[zer00318]

Description

Summary

Adds festim.StressDriftTerm, stress-assisted diffusion of an interstitial species along a hydrostatic stress gradient (Sofronis & McMeeking, 1989):

$$J = -D \nabla c + \frac{D c V_H}{R T} \nabla \sigma_h$$

with partial_molar_volume $V_H$ in m³/mol and hydrostatic_stress $\sigma_h$ in Pa (positive in tension). It is a DriftTermBase subclass built exactly like SoretTerm and ElectromigrationTerm, so it reuses the conservative drift_form, the 0D handling and the structural-zero warning. FESTIM does not solve for the stress: like the potential of ElectromigrationTerm, it is an input (float, callable of x/t/T, or a fenics object), typically the hydrostatic part of a separate dolfinx mechanics solve on the same mesh.

Related Issues

Closes #1261

Motivation and Context

Hydrogen accumulation in tensile regions matters for crack tips and for thermally mismatched joints such as W/CuCrZr plasma-facing components, where the cool-down after joining leaves large residual stresses at the interface.

Type of Change

• ✨ New feature (non-breaking change which adds functionality)
• 📝 Documentation update
• ✅ Test update (adding missing tests or correcting existing tests)

Testing

• All existing tests pass locally (pytest) — dolfinx 0.10.0, macOS arm64, conda-forge env: 1206 passed, 66 skipped; the one failure, test_heat_transfer_problem.py::test_sympify, is ModuleNotFoundError because sympy is not installed in my env
• I have added new tests that prove my fix is effective or that my feature works

New tests in test/test_drift.py: the term joins the existing parametrisations (0D warning, subdomain/species validation, shared base class); a uniform stress gives a structural-zero velocity and the "no effect" warning; the velocity equals D V_H / (R T) grad(sigma_h); the drift points towards tension for V_H > 0 and reverses with the sign of V_H.

New system test in test/system_tests/test_drift.py: a closed slab under a linear stress gradient relaxes to the closed form c ∝ exp(V_H σ_h / RT) (max relative error < 1e-5), with the stress span chosen so that the tensile end holds e times the unstressed end.

I also checked the fenics-object path end to end (not added to the suite, script below): plane-strain elasticity of a bent beam in dolfinx, σ_h = tr(σ)/3 interpolated into P1 and passed to StressDriftTerm on the same mesh. The closed transient run gives c_top / c_bottom = 1.0030 against the Boltzmann value 1.0030.

bent-beam check

# plane-strain elasticity on a 10 x 2 mm beam, end moments imposed as a linear axial displacement
uh = LinearProblem(a, L, bcs=bcs, petsc_options={"ksp_type": "preonly", "pc_type": "lu"},
                   petsc_options_prefix="elastic").solve()
s = sig(uh)
sigma_h = fem.Function(fem.functionspace(msh, ("Lagrange", 1)))
sigma_h.interpolate(fem.Expression((1 + nu) * (s[0, 0] + s[1, 1]) / 3,
                                   sigma_h.function_space.element.interpolation_points))
model = F.HydrogenTransportProblem(
    mesh=F.Mesh(msh), subdomains=[vol], species=[H], temperature=600.0,
    initial_conditions=[F.InitialConcentration(value=1.0, volume=vol, species=H)],
    drift_terms=[F.StressDriftTerm(species=H, partial_molar_volume=2e-6,
                                   hydrostatic_stress=sigma_h, subdomain=vol)],
    settings=F.Settings(atol=1e-12, rtol=1e-10, transient=True, final_time=200.0),
)

Code Quality Checklist

• My code follows the code style of this project (Ruff formatted: ruff format .)
• My code passes linting checks (ruff check .)
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas

Documentation

• I have updated the documentation accordingly (user guide section, theory equation eq_stress_drift, Sofronis1989 reference)
• I have added docstrings to new functions/classes following the project conventions

Additional Notes

• V_H = 2e-6 m³/mol in the examples is the classic value for H in iron used by Sofronis & McMeeking; the docstring says so rather than suggesting a tungsten value.
• Only the hydrostatic part of the stress enters, as in the standard model; trap-binding changes with plastic strain are out of scope.
• Written with AI assistance (Claude); I reviewed and ran everything above.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 96.15%. Comparing base (1bd7a80) to head (57a9d95).

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1288      +/-   ##
==========================================
+ Coverage   96.14%   96.15%   +0.01%     
==========================================
  Files          57       57              
  Lines        5165     5181      +16     
==========================================
+ Hits         4966     4982      +16     
  Misses        199      199              

Flag	Coverage Δ
dolfinx-v0.10.0	93.28% <100.00%> (+0.02%)	⬆️
dolfinx-v0.11.0	96.02% <100.00%> (+0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[RemDelaporteMathurin]

Thanks for this @zer00318 i’ll try to give it a look later today

[RemDelaporteMathurin]

Thanks @zer00318 for this contribution and congrats on your first PR! 🎉

@jhdark do you want to have a look at this or shall I merge?