Pass schedule algorithms for hot open die forging

Forging pass schedule algorithms on billet to square bar and billet to wide bar are developed by neural network with systematic procedures; data acquisition, mechanical parameter search and forging schedule. To analyze mechanical properties, forging pressure, ram position, manipulator position, tong angle and temperature of thermometer signals are processed with data acquisition system. Mechanical properties including deformation resistance, width spread, forging load, forming velocity and temperature are found by functional neural network through the parameter search and optimization. Developed forging pass schedule algorithms are mainly composed of calculation of optimum number of passes and reduction in each pass to economize power and minimize forging cycle time. The algorithms based on the void closure in forged products provide reasonable reliability in the application of industrial environments. (C) 2002 Elsevier Science B.V. All rights reserved.