CFD Online Discussion Forums

CFD Online Discussion Forums (
-   STAR-CCM+ (
-   -   Initialization problem in open channel flow simulation (

swtbkim April 11, 2015 08:18

Initialization problem in open channel flow simulation
1 Attachment(s)
Hi all.

I'm analyzing a propeller in a tank. I attached simple schematic figure.

Behind the propeller is a pipe, so air will be sucked because of a negative pressure there and rise to the surface.

I made a field function for a hydrostatic pressure distribution
--> hydro distribution = ($$Position[1]<=1)?{$$DensityofWater}*9.81*(1-$$Position[1]):0 (i.e. reference altitude is y=1.0m and gravity acts along -y direction)

Problem is that I can not initialize the hydrostatic pressure. star-ccm+ gave me an error.
"Unable to compute field function a hydrostatic pressure on "region", Please check that function is defined here"

If I don't initialize the pressure, then hydrostatic pressure distribution should be calculated in solver and it will take too long, and it didn't even converge.

I'm using MRF and multiphase - VOF.

Thank you in advance!

fluid23 April 13, 2015 16:35

I have have seen it do something similar resulting from poorly defined region names that had extra '.' separators in the region name. So if I had a region named 'Region' and a boundary named 'Boundary One' I named it as 'Region.Boundary.One' So it was looking for a region boundary named Boundary not Boundary.One. It doesn't appear that this is your problem though.

I will say that it looks like maybe your DensityofWater field function maybe defined or called wrong. Perhaps I am mistaken, but I thought that a double $$ is for a vector, not a scalar.

Also, have you tried expert initialization (grid sequencing)? That has cut (literally) days off my analyses in the past. It isn't always appropriate, but when it works it works well.

All times are GMT -4. The time now is 16:38.