[FSCFD01]

I am a bachelor student currently working on a CFD Simulation of a Formula Student Vehicle at maximum speed (40 m/s).
With my current mesh settings, and the selection of k-w SST model, the residuals remain fluctuating at 10^-3 for continuity, x, y and z momentum, 10^-2 for Tke, and 10^-5 for Sdr.

My current mesh settings are as follows:
Default Controls: triangle mesh, trimmed cell mesher.
Base Size 1.0 m
Target Surface Size 130%
Minimum Surface Size 0.001m
Surface Curvature:
10 pts/circle
200 pts/circle maximum
6E-4 m
Surface Proximity: 32 points in gap (Rest is default)
Surface growth rate: 1.65
Maximum number of connected surfaces: 10000000
10 prism layers 1.3 growth rate, thickness of 0.15m
1 Volume Growth layer
Custom Controls: Enhanced quality triangle surface mesh.
Target surface size: 2%
Prism Layers: 24 layers, thickness ratio 1.3, thickness 0.045m
Minimum cell size 0.003m
Curvature: 48 pts/circle, 480 max pts per circle, 7E-4 curvature deviation distance
Edge proximity of 32 no of faces
Surface growth rate 1.2
12 trimmer surface growth layers.
This results in cell count of 21.5 million cells.

These settings were determined by using the equations for skin friction coefficient, shear stress and y+, along with boundary layer thickness equation for a flat plate.

I was wondering if the present process to determine minimum cell size is erroneous. What is the process used to determine minimum cell size, number and thickness of prism layers, growth ratio and prism layer thickness ratio for a model with steady incompressible turbulent flow?

[Aeronautics El. K.]

Originally Posted by FSCFD01

I was wondering if the present process to determine minimum cell size is erroneous. What is the process used to determine minimum cell size, number and thickness of prism layers, growth ratio and prism layer thickness ratio for a model with steady incompressible turbulent flow?

What makes you think that this process might be erroneous?

What is the desired outcome of your simulations (i.e., what are you studying this for? Drag? Detection of areas of separated flow?...)

Are you simulating a full car with rotating wheels? Non rotating wheels?

Why have you assumed this is a steady state problem?

Have you done a mesh sensitivity study?

These are just some questions to hopefully help you investigate your assumptions and your method.