Shape and process optimization of an industrial porthole extrusion die under multi-objective functions

The increasing demand for complex and of high quality extruded profiles has led to the development of even more intricate and difficult to optimize process conditions and dies shapes. The consequence is that different and potentially conflicting objectives are involved in the optimization of the process as the compliance with the required profile tolerances and mechanical proprieties and the endurance of the die itself. The fulfilment of all these requirements necessary demands for an automatic, robust and comprehensive methodology to investigate the problem. In the present work, a multi-objective optimization of the extrusion process is presented by applying it to an industrial round tube profile. Six geometric die parameters and the ram velocity were set as input variables. The objective functions were the peak pressure in the welding chamber, the minimum mandrel deflection and peak principal stress and the minimum flow unbalance at the die exit. The optimization was performed by combining a response surface methodology with a multi-objective genetic algorithm based on a 2-level full factorial numerical and experimental designs. The optimal solution extracted from the Pareto front was compared to experimental data of manually optimized solutions achieving higher product performance in terms of welds quality and die solicitation.