Predicting Pallet Part Yields From Hardwood Cants

• Main Author: Mitchell, Hal Lee
• Other Authors: Wood Science and Forest Products
• Format: Others
• Published: Virginia Tech 2014
• Subjects: yields; hardwood; pallet lab; crates; skids; cants; wood; cant grades; cant; pallet
• Online Access: http://hdl.handle.net/10919/41288; http://scholar.lib.vt.edu/theses/available/etd-022499-194951/

Pallet cant quality directly impacts pallet part processing and material costs. By knowing the quality of the cants being processed, pallet manufacturers can predict costs to attain better value from their raw materials and more accurately price their pallets. The study objectives were 1) to develop a procedure for accurately predicting hardwood pallet part yield as a function of raw material geometry and grade, processing equipment, and pallet part geometry, 2) to develop a model for accurately predicting raw material costs for hardwood pallet parts as a function of yield, 3) to examine current pallet industry methods of determining hardwood cant quality, and 4) to develop and evaluate hardwood cant grading rules for use in the pallet industry. Yield studies were necessary to accurately quantify the relationship between yield and cant quality. Thirty-one yield studies were conducted throughout the Eastern United States at pallet mills producing pallet parts from hardwood cants. 47, 258 board feet of hardwood cants were graded, and the usable pallet part yield and yield losses were determined for each grade. Yield losses were separated into three components: kerf loss, dimension loss, and defect loss. Kerf and dimension losses are a function of raw material and part geometry and were calculated without regard to cant quality. Defect loss is dependant on cant quality and was calculated for each cant grade as a function of total yield, kerf loss, and dimension loss. Mathematical models were developed from twenty-eight mill studies to predict each yield loss component as a function of cant dimensions, grade, and orientation, cutting bill parameters, pallet part dimensions, and kerf. Dimension and kerf losses were predicted geometrically. Regression analysis was used to predict defect loss. Results indicated that these models accurately predicted the total yield of usable pallet parts and pallet part material costs as a function of cant quality and price. Results also indicated that the pallet industry's current method of counting the number of "bad" ends per cant bundle to determine cant quality is not adequate. The effectiveness of the proposed cant grading rules was determined by grading cants and analyzing the cant grade distributions and corresponding pallet part yields. The grade rules produced statistically different quality divisions between grades. However, a more practical single cant grade based on the minimum quality for the proposed grade 2 rules is recommended. === Master of Science