Machining process modeling by means of FEM and the use of metamodels with fuzzy logic and linear regression

Abstract

In this work it was implemented a numerical model that simulates the process of orthogonal machining of metallic materials by means of the Finite Elements Method (FEM). For the development of the model, the program Ansys Flotran was used. The material was simulated as a high viscosity fluid crashing against a solid with the cutting edge geometry. From the numerical model it was made a factorial fractionated design of experiments using Minitab software, where the selected input variables were: cutting speed, depth of cut, rake angle, density and viscosity of the fluid. The considered output variables were the velocity of chip and the position of the stagnation point. Two functional metamodels of the FEM model were made, the first one by determining the empirical equations by means of a linear regression. The second was made by establishing functions based on fuzzy logic. The experiments with both metamodels showed that the depth of cut and the rake angle have the greater influence in the chip speed and in the position of the stagnation point. The metamodel based on fuzzy logic presented a better representation between the input and output variables.