GMSH Meshing Skill

Specialized expertise for finite element mesh generation and manipulation using GMSH with quality assessment and engineering tool integration.

Version Metadata

version: 1.0.0
python_min_version: '3.10'
compatibility:
  tested_python:
  - '3.10'
  - '3.11'
  - '3.12'
  - '3.13'
  os:
  - Windows
  - Linux
  - macOS

Changelog

[1.0.0] – 2026-01-07

Added:

• Initial version metadata and dependency management
• Semantic versioning support
• Compatibility information for Python 3.10-3.13

Changed:

• Enhanced skill documentation structure

When to Use

• 1D/2D/3D mesh generation for analysis
• Geometry import and healing (STEP, IGES, STL)
• Mesh quality assessment and optimization
• Panel mesh generation for hydrodynamics
• Structural mesh for FEM analysis
• Mooring line discretization
• Batch mesh processing
• Integration with OrcaFlex, ANSYS, OpenFOAM

Agent Capabilities

This skill integrates agent capabilities from /agents/gmsh/:

Mesh Generation

Geometry Formats

Quality Metrics

Optimization Methods

• Laplacian smoothing
• Adaptive refinement
• Remeshing
• Parallel optimization (up to 100 iterations)

Prerequisites

• Python 3.10+
• GMSH Python bindings (pip install gmsh)
• NumPy, SciPy, PyVista

Configuration

Default Settings

{
  "configuration": {
    "default_algorithm": "frontal-delaunay",
    "default_element_size": 1.0,
    "quality_thresholds": {
      "min_jacobian": 0.3,
      "max_aspect_ratio": 5.0,
      "max_skewness": 0.7
    },
    "performance": {
      "parallel_workers": 4,
      "chunk_size": 1000,
      "memory_limit_gb": 4
    }
  }
}

Python API

Basic Mesh Generation

from digitalmodel.agents.gmsh import GMSHAgent

# Initialize agent
agent = GMSHAgent(config="agent_config.json")

# Generate mesh from geometry
mesh = agent.generate_mesh(
    geometry="platform.step",
    config={
        "algorithm": "frontal-delaunay",
        "element_size": 0.5,
        "element_type": "tetrahedron"
    }
)

# Save mesh
agent.save_mesh(mesh, "platform.msh")

Quality Assessment

# Assess mesh quality
quality = agent.assess_quality(mesh)

print(f"Elements: {quality['element_count']}")
print(f"Min Jacobian: {quality['min_jacobian']:.3f}")
print(f"Avg Aspect Ratio: {quality['avg_aspect_ratio']:.2f}")
print(f"Max Skewness: {quality['max_skewness']:.3f}")

# Check if optimization needed
if quality['min_jacobian'] < 0.3:
    print("Mesh requires optimization")

Mesh Optimization

# Optimize poor quality mesh
optimized = agent.optimize_mesh(
    mesh,
    method="laplacian_smoothing",
    iterations=10,
    quality_targets={
        "min_jacobian": 0.4,
        "max_aspect_ratio": 3.0
    }
)

# Verify improvement
quality_after = agent.assess_quality(optimized)
print(f"Jacobian improved: {quality['min_jacobian']:.3f} -> {quality_after['min_jacobian']:.3f}")

Panel Mesh for Hydrodynamics

# Generate panel mesh for OrcaFlex/OrcaWave
panel_mesh = agent.generate_panel_mesh(
    geometry="hull.step",
    panel_size=2.0,
    refinement_zones=["waterline", "appendages"]
)

# Export to OrcaFlex format
agent.export_to_orcaflex(panel_mesh, "hull_panels.dat")

Structural Mesh for ANSYS

# Generate structural mesh with boundary layers
struct_mesh = agent.generate_mesh(
    geometry="structure.step",
    element_type="hexahedron",
    boundary_layers={
        "walls": {"layers": 5, "growth_rate": 1.2}
    }
)

# Export to ANSYS with named selections
agent.export_to_ansys(
    struct_mesh,
    output="model.cdb",
    named_selections=["inlet", "outlet", "walls"]
)

Batch Processing

from digitalmodel.agents.gmsh.batch import GMSHBatch

# Initialize batch processor
batch = GMSHBatch(parallel=True, max_workers=4)

# Process multiple geometries
results = batch.process_directory(
    input_directory="./geometries",
    output_directory="./meshes",
    config={
        "algorithm": "frontal-delaunay",
        "element_size": {
            "min": 0.1,
            "max": 1.0
        },
        "quality_targets": {
            "min_jacobian": 0.4,
            "max_aspect_ratio": 4.0
        }
    }
)

# Generate quality report
batch.generate_report(results, "mesh_quality_report.html")

Command Line Interface

# Generate mesh from geometry
python run_gmsh_agent.py generate \
    --input geometry.step \
    --output mesh.msh \
    --element-size 0.5

# Batch processing
python run_gmsh_agent.py batch \
    --input-directory ./geometries \
    --output-directory ./meshes \
    --config batch_config.yml \
    --parallel 4

# Quality assessment
python run_gmsh_agent.py assess \
    --mesh mesh.msh \
    --report quality_report.html

# Mesh optimization
python run_gmsh_agent.py optimize \
    --input mesh.msh \
    --output optimized.msh \
    --method laplacian \
    --iterations 10

Integration Examples

OrcaFlex Integration

# Panel mesh for hydrodynamic analysis
panel_mesh = agent.generate_panel_mesh(
    geometry="hull.step",
    panel_size=2.0,
    refinement_zones=["waterline", "appendages"]
)

# Mooring line discretization
mooring_mesh = agent.discretize_mooring(
    length=800.0,
    segments=50,
    target_element_length=16.0
)

# Export to OrcaFlex
agent.export_to_orcaflex(panel_mesh, "hull_panels.dat")

ANSYS Integration

# Structural mesh with named selections
agent.export_to_ansys(
    mesh,
    output="model.cdb",
    format="CDB",
    boundary_conditions=True,
    named_selections=["fixed", "load", "contact"]
)

# Alternative INP format
agent.export_to_ansys(mesh, output="model.inp", format="INP")

Templates

Pre-configured mesh templates:

• offshore_platform.geo – Platform mesh template
• mooring_line.geo – Mooring discretization
• seabed_terrain.geo – Seabed mesh

MCP Tool Integration

Swarm Coordination

// Initialize meshing swarm
mcp__claude-flow__swarm_init { topology: "ring", maxAgents: 4 }

// Spawn specialized agents
mcp__claude-flow__agent_spawn { type: "code-analyzer", name: "mesh-generator" }
mcp__claude-flow__agent_spawn { type: "reviewer", name: "quality-checker" }

Memory Coordination

// Store mesh configuration
mcp__claude-flow__memory_usage {
  action: "store",
  key: "gmsh/mesh/config",
  namespace: "meshing",
  value: JSON.stringify({
    geometry: "hull.step",
    elements: 50000,
    quality: "verified"
  })
}

// Share mesh status with analysis agents
mcp__claude-flow__memory_usage {
  action: "store",
  key: "gmsh/mesh/ready",
  namespace: "shared",
  value: JSON.stringify({
    file: "hull_panels.dat",
    for_tool: "orcaflex",
    panels: 5000
  })
}

Performance

Standard Workflows

Standard Meshing Workflow

1. Import geometry
2. Heal geometry defects
3. Generate mesh
4. Assess quality
5. Optimize if needed
6. Export to desired format

Batch Processing Workflow

1. Configure batch settings
2. Process files in parallel
3. Generate quality reports
4. Optimize poor meshes
5. Export all results

Error Handling

Memory Issues with Large Meshes

# Use chunked processing
agent = GMSHAgent(config={
    "performance": {
        "chunk_size": 10000,
        "memory_limit_gb": 2
    }
})

Poor Quality Meshes

# Increase optimization iterations
optimized = agent.optimize_mesh(
    mesh,
    method="remeshing",
    iterations=20,
    quality_targets={
        "min_jacobian": 0.4,
        "max_aspect_ratio": 3.0
    }
)

Related Skills

• freecad-automation – CAD geometry creation
• orcawave-analysis – Panel mesh application
• orcaflex-modeling – Hydrodynamic analysis
• structural-analysis – FEM application

References

• GMSH Documentation: https://gmsh.info/doc/texinfo/gmsh.html
• Python API Reference: https://gitlab.onelab.info/gmsh/gmsh/-/blob/master/api/gmsh.py
• Agent Configuration: agents/gmsh/agent_config.json

Version History

• 1.0.0 (2025-01-02): Initial release from agents/gmsh/ configuration