Thirty full-length sample trees from the B.C. Interior were selected for a study to determine whether external log characteristics can predict internal log quality. The sample trees were also used to create 3-dimensional log images for sawmill simulation purposes. "LogSaw", a simulation tool with internal log defect detection capabilities, was used to explore the extent to which internal and external log quality information can improve log breakdown optimization. A model of a hypothetical sawmill producing lumber for the standard North American dimension market was created to study how lumber value recovery depends on different sawing optimization scenarios.
Three sawing optimization scenarios using different levels of knowledge of internal log defects were compared to currently used sawing optimization technique:
Ideal sawing optimization - all defects within log interior are known.
Sawing optimization using only the knowledge of surface knots.
Sawing optimization using log rotation instructions based on zones of least external knot density.
Simulation results have shown that it is worthwhile to “look into the log”. When compared with the current optimization technique, the sawing optimization, including the full knowledge of log interior, has increased the value recovery by 6.2%. When only the surface knots were projected into the log interior and included in the optimization, the value recovery had increased by 4.3%. Even this 4.3% increase is still a big improvement because this sawing optimization could be implemented using currently available scanning technologies and optimization software enhanced to include log surface knots. The scenario of using log rotation instructions based on predicted zones of least internal knot density did not show value recovery improvement.
Including surface knots in the log breakdown optimization has considerably increased sawmill revenue; the hypothetical sawmill considered in this study, processing 400,000 m3 of log per year, has increased its revenue by $2.2 million.
Forest companies across Canada are interested in using laser scanners for scaling logs because it has potential for reducing scaling costs. Scanning logs over bark requires a method to obtain the under-bark diameter in order to calculate the solid wood volume. This report evaluates the methods of applying a bark factor to determine under-bark diameter. It also identifies new scanner scaling technologies for measuring bark thickness.
Using automation to maximise yield from increasingly rare and costly raw materials is a solution that can help secondary wood producers improve their profitability. By integrating an automated defect detection system, lumber producers can potentially increase production output and grade recovery, helping them to strengthen their strategic business advantage.
To develop a reference tool to assist in the choice of an appropriate defect detection system, Forintek conducted a detection capacity evaluation of commercially available equipment. Nineteen (19) manufacturers who work in the area of defect detection in lumber were contacted; of these, four agreed to participate in the study.
The project objectives were based on requests from the producers: the evaluation focussed on the detection capacity of specific defects and not on the performance of the overall system. Defects were identified and an experimental evaluation was conducted to determine if the equipment recognised the defects or not. A decision tool based on a multi-criteria analysis has been proposed in the completed project report, to help producers identify the most appropriate defect detection system. However, no evaluation can be offered for the overall performance of the systems assessed, as production needs differ from producer to producer.
A software, by which virtual 3-D subalpine fir (Abies lasiocarpa [Hook] Nutt) logs can be re-created, visualized, and theoretically sawn an infinite number of times, was developed. The software also facilitates obtaining data for determining quantitative variation of clear wood, wet-wood, and knot patterns within the tree stems.
Results based on the quantitative calculations showed that there are two general patterns of wet-wood within the sub-alpine fir stems. The first pattern is called wet-pocket and the second pattern is called wet-streak. Wet-streak patterns are generally confined to the medullary-inner heartwood regions in the outer heartwood and heartwood-sapwood transition zones of the tree stems, mostly associated with dead knots. Wet-pocket patterns consist of portions occurring in mid regions in close proximity to the base and regions mostly around partially dead knots of the tree stems. Both wet-wood patterns usually converge at the nodes and extend along the branch axes, forming a connection with the exterior boundary only around branches.
Numerical analysis of the results showed that the volume of both types is more prevalent in the lower-stem regions, becoming less prevalent towards the living crown. The radial extension of wet-wood types with radial distance from the tree centre was variable, with a maximum diameter of 22 cm. Both wet-wood volumes increased with increasing tree age and diameter class independent of age. However, the percentage of total wet-wood volume decreased with increasing DBH, increased stem height and showed no clear trend with age class. Total amounts of wet-wood ranged up to 27 per cent in individual stems. A weak relationship was found between dead knot-pattern and wet-streak pattern volumes, while a moderate high relationship was found between partially dead knot and wet-pocket volumes. A weak relationship was found between external tree characteristics and both wet-wood distributions.
As a result, some promising trends emerged for a better understanding of wet-wood and knot pattern variations as influenced by tree stem locations, DBH, and age. The developed software may offer a compelling technique for assisting subalpine fir log processing decisions. However, the destructive data collection method used in this study is “error-prone”. Therefore, an interesting alternative would be the use of more accurate non-destructive scanning techniques, such as CT-scanning, to verify the trends identified here through more deliberate sampling at other forest sites. A new study is already underway to meet this need.
Computer simulation was used to evaluate the performance of three scenarios of sawmill operation: 1) External Scanning at the bucking station, 2) X-ray Scanning with external plus partial internal knot scanning at the bucking station and 3) Perfect Scanning with external plus full internal knot scanning throughout the mill.
Forty sample stems were scanned and stem models developed to provide input for Forintek’s sawing simulation program, OPTITEK®. Input files of both sawmill machinery and their products were developed based upon the operation of a typical sawmill in the Interior of B.C. Optimized bucking solutions were generated, and sample stems were sawn accordingly. Lumber value and volume recovery data were obtained and enabled a comparison of the performance of the three operational scenarios.
The X-ray Scanner Scenario provided a 2% value recovery increase compared to that of External Scanner Scenario. The Perfect Scanner Scenario added an additional 5.5% to that of X-ray Scanner Scenario. This was due partly to improved scanning at the bucking station but more so to the internal defect detection at all machine centers in the mill.
Conclusions of the study should help sawmillers in their investment decisions regarding sawing optimization improvements.