Error Compensation in Turning Process with Tool Wear

Abstract

An error compensation system for turning operations based on prediction of machining error is described. The machining error in longitudinal turning processes is mainly dependent on the deflection of the workpiece due to the cutting force and the thermal expansion of the cutting tool due to cutting heat. Furthermore the cutting force and cutting temperature vary with flank wear. The machining error can, therefore, be predicted with cutting force, cutting temperature and flank wear. A neural network which uses input information including cutting force and cutting temperature enables us to predict the machining error with distance from chuck to cutting point. The machining error is then compensated for by modification of the depth of cut at each cutting location. The compensation system is proved to be effective in the turning of cylindrical products.