Application of Partial Least-Squares Quantitative Analysis of Infrared Spectroscopic Data to Low-Vulnerability Ammunition Propellant Powders

Abstract

Reported herein is an application of partial least-squares (PLS) quantitative analysis to Fourier transform infrared spectra of a low-vulnerability ammunition (LOVA) propellant powder. The propellant components analyzed included a highly-energetic material, 1,3,5-trinitro-1,3,5-triazacyclohexane (RDX), an energetic binder, nitrocellulose, a non-energetic binder, cellulose acetate butyrate, and a plasticizer agent, acetyl triethyl citrate. The propellant system consisted of one major (RDX, 76% nominally by weight) and four minor components. Consequently, separate calibrations were developed to cover the two different concentration ranges. Samples were analyzed with the use of a simple KBr transmission cell over the 1800–1500 cm−1 region. The optimized PLS calibration models created were found to provide estimations within 1% error for both validation and real propellant samples without any separation or purification step. The mathematical model built by such an algorithm could serve to develop a quality control method to monitor the composition of gun powders.