OPTIMIZATION OF MATERIAL SELECTION IN GUN BARREL PRODUCTION WITH ARTIFICIAL INTELLIGENCE

Abstract Choice of gun barrel material is the secret to the provision of cost-effectiveness, performance, and structural integrity in gun manufacturing. This study discusses the application of artificial neural networks (ANNs) for the evaluation of the matching compatibility of materials against major attributes such as fatigue life, yield strength, heat resistance, and wear resistance. Quantitative values were found for steel alloys and titanium with yield strength ranging from 750 MPa to 1300 MPa, fatigue resistance ranging from 0.75 to 0.92, thermal conductivity ranging from 18 W/mK to 50 W/mK, and wear resistance indices ranging from 0.75 to 0.95. The ANN model was trained on a database of 100 material samples and achieved a mean absolute error (MAE) of 3.2% in terms of predicting suitability scores. The model's architecture consisted of two hidden layers of 32 and 16 neurons, respectively, with ReLU activation and Adam optimization. The ANN-based approach was high in predictive accuracy in classifying top material selections for gun barrel manufacturing, encompassing a systematic and data-driven approach towards material selection. These findings confirm that ANNs can optimize material selection processes effectively, reduce production costs, and enable maximum manufacturing efficiency in defense production.