CFD simulation results

thanhtruc · Posts: 11

Hi,
I am using DB trial version 236005. I have built up several atrium models for CFD simulation. After simulation is finished, i always get the message "results are available but not coverged" (attached model with this post). The CFD output is quite ok. I wonder is it the CFD result reliable?
Look forward to your feedback.
Thanks in advance.
Best regards,
Truc Pham

Steve Potter · Posts: 572 Location: newcastle

Hi Truc,

A CFD simulation is considered converged when the dependent variable residuals (a quantity indicating how well each equation balances) have reached the specified termination values. If you pause the calculation or the specified maximum number of iterations have been reached before the residuals have decreased to their terminal values, you will see the message that you describe. However, the calculation may reach an acceptably accurate solution before the termination residuals have been reached. One way of assessing how well the solution has converged is to set up monitor points for say temperature or velocity at regions likely to undergo rapid change and see if the magnitudes of these monitored variables have reached a relatively static value.

Regards,

Steve

thanhtruc · Posts: 11

Hi Steve,
Thank you very much for your quick response. yes, your suggestion is great Wink

. i have another question regarding to the DB meshes method. I read some article mentioned that the cartesian meshes are good for the rectangular structure, but for the circle structure like Atrium will not be good or accurate enough. Could you maybe give me some comments on this?
However, i really enjoy to work with DB especially to build up the geometry of the building Smile

.
Best regards,
Truc Pham

Steve Potter · Posts: 572 Location: newcastle

Hi Truc,

blas__@hotmail.com · Posts: 5

Hello, I have some trouble with the CFD results. I've made an internal CFD of a 20x20x20 box. There is a big hole in the bottom and small hole in the top of each wall. 2 holes per wall.

The roof is a glass surface. I've made the normal simulation, and the temperatures are around 24ºC. But when I see the CFD results of the same model, while Velocity and Pressure contours seem correct, in the temperature contour plane, the program is telling me that I have between 169ºC and 177ºC.

The version I'm using is the 2.3.5.036. Is there a known issue about this?

thank you

Steve Potter · Posts: 572 Location: newcastle

Hi,

From the slice contours image that you have uploaded, it looks as if it may be the incoming air temperature that is causing the problem. I suggest that you check the setting for the incoming air temperature for the surface inlet opening. Did you import the boundary conditions from an E+ calculation? If so, there was a bug in an earlier beta that could misplace the decimal point for imported air temperatures. Alternatively, the air temperature may have been inadvertently mis-keyed.

Regards,

Steve

blas__@hotmail.com · Posts: 5

Hello Steve,

thank you for your quick answer. I did import the E+ boundary conditions so I guess that the problem can be the bug you are talking about.

Thanx again

oraleks · Posts: 3

I am using version 2.4.2.016.
CFD boundary conditions were imported from the energy plus simulation done in DesingBuilder. While the internal temperatures according to the simulation range between 27-30 C deg, the CFD result viewer indicate a rather strange range of 161-169 C deg.
Outdoor conditions were simulated through a weather file.
The CFD procedure ran until full convergence.
Indoor velocities range seems to be normal - between 0 and 4 m/s.
What is wrong?

oraleks · Posts: 3

Just forgot to mention - when checking the incoming air temperatures of each external wall values seem right. When checking the incoming air temperature at each window (for the same wall), I get a value which is exactly 10 times higher (298 instead of 29. Cool

. So I guess there is still a bug in the import process of the simulation results...

Shortcuts

Login