A comparison of twenty simulated strategies for achieving maximum parts value recovery in optimizing chop saw operations

A simulation model, using a dynamic program and other functions, successfully emulated an automated chop saw that cuts, from a sample mill, full length lumber into usable parts according to a cutting bill. The program determined i f value recovery per board foot (value/fbm) from a mill run could be increased by modifying the chop saw's priority settings which forced parts into a cutting solution. Twenty different cutting algorithms were employed of which 18 prioritized up to three different parts, one was based on the sample mill's priority settings and the remaining one was a control or regular optimization. Each algorithm scanned the cutting bill to dynamically update priority settings, using the value of pieces remaining to be cut as a deciding factor each time a full length board was processed. Production was stopped on parts which had reached a certain production point or cutoff percentage. Data, consisting of grades and lengths of all lumber sections from two dimensions, 3" width and 3.75" width, were put into a simulation model for processing. Students-Newman-Keuls tests showed that there was no single algorithm that performed decisively better than others. The highest rated cutting algorithm prioritized three different parts and forced one of each part into a cutting solution with an 80% cutoff percentage. Further data analysis proved that good value/fbm results were not obtained by using more infeed lumber or sacrificing parts recovery percentage. Key Words: chop saw, cutting algorithms, dynamic program, optimization, priority settings, simulation.Forestry, Faculty ofGraduat