Researches On The Quantitative Relationship Models Between Structures And Impact Sensitivity And Other Safety Parameters Of Energetic Materials

Energetic materials are widely used in weapon system and civil explosive industry.The research on the safety parameters,such as impact sensitivity,electrostatic sensitivity and detonation velocity,is of great significance to the design,production and secure use of energetic materials.Based on the basic principle of Quantitative Structure-Property Relationship(QSPR),the impact sensitivity,electrostatic sensitivity and detonation velocity of energetic materials had been obtained systematic studies and the main research contents and conclusions were as follows:(1)The molecular structure descriptors were calculated by Dragon 2.1 software.Genetic Algorithm(GA)was used to sift the moleluclar structure descriptors,and then Multiple Linear Regression(MLR)and Artificial Neural Network(ANN)were employed to establish GA-MLR and GA-ANN models respectively.The validation of the models had been fully assessed through evaluating parameters of the model.Meanwile,the application domain of the model was studied with a Williams graph.(2)The QSPR results of the impact sensitivity of polynitro compounds were in the following.The R2 of two models were 0.854 and 0.974 respectively and the RMSE were 0.195 and 0.071 respectively.The two models all showed stable and reliable performance and good predictive ability.By comparing the models,the GA-ANN model was of better performance than GA-MLR,which means the impact sensitivity of polynitro compounds had strong nonlinear relationship with their molecular structures and the two models were better than the models in the literature.(3)The QSPR studies had been made on the electrostatic sensitivity of nitroaromatic compounds and nitramine compounds respectively.For nitroaromatic compounds,the R2 of two models reached 0.89 while the R2 of two models were higher than 0.96 for nitramine compounds.The fact that GA-ANN model was better than GA-MLR model means ANN can be an effective tool to predict electrostatic sensitivity accurately.The molecular structure descriptors obtained by GA illuminated that atomic weight and electronegativity of the molecule affected on the electrostatic sensitivity.By comparing models,there were a lot of advantages in terms of the selection of characteristic parameters,modeling and model evaluation.(4)The QSPR studies had been made on the detonation velocity of polynitro compounds and rich nitrogen heterocyclic compounds respectively.For polynitro compounds,the R2 of two models reached 0.90 while the R2 of two models were even higher than 0.96 for rich nitrogen heterocyclic compounds.All models built showed satisfactory predictive results and generalization ability.The molecular structure descriptors obtained by GA indicate that polarity and number of nitro and hydrazine group were the main factors affecting the detonation velocity of rich nitrogen heterocyclic compounds.By comparing models,the models here were better than the existing models.