Modeling Mechanical Properties of 25Cr-20Ni-0.4C Steels over a Wide Range of Temperatures by Neural Networks

Abstract

From the point of view of designing materials, it is important to study the complex correlational research that involves measuring several variables and assessing the relation among them. Hence, the notion of machine-oriented data modeling is explored. Among various machine-learning tools, artificial neural networks (ANN) have been used as a stimulating tool to solve engineering-related issues. In this study, the ANN model is designed and trained to correlate the complex relations among composition, temperature and mechanical properties of 25Cr-20Ni-0.4C austenitic stainless steel. The developed model was exploited to estimate the composition−property and temperature−property correlations. The ANN predictions are well suitable for experimental results. The model was able to correlate the complex nature among input and output variables. The model was used to investigate the effect of service temperature on the mechanical properties of 25Cr-20Ni-0.4C steels over a wide temperature range. The effective response of the alloying elements on the mechanical properties of ambient as well as elevated temperatures was quantitatively estimated with the help of the index of relative importance (IRI) method. Hence, this handy technique is the best tool to overcome the designing complications and to develop the components having remarkable properties.

Keywords

• 25cr-20ni-0.4c steels
• mechanical properties
• artificial neural networks
• simulation and modeling
• index of the relative importance