[cfdan]

I'm trying to model a water sump using Interfoam. Attached is the link to my model:

https://www.dropbox.com/sh/9bhhbvj5n1yhov1/tboAYJyAg_

I've got the model to run until about 0.8 or 1.6s before crashing. I think it occurs when the water plunges down to the sump.

The model can also run without utilizing setfields, in most situations.

Can anyone please help check whether my boundary conditions are alright? I've tried numerous iterations, but have so far got a viable solution.

Thanks!

[Tobi]

Hi,


1. polyMesh/boundary

inlet, outlet, atmosphere = patch not wall

2. 0/files

U_outlet why fixed?
alpha_outlet outlet inlet value = 1? 
.
.
.

?

Regards
Tobi

[cfdan]

Hi Tobi,

Thanks! Model is running fine!

[Tobi]

Hi,

what did you change?

[cfdan]

alpha 1:

outlet : inletoutlet 0

p_rgh:

fixedFluxPressure uniform 0 for all except inlet

U outlet is specified because I'm trying to model a pump. Do you have any suggestions?

[Franz_J]

Hi
you could try FlowRateInletVelocity
I noticed that the level of refinement around pump inlet is probably too course to model swirl
cheers
Franz

[cfdan]

Hi Franz!

Thanks! Yup I'm running a model with a refined mesh. Any way to make it solve faster?

[Franz_J]

you can eliminate the upstream channels, I don't think it's necessary for what u want to do

[cfdan]

ok i'll try that too.

One more thing, I've tried the different filters (contour, threshold, smoothing, etc.) to capture the free surface vortex, but so far no luck.

http://www.youtube.com/watch?v=bR3rulCQS08

was the video you posted post-processed by Paraview or Blender? How many cells were you using too?

Thanks!

[Franz_J]

that video was created with paraview while this one;
http://www.youtube.com/watch?v=oqtNdocCaJo
was created using blender, and has about 1M cells. Glad someone has watched the videos :-)

[cfdan]

Those videos are exactly what I want Any tips on how to do it on paraview, if not blender?

[Franz_J]

Hi, excuse my late reply, I ran your model with some modifications, as mentioned above. The outlet is sucking in air, as shown in the attachment . Vortices would probably be obscured by the disturbance caused by the cascade. This common type of wet well pump station would typically feature additional physical elements in the arrangement, to mitigate these undesirable flow characteristics.
The videos from paraview are created by pressing "Save Animation" in the file menu.
cheers
Franz

[cfdan]

Hi Franz,

Thanks for taking the time to run the model.

Speaking of which, how long did you took? And what was your setup (mesh size, and computer specs?)

Also, did you used 'contours' filter to produce the left side picture? perhaps with 'smoothing'? or is it 'threshold'?

And for the import in blender, was it done in x3d? and I have to individually import each time steps?

Appreciate the help!

[Franz_J]

I used the Helyx-OS GUI for setting up the model and running it, as well as playing around with mesh settings. It did not take long.
I had to make some assumptions regarding flow rate that would result in vortices, using engineering formula.
The water surface is produced using the contours filter, give it a try.
You can also import to blender using scripting.
The model is still too basic for accurate assessment but adequate to show qualitatively that this pump station layout would contribute to pump failure in a short time.

[Tobi]

Hey Franz,

can you explain or give more informations due to paraview -> scripting -> blender?

[cfdan]

Originally Posted by Franz_J

I used the Helyx-OS GUI for setting up the model and running it, as well as playing around with mesh settings. It did not take long.
I had to make some assumptions regarding flow rate that would result in vortices, using engineering formula.
The water surface is produced using the contours filter, give it a try.
You can also import to blender using scripting.
The model is still too basic for accurate assessment but adequate to show qualitatively that this pump station layout would contribute to pump failure in a short time.

Hi Franz,

Thanks for the information.

I'm running a case with about 1.5 million cells, parallel with 2 cores.
Still taking me at least a day to develop a closed loop system. By close loop, I mean the water reaching the 'plunging' stage and at the same time being pumped out. I've set the 2 CFL conditions to 1.0 and 0.5.

I understand that to create the vortices the initial levels of the water should not be too high from the bellmouth, and the suction from the pumps have to be strong enough to create them. Am I missing anything?

I've been able to create similar illustrations using 'contour' filter, but I am also interested in the blender scripting too, if you are willing to share. Thanks!

[bzhao]

Hi cfdan,

I am working on similar problem, but still having some trouble with BCs. Would you mind to share a working copy of your model. The dropbox files seem have been removed.

Thanks!

[mwmalkawi]

Hi Cfdan


I am doing a similar to the sump pump model you did and am facing issues most probably with my boundary conditions i am wondering if it is possible to share the case file please as am struggling for some time



many thanks