A knowledge-based system for process planning in cold forging using the adjustment of stepped cylinder method

We report on a knowledge-based system for process planning in cold forging. In this system, a forged product is represented as an aggregation of forming patterns that consists of cylindrical pillar parts. The basic principle of the inference method we propose is to adjust the diameters of neighboring stepped cylinders so that they are identical. Plural deformable process plans are generated using expert knowledge about working limits, die configurations, and metal flow. This system can eliminate ineffective plans by using the knowledge of how to combine plural forming patterns into a single process. Moreover, it can evaluate process plans and interactively select the optimal one by considering production costs, the forming load, the effective strain in the product, the equipment, and other factors. We applied this system to actual forged products. As a result, this system is widely applicable to various shapes and types of equipment and can improve both maintenance and operation.