Prediction of the main degradation mechanisms in a hot forging steel die: Optical scanning, simulation, microstructural evolution, and neural network modeling

This paper presents a framework for assessing degradation mechanisms and life service of an H21 (ISO-EN X30WCrV9-3; 3Cr2W8V Chinese standard) carbon steel die for hot forging. Four main failure mechanisms are considered: abrasive wear, thermal cracking, plastic deformation, and mechanical cracking. Optical scanning, finite element method (FEM), nano-indentation, scanning electron microscopy (SEM), optical micrography, and electron backscatter diffraction (EBSD) are used to identify and analyze failure mechanisms in the exposed areas. Accordingly, computational models employing artificial neural networks (ANN) simulate each failure mechanism. The experimental data gathered from optical scanning and microstructure analysis show that the three regions of the die surface are subject to major failure mechanisms. Notably, ANN models developed for each degeneration/failure mechanism are accurate and reliable, and their outputs agree with experimental data. Since unexpected tool failures can increase final manufacturing costs from 15 to 30%, using the current ANN prediction models developed here may help reduce costs by up to 15%. This financial benefit can be achieved by preventing sudden stops of product lines and dedicating the chance to wear tools for treatment before breaking.