Use of artificial neural networks to optimize stacking sequence in UHMWPE protections

Abstract

The aim of the present work is to provide a methodology to evaluate the influence of stacking sequence on the ballistic performance of ultra-high molecular weight polyethylene (UHMWPE) protections. The proposed methodology is based on the combination of experimental tests, numerical modelling, and Artificial Neural Networks (ANN). High-velocity impact experimental tests were conducted to validate the numerical model. The validated Finite Element Method (FEM) model was used to provide data to train and to validate the ANN. Finally, the ANN was used to find the best stacking sequence combining layers of three UHMWPE materials with different qualities. The results showed that the three UHMWPE materials can be properly combined to provide a solution with a better ballistic performance than using only the material with highest quality. These results imply that costs can be reduced increasing the ballistic limit of the UHMWPE protections. When the weight ratios of the three materials remain constant, the optimal results occur when the highest-performance material is placed in the back face. Furthermore, ANN simulation showed that the optimal results occur when the weight ratio of the highest-performance material is 79.2%.