Optimization of AISI 1045 end milling using robust parameter design

Abstract

AISI 1045 steel end milling, which enables manu- facturers to machine parts with low-cost tools, has been gaining prominence in the industry. To ensure the quality of the final products though, it is important to properly adjust the process parameters so as to avoid premature tool wear while providing good levels of productivity along with zero defects. This study aims to optimize the end milling of AISI 1045 steel, using carbide inserts coated with titanium nitride (TIN). The objective—to produce the best surface finishing for machined parts—is achieved by identifying the optimal combination of input parameters and output variables. While the responses analyzed consist of surface roughness, Ra and Rt, the study also considers how Ra and Rt are impacted by the cutting fluid and tool wear during the process.The process parameters analyzed include cutting speed (vc), feed per tooth (fz), axial depth of cut (ap), and radial depth (ae). The noise variables considered are tool wear (z1), cutting fluid concen- tration (z2), and flow rate (z3). To obtain optimal results, 82 experiments of a combined response surface array are con- ducted to collect data and analyze the effects of the parame- ters. In such a design, noise factors are used to generate vari- ation for the responses, allowing the estimation of a mean and a variance equation for Ra and Rt. To optimize the process, a weighted mean square error (MSE) approach is used to form a set of optimal and non-dominated solutions through a Pareto frontier. In this manner, depending on the weight assigned to the mean or variance equation, the algorithm leads to a feasi- blesolution.Theoreticalandpracticalresultsobtainedconfirm the adequacy of this proposal; a minimal surface roughness is achieved with the smallest possible influence from tool wear, cutting fluid concentration, and flow rate.