A microstructure-sensitive electro-chemo-mechanical phase-field model of pitting and stress corrosion cracking

Read original: arXiv:2403.20301 - Published 4/1/2024 by M. Makuch, S. Kovacevic, M. R. Wenman, E. Mart'inez-Pa~neda

A microstructure-sensitive electro-chemo-mechanical phase-field model of pitting and stress corrosion cracking

Introduction

The paper presents an electro-chemo-mechanical phase-field model for assessing pitting and stress-assisted corrosion in polycrystalline materials. Corrosion is a common failure mode in engineering components, with pitting corrosion being particularly damaging and difficult to predict. The onset and propagation of pitting corrosion strongly depend on the crystallographic orientation of the material.

Existing numerical models for corrosion have limitations in coupling with other physico-mechanical phenomena, handling geometric interactions in arbitrary dimensions, and accounting for the formation of an electric double layer (EDL) at the metal-electrolyte interface. The phase-field method is proposed as a promising approach to overcome these limitations by implicitly tracking the metal-environment interface and capturing topological changes.

The developed model incorporates the influence of crystallographic orientation on mechanical properties and corrosion potential, as well as a general boundary condition for the solution potential that accounts for the formation and charging dynamics of the EDL. This boundary condition is derived from a resistor-capacitor equivalent circuit model.

The paper outlines the underlying electrochemistry and corrosion mechanisms, develops the phase-field model based on a generalized thermodynamic free energy functional, calibrates and validates the model against experimental measurements, and presents case studies demonstrating the model's ability to capture pitting and stress-assisted corrosion in polycrystalline materials.

Phase-field formulation

The provided text summarizes the underlying electrochemistry and governing equations for a phase-field model of pitting corrosion in polycrystalline materials. The key points are:

The corrosion process is described by metal dissolution, primary hydrolysis, and water dissociation reactions. The concentrations of ionic species are modeled variables.
The free energy functional includes chemical, interfacial, electric, and mechanical contributions. The chemical free energy depends on ionic concentrations.
Evolution equations are derived for the phase-field parameter tracking the metal-electrolyte interface, ionic concentrations, mechanical deformation, and solution potential.
The solution potential accounts for the electric double layer using a resistor-capacitor circuit model, capturing transient effects like surface polarization.
The corrosion kinetics follows a mechano-electro-chemical coupling based on Butler-Volmer and Gutman's theories. The rate depends on crystallographic orientation via the equilibrium corrosion potential.
Numerical implementation details like the finite element discretization, time-stepping, and monolithic solution strategy are provided.

The model enables predicting corrosion damage evolution in polycrystalline materials by coupling electrochemistry, mechanics, microstructure, and transient electrical effects.

Model calibration and validation

The experimental setup described involves a 304 stainless steel wire with a 50 μm diameter immersed in 1 M NaCl solution. The wire was coated in epoxy resin, exposing only the cross-section to the corrosive environment. A potentiostatic loading approach was used, where a current was supplied to maintain a 600 mV potential difference between the wire and a reference electrode.

Experimental measurements of pit depth and current density over time were used to calibrate model parameters, including the kinematic mobility coefficient controlling the interface motion, and equivalent circuit parameters describing the ratio between the electrical double layer resistance and bulk electrolyte resistance, and the capacitor charging time.

Numerical simulations considered an axisymmetric domain replicating the experimental setup. Appropriate boundary conditions were applied to model an unbounded domain, with no flux across exterior boundaries and a constant potential at the metal-electrolyte interface. Initial concentrations of ions were set based on the experimental conditions and material properties.

The simulation did not account for mechanical effects or variations in the corrosion potential.

Figure 4: Schematic disposition of the experimental setup used in Ref. ERNST2002a (left) and the corresponding computational domain (right) for the 304 stainless steel (SS) wire immersed in 1 M NaCl solution.

The paper discusses a phase-field model for simulating pitting corrosion in stainless steel. Material properties, diffusivities, and reaction constants are obtained from literature. Equilibrium conditions are enforced using large backward reaction constants. Equilibrium concentrations and potentials are determined based on solubility, molar fractions, and charge numbers. Phase-field parameters are linked to interfacial energy and thickness.

The model predictions for pit depth and current density agree well with experimental data. Initially, pit growth is activation-controlled, governed by interface reactions. As the electric double layer (EDL) charges up, there is a transition to diffusion-controlled growth, with long-range ion transport dictating the corrosion rate. The current density increases rapidly during EDL charging, then decreases and stabilizes.

The model is calibrated using interfacial mobility, capacitor charging half-time, and EDL resistance parameters. The values agree with experiments using the same equivalent circuit model for stainless steel in aqueous environments. Previous models validated only pit depth, while this framework captures both pit depth and current density measurements.

Figure 5: Comparison between experimental measurements ERNST2002a and phase-field predictions of (a) the evolution of the pit depth and (b) current density as a function of immersion time in 1 M NaCl solution.

The text discusses the distribution of ion concentrations and solution potential at the final computational time in a corrosion simulation. Key points:

The metal ion concentration gradually reaches equilibrium at the liquid-metal interface, indicating a diffusion-controlled process. This agrees with previous pit kinetics results.

Hydrolysis reactions reduce pH near the interface but not below 6. This decreases OH- concentration and increases M(OH)(z1-1)+ concentration to maintain water dissociation equilibrium.

The simulation properly satisfies equilibrium conditions like pH + pOH = 14 by choosing appropriate backward reaction rate constants.

Cl- ions move towards the interface due to electromigration from their negative charge, while Na+ ions exhibit opposite behavior from their positive charge.

The solution potential gradient is highest near the interface and diminishes in the bulk electrolyte.

Varying the applied potential captures the transition from activation- to diffusion-controlled corrosion regimes, with higher potentials leading to faster diffusion-controlled pit kinetics, consistent with literature.

$Figure 6: Distribution of (a) metal ions Mz1+superscriptMlimit-fromsubscript𝑧1\text{M}^{z_{1}+}M start_POSTSUPERSCRIPT italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + end_POSTSUPERSCRIPT, (b) pH, (c) chloride ions Cl−{}^{-}start_FLOATSUPERSCRIPT - end_FLOATSUPERSCRIPT, and (d) solution potential at the final computational time. The legend bars apply to the whole computational domain.$

Figure 6: Distribution of (a) metal ions Mz1+superscriptMlimit-fromsubscript𝑧1\text{M}^{z_{1}+}M start_POSTSUPERSCRIPT italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + end_POSTSUPERSCRIPT, (b) pH, (c) chloride ions Cl−{}^{-}start_FLOATSUPERSCRIPT - end_FLOATSUPERSCRIPT, and (d) solution potential at the final computational time. The legend bars apply to the whole computational domain.

$Figure 7: Electrochemical predictions. (a) Concentration profile of M(OH)(z1−1)+limit-fromsubscript𝑧11{}^{(z_{1}-1)+}start_FLOATSUPERSCRIPT ( italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT - 1 ) + end_FLOATSUPERSCRIPT, H+{}^{+}start_FLOATSUPERSCRIPT + end_FLOATSUPERSCRIPT, and OH−{}^{-}start_FLOATSUPERSCRIPT - end_FLOATSUPERSCRIPT as a function of distance from the interface at the final computational time for 600 mV (vs. SCE). Concentration values c¯isubscript¯𝑐𝑖\bar{c}_{i}over¯ start_ARG italic_c end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT are normalized with 1 mol/L. (b) Pit depth as a function of immersion time for different applied potentials. The potentials are expressed vs saturated calomel electrode.$

Figure 7: Electrochemical predictions. (a) Concentration profile of M(OH)(z1−1)+limit-fromsubscript𝑧11{}^{(z_{1}-1)+}start_FLOATSUPERSCRIPT ( italic_z start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT - 1 ) + end_FLOATSUPERSCRIPT, H+{}^{+}start_FLOATSUPERSCRIPT + end_FLOATSUPERSCRIPT, and OH−{}^{-}start_FLOATSUPERSCRIPT - end_FLOATSUPERSCRIPT as a function of distance from the interface at the final computational time for 600 mV (vs. SCE). Concentration values c¯isubscript¯𝑐𝑖\bar{c}_{i}over¯ start_ARG italic_c end_ARG start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT are normalized with 1 mol/L. (b) Pit depth as a function of immersion time for different applied potentials. The potentials are expressed vs saturated calomel electrode.

The experimental measurements and phase-field predictions align for pit depth and current density, demonstrating the model's accuracy in simulating uniform corrosion. The model effectively predicts pit kinetics, electrochemical measurements, and exhibits sensitivity to externally applied potentials. The subsequent section utilizes the model to investigate the influence of microstructure on pitting and stress corrosion cracking in polycrystalline materials.

Microstructure-sensitive phase-field

simulations and discussion

The provided text discusses two case studies assessing the evolution of corrosion in polycrystalline materials using phase-field simulations.

The first case study examines pitting corrosion associated with the local rupture of a protective layer. Simulations are performed for synthetic microstructures with varying grain sizes and degrees of variation in equilibrium corrosion potentials. The results show that nonuniform corrosion potentials promote nonuniform pit growth and faceted pit shapes, with larger grain sizes leading to more severe damage compared to finer microstructures. Diffusion-controlled corrosion remains unaffected by variations in corrosion potential and grain size.

The second case study involves stress-assisted corrosion, where the polycrystalline material is subjected to tensile strain while immersed in a corrosive environment. The results demonstrate that the underlying microstructure with anisotropic corrosion potentials and mechanical properties plays a significant role in determining defect growth, morphology, and distributions of hydrostatic stress and plastic strain. Increasing variations in corrosion potential and applied tensile strain enhance defect kinetics, leading to more extensive and irregular crack shapes.

A statistical analysis is performed to quantitatively compare the results for polycrystalline materials and homogeneous materials. The analysis indicates that excluding microstructural effects would underestimate localized corrosion damage, particularly for high variations in corrosion potential and coarse microstructures. The model can predict the long-term durability of materials under various corrosive environments and aid in developing more corrosion-resistant microstructures.

Conclusions and future work

The developed computational framework integrates electrochemical, mechanical, and microstructural aspects to predict pitting and stress corrosion cracking in polycrystalline materials. The model captures the role of microstructure by including dependencies of corrosion potential and mechanical properties on crystallographic orientation. It also considers the formation and charging dynamics of an electric double layer.

The results demonstrate that microstructural features significantly influence the growth, morphology, stress distribution, and plastic strain distribution of defects. The inclusion of microstructural effects leads to more extensive defects, faster defect kinetics, irregular pit and crack shapes, and highly non-uniform stress and plastic strain distributions compared to homogeneous materials without microstructural effects.

Localized corrosion damage is quantified using two metrics: maximum defect depth and dissolved metal area. The presence of microstructural features results in more pronounced pitting and stress corrosion cracking, leading to longer pits, cracks, and larger dissolved metal areas. The crystallographic dependence and variations in corrosion potential have a more dominant impact on maximum pit and crack depth than grain size and mismatch in mechanical properties.

The study highlights the importance of considering underlying microstructure in simulating pitting corrosion and stress corrosion cracking, as excluding it would underestimate defect kinetics and corrosion damage. Future work should investigate the role of grain boundaries, surface roughness, grain orientation dependence of interfacial energy, and incorporation of crystal plasticity into the framework.

Acknowledgments

The authors M.M., M.R.W., and E.M.-P. received funding from the EPSRC Centre for Doctoral Training in Nuclear Energy Futures through Grant EP/5023844/1. Additionally, S.K. and E.M.-P. were financially supported by UKRI's Future Leaders Fellowship program under Grant MR/V024124/1.

Data Availability

There is no text to summarize in this section. It simply provides a link to access the code developed for the research work after the article is accepted for publication.

Appendix A. Mesh sensitivity analysis and grain morphology distributions

The text describes a mesh sensitivity analysis for a pencil electrode simulation. Three maximum element sizes (ℓ/2, ℓ/4, and ℓ/6) relative to the interface thickness ℓ were considered. Second-order Lagrangian triangular finite elements were used with a maximum time step of 0.2 s and a relative solver tolerance of 10^-4. The results showed that all three cases gave the same pitting kinetics, and the ℓ/4 and ℓ/6 cases gave nearly identical interface profiles at the final time, while ℓ/2 showed a slight discrepancy. However, using ℓ/4 significantly reduced computational time compared to ℓ/6. Therefore, ℓ/4 was chosen as the maximum element size across the interface thickness for all simulations.

The text also provides grain morphology distributions for synthetic microstructures with average grain sizes of 20 μm, 40 μm, and 60 μm. Ten different distributions are shown for each grain size. The color represents variation in equilibrium corrosion potential based on crystallographic orientation, with a maximum variation of 10% from the macroscopic equilibrium corrosion potential of -0.729 V vs. SCE. The microstructure dimensions are 300 μm width and 150 μm height.

This summary was produced with help from an AI and may contain inaccuracies - check out the links to read the original source documents!

Follow @aimodelsfyi on 𝕏 →

Related Papers

A microstructure-sensitive electro-chemo-mechanical phase-field model of pitting and stress corrosion cracking

M. Makuch, S. Kovacevic, M. R. Wenman, E. Mart'inez-Pa~neda

An electro-chemo-mechanical phase-field formulation is developed to simulate pitting and stress corrosion in polycrystalline materials. The formulation incorporates dependencies of mechanical properties and corrosion potential on crystallographic orientation. The model considers the formation and charging dynamics of an electric double layer through a new general boundary condition for the solution potential. The potential of the model is demonstrated by simulating corrosion in polycrystalline materials with various grain morphology distributions. The results show that incorporating the underlying microstructure yields more extensive defects, faster defect kinetics, and irregular pit and crack shapes relative to a microstructurally-insensitive homogeneous material scenario.

4/1/2024

$A spatiotemporal deep learning framework for prediction of crack dynamics in heterogeneous solids: efficient mapping of concrete microstructures to its fracture properties$

A spatiotemporal deep learning framework for prediction of crack dynamics in heterogeneous solids: efficient mapping of concrete microstructures to its fracture properties

Rasoul Najafi Koopas, Shahed Rezaei, Natalie Rauter, Richard Ostwald, Rolf Lammering

A spatiotemporal deep learning framework is proposed that is capable of 2D full-field prediction of fracture in concrete mesostructures. This framework not only predicts fractures but also captures the entire history of the fracture process, from the crack initiation in the interfacial transition zone to the subsequent propagation of the cracks in the mortar matrix. In addition, a convolutional neural network is developed which can predict the averaged stress-strain curve of the mesostructures. The UNet modeling framework, which comprises an encoder-decoder section with skip connections, is used as the deep learning surrogate model. Training and test data are generated from high-fidelity fracture simulations of randomly generated concrete mesostructures. These mesostructures include geometric variabilities such as different aggregate particle geometrical features, spatial distribution, and the total volume fraction of aggregates. The fracture simulations are carried out in Abaqus, utilizing the cohesive phase-field fracture modeling technique as the fracture modeling approach. In this work, to reduce the number of training datasets, the spatial distribution of three sets of material properties for three-phase concrete mesostructures, along with the spatial phase-field damage index, are fed to the UNet to predict the corresponding stress and spatial damage index at the subsequent step. It is shown that after the training process using this methodology, the UNet model is capable of accurately predicting damage on the unseen test dataset by using 470 datasets. Moreover, another novel aspect of this work is the conversion of irregular finite element data into regular grids using a developed pipeline. This approach allows for the implementation of less complex UNet architecture and facilitates the integration of phase-field fracture equations into surrogate models for future developments.

7/25/2024

🗣️

Predicting Peak Stresses In Microstructured Materials Using Convolutional Encoder-Decoder Learning

Ankit Shrivastava, Jingxiao Liu, Kaushik Dayal, Hae Young Noh

This work presents a machine learning approach to predict peak-stress clusters in heterogeneous polycrystalline materials. Prior work on using machine learning in the context of mechanics has largely focused on predicting the effective response and overall structure of stress fields. However, their ability to predict peak stresses -- which are of critical importance to failure -- is unexplored, because the peak-stress clusters occupy a small spatial volume relative to the entire domain, and hence requires computationally expensive training. This work develops a deep-learning based Convolutional Encoder-Decoder method that focuses on predicting peak-stress clusters, specifically on the size and other characteristics of the clusters in the framework of heterogeneous linear elasticity. This method is based on convolutional filters that model local spatial relations between microstructures and stress fields using spatially weighted averaging operations. The model is first trained against linear elastic calculations of stress under applied macroscopic strain in synthetically-generated microstructures, which serves as the ground truth. The trained model is then applied to predict the stress field given a (synthetically-generated) microstructure and then to detect peak-stress clusters within the predicted stress field. The accuracy of the peak-stress predictions is analyzed using the cosine similarity metric and by comparing the geometric characteristics of the peak-stress clusters against the ground-truth calculations. It is observed that the model is able to learn and predict the geometric details of the peak-stress clusters and, in particular, performed better for higher (normalized) values of the peak stress as compared to lower values of the peak stress. These comparisons showed that the proposed method is well-suited to predict the characteristics of peak-stress clusters.

5/10/2024

Comparison of Different Elastic Strain Definitions for Largely Deformed SEI of Chemo-Mechanically Coupled Silicon Battery Particles

Raphael Schoof, Giuseppe Fabian Castelli, Willy Dorfler

Amorphous silicon is a highly promising anode material for next-generation lithium-ion batteries. Large volume changes of the silicon particle have a critical effect on the surrounding solid-electrolyte interphase (SEI) due to repeated fracture and healing during cycling. Based on a thermodynamically consistent chemo-elasto-plastic continuum model we investigate the stress development inside the particle and the SEI. Using the example of a particle with SEI, we apply a higher order finite element method together with a variable-step, variable-order time integration scheme on a nonlinear system of partial differential equations. Starting from a single silicon particle setting, the surrounding SEI is added in a first step with the typically used elastic Green--St-Venant (GSV) strain definition for a purely elastic deformation. For this type of deformation, the definition of the elastic strain is crucial to get reasonable simulation results. In case of the elastic GSV strain, the simulation aborts. We overcome the simulation failure by using the definition of the logarithmic Hencky strain. However, the particle remains unaffected by the elastic strain definitions in the particle domain. Compared to GSV, plastic deformation with the Hencky strain is straightforward to take into account. For the plastic SEI deformation, a rate-independent and a rate-dependent plastic deformation are newly introduced and numerically compared for three half cycles for the example of a radial symmetric particle.

4/3/2024