Issues with Wind Loads on Isolated Building (weuq-0013)

[Imad_Alhayik]

Hello,

I faced multiple problems with initiating the CFD case for the Isolated Building while following the instructions on this case. I am using WE-UQ version 3.0.1. I was able to create the BlockMesh and snappyHexMesh locally. I am using the LES with the Samagrinsky model. Once I finished defining all the variables in this case, I ran the case using DesignSafe.

I received an error, which I have attached here. Basically, the error says that in k file, the building is defined as symmetry, which needs to be defined as a wall (i.e., wall function). I am not even sure why I need k file if I use the Samagrinsky model, but this is not the issue I want to address here.

I tried to modify it manually to run the case to avoid the issue.  However, once I save the OpenFOAM files using the icon or run the model in DesingSafe, the file keeps overwriting the 0/ file, which is the problem here.

Please let me know how to fix this issue so that I can run the model, and please let me know if there is any way to run the model again after I modify the case in the server.

I found more bugs as well related to this case maybe they can be addressed with the release of the new version.

1 - I can't run this case locally. I have tried to run it in series and parallel. Other cases can run fine except the weuq-0013.
2- In EVT --> CFD - wind loads on Isolated building  --> Numerical setup, If I changed the simulation type from LES, I got the wrong model for the LES and RANS. The RANS model seems to take over the options for LES and vice versa. photo attached

Thanks!
Imad

[abiy]

Hi Imad,

Thank you for the feedback and for being interested in SimCenter tools. The two attached images are the same. Can you please attach the screenshot of the error you see for the k files?

Considering that high-fidelity CFD simulation takes significant computational time, weuq-0013 and all the CFD runs are supposed to be done remotely using the resources at DesignSafe.

The bug related to LES/RANS model setup will be addressed in the next release, we appreciate your comment on that.

Abiy   

[Imad_Alhayik]

Hi abiy,

I appreciate your taking care of my case. I have attached the full error with this replay with a screenshot of the error that I am getting in the k file when I send the case remotely. I have tried to change it manually but however, but when I send the case, it overwrites what I did. However, I was able to run the OpenFOAM case locally without using the WE-UQ tool for just a couple of iterations, and I didn't get an issue.

Thanks

[abiy]

Hi Imad,

Thanks for noting this issue. It seems that there is a mismatch between the BC used for mesh generation and for the flow fields (depending on the version of OpenFOAM used). This will be addressed in the coming release of the WE-UQ. For now, please use a "slip" boundary for the side and top faces of the domain.  In this context, the "slip" is similar to "symmetry" BC in implementation.


Best,
Abiy