Olav Høibø

Modelling moisture content and dry matter loss during storage of logging residues for energy

Abstract To achieve optimal utilisation of logging residues for energy, it is important to know how different handling and storage methods affect fuel properties. The aim of this study was to model how the moisture content and dry matter losses of logging residues develop during storage. Logging residues were collected from five different stands of spruce and pine during different seasons of the year and stored in the same location. The logging residues were stored in covered piles of bundled residues and loose residues. Only minor differences were found in the moisture content profiles between piles of bundles and loose residues. Logging residues located in the centre of both types of piles had considerably lower moisture content than the outer parts. The moisture content significantly affected dry matter loss, with the highest dry matter losses being found in the samples with the least favourable drying conditions. The dry matter losses varied between 1 and 3% per month. Significantly higher dry matter losses were found in the spruce bundles than in the pine bundles. Seasoned logging residues had the lowest dry matter loss, while the logging residues harvested and piled in the autumn had the highest loss.

Consumer perception of wood surfaces: the relationship between stated preferences and visual homogeneity

Information about people’s preferences as to wood products is of relevance to several decision makers in the forest sector. Studies revealing consumer preference provide information that can be used for marketing and manufacturing of wood products, but these also provide information of relevance to designers and decision makers involved in building design and construction processes. Previous studies show that the overall harmony of the visual surface is correlated with preference. In this study, perceived visual homogeneity is modeled for five copper-impregnated and five organic biocide-impregnated decking materials with different visual quality. The models are based on visual variables. Homogeneity is a function of material-dependent variables (dry knots, knot shape, and splay knot), production-dependent variables (stain), and surplus color, which is a combination of both wood property and treatment. The results imply that homogeneity is influenced by both wood properties and treatment. Producers of decking should, while maintaining a focus on using high-quality raw material, also focus on producing a product with an unstained appearance.

Development of chemometric models based on near infrared spectroscopy and thermogravimetric analysis for predicting the treatment level of furfurylated Scots pine

The use of furfuryl alcohol (FA) as a wood modification agent has been known for decades. An independent and reliable analytical method to determine the level of furfurylation is not available. This article reports the use of near infrared spectroscopy (NIR) and thermogravimetric analysis (TGA) to make partial least square prediction models for determining the furfurylation level (the percentage of FA polymer formed within the wood structure). A total of 115 individual samples of furfurylated Scots pine (Pinus sylvestris) originating from 115 production batches were used for modelling. As much as 81 samples were randomly selected for the calibration set and 34 samples for the validation set. Both NIR and TGA gave good predictions when validated by a separate test set. The r2 for NIR and TGA are 0.93 and 0.94, respectively, and the root mean square errors of predictions are 1.025 and 0.958, respectively. However, the number of principal components for the NIR and TGA models is two and six, respectively. The NIR method is preferred because only two principal components are used and sampling is fast.

Variation in treatability of Scots pine ( Pinus sylvestris ) by the chemical modification agent furfuryl alcohol dissolved in water

Scots pine (Pinus sylvestris) sapwood was investigated for variation in treatability using the wood modifying agent, furfuryl alcohol (FA) in water. The variation in treatability within trees, between trees and between different stands of Scots pine was studied. Investigated variables that reduced the residual variance significantly were: site location, latitude of site, height of trees, annual ring width, vertical and horizontal position in the tree and method of drying. Linear mixed model statistics were used and tree number was handled as a random variable. The best model reduced the treatability residual variance by 67%. Location was the single factor affecting treatability most. Differences in latitude between locations may be the reason for that. Latitude correlated negatively with the treatability. Within the trees, the treatability of sapwood increased with distance from ground and with distance from heartwood border. A small, but significantly better treatability was found for kiln dried wood (60°C) compared to air dried wood (20°C).

Factors influencing surface mould growth on wooden claddings exposed outdoors

Abstract Development in surface mould growth on painted/unpainted wooden claddings and acting climatic factors were investigated over a period of 3 years. Eight wood substrates, including modified, preservative-treated and untreated wood, were tested in combination with three types of paint: (1) water-borne alkyd modified acrylic paint without fungicide; (2) solvent-borne alkyd paint without fungicide; and (3) ICP (internal comparison product). One set of samples was exposed unpainted. The samples were tested according to a modified version of EN 927-3. A logistic regression model was fitted to the data. The degree of mould growth varied with exposure time, coating typology, wood substrate, temperature and relative humidity. Exposure time and coating typology contributed most to the model. After 3 years of outdoor exposure unpainted panels and panels coated with solvent-borne paint without fungicide had more mould growth than panels coated with ICP and water-borne paint without fungicide. Unpainted oil/copper–organic preservative-treated claddings had higher resistance to mould growth than other unpainted wood substrates. Coated untreated pine and coated acetylated pine were more susceptible to mould growth than other coated wooden substrates.

Simulated yield in a sawmill using different measurement technologies

The potential of using different scanning technologies for measuring the shape of sawlogs is evaluated. The operation of 1- and 2-axis shadow scanners, a 3-d scanner and a 3-sources X-ray scanner is simulated, however, without taking measurement inaccuracy into account. The cross-sectional shape at the top end of 100 Scots Pine (Pinus silvestris) logs and 106 Norway spruce (Picea abies) logs was used in the study. These shapes are measured manually and the operation of the different scanners, that is, the measurements and the shape predictions leading to a decision on the sawing pattern, is simulated. Different technological situations as whether the bark has or has not been removed by the time of scanning, and whether the rotation of the logs at sawing is random, fixed or optimized, are considered. The performances of the scanners in the different situations are given in terms of the average main yields. The results show that there is a yield increase of ca. 0.5 percentage points when replacing a 1-axis shadow scanner with a 2-axis shadow scanner. There is a further potential yield increase in introducing a 3-d scanner or an X-ray scanner, however, in order to get a substantial improvement using these scanners one has to control the orientation at sawing. The difference in yield of the 3-d scanner between random and optimized orientation is ca. 2 percentage points when measuring on bark and ca. 4 percentage points when measuring under bark.ZusammenfassungDie Nutzungsmöglichkeit verschiedener Meßtechnologien zum Erfassen von Rundholzquerschnitten in einem Sägebetrieb werden bewertet. Die Verfahren mit ein- und zweiachsigen Scannern, einem 3D-Scanner sowie einem dreiachsigen Röntgenscanner wurden simuliert, allerdings ohne Berücksichtigung ihrer Fehlergrenzen. Für die Untersuchung wurden die oberen Querschnitte von 100 Kiefernstämmen und 106 Fichtenstämmen verwendet. Deren Querschnittsformen wurden zunächst handvermessen, danach die verschiedenen Verfahren simuliert, mit denen die Querschnittsform bzw. dieEinschnittspläne berechnet werden. Dabei wurden auch verschiedene Situationen berücksichtigt, je nachdem, ob mit oder ohne Rinde gescannt wird oder ob die Stämme zum Sägen frei drehbar, fixiert oder optimal ausrichtbar sind. Die Ergebnisse zeigen, daß die Ausbeute um 0.5% gesteigert werden kann, wenn einachsige Scanner durch zweiachsige ersetzt werden. Eine weitere Steigerung ermöglicht der Einsatz eines SD-Scanners, allerdings nur, wenn das Rundholz zum Sägen optimal ausgerichtet werden kann. Die Ausbeutesteigerung beträgt ca. 2% bei Messen mit Rinde und bis zu 4% bei Messen unter der Rinde.

Variability of density and bending properties of Picea abies structural timber

Abstract Density, modulus of elasticity and bending strength were measured on 333 boards from 54 Picea abies trees, sampled from six sites in a small area in southern Norway. The variability of the properties was analysed in a linear mixed model with relative tree size and log type as fixed effects, and with site and tree as random variables. Relative tree size was the most important fixed effect, and both density, MOE and MOR all decreased with increasing relative tree size. Log type explained a smaller part of the variance, but it was important when predicting MOR from MOE since these had different longitudinal variations.

Pre-sorting of Norway spruce structural timber using acoustic measurements combined with site-, tree- and log characteristics

This study assesses whether acoustic velocity alone or in combination with information about the timber obtained in the forest or on the log-yard can be used for pre-sorting of Norway spruce (Picea abies [L.] Karst) structural timber. The study is based on 1235 boards from 205 trees sampled from 14 sites in Norway. In addition to various site-, tree- and log variables, acoustic velocity was obtained by Fibre-gen tools, the ST300 on standing trees and the HM200 on logs. Both ST300 and HM200 explained parts of the variability of the indicating property measured by Dynagrade strength grading machine, but the accuracy was better for HM200 than for ST300. The model with ST300 was substantially improved by introducing forest data, i.e., height to diameter ratio, age and relative longitudinal position in the tree while the improvement by combining HM200 with log data, i.e., log tapering, was minor. Grade yield after pre-sorting based on the developed models was simulated, and the results showed a possibility of increasing the C30 yield. The results show that acoustic velocity could be used either as an inventory tool, to give information about the available material from a site, or as a pre-sorting tool before sawing the logs to certain products.

A high-throughput X-ray-based method for measurements of relative wood density from unprepared increment cores from Picea abies

We have described a high-throughput method for ring width and relative wood density analysis of unprepared increment cores based on conventional computer tomography scanners (CT scanners) used in health care. The method was developed for analysis of 5-mm increment cores collected in forest genetic and tree breeding trials. For these studies, the relative differences between individuals are more important than the exact value. The method is highly efficient in terms of quantitative data collection from large number of samples. There are, however, limitations in terms of resolution and accuracy on very small volumes of wood that must be considered when using the method to measure individual growth rings or increment cores thinner than 5 mm. However, in studies comprising large number of samples, the use of cost-efficient CT scanning on the major part of the material, combined with more laborious high-resolution scanning on sub-samples can be a good combination for efficient determination of radial variation in growth rate and wood density.

Between- and within-site variation of density and bending properties of Picea abies structural timber from Norway

ABSTRACT This study provides an analysis on the variability of structural timber of Norway spruce (Picea abies) grown in Norway. Density, modulus of elasticity (MOE) and bending strength were measured on 1188 boards from 205 trees, sampled from 14 sites throughout Southern Norway, Eastern Norway and Trøndelag. The area represents the procurement area for the majority of Norwegian sawmills. The variability of the timber properties was analysed in a linear mixed model where the random variance was divided into variance due to site, variance due to trees and within-tree variance. Models describing variance due to site based on site index, altitude and latitude were developed, and combined with data from the Norwegian National Forest Inventory to estimate mean values and variability of the timber properties. The results showed that major parts of the variance due to site are explained by altitude and site index, and for density also by latitude. Major parts of the variance due to site and the variance due to trees in bending strength and MOE were explained by density.