Statistical Methods Applied in Modeling Corrosion Depth Measurement on Three Different Shape Memory Alloys

Abstract

This chapter presents a study focused on the corrosion behavior of three distinct shape memory alloys (CuAlNi and two types of NiTi alloys) in varied marine environments—air, tide, and seawater. The research documents corrosion damage after 6, 12, and 18 months, utilizing focused ion beam. Scanning electron microscopy and energy dispersive X-ray analyses were employed to detect the chemical alterations. This study includes both deterministic and stochastic frameworks for modeling corrosion processes. Employing a range of statistical techniques, including linear and multivariate regression, principal component analysis, and correlation analysis (linking corrosion depth with oxygen presence), the research provides an in-depth understanding of corrosion dynamics. The study explores fitting standard two-parameter and advanced multi-parameter distributions to the observed data. The dual treatment of corrosion parameters via linear and non-linear models enhances the robustness and applicability of our findings, offering more precise and effective corrosion management in marine engineering applications.