EXTERNAL CFD with multiple buildings

Vi_06 · Posts: 2

Hi,

I have a large external CFD model which I am simulating for wind from 1 direction (45° from North)...in all cases it doesn't reach a satisfactory solution.
I've changed again and again grid type and domain size to work on it. Now I'm trying with default domain size (3,3,2) , a 10 m grid default spacing and a grid line merge tolerance of 1m.

Choosing turbulence K-E as dependent variable, it seems to be not running…

Am I “corrupting” the simulation with wrong dependent variable residuals or wrong specified termination values?? Is it possible that simulation is not converging because turbulence isn’t right variable dependent in my case??

I’m uploading last simulation and a CFD section image

I’ve also tried to simplify calculations reducing the entire model into one with fewer buildings. In this case turbulence seems to getting wild. I attache also this simulation.

Regards

Steve Potter · Posts: 615 Location: newcastle

Hi,

From the residual plot, it appears that the solution is generally proceeding quite well. However, the velocity z-component residual seems to remain at a fairly high value which may indicate that you just need to increase the number of cells in the Z dimension (with such a coarse grid, only one or two grid lines may be being generated for the z-dimension).

The most likely reason for the dependent variable plot for turbulence kinetic energy remaining at a fixed value of 0.01 is that the monitored cell is located inside a solid region. If you select a monitor cell somewhere in the fluid region, I think you'll find that the dependent variables will tend towards a final value as expected.

Hope this is of help.

Regards,

Steve

Vi_06 · Posts: 2

That was helpful thank you so much!!

Regards

Shortcuts

Login