cardiacFoam is an OpenFOAM toolbox for cardiac electrophysiology and electro-mechanics. It keeps the solids4foam runtime-selection style, and can run in two modes:

• Full mode: with a full solids4foam installation (electro + solid + FSI workflows).
• Electro-only mode: with the lightweight fallback modules/physicsModel shipped in this repository.

cardiacFoam/
├── applications/
│   ├── solvers/cardiacFoam/                 # Main executable
│   ├── utilities/                           # sweepCurrents, setFibreField, ...
│   └── scripts/
│       ├── driverFoam/openfoam_driver/      # Python tutorial automation engine
│       └── cellML2foam/                     # CellML -> ionic model generation pipeline
├── src/
│   ├── electroModels/                       # Runtime-selectable electro solvers
│   ├── ionicModels/                         # Runtime-selectable ionic ODE models
│   ├── genericWriter/                       # Shared I/O and stimulus parsing helpers
│   └── couplingModels/                      # Electro-mechanics coupling signal interfaces
├── modules/
│   └── physicsModel/                        # Lightweight fallback physicsModel
├── tutorials/                               # Reference cases and regression cases
└── etc/resolveSolids4Foam.sh                # Backend selection helper

At runtime, solver/model selection is fully dictionary-driven:

1. applications/solvers/cardiacFoam/cardiacFoam.C creates physicsModel::New(runTime).
2. physicsModel type is selected from constant/physicsProperties (type).
3. For electro runs, src/electroModels/electroModel::New(...) selects electroModel from constant/electroProperties (electroModel).
4. Electro models (MonoDomainSolver, SingleCellSolver, EikonalSolver) select ionic models through ionicModel::New(...) (ionicModel in electro coefficients).

This gives one stable executable (cardiacFoam) with pluggable electro and ionic sub-models.

• MonoDomainSolver: tissue PDE-ODE model, explicit/implicit stepping, activation-time tracking.
• SingleCellSolver: single integration-point ODE workflow (no spatial PDE solve).
• EikonalSolver: reduced-order activation-time model.

Compiled in libionicModels:

• AlievPanfilov
• BuenoOrovio
• Courtemanche
• Fabbri
• Gaur
• Grandi
• ORd
• Stewart
• TNNP
• ToRORd_dynCl
• Trovato
• monodomainFDAManufactured (manufactured monodomain verification model)
• bidomainFDAManufactured (manufactured bidomain verification model)

Manual tutorial entry points are in tutorials/*/Allrun. For parameter sweeps and post-processing automation, use:

• applications/scripts/driverFoam/openfoam_driver

Current tutorial specs in the Python driver:

• singleCell
• niederer2012
• manufacturedFDA
• manufacturedFDABidomain
• restitutionCurves

The driver writes run manifests and artifact manifests (run_manifest.json, plots.json) for reproducibility.

./Allwmake

Run examples:

cd tutorials/singleCell
./Allrun

cd ../NiedererEtAl2012
./Allrun parallel

Run automation from repository root:

foamctl all --tutorial singleCell
foamctl all --tutorial niederer2012

• Tutorial regression entrypoint: tutorials/Alltest-regression
• Driver architecture contract tests: applications/scripts/driverFoam/openfoam_driver/tests/test_tutorial_architecture_contract.py

This is active research software. APIs and dictionaries evolve as models and workflows are expanded.