​

CFD Best Practices

​

Simulation Planning

​

Define Objectives Clearly

​

Understand Your Physics

​

Geometry Best Practices

​

CAD Preparation

• Remove small holes and fillets (< 1% of characteristic length)
• Simplify complex assemblies to essential components
• Use symmetry when applicable to reduce domain size
• Close gaps between components if flow-through is not critical

​

File Quality Standards

Pre-Simulation Checklist:
✓ Watertight geometry (no holes)
✓ Proper normal orientation
✓ Reasonable triangle quality
✓ Consistent units (meters recommended)
✓ Proper coordinate system alignment
✓ Scale verification (realistic dimensions)

​

Mesh Strategy

​

Hierarchical Approach

1. Start Coarse: Validate setup with ~1M cells
2. Refine Strategically: Add refinement only where needed
3. Check Independence: Verify results converge with mesh density
4. Optimize Performance: Balance accuracy with computational cost

​

Critical Refinement Areas

​

Mesh Quality Guidelines

​

Boundary Condition Guidelines

​

Inlet Conditions

​

Outlet Conditions

• Zero gauge pressure
• Natural pressure boundary
• Prevents artificial pressure drops

• When outlet velocity is known
• Mass flow conservation critical
• Confined flow channels

​

Wall Conditions

​

Convergence and Monitoring

​

Convergence Criteria

• Target: < 1e-4 for momentum and continuity
• Target: < 1e-5 for turbulence equations
• Trend more important than absolute values

• Drag/lift coefficients stable within ±1%
• Running average over last 100 iterations
• Check for oscillatory behavior

• Inlet mass flow = Outlet mass flow
• Imbalance < 0.1% of inlet flow
• Check for leaks or boundary condition errors

​

Recognizing Convergence Issues

1. Improve mesh quality (orthogonality, aspect ratio)
2. Reduce relaxation factors (more conservative)
3. Check boundary conditions (compatibility, physical realism)
4. Enable potential flow initialization
5. Extend iteration count (sometimes needs more time)

​

Result Validation

​

Sanity Checks

• Pressure higher at stagnation points
• Flow separation in adverse pressure gradients
• Wake formation behind bluff bodies
• Symmetry preservation (if geometric symmetry exists)

• Mass conservation (continuity equation)
• Momentum conservation (Newton’s second law)
• Energy conservation (first law of thermodynamics)

• Drag coefficients: 0.1-1.5 for most bodies
• Pressure coefficients: -3 to +1 typically
• Velocities: Within expected physical bounds

​

Comparison Standards

​

Performance Optimization

​

Computational Efficiency

• Start with coarse mesh for setup validation
• Use adaptive refinement when available
• Focus high resolution on critical regions only
• Document mesh independence studies

• Use potential flow initialization for external flows
• Monitor convergence and stop when adequate
• Choose appropriate turbulence model for application
• Balance accuracy requirements with time constraints

​

Resource Planning

Mesh Size vs. Runtime (8-core system):
1-3M cells: 30-60 minutes
3-8M cells: 1-3 hours
8-15M cells: 3-8 hours
15M+ cells: 8+ hours

Rule of Thumb:
~1GB RAM per 1M cells
Safety factor: 2× for comfort
Example: 10M cells → 20GB RAM recommended

​

Common Pitfalls to Avoid

​

Geometry Issues

• Scale Errors: Always verify dimensions after import
• Orientation Problems: Ensure proper flow direction alignment
• Poor Quality: Check for holes, inverted normals, bad triangles

​

Mesh Problems

• Under-Refinement: Missing critical flow features
• Over-Refinement: Wasting computational resources
• Poor Quality: High skewness, negative volumes, bad aspect ratios

​

Boundary Condition Errors

• Unphysical Values: Unrealistic velocities or pressures
• Incompatible Conditions: Mass flow imbalance
• Wrong Types: Inappropriate boundary condition selection

​

Convergence Mistakes

• Premature Stopping: Not running enough iterations
• Ignoring Oscillations: Accepting oscillatory “convergence”
• Poor Monitoring: Not watching the right variables

​

Documentation and Workflow

​

Simulation Documentation

• Geometry source and modifications
• Mesh settings and refinement strategy
• Boundary conditions and material properties
• Solver settings and convergence criteria
• Key results and validation metrics

​

Iterative Improvement Process

1. Initial Setup: Quick validation run
2. Mesh Study: Verify grid independence
3. Validation: Compare with known data
4. Optimization: Refine critical areas
5. Production: Final high-quality simulation

​

Quality Assurance Checklist

Pre-Submission Checklist:
□ Geometry validated and properly oriented
□ Mesh quality meets standards
□ Boundary conditions physically realistic
□ Simulation converged to target criteria
□ Results pass sanity checks
□ Key metrics documented
□ Comparison with validation data (if available)