[Adamcsyou]

Hi all,

I'm running the student version of CFX so I have a 512k cell/node limit.
I'm trying to perform a simple flutter analysis on an axial turbine rotor so I have modeled two passages, meaning I only really have a 256k limit. Furthermore, my professor has requested a fine and a coarse mesh, meaning I'll likely need to half the limit once more.

My question is about the guideline given by CFX stating that in order for it to be properly resolved, there should be 10 cells in the boundary layer.
I'm dealing with a trade-off in that if I put enough cells in the boundary layer, I sacrifice the quality of the rest of my mesh, getting fairly significant error (mostly edge length ratio and skewness).

I'm using Turbogrid to make the mesh. The desired outcomes of the simulation are max blade amplitudes and aerodynamic damping.

Does anyone have any knowledge of the reasoning behind the 10-cell guideline, and know if it's worth sacrificing the rest of the mesh?


Thanks in advance!

[ghorrocks]

The best way to answer this is to run a benchmark simulation of a similar flow to your case and test a 10 node BL mesh versus the alternatives. Then you will see the difference in your case specifically.