Solver Benchmark

When to Use This Skill

Use solver benchmarking when:

• Validating solver setups - Confirm all solvers produce consistent results
• Cross-validating results - Compare different solver outputs for the same hull
• Identifying solver quirks - Find solver-specific behaviors or numerical differences
• Quality assurance - Verify analysis methodology across tools
• Reporting - Generate comparison reports for documentation

Quick Start

# Run 3-way benchmark on Unit Box (validation case)
uv run python scripts/benchmark/run_3way_benchmark.py \
    specs/modules/benchmark/unit_box_spec.yml

# Dry run (generate files, skip solver execution)
uv run python scripts/benchmark/run_3way_benchmark.py \
    specs/modules/benchmark/unit_box_spec.yml --dry-run

# Run specific solvers only
uv run python scripts/benchmark/run_3way_benchmark.py \
    spec.yml --solvers orcawave,aqwa

# Custom output directory
uv run python scripts/benchmark/run_3way_benchmark.py \
    spec.yml -o results/my_benchmark

Related Skills

• mesh-utilities - Mesh inspection and conversion (/mesh)
• hydrodynamic-analysis - BEM theory and RAO analysis
• orcaflex/specialist - OrcaFlex integration

Sub-Skills

• Troubleshooting

Sub-Skills

• Solver Paths
• Benchmark Workflow
• Output Structure
• Consensus Levels
• Creating Benchmark Specs
• Available Hull Geometries
• Programmatic Usage
• MultiSolverComparator (+1)
• OrcaWave (via OrcFxAPI) (+3)
• Pre-Flight Validation Checklist
• Correlation Thresholds (+1)
• Mandatory Checks Before Committing Reports (+2)
• Related Work Items