A while some one asked a question about Reynolds number calcualtion in CFX, unfortunately there was no answer, so I thought I will try again. I have been running an external flow simulation. I made 2 models, one a tet dominated mesh without any hex's, and another model with hex dominated meshes, I am getting a higher Reynolds number from the Hex simulated meshes, by the way, I have more volume mesh from the tet dominated model, about 5:1. I am waiting for the wind tunnel report to compare results. Perhaps some one knows how CFX is computing the Re number, perhaps this might help me to figure out the influence of the mesh. Secondly, I am told that Hex mshes for external flows are generally more accurate, has any one tried this or know about it ? Thanks

Hi,

The Reynolds number reported in the output file is not a good way to compare results. You should compare results based on the physical quantity you are interested in, such as pressure drop, flow rate, forces on surfaces or whatever.

Hex meshes are in general better than tet meshes, as long as you can generate a good quality hex mesh. A hybrid grid with tets in the domain and prisms at the boundaries is almost as good as a good hex mesh. The problem is that many geometries are very difficult to generate good hex meshes, so in general the hybrid grid method is recommended.

As for comparing the quality of hex, prism and hybrid grids - there is a paper on the CFX-Community site which discusses this.

Regards, Glenn Horrocks

Hi Ogbeni,

The Reynolds number written by the solver is a rough estimate which uses the the cube root of the mesh volume as the length scale and either the average velocity of the boundary conditions or the RMS averaged velocity of the solution as the velocity scale. If density and viscosity vary, then these too are averaged.

Regards, Robin