Semi-Empirical Prediction of Residual Stress Distributions Introduced by Turning Inconel 718 Alloy Based on Lorentz Function

Abstract

The residual stress of machined surface has a crucial influence on the performance of parts. It results in large deviations in terms of the position accuracy, dimension accuracy and service life. The purpose of the present study is to provide a novel semi-empirical residual stress prediction approach for turning Inconel 718. In the method, the bimodal Lorentz function was originally applied to express the residual stress distribution. A statistical model between the coefficients of the bimodal Lorentz function and cutting parameters was established by the random forest regression, in order to predict the residual stress distribution along the depth direction. Finally, the turning experiments, electrolytic corrosion peeling, residual stress measurement and correlation analysis were carried out to verify the accuracy of predicted residual stress. The results show that the bimodal Lorentz function has a great fitting accuracy. The adjusted R2 (Ad-R2) are ranging from 95.4% to 99.4% and 94.7% to 99.6% in circumferential and axial directions, respectively. The maximum and minimum errors of the surface residual tensile stress (SRTS) are 124.564 MPa and 18.082 MPa, those of the peak residual compressive stress (PRCS) are 84.649 MPa and 3.009 MPa and those of the depth of the peak residual compressive stress (DPRCS) are 0.00875 mm and 0.00155 mm, comparing three key feature indicators of predicted and simulated residual stress. The predicted residual stress is highly correlated with the measured residual stress, with correlation coefficients greater than 0.8. In the range of experimental measurement error, the research in the present work provides a quite accurate method for predicting the residual stress in turning Inconel 718, and plays a vital role in controlling the machining deformation of parts.

Cutting Forces and Tool Wear Investigation during Turning of Sintered Nickel-Cobalt Alloy with CBN Tools

“…Before starting the tests, experiments were carried out to determine the wear of the cutting edge. In the statistical analysis of the test results, the model of the matching function according to Equation (1) was adopted [ 13 ]. …”

“…The S/N (signal-to-noise) ratio analysis strategy was adopted as “the lowest-best” according to the Equation (2) [ 13 ]. where y i is the respective characteristic and n is the number of observations.…”

“…Figure 4 shows the microstructure of the cutting edge surface. The research plan was developed according to the Taguchi method [13]. Based on the analysis of the problem, the research plan was adopted in accordance with Table 4.…”

“…The S/N (signal-to-noise) ratio analysis strategy was adopted as "the lowest-best" according to the Equation (2) [13].…”

“…Taking into account a significant number of limitations in the cutting process (e.g., related to a machine tool, a workpiece, and the required condition of the surface layer of the material), the problem of selecting optimal cutting parameters is very complex, as demonstrated by Pawade et al [ 12 ]. An analysis of stresses and deformation during the cutting process was performed using the finite element method and verified experimentally by Peng et al [ 13 ]. Struzikiewicz et al [ 14 ] used the TGRA method to optimize the cutting process and to select the optimal parameters for turning sintered alloys made by a laser sintering method.…”