Machine Learning-Based Flexural Strength Prediction of Layered Ceramic Materials

Abstract

The flexural strength of layered ceramic materials plays an important role in its application, and are affected by many structural parameters and process factors. At present, the experimental methods of layered ceramic materials are inefficient, and it is impossible to systematically analyze the effect of structural parameters and process factors on the bending strength of silicon nitride ceramics. In this paper, a machine learning model based on 3 classical algorithms is successfully established, and its bending strength is predicted with 5 indexes. According to the SHAP diagram, the material of layered ceramics has the greatest influence on the bending strength, followed by the sintering temperature, this article provides the basis and reference for the preparation of layered ceramic materials with high bending strength.