Determination of the state parameters of explosive detonation products by computational inverse method

Abstract

A computational inverse method is presented to determine the state parameters of Jones-Wilkins-Lee (JWL) equation of explosive detonation products based on cylinder test. In this method, the inverse problem of identifying the JWL parameters is formulated through minimizing the errors of the measured radial displacements on the cylinder surface and those computed by the forward solver. An available numerical model simulated by finite element method is built for the sake of obtaining results using the given JWL parameters. Because of the difference of coordinate systems between experiment and numerical model, it is necessary to conduct the transformation between the two coordinate systems. The sensitivity analysis method is adopted to evaluate the influence of the JWL parameters on the responses and find out the parameters those are suitable to be identified. In order to improve the computational efficiency, radial basis function approximate model is constructed to replace the time-consuming numerical model. In the process of constructing approximate model, the truncated singular value decomposition method is used to deal with the ill-condition of the model. At last, the intergeneration projection genetic algorithm is adopted to identify the parameters. The numerical results demonstrate that the proposed inverse method is potentially available to effectively identify the JWL parameters.

Determination of the JWL Parameters of Detonation Products Using Adaptive Genetic Algorithm

“…However, the manualfitting process can be complicated and time-consuming. Hence, other fitting methods combining experimental data and theoretical calculations, i. e., taking advantage of fitting function [3], computational inverse [4], analytical models [5][6], and optimizing code [7], have also been developed. JWL parameters can be obtained efficiently with high precision by substituting the experimental data into these theoretical models.…”