Agent Skills: CFD Analysis Skill

Deep integration with computational fluid dynamics tools for internal and external flow analysis

thermal-fluid-analysisID: a5c-ai/babysitter/cfd-fluids

Install this agent skill to your local

pnpm dlx add-skill https://github.com/a5c-ai/babysitter/tree/HEAD/plugins/babysitter/skills/babysit/process/specializations/domains/science/mechanical-engineering/skills/cfd-fluids

Skill Files

Browse the full folder contents for cfd-fluids.

Download Skill

Loading file tree…

plugins/babysitter/skills/babysit/process/specializations/domains/science/mechanical-engineering/skills/cfd-fluids/SKILL.md

Skill Metadata

Name
cfd-fluids
Description
Deep integration with computational fluid dynamics tools for internal and external flow analysis

CFD Analysis Skill

Purpose

The CFD Analysis skill provides deep integration with computational fluid dynamics tools for internal and external flow analysis, enabling systematic setup, execution, and post-processing of fluid simulations.

Capabilities

  • ANSYS Fluent, CFX, OpenFOAM workflow automation
  • Mesh generation for complex geometries (structured, unstructured)
  • Turbulence model selection (k-epsilon, k-omega, SST, LES)
  • Boundary condition specification (inlet, outlet, wall, symmetry)
  • Steady-state and transient flow simulations
  • Post-processing for pressure, velocity, and flow visualization
  • Mesh independence studies and validation
  • Pressure drop and flow coefficient calculations

Usage Guidelines

Pre-Processing

Geometry Preparation

  1. CAD Cleanup

    • Remove small features (< 3 cells)
    • Fill gaps and holes
    • Create smooth transitions
    • Define fluid domain boundaries
  2. Domain Definition

    • Internal flow: Extract fluid volume
    • External flow: Create far-field boundary
    • Symmetry: Identify planes of symmetry
    • Periodic: Define periodic pairs

Mesh Generation

  1. Mesh Types | Type | Application | Pros/Cons | |------|-------------|-----------| | Structured hex | Simple geometries | High quality, more effort | | Unstructured tet | Complex geometries | Flexible, more cells | | Polyhedral | Complex internal | Good quality, moderate count | | Hybrid | Mixed regions | Optimized for accuracy |

  2. Boundary Layer Mesh

    First cell height: y+ = 1 (wall-resolved)
                      y+ = 30-300 (wall functions)
    
    y = y+ * mu / (rho * u_tau)
    u_tau = sqrt(tau_w / rho)
    
  3. Mesh Quality Criteria

    Orthogonality: > 0.1 (> 0.3 preferred)
    Skewness: < 0.95 (< 0.8 preferred)
    Aspect ratio: < 100 (< 20 near walls)
    

Solver Configuration

Turbulence Models

| Model | Application | Wall Treatment | |-------|-------------|----------------| | k-epsilon Standard | General industrial | Wall functions | | k-epsilon Realizable | Rotation, separation | Wall functions | | k-omega SST | Aerospace, separation | Low-Re or wall functions | | Spalart-Allmaras | External aero | Low-Re | | LES/DES | Unsteady, vortex shedding | Wall-resolved |

Boundary Conditions

  1. Inlet Conditions

    • Mass flow rate or velocity
    • Turbulence intensity (1-5% typical)
    • Hydraulic diameter or length scale
    • Temperature (if energy equation)
  2. Outlet Conditions

    • Pressure outlet (most common)
    • Outflow (fully developed)
    • Mass flow outlet (specified)
  3. Wall Conditions

    • No-slip (default)
    • Roughness (if significant)
    • Thermal (adiabatic, fixed T, heat flux)

Solution Settings

  1. Discretization Schemes

    Convection: Second-order upwind (accuracy)
                First-order (stability)
    Pressure: PRESTO (complex geometry)
              Standard (simple geometry)
    
  2. Convergence Criteria

    Residuals: < 1e-4 (typical)
               < 1e-6 (high accuracy)
    
    Monitor: Mass imbalance < 0.1%
             Force convergence
    

Post-Processing

  1. Flow Visualization

    • Streamlines and pathlines
    • Velocity vectors
    • Contour plots (P, V, T)
    • Surface integral reports
  2. Quantitative Results

    • Pressure drop
    • Flow coefficient (Cv)
    • Heat transfer coefficient
    • Force and moment

Process Integration

  • ME-010: Computational Fluid Dynamics (CFD) Analysis

Input Schema

{
  "geometry": "CAD file path",
  "flow_type": "internal|external",
  "fluid": {
    "name": "string",
    "density": "number (kg/m3)",
    "viscosity": "number (Pa.s)",
    "specific_heat": "number (J/kg.K, if thermal)"
  },
  "inlet": {
    "type": "velocity|mass_flow|pressure",
    "value": "number",
    "temperature": "number (K, if thermal)"
  },
  "outlet": {
    "type": "pressure|outflow",
    "value": "number (if pressure)"
  },
  "analysis_type": "steady|transient",
  "turbulence_model": "k-epsilon|k-omega-sst|spalart-allmaras|laminar"
}

Output Schema

{
  "flow_results": {
    "pressure_drop": "number (Pa)",
    "flow_coefficient": "number (Cv)",
    "max_velocity": "number (m/s)",
    "reynolds_number": "number"
  },
  "forces": {
    "drag": "number (N)",
    "lift": "number (N)",
    "moment": "array [Mx, My, Mz]"
  },
  "thermal_results": {
    "heat_transfer_rate": "number (W)",
    "average_htc": "number (W/m2.K)",
    "outlet_temperature": "number (K)"
  },
  "mesh_statistics": {
    "cell_count": "number",
    "y_plus_range": [min, max],
    "orthogonality_min": "number"
  },
  "convergence": {
    "iterations": "number",
    "residuals": "object",
    "mass_imbalance": "number"
  }
}

Best Practices

  1. Always perform mesh independence study
  2. Verify y+ values match turbulence model requirements
  3. Monitor mass and energy imbalance
  4. Validate with experimental data when available
  5. Start with steady-state before transient
  6. Use appropriate turbulence model for flow physics

Integration Points

  • Connects with CAD Modeling for geometry
  • Feeds into Thermal Analysis for conjugate heat transfer
  • Supports Heat Exchanger Design for performance prediction
  • Integrates with Test Correlation for validation