AI-SUPPORTED COMPUTATIONAL COMBUSTIBLE CARTRIDGE CASE DESIGN

Abstract

AI-supported combustible cartridge case (CCC) design by expert system represents an innovative approach that addresses the inefficiency and inaccuracy in present CCC production design practices. Traditional CCC production design focuses mainly on its dimensions and mechanic strength, and few considerations are in its comprehensive properties, such as the relationship between composition, dimensions, microstructure and burning rate. This paper proposes a novel AI-supported CCC design expert system based on a computational model on the CCC production parameters and its burning rate. Initially, principal component analysis was employed to combine the chemical components of CCC to reduce the dimensionality of CCC production parameters, thereby facilitating the establishment of AI-supported computational match model of CCC design. Subsequently, the data acquired during the CCC production process was input into the K-means clustering algorithm to cluster the data into several clusters in order to screen necessary data used in the match model. Ultimately, the trilinear/ Radial Basis Function (RBF) interpolation method integrated in the expert system was used to calculate the production parameters according to the means of the clusters obtained from the K-means clustering algorithm. Thus, the established AI-supported computational CCC design method can streamline the experimental process, elucidate the relationship between CCC production parameters and its burning rate and facilitate the inverse design of CCC manufacturing parameters. This research provides a data-driven method for accurately and efficiently computating CCC manufacturing parameters, offering valuable insights for the design and development of high-performance CCC. It paves a transformative way for CCC formulation and structural design and is of great scientific significance.1