A Multiscale Framework for Predicting Alloy Tribocorrosion Behavior

Thesis Abstract

In modern engineering systems, the tribocorrosion behavior of alloys under combined mechanical wear and chemical degradation remains inadequately predictable due to the coupled, multiscale phenomena spanning atomic to macroscopic scales. This study addresses the gap by developing a multiscale framework that integrates atomistic, mesoscopic, and continuum perspectives to predict alloy tribocorrosion response under realistic service conditions. The aim is to establish a transferable predictive model linking microstructural features, environmental variables, and wear regimes to tribocorrosion life and damage evolution. Specific objectives are (i) to quantify the influence of alloy composition, grain size, and interfacial phase distributions on wear-corrosion synergism; (ii) to couple first-principles calculations, specifically density functional theory (DFT) and CALPHAD-based thermodynamic assessments, with mesoscopic dislocation-based models to predict defect generation and local corrosion tendencies; (iii) to develop a probabilistic, multiscale surrogate model that outputs critical wear-accelerated corrosion rates and critical wear depths for predefined service cycles; (iv) to validate the framework against experimental tribocorrosion data from aset of representative alloys (stainless steels, nickel-based superalloys, and aluminum alloys) under varied environments (neutral, acidic, and chloride-containing solutions) and sliding conditions (load, speed, and humidity); and (v) to perform sensitivity and uncertainty analyses to identify dominant parameters and propagate uncertainties through the framework. The methodology comprises a mixed-methods design anchored in mechanistic modeling and empirical validation. The population includes commercially relevant alloys (e.g., AISI 304 stainless steel, IN718 nickel-based alloy, and Al2024-T3), with sample sizes of n = 12 specimens per material per environment for statically controlled tribocorrosion tests and n = 6 for high-cycle fatigue-wear experiments. Data collection employs in-situ electrochemical impedance spectroscopy and potentiodynamic polarization to quantify corrosion kinetics, nanoindentation and electron backscatter diffraction (EBSD) for local mechanical properties and microstructure, and tribometer-based wear tests under controlled environmental chambers to elicit real-time wear rates and potential shifts. Analytical instruments also include X-ray photoelectron spectroscopy (XPS) and transmission electron microscopy (TEM) for interfacial chemistry and defect structure characterization. The analysis integrates first-principles calculations to obtain adsorption energies of corrosive species on specific alloy facets, a dislocation-based plasticity model to simulate wear-induced defect evolution, and a Monte Carlo-based surrogate model to capture stochastic tribocorrosion interactions. Regression analysis and multivariate ANOVA will assess statistically significant effects of composition, grain size, and environment on measured wear-corrosion coupling parameters, while structural equation modeling will test hypothesized causal pathways between microstructural features and tribocorrosion outcomes. The expected findings include quantifiable relationships between microstructural attributes and tribocorrosion resistance, validated multiscale predictions of wear-assisted corrosion rates, and uncertainty bounds for service-life estimates. The study seeks to demonstrate that a coherent multiscale framework can accurately predict tribocorrosion behavior across material classes and environments, enabling material selection and design optimization for enhanced durability. Contributions to knowledge encompass (i) a novel integrative multiscale framework bridging atomistic to continuum scales for tribocorrosion prediction, (ii) quantified links between microstructure and synergistic degradation under realistic wear conditions, and (iii) a validated predictive tool with practical implications for lifecycle assessment and maintenance planning. The main conclusion is that predictive accuracy improves markedly when interfacial chemistry, mechanical defect evolution, and environmental factors are coherently coupled within a probabilistic multiscale model. Recommendations include extending the framework to additional alloy systems, incorporating machine learning-based surrogate models for rapid scenario analysis, and developing standardized experimental protocols to harmonize tribocorrosion data for cross-study comparability.