Transfer Function to Predict Warhead Fragmentation In-Flight Behavior from Static Data

Abstract

In recent years, static warhead arena tests have used stereoscopic cameras to count fragments in real time and correlate their individual velocity and mass. This new technique provides the ability to generate higher confidence dynamic fragmentation data that in turn can be used for code validation and much more realistic lethality and collateral damage calculations. Arena tests involve detonating static warheads; whereas, in reality, warheads arrive and detonate at high speeds. State-of-the-art simulations for high-speed warhead detonations can be challenging and time consuming while potentially missing the relevant physics of real-world detonations. In this investigation, a framework to predict warhead fragment track characteristics from real-world static arena experimental data and dynamic simulation data is explored. A model is trained on simulation and experimental data to predict the number of fragments that pass through a defined surface of interest given warhead in-flight terminal conditions. Distributions of fragment–surface intersections are modeled by Gaussian mixture models (GMMs), and random forest regressors are trained to predict these GMMs. Monte Carlo methods are used to show that random forests trained by both simulation and experimental data can predict fragment–surface intersection distributions of both static and dynamic high-speed warhead configurations.