Optimization of forging processes with a concurrent approach

Abstract

With the increasing complexity of manufacturing processes and the interrelations between the different phases, the possibility of a phase being affected by a preselected design variable is very high. In that case, the influence of a design variable not only on the part performance of but also on the feasibility and efficiency of its manufacturing should be considered. Concurrent engineering approaches should then be adopted to take the control of such complicated relations among the subprocesses. In this thesis, a concurrent design optimization methodology was proposed to minimize the cost of a cold forged manufacturing process using the design parameters as optimization variables. An objective function was defined combining material cost, manufacturing cost, and post manufacturing (shearing) cost of the product. The part to be optimized was a simply supported I-beam under centric load. Nelder-Mead was selected as the search algorithm and analysis was carried out using commercial finite element software, ANSYS. Some geometric parameters were chosen as optimization variables: the rib radiuses, die land, and burr thickness. After finding the optimum values for these parameters within the feasible domain, some design guidelines were proposed. The most effective parameter on the overall cost was found to be the die land and the most dominant term among the three cost terms was found to be the material cost.