Fix BPX 1.1 hysteresis parameter mapping

[rtimms]

Summary

Follow-up to #5469 (which added BPX 1.x support). The newly-introduced hysteresis fields on BPX 1.1's Particle block weren't being translated by parameters/bpx.py::_get_pybamm_name, so:

• OCP (lithiation) [V] ended up as Negative electrode OCP (lithiation) [V] (PyBaMM expects Negative electrode lithiation OCP [V])
• OCP (delithiation) [V] similarly
• OCP hysteresis decay constant ended up under its raw BPX alias on the electrode pre-name; PyBaMM's Axen one-state hysteresis OCP submodel (OneStateHysteresisOpenCircuitPotential) evaluates two FunctionParameters — <Domain> particle lithiation hysteresis decay rate and <Domain> particle delithiation hysteresis decay rate — both on the particle pre-name. The BPX scalar now fans out to both with the same value.

While auditing the State section I found three adjacent gaps and fixed them in the same PR:

• Initial hysteresis state: Positive/Negative electrode were extracted into bpx.state but never copied to the PyBaMM dict, so the FunctionParameter was always falling back to 0. Now mapped to [<Phase>: ]Initial hysteresis state in <domain> electrode. Handles both the scalar (single-phase) and the per-particle-phase dict (blended) shapes by routing the blended-ness decision through the existing _get_phase_names helper and matching BPX phase keys to PyBaMM Primary: / Secondary: by particle declaration order.
• Total heat transfer coefficient [W.m-2.K-1] from bpx.state.thermal_environment was being silently clobbered to 0 a few lines later by an unconditional pybamm_dict.update(...). Replaced with setdefault.
• target_soc defaulted to 1.0 regardless of what the BPX file declared in State.initial_conditions.initial_soc. The default is now None, resolved to the BPX value when a State section is present and 1.0 otherwise.

Implementation note: _get_pybamm_name now returns list[str] instead of str to support the 1→2 fan-out for the decay constant. The two call sites in _bpx_to_domain_param_dict iterate the returned list. Missing/None-valued optional hysteresis fields are skipped so they don't pollute the parameter dict.

Refs #5463.

Test plan

• All existing BPX tests in tests/unit/test_parameters/test_bpx.py still pass (13/13)
• New unit tests (8) added to the same file:
  • Single-phase BPX hysteresis fields → expected PyBaMM names; legacy names not present
  • Missing hysteresis fields are skipped, not added as None
  • Blended-electrode hysteresis fields → phase-prefixed PyBaMM names
  • Initial hysteresis state (scalar form)
  • Initial hysteresis state (per-particle-phase dict form, blended)
  • Heat transfer coefficient regression test (BPX-supplied value preserved)
  • target_soc=None falls back to BPX Initial state-of-charge
  • End-to-end: BPX-derived parameter set runs the Axen one-state hysteresis SPM through a Discharge experiment

Notes for reviewer

• BPX models only the Axen one-state hysteresis OCP, so the decay constant intentionally does NOT fan out to the Wycisk differential-capacity variant's <Domain> particle hysteresis decay rate (without lithiation/delithiation prefix). Users on that submodel will still need to add it manually.
• BPX State.degradation (LLI, LAM: Positive electrode, LAM: Negative electrode) is still unhandled — it has no direct PyBaMM equivalent and is out of scope here. Worth a separate issue if/when there's demand.

🤖 Generated with Claude Code

[rtimms]

@ejfdickinson could you please take a look and see if any issues remain with any BPX files you have?

[codecov]

Codecov Report

❌ Patch coverage is 97.72727% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 98.09%. Comparing base (ce1a240) to head (a7ab0d0).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
src/pybamm/parameters/bpx.py	97.56%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #5507      +/-   ##
==========================================
- Coverage   98.09%   98.09%   -0.01%     
==========================================
  Files         338      338              
  Lines       31246    31272      +26     
==========================================
+ Hits        30650    30675      +25     
- Misses        596      597       +1     

☔ View full report in Codecov by Sentry.
📢 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.

[ejfdickinson]

@rtimms Thanks, everything BPX-related is working for me for the test cases we have.

Notes:

1. There is an oddity around initial hysteresis state parameters being created for materials that don't have an attached hysteresis model, but this is inherited from the BPX limitations, right? I assume these parameters don't do anything, and you don't know what model is targeted.

2. I did one test case (full DFN) with a single-material (blended description) graphite-SiOx electrode but setting
   options={"open-circuit potential": ("one-state hysteresis", "single")}
   This fails at discretisation stage with ModelError: model is overdetermined (extra algebraic keys) from the well-posedness checks. Is this something you're familiar with? It seems like it should be a separate issue but let me know if you'd like a repeat case, if you don't have any similar parameter sets.

[rtimms]

@ejfdickinson thanks for testing!

1. Yes, exactly. These will just be ignored if they don't show up in the model.
2. I can't reproduce this. Could you share an example?

[ejfdickinson]

@rtimms I diagnosed the problem - it was due to some placeholder zeroes for electrolyte properties, so no issues here. Just the runtime error path had changed compared to expectations from previous versions.

Good to go for this PR as far as I can see! Thank you!

[Daniel-Nicolae23]

Hi @rtimms, can we please have the initial concentrations calculated via ESOH and added in for blended electrodes too?

[ejfdickinson]

This would be great, but it's a known limitation also for beginning-of-life preceding the degradation content in BPX, so I'd ask this not be at the expense of delaying a PyBaMM release covering the schema support!

@Daniel-Nicolae23 @rtimms You are welcome to inspect some of the tools we use to implement this after BPX import in our AEPyBaMM toolkit. Currently the code supports this for negative electrode blends only, but we are also working on some general-purpose utilities to be released soon.

https://github.com/About-Energy-OpenSource/AEPyBaMM/blob/main/src/aepybamm/sci_tools.py

see compute_lithiation_bounds and add_initial_concentrations