aesthetically speaking, color is an important factor when choosing a wood for most of the final products. Usually darker shades which are mostly related to expensive tropical hardwoods are perceived to be luxurious. However, using thermal modification, inexpensive wood species could easily gain darker shades and imitate the exotic color of expensive species (Korkut et al. 2013). In this case, staining the wood is no longer required and the change in color could be controlled and adjusted to the final application by changing the treatment parameters. My experimental results showed that as treatment temperature raised, samples became darker and the color parameters (a*, b*, and L*) modified in a way that resulted in increased color changes. Considering the average color parameters reported by Cheng (2015) for untreated ipe (L*~51, a*~10.2, b*~17.5) and western red cedar (L*~71, a*~9.8, b*~23), the altered color of thermally modified western hemlock at 212°C and 230°C  makes it visually comparable with such expensive wood species. Determined average L* of 48 and 41 for treatment temperatures of 212°C and 230°C show that thermally modified western hemlock is much darker than western red cedar. However, it is almost as dark as ipe. Taking standard deviations into account, the results indicated that treated hemlock at 230 is almost as red (a*~9.5) and as yellow (b*~17.9) as ipe, and truly mimicking its color. This dark color of treated western hemlock can open up various potential non-structural markets for this species such as interior decorative panels, flooring, and shelves in which the cost 58  of using treated western hemlock would be much lower than using valuable dark hardwoods (Hoang 2009).  In brief, when comparing the treatment temperatures, the first noticeable change in properties appeared when samples were treated at 212°C. Moreover, the cut of samples did not significantly affect any of the tested properties. Generally, quartersawn wood is less prone to deformation, dimensional changes, surface peeling and crack compared to flatsawn wood (Flaete et al. 2000, Sandberg and Söderström 2006) and that is one of the reasons for its higher price. The same has been reported for thermally modified quartersawn wood (Syrjänen 2001). So, knowing that the cut of thermally modified western hemlock does not grossly affect its tested properties, cheaper treated flatsawn hemlock could be used in applications where it is not exposed to weather effects. In retrospect, within the scope and limitations of this study, to achieve enhanced properties of western hemlock in a cost, energy, and time–effective manner, using the mild treatment temperature 212°C and 2 hours of holding time, without considering the cut (depends on the application) is the treatment combination recommended when the loss of strength is important. Since this research was done in an industrial-size thermal modification unit, the results are representative of the full-scale commercial conditions and they are immediately transferable to industrial conditions.  As a conclusion, thermally modified western hemlock could be considered as a domestic, sustainable, and environmentally safe product with exotic dark colors which is suitable for both interior and exterior applications. However, further research is required to find the best applications for it. 6. Conclusion The ThermoWood modification process is suitable for most wood species but it must be optimized individually by species. In this thesis, ThermoWood process was explored as a means of enhancing western hemlock properties. As expected, higher temperatures resulted in more intense changes in properties. But, the cut did not significantly affect any of the properties. Thermal modification reduced basic density and equilibrium moisture content as a result of degradation of cell wall components consequently decreasing water absorption. Significant improvement in dimensional stability was observed which was directly correlated with the decrease in water absorption. Moreover, mechanical properties including side hardness, end hardness, and stiffness slightly decreased as purported. Finally, the color changes of treated wood compared to untreated 59  wood was determined to be intensified by increasing the treatment temperature. The exotic darker color after modification could allow western hemlock to visually compete with valuable wood species. Yet, much more tests need to be done (i.e. durability, mechanical tests, and workability) in order to confirm treated western hemlock’s superiority. It was also concluded that among all the three schedules, the one with the mild temperature of 212°C and 2 hours of holding time, without considering the cut, significantly enhanced the tested properties of western hemlock to the extent that small but almost statistically significant change was observed in the hardness and stiffness. These findings illustrate that the development of treatment schedules for western hemlock is an opportunity for local industry to add more value to this species and will help them to define the best practice guidelines for the application of treated western hemlock. 7. Further Research Extensive research on varying process parameters namely treatment temperature and holding time could be done in order to tailor the properties of thermally modified western hemlock to be used as an improved material for building industry such as flooring, cladding, garden furniture, window frames, doors and specialty products (i.e. sauna, bathrooms). Accordingly, much more physical, mechanical, and biological tests along with exterior weathering evaluation needs to be performed on thermally modified western hemlock for gaining a better understanding of its altered properties and where it could be applied. Loss of strength has been reported to be one of the biggest disadvantages of thermally modified wood, which limits its application in load-bearing uses and construction. A great reduction in static bending strength (10-20%) and static modulus of elasticity (5-20%) has been reported in literature depending on the species and the treatment conditions. (Jämsä and Viitaniemi 2001, Sandberg et al. 2013). In this regard, although the results from this thesis indicated only a slight decrease in stiffness and hardness values after the treatment, determining the static modulus of elasticity (MOE) and static modulus of rupture (MOR) of thermally modified western hemlock would be advantageous in finding a commercial use for it. Moreover, finding the correlation between static modulus of elasticity (MOE) and dynamic modulus of elasticity (ED) would allow to avoid the problems in the destructive measuring of MOE and provide us a quality control method using non-destructive measurements.  60  Alternatively, the high dimensional stability of thermally modified wood makes it suitable for outdoor uses. ThermoWood® products have been reported to be sufficiently resistant to biological damages namely fungi decay (Finnish ThermoWood Association 2003). This resistance occurs because of degradation of hemicellulose and mass loss after thermal modification, which interferes with the metabolism of the decay fungi (Syrjänen 2001, Metsä-Kortelainen et al. 2006, Welzbacher et al. 2007). Consequently, the span life of wood used for outdoor applications increases from 5 to 20 years if it is thermally modified and used above the ground (Dagbro 2016). Therefore, investigating the mass loss of western hemlock after thermal modification and the extent of its durability against fungi attack could be beneficial for finding the best outdoor applications for it. Another key factor for selecting the outdoor applications for thermally modified western hemlock is its weathering behavior. Generally, exposure to outdoor conditions results in dimensional changes of wood and release of stresses that cause checking (Feist and Sell 1987). Some studies showed that denser woods are more susceptible to checking (Cheng 2015). However, despite the lower density and higher dimensional stability of thermally modified wood, the occurrence of checking during weathering has been reported to be the same as in un-modified wood (Jämsä et al. 2000, Finnish ThermoWood Association 2003, Altgen et al. 2012). Therefore, determining the vulnerability of thermally modified western hemlock to checking, and finding a solution to reduce checking would be beneficial for expanding its application to outdoors. Moreover, it has been reported that wood with quartersawn cut is less prone to checking when used in the exterior (Sandberg and Söderström 2006). According to Navi and Sandberg (2012), no literature focuses on sawing patterns of thermally modified woods and its relation to the occurrence of checking. Therefore, the data from this thesis, namely the physical properties of quartersawn and flatsawn samples could be coupled with weathering tests in order to compensate for the literature discrepancies. Exterior exposure also changes the wood’s color. The ultraviolet (UV) radiation from sunlight photodegrades lignin and creates a nutrition source for blue stain fungi, resulting in a grey appearance of wood (Hill 2006).  Thermally modified wood is not much different from natural wood when exposed to outdoor conditions (Sandberg et al. 2013). Some studies related the slightly increased weathering resistance of thermally modified wood to its decreased moisture content and water absorption (Feist and Sell 1987, Nuopponen et al. 2004). Nonetheless, the exotic brown color of thermally modified wood eventually fades away and becomes grey when used outdoors (Jämsä 61  et al. 2000, Syrjänen and Kangas 2000). Therefore, to use in outdoor conditions, artificial and natural weathering tests on thermally modified western hemlock would be necessary in order to determine its resistance against photodegradation and finding the best finishes that protect it from the color change.  It should be noted that the normal finishes are not applicable for thermally modified woods due to their high hydrophobicity (Vernois 2001, Petrissans et al. 2003). The same is true for glues. Water-based glues such as PVAc could not be easily absorbed by thermally modified wood and require longer pressing times (Rapp and Sailer 2001, Hakkou et al. 2005). Therefore, studying the wettability, glueability, paintability and generally coating of western hemlock after thermal modification is an asset.   When sought to use in cladding, outer doors, windows, and saunas, having a lower thermal conductivity is an advantage.  According to Finnish ThermoWood Association (2003), thermal conductivity of ThermoWood® products decreases 20-25% compared to natural wood. So, thermal conductivity of thermally modified western hemlock better to be determined for specific further uses.  Overall, since the properties of thermally modified wood are correlated with process conditions, namely temperature and holding time, it is possible to attune the modification process of western hemlock for the purpose of different end uses. For instance, when higher strength is desired, less intense treatment conditions could result in a smaller increase in dimensional stability, but higher strength of the wood. The increasing ongoing process optimization efforts demonstrate that there is a great scope to improve our perception of the properties of thermally modified wood and its appropriate end uses.      62  References Akyildiz, M. H. & Ates, S. (2008). Effect of heat treatment on equilibrium moisture content (EMC) of some wood species in Turkey. Research Journal of Agriculture and Biological Sciences, 4(6): 660-665.  Albrektas, D. & Navickas, P. (2017). An evaluation of modulus of elasticity, dimensional stability and bonding strength of bonded heat-treated wood. Wood Industry/Drvna Industrija, 68(2): 137-144. Alden, H. (1995). Softwoods of North America. US Department of Agriculture, Forest Service Research Paper FPL-GTR-102, Forest Products Laboratory, Madison, W.I., U.S.A.   Altgen, M., Adamopoulos, S., Ala-Viikari, J., Hukka, A., Tetri, T., & Militz, H. (2012). Factors influencing the crack formation in thermally modified wood. In: The Sixth European Conference on Wood Modification, Ljubljana, Slovenia, 17-18 September, pp. 149-158.  ASTM D1037-12. (2012). Standard test methods for evaluating properties of wood-base fiber and particle panel materials. ASTM International, West Conshohocken, PA.  ASTM D143-14 (2014). Standard test methods for small clear specimens of timber. ASTM International, West Conshohocken, PA. ASTM D4442-16. (2016). Standard test methods for direct moisture content measurement of wood and wood-based materials. ASTM International, West Conshohocken, PA. ASTM D2244-16. (2016). Standard practice for calculation of color tolerances and color differences from instrumentally measured color coordinates. ASTM International, West Conshohocken, PA, 2016.  ASTM D2395-17. (2017). Standard test methods for density and specific gravity (relative density) of wood and wood-based material. ASTM International, West Conshohocken, PA. Awoyemi, L. & Jones, I. P. (2011). Anatomical explanations for the changes in properties of western red cedar (Thuja plicata) wood during heat treatment. Wood Science and Technology, 45(2): 261-267. Aydemir, D., Gunduz, G., & Ozden, S. (2012). The influence of thermal treatment on color response of wood materials. Color Research & Application, 37(2): 148-153. Aytin, A. & Korkut, S. (2016). Effect of thermal treatment on the swelling and surface roughness of common alder and wych elm wood. Journal of Forestry Research, 27(1): 225-229. Bekhta, P. & Niemz, P. (2003). Effect of high temperature on the change in color, dimensional stability and mechanical properties of spruce wood. Holzforschung, 57(5): 539-546. Bengtsson, C., Jermer, J., & Brem, F. (2002). Bending strength of heat-treated spruce and pine timber. The International Research Group on Wood Preservation. IRG/WP/02-40242. Stockholm, Sweden. 63  Bois Perdure (2016). Manufacturer promotional information. Retrieved on May 2017 from:     http://www.perdure.com/PerdurePortail/DesktopDefault.aspx?tabindex=1&tabid=2 Bhuiyan, T., Hirai, N., & Sobue, N. (2000). Changes of crystallinity in wood cellulose by heat treatment under dried and moist conditions. Journal of Wood Science, 46(6): 431-436. Boonstra, M. J. (2008). A two-stage thermal modification of wood (Doctoral dissertation) in co-supervision Ghent University and Université Henry Poincaré, Nancy 1, 297 p. ISBN 978-90-5989-210-1. Boonstra, M. J. & Tjeerdsma, B. F. (2006). Chemical analysis of heat treated softwoods. Holz als Roh- und Werkstoff, 64(3): 204-211. Borůvka, V., Zeidler, A., & Holeček, T. (2015). Comparison of stiffness and strength properties of untreated and heat-treated wood of Douglas fir and alder. BioResources, 10(4): 8281-8294. Bourgois, J. & Guyonnet, R. (1988). Characterization and analysis of torrefied wood. Wood Science and Technology, 22(2): 143-155. Bourgois, J., Janin, G. & Guyonnet, R. (1991). Measuring colour: a method of studying and optimizing the chemical transformations of thermally-treated wood. Holzforschung, 45(5): 377-382. Brauner, A. B. & Conway, E. M. (1964). Steaming walnut for color. Forest Products Journal, 14(11): 525- 527. Burmester, A. (1973). Investigation on the dimensional stabilization of wood. Bundesanstalt für Materialparüfung, Berlin, Dahlem, pp. 50-56. Burtscher, E., Bobleter, O., Schwald, W., Concin, R., & Binder, H. (1987). Chromatographic analysis of biomass reaction products produced by hydrothermolysis of Poplar wood. Journal of Chromatography, 390(2): 401-412 Candelier, K., Thevenon, M. F., Petrissans, A., Dumarcay, S., Gérardin, P., & Petrissans, M., (2016). Control of wood thermal treatment and its effects on decay resistance: a review. Annals of Forest Science, 73(3): 571-583. DOI: 10.1007/s13595-016-0541-x Chang, C. I. & Keith, C. T. (1978). Properties of heat-darkened wood. II – Mechanical properties and gluability. Ottawa: Eastern Forest Products Laboratory, Fisheries and Environment, Canada. Chaouch, M., Pétrissans, M., Pétrissans, A., & Gérardin, P. (2010). Prediction of decay resistance of different softwood and hardwood species after heat treatment based on analysis of wood elemental composition. Proceedings of the Fifth European Conference on Wood Modification, Riga, Latvia, pp.135-142. 64  Chen, A. (2013). Comparisons of heat treated wood to chemically treated and untreated wood in commercial usages (Unpublished undergraduate essay), The University of British Columbia, Vancouver, Canada. Chen, Y., Fan, Y., Gao, J., & Li, H. (2012). Coloring characteristics of in situ lignin during heat treatment. Wood Science and Technology, 46(1-3): 33-40. Cheng, K. J. (2015). Reducing the surface checking of deck-boards exposed to natural weathering: Effects of wood species and surface profiling (Master of Science Thesis). The University of British Columbia, Vancouver, B.C., Canada. Chotikhun, A. & Hiziroglu, S. (2016). Measurement of dimensional stability of heat treated southern red oak (Quercus falcata Michx.). Measurement, 87(2016): 99-103.  Chirkova, J., B. Andersons, I. Andersone & H. Militz (2005). Water Sorption Properties of Thermos-modified Wood. In: The Second European Conference on Wood Modification, Göttingen, Germany. 6-7 October. Christmas, J., Sargent, R., & Tetri, T. (2005). Thermal modification of New Zealand radiata pine. In: Proceedings of the second European Conference on Wood Modification. Militz, H., and Hill, C. (ed). Gottingen, Germany, pp. 83-86. Dagbro, O. (2016). Studies on Industrial-Scale Thermal Modification of Wood (Doctoral Dissertation), Luleå University of Technology, Luleå, Sweden. Del Menezzi, C. H. S., Amorim, M. R., Costa, M. A., & Garcez, L. R. (2014). Evaluation of thermally modified wood by means of stress wave and ultrasound nondestructive methods. Journal of Materials Science, 20(1): 61-66.  Dunn, D. (1992). A preliminary assessment of the Metriguard 239A stress wave timer (Dissertation). School of Forestry, University of Canterbury, New Zealand. Esteves, B. & Pereira, H. (2009). Wood modification by heat treatment: A review. BioResearch, 4(1): 370-404. Esteves, B., Marques, A. V., Domingos, I., & Pereira, H. (2007). Influence of steam heating on the properties of pine (Pinus pinaster) and eucalypt (Eucalyptus globulus) wood. Wood Science and Technology, 41(3): 193-207.  Feist, W. C. & Sell, J. (1987). Weathering behavior of dimensionally stabilized wood treated by heating under pressure of nitrogen gas. Wood and Fiber Science, 19(2): 183-195. Fengel, D. & Wegener, G. (1984). Wood: chemistry, ultrastructure, reactions. Walter de Gruyter, Berlin.  Ferrari, S., Allegretti, O., Cuccui, I., Sandak, J., & Sandak, A. (2012). Thermo-vacuum process for wood thermal modification: Results for some European softwood and hardwood species treated at different conditions. In: The Sixth European Conference on Wood Modification, Ljubljana, Slovenia, 17-18 September, pp. 159-166. 65  Finnish ThermoWood Association. (2003). ThermoWood® Handbook. Helsinki. Finland. Flaete, P. O., Høibø, O. A., Fjærtoft, 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(3): 135-139. Garratt, G.A. (1931). The Mechanical Properties of Wood. John Wiley & Sons, Inc., New York. Garcia, R. A., de Carvalho, A. M., de Figueiredo Latorraca, J. V., de Matos, J. L. M., Santos, W. A., & de Medeiros Silva, R. F. (2012). Nondestructive evaluation of heat-treated Eucalyptus (grandis Hill ex Maiden) wood using stress wave method. Wood Science and Technology, 46(1-3): 41-52.  Gérardin, P. (2016). New alternatives for wood preservation based on thermal and chemical modification of wood— a review. Annals of Forest Science, 73(3): 559-570. Ghalehno, M. D. & Nazerian, M. (2011). Changes in the physical and mechanical properties of Iranian hornbeam wood (Carpinus betulus) with heat treatment. European Journal of Scientific Research, 51(4): 490-498. Giebeler, E. (1983). Dimensional stabilization of wood by moisture-heat-pressure-treatment. Holz als Roh-und Werkstoff, 41(3): 87-94. Green, D. W., Begel, M., & Nelson, W. (2006). Janka hardness using nonstandard specimens. US Department of Agriculture, Forest Service Research Note FPL-RN-0303, Forest Products Laboratory, Madison, W.I, USA. González-Peña, M. M. & Hale, M. D. C. (2009) Colour in thermally modified wood of beech, Norway spruce and Scots pine. Part 2: Property predictions from colour changes. Holzforschung, 63(4): 394-401. Gonzalez de Cademartori, P. H., Schneid, E., Gatto, D. A., Martins Stangerlin, D., & Beltrame, R. (2013). Thermal modification of Eucalyptus grandis wood: variation of colorimetric parameters. Maderas. Ciencia y tecnología, 15(1): 57-64. Gray, J. D., Grushecky, S. T., & Armstrong, J. P. (2008). Stress wave velocity and dynamic modulus of elasticity of yellow-poplar ranging from 100 to 10 percent moisture content. Proceedings of the 16th Central Hardwoods Forest Conference, West Lafayette, pp. 139-142. Guller, B. (2012). Effects of heat treatment on density, dimensional stability and color of Pinus nigra wood. African Journal of Biotechnology, 11(9): 2204-2209.  Gunduz, G., Korkut, S., & Sevim Korkut, D. (2008). The effects of heat treatment on physical and technological properties and surface roughness of Camiyanı Black Pine (Pinus nigra Arn. subsp. pallasiana var. pallasiana) wood. Bioresource Technology, 99(7): 2275–2280, ISSN: 0960-8524. 66  Gunduz, G., Aydemir, D., & Karakas, G. (2009). The effects of thermal treatment on the mechanical properties of wild Pear (Pyrus elaeagnifolia Pall.) wood and changes in physical properties. Materials & Design, 30(10): 4391-4395. Hakkou, M., Pétrissans, M., Zoulalian, A., & Gérardin, P. (2005). Investigation of wood wettability changes during heat treatment on the basis of chemical analysis. Polymer Degradation and Stability, 89(1): 1-5. Hartley, C., Davidson, R. W., & Crandall, B. S. (1961). Wetwood, bacteria, and increased pH in trees. US Department of Agriculture, Forest Service Research Note FPL-RN-2215, Forest Products Laboratory, Madison, W.I, USA. Han, G., Wu, Q., & Wang, X. (2006). Stress-wave velocity of wood-based panels: Effect of moisture, product type, and material direction. Forest Products Journal, 56(1): 28-33. Hill, C. (2006). Wood modification-chemical, thermal and other processes, Wiley Series in Renewable Resources, John Wiley & Sons, Ldt, pp. 99-127, ISBN: 0-470-02172-1. Hiltunen, E., Pakkanen, T. T., & Alvila, L. (2006) Phenolic compounds in silver birch (Betula pendula Roth) wood. Holzforschung, 60(5): 519-527. Hillis, W. E. (1984). High temperature and chemical effects on wood stability. Wood Science and Technology, 18(4): 281-293. Hoadley, R. B. (1990). Identifying wood, accurate results with simple tools. Newtown, Taunton Press. Hoang, D. (2009). Thermally modified wood: from preservative to potential substitute (Unpublished undergraduate essay). Department of Wood Science, University of British Columbia, Vancouver, Canada.  Inari, G. N., Pétrissans, M., Dumarcay, S., Lambert, J., Ehrhardt, J. J., Šernek, M., & Gérardin, P. (2011). Limitation of XPS for analysis of wood species containing high amounts of lipophilic extractives. Wood Science and Technology, 45(2): 369-382. Jämsä, S., Ahola, P. & Viitaniemi, P. (2000). Long-term natural weathering of coated thermowood. Pigment & Resin Technology, 29(2): 68-74. Jämsä, S. & Viitaniemi, P. (2001). Heat treatment of wood: Better durability without chemicals. In: Review on heat treatment of wood. Proceedings of special seminar of Cost Action E22, Rapp, A. (ed.). Antibes, France, pp. 21-26. Johansson D. & Moren, T. (2006). The potential of color measurement for strength prediction of thermally treated wood. Holz als Roh- und Werkstoff, 64(2):104-110. Johansson, D. (2008). Heat Treatment of Solid Wood, Effects on Absorption, Strength, and Colour (Doctoral Dissertation), Luleå University of Technology, Luleå, Sweden.  67  Johnson, R. P. A. & Gibbons, W. H. (1929). Properties of Western hemlock and their relation to uses of the wood. U.S. Dept. of Agriculture, Technical Bulletin 139, Washington, D.C. Källander, B. (2016). Drying and Thermal Modification of Wood- Studies on Influence of Sample Size, Batch Size, and Climate on Wood Response (Doctoral dissertation), Luleå University of Technology. Luleå, Sweden. ISBN 978-91-7583-629-4. Kim, J. Y., Hwanga, H., Oha, S., Kim, Y. S., Kim, U. J., & Choi, J. W. (2014). Investigation of structural modification and thermal characteristics of lignin after heat treatment. International Journal of Biological Macromolecules, 66: 57-65. Kim, D. Y., Yochiharu, N., Masahisa, W., Shigenori, K., & Takeshi, O. (2001). Thermal decomposition of cellulose crystallites in wood. Holzforschung, 55(5): 521-524. Klauditz, W. & Stegmann, G. (1955). Contributions to the knowledge of the sequence and the effect of thermal reactions in the formation of wood materials.) Holz als Roh- und Werkstoff, 13(11): 434-440. Kollmann, F. (1936). Technologie des Holzes und der Holzwerkstoffe, (Wood Technology), Verlag von Julius Springer, Berlin. Kollmann, F., & Fengel, D., (1965). Changes in the chemical composition of wood by heat treatment, Holz als Roh- und Werkstoff, 23(12): 461-468. Konica Minolta. (2007). Colors also look different according to the subject conditions and environment. In: Precise colour communication color control from perception to instrumentation. Konica Minolta, Japan. Retrieved March 3, 2017, from:  https://www.konicaminolta.com/  Korkut, D. S., Hiziroglu, S., & Aytin, A. (2013). Effect of heat treatment on surface characteristics of wild cherry wood. BioResources, 8(2): 1582-1590. Korkut, D. S., Korkut, S., Bekar, I., Budakci, M., Dilik, T., Cakicier, N. (2008). The effects of heat treatment on the physical properties and surface roughness of Turkish hazel (Corylus colurna l.) wood. International Journal of Molecular Sciences, 9(9): 1772-1783. Lazarescu, C., & Avramidis, S. (2012). Heating characteristics of western hemlock (Tsuga heterophylla) in a high frequency field. European Journal of Wood and Wood Products, 70(4): 489-496.  Lekounougou, S., Kocaefe, D., Oumarou, N., Kocaefe, Y., & Poncsak, S. (2011). Effect of thermal modification on mechanical properties of Canadian white birch (Betula papyrifera). International Wood Products Journal, 2(2): 101-107. Majano-Majano, A., Hughes, M., & Fernandez-Cabo, J. L. (2012). The fracture toughness and properties of thermally modified beech and ash at different moisture contents. Wood Science and Technology, 46(1-3), 5-21. 68  Metsä-Kortelainen, S., Antikainen, T., & Viitaniemi, P. (2006). The water absorption of sapwood and heartwood of Scots pine and Norway spruce heat-treated at 170 °C, 190 °C, 210 °C and 230 °C. Holz als Roh-und Werkstoff, 64(3): 192-197.  Middleton, G. R., & Munro, B. D. (2001). Second-growth western hemlock product yields and attributes related to stand density. Forintek Canada Corporation, Western Lab Special Publication SP-41, Vancouver, BC. Militz, H. (2002). Heat Treatment Technologies in Europe: Scientific Background and Technological State-of-Art. In: Proceedings of Conference on “Enhancing the durability of lumber and engineered wood products”, Kissimmee, Orlando. Forest Products Society, Madison, USA. Militz, H., & Tjeerdsma, B. (2001). Heat treatment of wood by the Plato process. In: Review on heat treatments of wood, Proceedings of the special seminar of COST Action E22, Rapp, A.O. (ed.). Antibes, France, pp. 25-35. Millett, M. A. & Gerhards, G. C. (1972). Accelerated aging: residual weight and flexural properties of wood heated in air at 115 °C to 175 °C. Wood Science, 4(4): 193-201. Mitani, A. & Barboutis, I. (2014). Changes caused by heat treatment in color and dimensional stability of Beech (Fagus sylvatica L.) wood. Drvna Industrija, 65(3): 225-232. Mohareb, A., Sirmah, P., Desharnais, L., Dumarçay, S., Pétrissans, M., & Gérardin, P. (2010). Effect of extractives on conferred and natural durability of Cupressus lusitanica heartwood. Annals of Forest Science, 67(5): 504-504. Moliński, W., Roszyk, E., Jabłoński, A., Puszyński, J., & Cegiela, J. (2016). Mechanical parameters of thermally modified Ash wood determined by compression in radial direction. Maderas. Ciencia y Tecnología, 18(4): 577-586. Mullins, E. J. & McKnight, T. S. (1981). Canadian woods: their properties and uses. Published by University of Toronto Press in cooperation with Environment Canada and the Canadian Government Publishing Centre Supply and Services Canada.  Natural Resources Canada (2015). Western hemlock distribution map. Retrieved March 20, 2018, from:  https://tidcf.nrcan.gc.ca/en/trees/factsheet/119 Navi, P., & Sandberg, D. (2012). Thermo-Hydro-Mechanical Wood Processing. Switzerland, EPLF Press. Niemz, P., Hofmann, T. & Rétfalvi, T. (2010). Investigation of chemical changes in the structure of wood thermally modified. Proceedings of the 11th International IUFRO Wood Drying Conference, Skelleftea Sweden, pp. 18-22. Nishino, Y., Janin, G., Yamada, Y., & Kitano, D. (2000). Relations between the colorimetric values and densities of sapwood. Journal of Wood Science, 46(4): 267-272. 69  Nuopponen, M., Vuorinen, T., Jamsa, S., & Viitaniemi, P. (2003). The effects of a heat treatment on the behaviour of extractives in softwood by FTIR spectroscopic methods. Wood Science and Technology, 37(2): 109-115. Nuopponen, M., Vuorinen, T., Jamsä, S., & Viitaniemi, P. (2004). Thermal modification in softwood studied by FT-IR and UV resonance Raman spectroscopies. Journal of Wood Chemistry and Technology, 24(1):13-26. Ostergard, D. E. (1987). Bent Wood and Metal Furniture 1850–1946. The American Federation of Arts, New York, ISBN 0-295-96409-X. Packee, E. C. (1990). Western hemlock, Tsuga heterophylla (Raf.) Sarg. In: Silvics of North America. Vol.1. Conifers (ed. R.M. Burns and B.H. Honkala), US Department of Agriculture Handbook, 654: 613-622. Panshin, A. J. & Carl De Zeeuw. (1980). Textbook of wood technology: structure, identification, properties, and uses of the commercial woods of the United States and Canada. New York, McGraw-Hill. Patzelt, M., Emsenhuber, G., & Stingl, R. (2003). Colour Measurement as means of Quality Control of Thermally Treated Wood. In: Proceedings of the first European Conference on Wood Modification. Van Acker, J., and Hill, C. (ed). Ghent, Belgium, pp. 213-218. Percin, O., Peker, H., & Atilgan, A. (2016). The effect of heat treatment on some physical and mechanical properties of beech (Fagus Orientalis Lipsky) wood. Wood Research, 61(3): 443-456. Petrissans, M., Gérardin, P., El Bakali, I., Serraj, M. (2003). Wettability of heat-treated wood. Holzforschung, 57(3):301-307. Pleschberger, H., Teischinger, A., Müller, U., & Hansmann, C. (2014). Change in fracturing and colouring of solid spruce and ash wood after thermal modification. Wood Material Science and Engineering, 9(2): 92-101.  Pojar, J., Klinka, K., & Demarchi, D. A. (1991). Coastal Western hemlock zone. In Ecosystems of British Columbia. Special Reports Series 6, BC Ministry of Forests, Victoria, BC, Canada. pp. 95-111.  Poncsak, S., Kocaefe, D., Bouazara, M., & Pichette, A. (2006). Effect of high temperature treatment on the mechanical properties of birch (Betula papyrifera). Wood Science and Technology, 40(8), 647-663. Poncsak, S., Kocaefe, D., & Younsi, R. (2011). Improvement of the heat treatment of Jack pine (Pinus banksiana) using ThermoWood technology. European Journal of Wood and Wood Products, 69(2): 281-286.  Rapp, A.O. & Sailer, M. (2001). Oil heat treatment of wood in Germany—State of the art. In: Review on heat treatment of wood. Proceedings of special seminar of Cost Action E22: 70  Environmental Optimization of Wood Protection, Rapp, A. (ed.). Antibes, France, pp. 45-62. Rowell, R. M., Andersone, I., & Andersons, B. (2012). Heat Treatment. Handbook of Wood Chemistry and Wood Composites. 2nd ed. CRC Press. pp. 511-531.  Rohrbach, K. (2008). Schedule and post-drying storage effects on Western Hemlock squares quality (Doctoral dissertation), The University of British Columbia, Vancouver, BC, Canada. Ross, R. J. (2010). Wood handbook: Wood as an engineering material. US Department of Agriculture, Forest Service General Technical Report FPL-GTR-190, Forest Products Laboratory, Madison, W.I., USA. Sandberg, D. & Söderström, O. (2006). Crack formation due to weathering of radial and tangential sections of pine and spruce. Wood Material Science and Engineering, 1(1):12-20. Sandberg, D. & Kuntar, A. (2016). Thermally modified timber: Recent developments in Europe and North America. Wood and Fiber Science, 48 (2015 Convention, Special Issue): 28-39. Sandberg, D., Haller, P., & Navi, P. (2013). Thermo-hydro and thermo-hydro-mechanical wood processing: An opportunity for future environmentally friendly wood products. Wood Material Science & Engineering, 8(1): 64-88. Sanderman, W., & Augustin, H. (1964). Chemical investigations on the thermal decomposition of wood. Part III: chemical investigation on the course of decomposition. Holz als Roh-und Wekstoff. 22(10), 377–386  Seborg, R.M., Millet, M.A., & Stamm, A.J. (1945). Heat-stabilized compressed wood (Staypack). Mechanical Engineering, 67(1): 25-31. Seborg, R.M., Tarkow, H., & Stamm, A.J. (1953). Effect of heat upon dimensional stabilization of wood. Journal of Forest Products Research Society, 3(9): 59-67. Sehlstedt-Persson, M. (2003). Colour responses to heat-treatment of extractives and sap from pine and spruce. In: International IUFRO Wood Drying Conference. Faculty of Wood Industry, Transilvania University of Brasov, Romania. 25-29 August, pp. 459-464. Shi, J., Kokaefe, D., & Zhang, J. (2007). Mechanical behavior of Quebec wood species heat-treated using ThermoWood processes. Holz als Roh-und Werkstoff, 65(4): 255-259. Softwood Export Council (SEC). (2004). Hem-fir. Retrieved April 10, 2018, from:  http://www.softwood.org/publications-library/category/hemfir Stamm, A. J. & Baechler, R. H. (1960). Decay resistance and dimensional stability of five modified woods. Forest Products Journal, 10(1), 22–26. Stamm, A. J., Burr, H. K., & Kline, A. A. (1946). Staybwood—heat-stabilized wood. Industrial & Engineering Chemistry, 38(6): 630-634. 71  Stamm, A. J. & Hansen, I. A. (1937). Minimizing wood shrinkage and swelling: Effects of heating in various gasses. Industrial and Engineering Chemistry, 29(7): 831-833. Sundqvist, B.  & Morén, T. (2002). The influence of wood polymers and extractives on wood colour induced by hydrothermal treatment. European Journal of Wood and Wood Products, 60(5): 375-376.  Sundqvist, B., Karlsson, O., & Westermark, U. (2006). Determination of formic-acid and acetic acid concentrations formed during hydrothermal treatment of birch wood and its relation to colour, strength and hardness. Wood Science and Technology, 40(7), 549-561. Syrjänen, T. (2001). Production and classification of heat treated wood in Finland. In: Review on heat treatment of wood. Proceedings of special seminar of Cost Action E22: Environmental Optimization of Wood Protection, Rapp, A. (ed.). Antibes, France, pp. 7-15. Syrjänen, T. & Kangas, E. (2000). Heat treated timber in Finland. International Research Group on Wood Protection, IR G/WP/00-40158. ThermoWood® brochure. Retrieved March 20, 2017, from:  http://www.en.thermowood.kotisivukone.com/brochurestexts ThermoWood® Production Statistics (2016). Retrieved May 28, 2017, from: http://www.thermowood.fi/latestnews Tjeerdsma, B., Boonstra, M., & Militz, H. (1998a). Thermal modification of non-durable wood species. Part 2. Improved wood properties of thermally treated wood. International Research Group on Wood Protection, IRG/WP/98- 40124. Tjeerdsma, B. F., Boonstra, M., Pizzi, A., Tekely, P., & Militz, H. (1998b). Characterization of thermally modified wood: molecular reasons for wood performance improvement. Holz als Roh-und Werkstoff, 56(3): 149-153.  Tiemann, H. D. (1915). The effect of different methods of drying on the strength of wood. Lumber World Review. 28(7): 19-20. Twede, D. (2005). The cask age: The technology and history of wooden barrels. Packing Technology and Science. 18(5): 253-264. Unsal, O. & Ayrilmis, N.C. (2005). Variations in compression strength and surface roughness of heat-treated Turkish river red gum (Eucalyptus camaldulensis) wood. Journal of Wood Science, 51(4): 405- 409. Unsal, O., Buyuksari, U., Ayrilmis, N., & Korkut, S. (2009). Properties of wood and wood based materials subjected to thermal treatments under various conditions. Proceedings of International Wood Science and Engineering Conference in the Third Millennium (ICWSE), Transilvania University of Braşov, Romania. 72  Vernois, M. (2001). Heat treatment of wood in France – State of the art. In: Review on heat treatment of wood. Proceedings of special seminar of Cost Action E22: Environmental Optimization of Wood Protection, Rapp, A. (ed.). Antibes, France, pp. 37-44. Wallace, J. W., Hartley, I. D., Avramidis, S., & Oliveira, L. C. (2003). Conventional kiln drying and equalization of Western hemlock (Tsuga heterophylla (Raf.)[Sarg]) to Japanese equilibrium moisture content. Holz als Roh-und Werkstoff, 61(4): 257-263. Ward, J. C. & Pong, W. Y. (1980). Wetwood in Trees: A Timber Resource Problem. US Department of Agriculture, Forest Service, Pacific Northwest Forest and Range Experiment Station General Technical Report PNW-112. Retrieved from: http://www.fpl.fs.fed.us/documnts/misc/pnwtr112.pdf Weiland, J. J. & Guyonnet, R. (2003). Study of chemical modifications and fungi degradation of thermally modified wood using DRIFT spectroscopy. Holz als Roh-und Werkstoff, 61(3): 216-220.  Welzbacher, C. R., Brischke, C., & Rapp, A. (2007). Influence of treatment temperature and duration on selected biological, mechanical, physical, and optical properties of thermally modified timber. Wood Material Science and Engineering, 2(2): 66-76. Westermark, U., Samuelsson, B., & Lundquist, K. (1995). Homolytic cleavage of the β-ether bond in phenolic β-O-4 structures in wood lignin and in guaiacylglycerol-β-guaiacyl ether. Research on chemical intermediates, 21(3-5): 343-352. Williams, D. (2011). Thermally modified wood strategic analysis of the business potential for British Columbia. FPInnovations, Wood Products Division, Vancouver, British Columbia. Retrieved February 23, 2017, from:  https://fpinnovations.ca// Yang, T. H., Chang, F. R., Lin, C. J., & Chang, F. C. (2016). Effects of temperature and duration of heat treatment on the physical, surface, and mechanical properties of Japanese cedar wood. BioResources, 11(2): 3947-3963. Yildiz, S. (2002). Physical, mechanical, technological and chemical properties of beech and spruce wood treated by heating (Unpublished doctoral dissertation), Karadeniz Technical University, Trabzon.  

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.24.1-0368548/manifest

Comment

Related Items