Étude de la variabilité de la couleur du bois de bouleau à papier (Betula papyrifera Marsh.) et analyse de son impact sur la qualité et la valeur des sciages

Authors: Drouin, Myriam
Advisor: Beauregard, RobertDuchesne, Isabelle
Abstract: The availability of high quality hardwood timber has become critical for furniture and other appearance products industries over the past years. The broad distribution of paper birch (Betula papyrifera Marsh.), as well as the good aesthetic and physical characteristics of its wood, make it an interesting alternative to the high-value species traditionally used by the Québec hardwood sawmilling industry. Its pale and homogeneous sapwood is appreciated for many indoor uses. However, paper birch grows a false heartwood, also called discolored wood or red heartwood, contrasting in coloration with the surrounding sapwood. This difference in shade is not appreciated by the appearance wood products industry where in general homogeneous color products are desired. The main objective of this study is to define the limits of variability of paper birch wood color, to better understand its sources of variation, and to measure its impact on lumber products value and quality. More specifically, the aim of the research is to characterize objectively the color of its sapwood and discolored wood using L*a*b* values and to analyze the effects of tree age, diameter and vigor, as well as log height class and log quality, on these wood colorimetric values and on the proportion of discolored wood in boards. Another specific objective is to assess the impact of tree age, diameter and value on grade recovery, board color classification and lumber value. Results are based on 122 paper birch trees harvested in two different stands, from which logs of sawing quality have been sawn into 2284 boards. Trees were classified according to the MSCR tree classification system. The colorimetric analysis was performed on board images acquired by an industrial scanner developed for the appearance products industry. An image processing software, developed for the scanner was used to view these digital images on which defects have been automatically detected, to process them and to collect colorimetric information. The software was used to measure the proportion of every board surface belonging to sapwood and discolored wood regions. An average percentage area of 32.4 % of discolored wood on boards was obtained when considering all boards. Tree diameter and tree vigor significantly influenced the proportion of discolored wood in boards whereas the effect of tree age was not strong enough to have a significant influence in the model. Larger trees presented more board discoloration. Less vigorous trees showed a mean percentage area of 45.32 % compared to middle vigor classes and most vigorous trees which obtained a mean percentage area of 30.78% and 15.47 %, respectively. Neither log quality nor log height class had a significant effect on the proportion of discolored wood on the board surfaces. Results from this part of the study suggest shorter rotations as well as silvicultural treatments that can improve tree vigor to limit the presence of discolored wood in boards. Colorimetric results showed significant differences between L*a*b* values when comparing sapwood and discolored wood. The luminosity (L*) parameter appears as the best indicator of color changes in paper birch wood. These wood colorimetric values were mostly affected by tree age and tree diameter, but their effects on every colorimetric parameter were variable. Log quality and log height class also had a significant effect on some of the wood colorimetric variables, but once again their impact was variable which makes it hard to dress any clear general conclusions. Regarding board quality and value, paper birches of this study yielded a high proportion of low-grade lumber according to the NHLA rules; 60.3% of the total board surface area belonged to #2A Common and lower quality categories. Results showed that tree diameter was the most important variable affecting these outputs. Larger trees were associated with higher board quality and higher lumber value per tree. Lumber value per tree was as well influenced by tree vigor but not by tree age. Most vigorous trees (R) produced higher board values with an average of 316.62 $/m³, middle vigor S and C classes showed averages of 218.28$/m³ and 251.84 $/m³ while the less vigorous trees had the lowest average with 165.94$/m³. When selected for color, 50% of the board surface area fell under the sap category, while 28% was classified as regular presenting simultaneously both colorations and finally only 4% of the board area was classified as red. It was found that the most important variable affecting board color distribution were tree vigor and tree diameter whereas tree age had also a significant but lesser impact. In general, older, larger and less vigorous trees tended to present higher proportions of boards classified in the red category. Finally the results obtained in this study are favoring longer harvesting rotations in order to produce large trees that can be transformed in higher value and quality boards. The analysis of the random effects throughout the study demonstrated that most of the total random variance of the dependent factors came mostly from the between board variation, but also from the between tree variation and to a lesser extent from the between log variations. No site effect was found to be significant.
Document Type: Thèse de doctorat
Issue Date: 2010
Open Access Date: 16 April 2018
Permalink: http://hdl.handle.net/20.500.11794/21328
Grantor: Université Laval
Collection:Thèses et mémoires

Files in this item:
26969.pdf4.68 MBAdobe PDFView/Open
All documents in CorpusUL are protected by Copyright Act of Canada.