ENHANCED 155 MM ARTILLERY PROPELLING CHARGE DEVELOPMENT WITH NUMERICAL MODELLING AND A DEEP-LEARNING APPROACH

Abstract

Since the start of the war in Ukraine, one can realize the emphasis put on long-range artillery fires. Based on this observation, the EURENCO Group is developing new artillery propelling charges with improved interior ballistics performances and range. Developing and testings these charges may present a certain level of uncertainty because the dynamic behaviors of a new propellant and charge is difficult to accurately predict from closed vessel experiments and 0D numerical calculations. EURENCO is building models to tackle these issues in particular using a deep-learning based surrogate model coupled with EURENCO proprietary 0D internal ballistics numerical modelling code in order to increase numerical calculations representativeness. These modelling codes were used to support a long-range propelling charge development process. They allowed to determine optimal propellant grain dimensions and the velocities (deep-learning based model) and pressures (0D model) were fairly well predicted. Thus, development time and safety management were greatly improved.