Skip to content

CLAUDE.md

Latest commit

 

History

History
78 lines (58 loc) · 2.54 KB

CLAUDE.md

File metadata and controls

78 lines (58 loc) · 2.54 KB

CLAUDE.md

This file provides guidance for Claude Code when working with this repository.

Project Overview

FEM is a Python toolkit for simulating acoustic radiation force excitations and shear wave propagation in soft tissue using finite element methods. It integrates with LS-DYNA, a commercial finite element solver.

Build and Test Commands

# Install package in development mode
pip install -e .

# Run tests (uses pytest with parallel execution)
pytest -v tests/

# Build documentation
cd docs && make html

Project Structure

fem/
├── mesh/       # Mesh generation and boundary conditions
├── dyna/       # LS-DYNA mesh/material definitions (v9.0.0+ dataclass-based)
├── post/       # Post-processing LS-DYNA output to HDF5/MAT/VTK
├── field/      # Field II acoustic simulation integration
tests/          # Pytest tests with fixture data (nodes.dyn, elems.dyn, nodout)
examples/       # Complete workflow examples
docs/           # Sphinx RST documentation

Key Conventions

Coordinate System (LS-DYNA)

  • Axial: -z direction
  • Lateral: +y direction
  • Elevation: -x direction
  • Units: CGS (centimeters, grams, seconds)

Data Structures

  • Node arrays: [id, x, y, z, tc, rc] as numpy structured arrays
  • Element arrays: [id, pid, n1-n8] for 8-node solid elements
  • Common naming: nodeIDcoords (nic), sortedNodeIDcoords (snic)
  • Spatial sorting uses Fortran order ('F') for LS-DYNA compatibility

File Formats

  • .dyn - LS-DYNA keyword input files (comma-separated, * or $ prefix for comments)
  • nodout - LS-DYNA binary node output
  • Output formats: HDF5 (.h5), MAT (v5), VTR/PVD (Paraview)

Architecture Patterns (v9.0.0+)

  • Dataclass-based design with @dataclass decorator
  • Mixin pattern in fem/dyna/: _structure.py, _constraints.py, _loads.py, _writer.py
  • Material hierarchy with base Material class

Dependencies

Core: numpy, scipy, h5py, matplotlib, pyevtk Build: scikit-build, cmake (for C/SWIG acceleration in post-processing) Docs: sphinx>=7.2.0

Common Workflows

  1. Generate mesh: GenMesh.run(xyz_bounds, num_elements)
  2. Apply boundary conditions: bc.apply_face_bc_only(face_constraints)
  3. Define loads: TopLoad.generate_loads(loadtype='disp', ...)
  4. Run LS-DYNA (external)
  5. Extract results: create_disp_dat() (nodout → HDF5)
  6. Generate visualization: create_res_sim() (HDF5/MAT/PVD)

CI/CD

  • GitHub Actions workflows in .github/workflows/
  • python-package.yml - Runs pytest on Python 3.11
  • docs.yml - Builds and deploys Sphinx docs to GitHub Pages