Multivariate analysis applied to test procedure for determining gun propelling charge weight Part I. Preliminary analysis of the data set

A multivariate analysis was carried out in order to find the reasons for unsatisfactory performance of the test procedure for determining gun propelling charge weight discovered in the course of test firings for 155 mm full charge. The data set consisting of 2842 rounds fired on 67 firing occasions was analyzed simultaneously, which departs from the normal procedure of analyzing each firing occasion separately. A need to divide the data set into two groups based on two different barrel cone constructions of the gun was found by means of a correlation matrix and principal component analysis (PCA). In the further study it was found out that several additional changes and trends had taken place in variable values during the collection of the observations. The multivariate approach was found to be a suitable way to obtain an overview of the historical data set at hand. The PCA model developed could be used for the quality control of test procedure for determining charge weight.