Wood and Fiber Science (E-Journal)
Not a member yet
2592 research outputs found
Sort by
FIBER QUALITY PREDICTION USING NIR SPECTRAL DATA: TREE-BASED ENSEMBLE LEARNING VS DEEP NEURAL NETWORKS
Full text linkThe growing applications of near infrared (NIR) spectroscopy in wood quality control and monitoring necessitates focusing on data-driven methods to develop predictive models. Despite the advancements in analyzing NIR spectral data, literature on wood science and engineering has mainly uti- lizedthe classic model development methods, such as principal component analysis (PCA) regression or partial least squares (PLS) regression, with relatively limited studies conducted on evaluating machine learning (ML) models, and specifically, artificial neural networks (ANNs). This couldpotentially limit the performance of predictive models, specifically for some wood properties, such as tracheid width that are both time-consuming tomeasure and challenging to predict using spectral data. This study aims to enhance the prediction accuracy for tracheid width using deep neural networks and tree-based ensemble learning algorithms on a dataset consisting of 2018 samples and 692 features (NIR spectra wavelengths). Accord- ingly, NIR spectra were fed into multilayer perceptron (MLP), 1 dimensional-convolutional neural net- works (1D-CNNs), random forest, TreeNet gradient-boosting, extreme gradient-boosting (XGBoost), and light gradient-boosting machine (LGBM). It was of interest to study the performance of the models with and without applying PCA to assess how effective they would perform when analyzing NIR spectra with- out employing dimensionality reduction on data. It was shown that gradient-boosting machines outper- formed the ANNs regardless of the number of features (data dimension). Allthe models performed better without PCA. It is concluded that tree-based gradient-boosting machines could be effectively used for wood characterization utilizing a medium-sized NIR spectral dataset
Dynamic Properties of Wood Obtained by Frequency Resonance Technique and Dynamic Mechanical Analysis
Full text linkThe study of the visco-elastic properties of wood involves the determination of dynamic parameters using non-destructive methods. These methods differ in the possibilities of use (size and geometry of tested specimens, control of environmental conditions such as temperature or humidity), in the field of use (laboratory only, in-situ), and also in the way of determining dynamic parameters. This brings the question of whether the same parameters determined in different ways can be considered identical. This study aims to compare the Frequency Resonance Technique (FRT) and Dynamic Mechanical Analysis (DMA) as methods used for the determination of bending dynamic modulus of elasticity (MOED) and damping coefficient (tan δ) of dry and green wood. Groups of specimens of European beech (Fagus sylvatica L.), Small-leaved linden (Tilia cordata Mill.), European oak (Quercus robur L.), and Norway spruce (Picea abies L.) wood with three levels of moisture contents (MC) were tested with both methods. Obtained results showed that MOED decreases with increasing moisture content until the fibre saturation point (FSP). Above FSP, there is not any significant change in dynamic modulus with increasing MC. A strong linear correlation between MOED obtained through both methods was found (r = 0.92, r2 = 0.84). For tan δ, the relationship was weaker (r = 0.57, r2 = 0.32), and each method shows a different influence of MC on damping tan δ above the FSP, which leads to resume that this damping coefficient is sensitive to the method of determination
CASE STUDY OF 3-PLY COMMERCIAL SOUTHERN PINE CLT MECHANICAL PROPERTIES AND DESIGN VALUES
Full text linkThis work elucidates on a case study of industrially manufactured cross-laminated timber (CLT). Two methods are used to calculate specimens section modulus: Sgross and Seffective. The first assumes that specimens behave as a continuous material, whereas the second considers the cross laminations (shear analogy method). Although the shear analogy method is indicated for construction purposes, applications, such as trench shoring, matting, and work platforms, could benefit from a simpler calculation method. There- fore, theobjective of this work was to conduct a case study of Modulus of Rupture (MOR) and Modulus of Elasticity (MOE) of southern pine CLT to compare the previously mentioned calculation methods. Both parametric and nonparametric fifth percentiles and associated Fb values are reported and were substantially higher than those of the constituent lumber. For MOE, empirical testing and calculation based on gross moment of inertia provided lower values as compared with the constituent lumber
EVALUATION OF TEST SPECIMEN SURFACE PREPARATION ON COMPUTER VISION WOOD IDENTIFICATION
No full textPrevious studies on computer vision wood identification (CVWID) have assumed or implied that the quality of sanding or knifing preparation of the transverse surface of wood specimens could influence model performance, but its impact is unknown and largely unexplored. As a preliminary step in investigating the possible effects of surface preparation quality, this study evaluates the predictive accuracy of a previously published 24-class model, trained on images from Peruvian wood specimens prepared at 1500 sanding grit, with testing images from specimens (not used for training) prepared across a series of progressively coarser sanding grits (1500, 800, 600, 400, 240, 180 and 80) and high-quality knife cuts. The results show that while there was a drop in performance at the lowest sanding grit of 80, most of the higher grits and knife cuts did not exhibit statistically significant differences in predictive accuracy. These results lay the groundwork for a future larger-scale investigation into how the quality of surface preparation in both training and testing data will impact CVWID model accuracy
Densification of New Zealand-grown Eucalyptus species: Effect of grain orientation and densification process on wood properties
Full text linkEucalyptus fastigata and Eucalyptus nitens were densified using a thermo-hydro-mechanical (THM) densification process. The THM treatment was applied either as surface densification of one wood surface or as bulk densification of the entire wood thickness. To understand the effect of grain orientation on final wood properties, both quarter-sawn and flat-sawn boards were densified. The Eucalyptus nitens boards were able to be compressed to a greater degree without being damaged compared to Eucalyptus fastigata boards. This led to substantial increases in surface hardness and surface density in Eucalyptus nitens. Additionally, levels of set-recovery (irreversible swelling from contact with water) for bulk densified Eucalyptus nitens were substantially lower than Eucalyptus fastigata and lower than literature values for other species with a similar density. The reason for this unusually low set-recovery is not known, but it is of potential interest for the commercial application of densified wood, where set-recovery is unacceptable and would need to be eliminated. Density profiles showed that the peak density was generally at, or very close to, the wood surface, giving the maximum increase in surface hardness for a given degree of densification. The properties following densification were not substantially different between the quarter-sawn and flat-sawn boards suggesting that densification was effective irrespective of grain orientation
THE GLOBAL WOOD SPECIES PRIORITY LIST: A LIVING DATABASE OF TREE SPECIES MOST AT RISK FOR ILLEGAL LOGGING, UNSUSTAINABLE DEFORESTATION, AND HIGH RATES OF TRADE GLOBALLY
Full text linkThe illegal timber trade is one of the most impactful natural wildlife crimes, affecting the live- lihood of local communities, naturalresource availability, and the associated carbon storage and biodiver- sity. Many timber species are highly sought after and are atrisk of exhaustion and subsequent extinction. Although several initiatives exist to indicate tree species risk and conservation status, there is no single resource, or prioritized list, that qualifies the most high-risk and highly traded species across the globe. Organizations end up creating their own priority species lists to meet this lack of aggregated information, requiring hours of independent research and resulting in the recreation of similar lists. To provide a one- stop-shop for similar initiatives, World Forest ID developed the Global Priority Wood Species List (GPWSL) to synthesize existing information. Currently, the GPWSLharbors 270 species most at risk for illegal logging, unsustainable deforestation, and high rates of international trade. The database contains rel- evant information on each species; such as natural distribution, conservation listings, and countries of import. Here,we present the list, the methods used in its development, and its potential applications for the wood industry as a whol
PRESERVATIVE TREATMENT OF TASMANIAN PLANTATION EUCALYPTUS NITENS USING SUPERCRITICAL FLUIDS
Full text linkShort rotation plantation forests in Tasmania, Australia, are dominated by Eucalyptus nitens (common name: shining gum). These forests were primarily planted to provide material for pulp and paper production, but the timber is increasingly sought after for higher value and more enduring applications. Plantation E. nitens has a high proportion of low durability heartwood that resists penetration by conventional fluid preservatives. This limits its use to indoor applications. One approach to overcoming the refractory nature of E. nitens is to modify the treatment fluid. We investigated the use of supercritical carbon dioxide to deliver biocides deep into the wood. Timbers varying in thickness from 19 to 35 mm and 900 mm long were treated with a multicomponent biocide under supercritical conditions in a commercial facility in Denmark. The resulting timber was cut into zones inward from the surface. Wood from these zones was ground and extracted for HPLC analysis for tebuconazole and propiconazole. Preservative was detected in the inner portion of every sample examined, indicating that the process resulted in treatment throughout the boards, with concentrations meeting and on average exceeding the targeted amounts