Ballistic Performance Analysis of Silicon Carbide Ceramic Body Armor Using Finite Element Method and Machine Learning Algorithms

This study presents a machine learning-based approach for predicting the residual velocity of projectiles impacting silicon carbide (SiC) ceramic body armor plates of varying thicknesses. Explicit dynamic simulations were performed using the ANSYS finite element software to model the ballistic response of the armor under high-velocity impact. Key input parameters included projectile type, bullet muzzle velocity, ceramic thickness, and mesh size. The output parameter of interest was the residual velocity of the projectile after impact. Simulation data were used to train and evaluate three different machine learning models: Linear Regression, ElasticNet, and Multilayer Perceptron (MLP). The predictive performance of each model was assessed using the coefficient of determination (R), mean absolute error (MAE), and root mean square error (RMSE) metrics across both training and testing datasets. Among the tested algorithms, the MLP model achieved the highest accuracy and lowest error values, demonstrating superior capability in capturing the complex nonlinear relationships governing ballistic impact phenomena.The findings indicate that machine learning techniques, when trained with high-fidelity simulation data, can serve as efficient predictive tools for estimating residual velocity in ballistic protection applications. This approach can significantly reduce the need for extensive physical testing and computationally expensive simulations during the preliminary design phase of protective armor systems, thereby accelerating the material selection and optimization process.