JWL equation of state parameters for ideal detonation: A comparative study of literature reports, theoretical derivation, and artificial neural network prediction

The Jones-Wilkins-Lee (JWL) equation of state (EOS) is widely used for interpreting the energy release during explosive detonation. Focused on the ideal detonation of high explosives, an artificial neural network (ANN) model was developed in this study for predicting the parameters of JWL EOS. During model establishment, the chemical composition, charge density and enthalpy of formation of high explosives were used as input data. The output results comprised the JWL EOS coefficients. For model training, a dataset of 66 publicly available samples was used, and normalization preprocessing was applied to these data. Using the isentropic expansion curve of detonation products as references, the validation of ANN model was conducted through a comprehensive comparison method incorporating public data, theoretical derivation and model predictions. The results demonstrate that isentropic expansion curves obtained from literature data, theoretical derivation and model prediction exhibit a high degree of coincidence for HMX, LX-19 and PBX9701 high explosives. This agreement is evidenced by discrepancies of error metrics remained below 5 %, correlation coefficients exceeding 0.99, similarity index surpassing 0.97 and statistical test hypothesis being satisfied. This ANN model achieves accurate prediction of JWL EOS parameters for ideal detonation. This study can provide a novel approach for predicting JWL EOS parameters of high explosives. Focused on the ideal detonation behavior of high explosives, an artificial neural network (ANN) model was developed in this study for predicting the parameters of JWL EOS. Taking HMX, LX-19 and PBX9701 high explosives as testing cases, the results demonstrate that isentropic expansion curves obtained from literature data, theoretical derivation and model prediction exhibit a high degree of coincidence, which indicates the validity of ANN model for prediction of JWL EOS parameters in ideal detonation modeling of high explosives. This study can offer a novel approach for predicting JWL EOS parameters of high explosives.