Chemometrics as a tool for monitoring corrosion degradation of the selected alloys in real conditions

Abstract

Monitoring of the corrosion process of alloys in real conditions often results in extensive data, which is characterized by complex interdependence, but by a large degree of mutual deviation. First of all, the large dispersion of the obtained results makes it very difficult to draw accurate conclusions about the real influence of the tested parameters on the corrosion behavior of alloys. On the other hand, in many cases, the high interdependence between the corrosion factors can also greatly burden the analyzed system and thus make it significantly difficult to recognize the main influence. Multivariate analysis, especially the principal component analysis, is becoming increasingly popular in processing of this type of data, due to its ability to recognize and eliminate redundant data. The aim of this study was to examine the possibility of using multivariate analysis methods in the processing of the corrosion test results obtained under real conditions. Based on the obtained results, it can be concluded that used multivariate method in combination with energy dispersive spectrometer analysis can be successfully used to identify the most important corrosion factors (type of corrosion environment, exposure time and technological production processes), as well as their influence on the degradation of the tested TiNi alloys under the given conditions.