A Framework for Multiscale Modeling of Metal-Polymer Interfaces in Composites

Thesis Abstract

Multiscale modeling of metal–polymer interfaces in composites presents a critical pathway to predict interfacial strength, durability, and failure mechanisms across length scales, addressing the gap between atomistic processes and macroscopic mechanical performance that conventional single-scale models overlook. The study aims to develop a comprehensive multiscale framework that integrates quantum-mechanical insights, molecular dynamics, and continuum finite element analyses to predict interfacial behavior under thermo-mechanical loading. Specific objectives include (1) delineating the synergistic coupling across scales via a hierarchical upscaling scheme, (2) calibrating interfacial potential functions and constitutive relationships against experimental and ab initio data, (3) implementing a computational workflow that links density functional theory (DFT) calculations with reactive force field (ReaxFF) molecular dynamics and cohesive zone models in FE simulations, (4) validating the framework against a curated dataset of metal–polymer laminates with varied metal (aluminum, steel) and polymer (epoxy, polyimide) systems, and (5) performing sensitivity and uncertainty analyses to quantify the influence of interface chemistry, roughness, and processing-induced residual stresses on debonding and fracture. The methodology adopts a mixed-methods design that combines computational modeling with experimental validation. The population comprises representative metal–polymer interface systems and their processing-induced states, drawing from published datasets and newly generated measurements. A sample of 24 material configurations—covering four metal–polymer pairs across three surface treatments and two curing conditions—will be analyzed. Data collection instruments include high-resolution transmission electron microscopy (HRTEM), X-ray photoelectron spectroscopy (XPS) for interfacial chemistry, nanoindentation and microcantilever testing for interfacial toughness, and pull-off shear tests for debonding strength. Computational data will be generated via density functional theory to characterize bond formation energies at the interface, followed by parameterization of ReaxFF force fields for interfacial reactions, and atomistic simulations to inform cohesive zone model parameters. Finite element models will integrate the multiscale outputs through a hierarchical coupling approach, enabling simulation of debond growth under thermo-mechanical loading, including temperature ramps from ambient to service temperatures and applied strain rates representative of service conditions. Analyses will employ a suite of statistical and computational techniques. Regression analysis and Bayesian updating will calibrate interfacial properties against experimental observations, while analysis of variance (ANOVA) will assess the effects of surface treatments and material pairings on measured toughness. Sensitivity analyses will be conducted using Sobol indices to quantify the contribution of interfacial chemistry, roughness, and residual stress to performance variability. Theoretical framing will draw on interfacial fracture mechanics and the Theory of Multiscale Coupling, with explicit reference to cohesive zone modeling and thermodynamically consistent coupling between atomistic and continuum scales. The expected results include a validated multiscale protocol capable of predicting debond initiation and growth as a function of temperature, loading rate, and interfacial chemistry, along with quantified uncertainty bounds for model predictions. The study is anticipated to contribute to knowledge by delivering a rigorously validated, transferable multiscale framework for metal–polymer interfaces in composites, complete with a parameterization repository for common metal–polymer systems and a configurable workflow that can be extended to other interfaces. It will illuminate the relative importance of chemical bonding versus mechanical interlocks and surface roughness in governing interfacial strength, and provide actionable guidance for material and processing choices to optimize durability. The main conclusion is that integrated multiscale modeling, underpinned by experimental validation, can accurately predict failure modes and guide design of robust metal–polymer composite architectures; recommendations include standardizing interfacial characterization protocols, prioritizing surface treatments that promote favorable chemical bonding at the interface, and adopting the proposed multiscale workflow as a design tool in composite manufacturing.