CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   Minimum Allowable Wall Distance (

rks171 January 21, 2012 12:07

Minimum Allowable Wall Distance
I setup a model in starccm+. I get a warning "wall distance limited in n cells". I ran the simulation anyway and my solution seems to converge nicely. The residuals all decrease. I read in the manual that if you get that warning about wall distance, you should remedy it. So I tried reducing the minimum allowable wall distance so that less cells are limited. Now my turbulent residuals and energy residuals seem to be oscillating, though the other physical parameters that I'm monitoring in my solution still seem to converge correctly. So does it matter if a small number of my cells have their wall distance limited (~1,000 in a 6.5 million cell mesh)?

cfdnewbie January 21, 2012 12:11

are you doing RANS?

rks171 January 21, 2012 12:16

yes, it is RANS

cfdnewbie January 21, 2012 12:21

what kind of model are you using then?

rks171 January 21, 2012 12:26

It is a steady simulation using segregated flow and segregated energy equations. My turbulence model is realizable k-epsilon two layer with all y+ wall treatment.

rks171 January 27, 2012 13:58

Okay, so after some time, I found out another difference between these two times I ran this case. The first time, my mesh had 2 prism layers adjacent the walls in the model. The second time, I somehow (accidently) used a different outdated mesh that did not have prism layers included. The absence of the prism layers seems to cause the funny behavior in the residuals. But despite the residual behavior, the actual physical solution parameters (flow velocity and temperature) does not seem to change between using a prism layer and not using a prism layer (for my modeling scenario, anyway).

fisicas1 June 16, 2012 15:18

I have this warning " wall distance limited to 1e-6 in 678640 cells in region A" ... What can I do?

rks171 June 16, 2012 15:39

You can change the minimum allowable wall thickness (reduce it) and not as many cells should be limited. Does it cause you a problem or does your simulation still converge?

All times are GMT -4. The time now is 13:48.