Enhancing accuracy in equivalent in-service-time assessment for homogeneous solid propellants: A novel temperature-independent predictive model utilizing PCA of FTIR data

Abstract

The present study was devoted to setting a universal T-independent predictive model of equivalent in-service-time (EIST) for homogenous solid propellant (HSP) to surpass the limits of the van't Hoff law particularly when high aging temperatures and/or extended aging durations are employed in artificial aging plans. To achieve this objective, four double base rocket propellants (DBRP) underwent artificial aging for 4 months at temperatures of 323.65 K, 338.65 K, 353.65 K, and 368.65 K, with sampling conducted every 20 days. Fourier Transform Infrared spectrometry (FTIR) showed that the homolytic scission of the ONO₂ bonds and the hydrocarbon chains of the nitrate esters are the main processes occurring during the chemical decomposition. With the heating temperature increase, the chemical decomposition becomes more predominant. Furthermore, the scatter plot from Principal Component Analysis (PCA) of FTIR spectra obtained from each aging temperature showed, respectively, that over than 88.9 %, 94.3 %, 97.4 %, and 98.6% of the variances were described by the first principal component. This latter value was found 97.6 % when PCA was applied to all FTIR spectra. Using the PCA/FTIR approach recently developed, EIST was assessed for all the investigated samples. Subsequently, an individual predictive model of EIST was set for each heating temperature, which were used to establish the T-independent predictive model. The final model computed the EIST with a relative deviation of 5.3 % compared to those from the PCA/FTIR experimental way. Moreover, two similar DBRPs aged at different temperatures and durations have been used to validate the predictive model, and the associated mean absolute percentage error (MAPE) was found 4.6 %. The conducted comprehensive statistical analysis highlighted the excellent goodness-of-fit of the model. Furthermore, all the error metrics were found to decrease with the increase of the natural aging of the propellant and the heating temperature.