Frequently Asked Question

Mesh quality has a direct and often underestimated impact on both convergence and accuracy of steady-state results.

General mesh guidelines:

• Cell quality matters more than cell count. A coarser mesh with good-quality cells (low skewness, reasonable aspect ratios) will often converge better and give more trustworthy results than a dense but poorly-shaped mesh.
• Boundary layer resolution: For wall-bounded flows, near-wall mesh resolution needs to match your turbulence model's requirements — this is governed by the y+ value, which depends on whether you're using wall functions or resolving the boundary layer directly.
• Refinement in regions of interest: Areas with high gradients — near walls, around sharp geometry features, in wake regions — need finer mesh resolution than open, uniform flow regions.

Mesh quality metrics to check before running:

• Skewness — how distorted a cell is from its ideal shape; high skewness (especially above ~0.95 for most solvers) can cause convergence problems or inaccurate results
• Aspect ratio — the ratio of a cell's longest to shortest dimension; very high aspect ratios can be appropriate in boundary layers but problematic elsewhere
• Orthogonality — how close cell faces are to being perpendicular to the line connecting adjacent cell centers; poor orthogonality slows convergence

Mesh independence study: Before trusting your results, it's good practice to run the same case at 2–3 mesh densities and confirm that your key output (drag, pressure drop, temperature, etc.) stops changing significantly as the mesh gets finer. If results are still shifting meaningfully between mesh levels, the coarser mesh isn't yet capturing the physics accurately, and results from it shouldn't be relied on for final decisions.

Practical tip: Start coarser than you think you need, confirm the case runs and converges, then refine incrementally — this catches setup errors early without burning compute time on a mesh that turns out to have a modeling mistake baked in.

Numerical error and convergence

A numerical solution approximates the continuous conservation laws on a finite set of cells or control volumes. Discretization error depends on characteristic size, stretching, skewness, alignment, interpolation, boundary treatment, and the output being measured. Iterative convergence is different from mesh convergence: a solver can reduce algebraic residuals while the engineering quantity continues to drift.

φ is the reported quantity and h is a characteristic mesh size. Use an analytical or benchmark reference when possible; otherwise report a systematic refinement study.

Worked example

A pressure drop is 104 Pa on a coarse grid, 101 Pa on a medium grid, and 100 Pa on a fine grid. The medium-to-fine change is approximately 1%, while the coarse-to-medium change is approximately 3%. Report mesh sizes, refinement ratio, residual state, and whether the 1% change meets the design tolerance.

Check: refine where gradients or the engineering output are sensitive, not only where the global cell count is convenient.