Multi-Objective Optimization for the Micro-Milling Process With Adaptive Data Modeling

Abstract

This paper presents a multi-objective optimization study for the micro-milling process with adaptive data modeling based on the process simulation. A micro-milling machining process model was developed and verified through our previous study. Based on the model, a set of simulation data was generated from a factorial design. The data was converted into a surrogate model with adaptive data modeling method. The model has three input variables: axial depth of cut, feed rate and spindle speed. It has two conflictive objectives: minimization of surface location error (which affects surface accuracy) and minimization of total tooling cost. The surrogate model is used in a multi-objective optimization study to obtain the Pareto optimal sets of machining parameters. The visual display of the non-dominated solution frontier allows an engineer to select a preferred machining parameter in order to get a lowest cost solution given the requirement from tolerance and accuracy. The contribution of this study is to provide a streamlined methodology to identify the preferred best machining parameters for micro-milling.