Estimation of CNC Machining Parameter Levels for Brass Union Using an Adaptive Constrained Response Surface Optimization Model

The selection of appropriate levels of machining parameters is an important consideration that determines machinability or other quality measures. In this study, the CNC machining process was designed to optimize the effects of machining parameters such as feed rate, spindle speed, and tool life in the production of brass unions, which are commonly used in air conditioners. The numerical design was carried out using a proposed adaptive constrained response surface optimization model (ACRSOM). The ACRSOM's evolutionary operations began with the conventional factorial design, which was used to identify the influential effects of the main and some selected parameter interactions on transformed proportion responses. The first phase was used to move quickly toward the optimum with adequate design points, whereas the second phase was used to minimize the standard deviation of transformed responses under the desired mean target. The ACRSOM was generated in linear or nonlinear forms, based on either unreplicated or replicated designed plans, which were then combined to generate the new operating condition. With the optimal setting of 0.08 feed rate; 2,300 spindle speed and 10,000 tool life obtained from the proposed model, the percentage of defects is reduced from 0.3371 to 0.0610. Furthermore, process variation is greatly reduced from the previous operating condition.