The primary objective in precision machining is usually to attain excellent dimensional accuracy and surface finish. In addition, complimentary objectives such as cost and production rate are also important. Proper selection of cutting parameters can profoundly affect both primary and secondary machining performance objectives. While simplified and/or empirical models exist for machining processes, none of those models provides accurate prediction of the dynamic cutting forces, which in turn govern the obtainable quality of the machined surfaces. Finite element analysis (FEA) via ABAQUS/Explicit is adopted in this paper for predicting the machining dynamic cutting forces. Rake and clearance angles, as well as cutting speed are set as the design variables for optimization. Since the machining model requires significant computational resources, economizing the number of FEA runs is desirable. The optimization approach adopted is based off Efficient Global Optimization (EGO), where Kriging models are trained to predict the underlying behavior of the machining process via a finite set of sample points. New sample points are then generated via a multi-objective genetic algorithm that seeks locations of optima and/or high uncertainty in the Kriging models. Machining performance of the new samples is then evaluated via FEA, the Kriging models are re-trained and the process is repeated until one of termination criteria is met. The application study presented is an orthogonal cutting test for ultra-precision micro-cutting using diamond tools.

This content is only available via PDF.
You do not currently have access to this content.