'Airflow Around Sphere' simulation project by jdunham


#1

I created a new simulation project called 'Airflow Around Sphere':

Airflow Around Sphere


More of my public projects can be found here.


#2

Hi @jdunham!

Interesting approach to simulate the flow around a sphere :slight_smile: Can you tell us a bit more about your study? Really interested in the outcome though :+1:

All the best!

Jousef


#3

Hi @jousefm ,

I was actually trying to run a simple model with thermals in order to play around with Paraview. I was having trouble with Paraview with some other models I previously ran in Simscale so thought I’d try a more simple model. So I’m not really looking for anything interesting in this model.

That being said, I’ve ran into some troubles running this model…


#4

Hi @jdunham,

Could you tell us what those issues were? @Get_Barried and I might help you out here!

Best,

Jousef


#5

@jousefm

On a coarse mesh, I get the following two errors after it runs for a few hundred iterations.
https://www.simscale.com/docs/content/support/eventLogErrors/maximumNumberOfIterationsExceeded.html
https://www.simscale.com/docs/content/support/eventLogErrors/fatalIOError.html

On the moderate mesh, the run fails before the first iteration, and I can’t tell why.

Thanks for the help!


#6

Hi @jdunham,

I’ve copied your simulation over to see if I can resolve the issue. Will let you know if I get any breakthroughs.

Cheers.

Regards,
Barry


#7

Hi @jdunham,

Seems like it was a meshing issue. I’ve used the “Hex-dominant automatic (only CFD)” instead of the “Hex-dominant (beta preview)” and re-ran it with your simulation parameters. No issues at all. Do refer to this project and under “copy of air only” for the mesh, “incompressible” > “run 5” for the successful run.

Moving forward from here, I would suggest re-meshing the way I did it with both your geometries and then attempting to simulate again.

Cheers.

Regards,
Barry


#8

Hi @Get_Barried,

Thanks for the help! I’d like to practice some conjugate heat transfer simulations. Any chance you know of a good way to mesh a multi-region mesh using hex-dominant parametric? I’ve always seemed to have trouble with that in the past and was hoping the beta would suffice, but looks like it still has some issues. No biggie, but if you happen to have some tips, I’d love to hear them.

Thanks a lot for your time!
Jonah


#9

Hey Jonah!

Indeed that is possible but since your domain is cylindrical, it may be difficult to just mesh internally using the parametric option due to the bounding box being well… a box.

A simple workaround would be to make the region “encompassing” the sphere a square. In essence, that would mean just changing the cylindrical surface to a rectangular one. Although I might have concerns about the quarter sphere itself, but you can try it out first.

As for wanting to maintain the cylindrical shape and still use the parametric option for internal meshing, I’ll need to call on @PowerUsers_CFD for suggestions.

Do try it out and let me know!

Cheers.

Regards,
Barry


#10

Thanks!