Development of a 3D log processing optimization system for small-scale sawmills to maximize profits and yields from central appalachian hardwoods

Abstract

The current status of log sawing practices in small hardwood sawmills across West Virginia was investigated and the effects of log sawing practices on lumber recovery evaluated. A total of 230 logs two species, red oak (Quercus rubra) and yellow-poplar (Liriodendron tulipifera), were measured in five typical hardwood sawmills in the state. Log characteristics such as length, diameter, sweep, taper, and ellipticality were measured. Additionally, the characteristics of sawing equipment such as headrig type, headrig kerf width, and sawing thickness variation were recorded. A general linear model (GLM) was developed using Statistical Analysis System (SAS) to analyze the relationship between lumber recovery and the characteristics of logs and sawing equipment for small sawmills in West Virginia. The results showed that the factors of log grade, log diameter, species, log sweep, log length, different sawmills, the interaction between log species and grade, and the interaction between log species and log length had significant impacts on volume recovery. Log grade, log species and headrig type had significant effects on value recovery. Hardwood lumber production includes a sequence of interrelated operations. Methods to optimize the entire lumber production process and increase lumber recovery are important issues for forest products manufacturers. Therefore, a 3D log sawing optimization system was developed to perform 3D log generation, opening face determination, headrig log sawing simulation, cant resawing, and lumber grading. External log characteristics such as length, largeend and small-end diameters, diameters at each foot, and external defects were collected from five local sawmills in central Appalachia. The positions and shapes of internal log defects were predicted using a model developed by the USDA Forest Service. 3D modeling techniques were applied to reconstruct a 3D virtual log that included internal defects. Heuristic and dynamic programming algorithms were developed to determine the opening face and grade sawing optimization. The National Hardwood Lumber Association (NHLA) grading rules were computerized and incorporated into the system to perform lumber grading. Preliminary results have shown that hardwood sawmills have the potential to increase lumber value by determining the optimal opening face and optimizing the sawing patterns. Our study showed that without flitch edging and trimming, the average lumber value recovery in the sawmills could be increased by 10.01 percent using a heuristic algorithm or 14.21 percent using a dynamic programming algorithm, respectively. An optimal 3D visualization system was developed for edging and trimming of rough lumber in central Appalachian. Exhaustive search procedures and a dynamic programming algorithm were employed to achieve the optimal edging and trimming solution, respectively. An optimal procedure was also developed to grade hardwood lumber based on the National Hardwood Lumber Association (NHLA) grading rules. The system was validated through comparisons of the total lumber value generated by the system as compared to values obtained at six local sawmills. A total of 360 boards were measured for specific characteristics including board dimensions, defects, shapes, wane and the results of edging and trimming for each board. Results indicated that lumber value and surface measure from six sawmills could be increased on average by 19.97 percent and 6.2 percent, respectively, by comparing the optimal edging and trimming system with real sawmill operations. A combined optimal edging and trimming algorithm was embedded as a component in the 3D log sawing optimization system. Multiple sawing methods are allowed in the combined system, including live sawing, cant sawing, grade sawing, and multi-thickness sawing. The system was tested using field data collected at local sawmills in the central Appalachian region. Results showedthat significant gains in lumber value recovery can be achieved by using the 3D log sawing system as compared to current sawmill practices. By combining primary log sawing and flitch edging and trimming in a system, better solutions were obtained than when using the model that only considered primary log sawing. The resulting computer optimization system can assist hardwood sawmill managers and production personnel in efficiently utilizing raw materials and increasing their overall competitiveness in the forest products market.