Thermodynamically consistent phase field model for hydrogen-assisted cracking

This paper presents a thermodynamically consistent phase field model that simulates hydrogen-assisted cracking in polycrystalline materials by coupling crack propagation with hydrogen segregation and interfacial energy reduction, successfully capturing the transition from transgranular to intergranular failure under hydrogen-enhanced decohesion mechanisms.

The Gist

Imagine you have a very strong metal structure, like the frame of a car or a bridge. You'd expect it to hold up under pressure, but sometimes, invisible hydrogen atoms sneak inside the metal and cause it to shatter unexpectedly. This phenomenon is called hydrogen embrittlement. It's like the metal is secretly being "poisoned" from the inside, making it brittle and prone to snapping.

Scientists have been trying to build computer models to predict exactly how and where this metal will break. However, previous models had a major flaw: they treated the hydrogen's behavior like a simple, uniform rule that applies everywhere, even though the metal's internal structure is actually a complex, patchwork quilt of different grains and boundaries.

The New "Smart" Model
The authors of this paper have created a new, more sophisticated computer simulation (called a "phase field model") that acts like a high-definition, thermodynamically consistent map. Here is how it works, using some everyday analogies:

• The Metal as a Crowd: Imagine the metal is a crowded room filled with people (the metal atoms). The "grain boundaries" are the invisible lines separating different groups of people. The "crack" is a growing gap in the crowd.
• The Hydrogen as a Sticky Guest: Hydrogen atoms are like sticky guests who love to hide in the empty spaces between the people. They have a special preference: they love to stick to the edges of the crack and the lines between the groups (grain boundaries) even more than they like the middle of the crowd.
• The "Glue" Problem: In a healthy metal, the "glue" holding the crack edges together is strong. But when these sticky hydrogen guests gather at the crack edges, they act like a slippery oil, weakening the glue. This makes the crack much easier to open.
• The Old vs. New Approach:
  • Old Models: Used a generic rulebook (the Langmuir-McLean isotherm) that assumed the hydrogen was evenly distributed and in perfect balance everywhere. This is like assuming everyone in the crowded room is standing still and evenly spaced, which isn't true when a crack is forming.
  • New Model: Uses a flexible, "variational" framework (based on the Kim-Kim-Suzuki formalism). Instead of forcing a rigid rule, it lets the hydrogen naturally "migrate" to where it wants to go (the crack edges and grain boundaries) based on the local conditions. It calculates exactly how much the "glue" weakens in real-time as the hydrogen gathers.

What They Discovered
The team tested their new model with two main scenarios:

1. The Single Crack Test: They simulated a crack in a single piece of metal. Without hydrogen, the crack grew exactly as physics predicts (following the "Griffith criterion"). When they added hydrogen, the model showed the crack grew much more easily because the hydrogen had weakened the surface energy. The results matched theoretical predictions perfectly, proving the model works.

2. The Polycrystalline Test (The Big Discovery): They simulated a metal made of many tiny crystals (grains) with boundaries between them.

   • Without Hydrogen: The crack preferred to smash straight through the grains (transgranular cracking). It was like a wrecking ball smashing through the walls of a house because the walls were weaker than the mortar between them.
   • With Hydrogen: The hydrogen gathered heavily at the boundaries between the grains, weakening the "mortar" significantly more than the "walls." Suddenly, the crack changed its path. Instead of smashing through the grains, it started snaking along the boundaries (intergranular cracking). It was as if the hydrogen turned the mortar into wet sand, causing the house to fall apart along the seams rather than through the bricks.

Why This Matters
This new model is a significant upgrade because it doesn't just guess where the hydrogen goes; it calculates it based on the actual thermodynamics of the system. It successfully captures the transition from one type of cracking to another, which is crucial for understanding why materials fail in the presence of hydrogen.

The authors note that while this model is a major step forward, it currently focuses on one specific mechanism (hydrogen weakening the glue). Future work will need to add other complex factors, like how the metal bends and twists (plasticity) and how other types of defects interact with the hydrogen. But for now, this model provides a clear, consistent, and accurate way to see how hydrogen turns a strong metal into a fragile one.

Technical Summary

Problem Statement
Hydrogen embrittlement (HE) represents a critical challenge for hydrogen technologies, causing catastrophic material degradation and unexpected fracture in decarbonized energy systems. While mechanisms such as hydrogen-enhanced decohesion (HEDE), hydrogen-enhanced localized plasticity (HELP), and hydrogen-enhanced strain-induced vacancies (HESIV) have been proposed, predicting HE remains difficult due to its multiphysics nature spanning atomic to macroscopic scales. Existing phase field (PF) models for hydrogen-assisted cracking typically couple brittle fracture mechanics with the HEDE mechanism via hydrogen diffusion. However, these models suffer from two primary limitations: they do not explicitly reproduce hydrogen segregation (the phenomenon responsible for reduced surface energy), and they rely on the Langmuir-McLean isotherm to calculate hydrogen coverage. The latter is a macroscopic equilibrium relationship between homogeneous bulk and surface phases, making it physically unjustified as a local law for modeling heterogeneous microstructures involving elastic interactions.

Methodology
The authors propose a thermodynamically consistent phase field model based on the Kim-Kim-Suzuki (KKS) formalism to simulate hydrogen-assisted cracking in polycrystalline materials. The model operates within a variational framework where the total free energy includes contributions from crack surface energy, elastic energy, grain boundary (GB) coupling, and chemical free energy.

Key methodological features include:

• Field Variables: The system is described by a damage field ($\eta$), hydrogen occupation probability ($c$), and a set of order parameters ($\phi_i$) representing crystallographic orientations.
• KKS Formulation: Instead of imposing a local Langmuir-McLean law, the model defines distinct bulk ($c_b$) and crack surface ($c_{ck}$) phases. The chemical free energy density is constructed as a weighted sum of bulk and crack contributions, constrained by the equality of chemical potentials ($\mu_b = \mu_{ck}$). This allows interfacial properties to be decoupled from concentration fields.
• Thermodynamic Consistency: The model assumes ideal solid solutions for chemical energy and derives the governing equilibrium equations (mechanical equilibrium and chemical potential balance) from the total free energy functional.
• Numerical Implementation: The equilibrium equations are solved using Fourier-spectral methods and a discrete FFT solver for mechanical equilibrium.

Key Contributions

1. Explicit Segregation Modeling: The model naturally accounts for hydrogen segregation on crack surfaces and grain boundaries without assuming a macroscopic isotherm as a local law.
2. Variational Derivation of Surface Energy Degradation: The effective degradation of crack surface energy induced by hydrogen segregation is analytically computed within the thermodynamic framework.
3. Transition Capture: The model demonstrates the ability to capture the transition from transgranular to hydrogen-assisted intergranular cracking via the HEDE mechanism.

Results

• Validation in Single Crystals: Numerical simulations of a pre-existing crack in a single crystal under mode I loading showed that the critical stress for crack growth ($\sigma_c$) follows the Griffith criterion. When the HEDE mechanism was activated, $\sigma_c$ decreased due to a reduction in effective surface energy ($\Delta \gamma_{ch} \approx -4.29$ J/m$^2$), consistent with analytical predictions derived from the model's equations.
• Polycrystalline Application: Simulations in a nickel-based polycrystalline material revealed a distinct transition in crack propagation paths. In a hydrogen-free environment, the crack propagated transgranularly due to high GB toughness. Upon introducing hydrogen, the model predicted a shift to intergranular cracking. This was attributed to a drastic reduction in the effective GB-to-bulk fracture energy ratio (from 0.5 to 0.07), caused by higher hydrogen segregation energies at GB-type surfaces compared to bulk-type surfaces. The hydrogen concentration map confirmed significant segregation at GBs and broken GB surfaces.

Significance
The paper claims that this work provides a rigorous and generic method for modeling the HEDE mechanism at the microstructural scale. By utilizing the KKS formalism, the model reproduces hydrogen segregation and the resulting surface energy reduction without relying on the thermodynamically limited Langmuir-McLean isotherm for heterogeneous situations. The study validates that the isotherm emerges naturally only in specific homogeneous, static limits, thereby justifying the new approach for complex microstructures. The model successfully captures the experimentally observed transition from transgranular to intergranular cracking using realistic input data for nickel, offering a robust tool for understanding hydrogen-assisted failure in polycrystalline metals. The authors note that future work will address surface energy anisotropy and the incorporation of other HE mechanisms, such as HELP, through dislocation modeling.