View Issue Details
|ID||Project||Category||View Status||Date Submitted||Last Update|
|0003574||OpenFOAM||Feature||public||2020-10-16 00:52||2020-10-17 08:28|
|Fixed in Version|
|Summary||0003574: denseParticleFoam - unaccounted continuous increase in velocities at mesh transition for snappyHexMesh mesh|
|Description||I am simulating droplets flowing out of mouth of a human body, using denseParticleFoam.|
The velcoities keeps on increasing at different locations based on mesh transition.
|Steps To Reproduce||Take a slice at the z direction and on comapring the velicities over different times, concludes that the velocity is increasing at the mesh transitions,|
and eventually increasing beyond the inlet velocity
|Additional Information||Same problem observed with MPPICFoam |
On completely unstructured meshes (tetras), this problem is noticed everywhere in the domain, but doesnot constantly increase.
Problem disappears on the same mesh with pimpleFoam solver.
|Tags||No tags attached.|
Is denseParticleFoam only for hexa meshes?
case file attached again below
Link for case file
denseParticleFoam is suitable for all meshes; e.g., the cyclone tutorial.
I suspect that the issue is that the case is largely still with a hydrostatic pressure gradient, but this isn't a p_rgh solver. Hence staggering patterns in the vertical velocity component and some noise on the refinement patterns. Can you try switching gravity off (i.e., set g to (0 0 0)) to confirm it?
It is a p solver. And yes, when you turn gravity off this problem disappears.
However, gravity is important for this case, as I would like to see when the droplets fall into the ground.
||Can you provide funding to further develop denseParticleFoam for your purpose?|
Unfortunately, I am a student and in this COVID times its hard to find funding to sustain my research work.
However just a feedback, this way of modelling gravity term, leaves serious shortcoming on the type of meshes one can have and can be discouraging.
Moreover the gravity effects are required for the particles only, I believe they are insignificant for the continuous field.
I see there is a particle force based on gravity, but I think it takes value of g from the common gravity field.
> However just a feedback, this way of modelling gravity term, leaves serious shortcoming on the type of meshes one can have and can be discouraging.
Only for some cases. We would be happy to further develop this solver if anyone is interested enough to fund the work.
||Sure. With this issue now identified, hopefully there can be someone who would be interested and can fund this work. On a personal capacity if I can find a work around, I will sure be willing to put forward a patch.|
|2020-10-16 00:52||sjohn2||New Issue|
|2020-10-16 00:54||sjohn2||Note Added: 0011615|
|2020-10-16 01:01||sjohn2||Note Added: 0011616|
|2020-10-16 08:42||will||Note Added: 0011617|
|2020-10-16 16:40||sjohn2||Note Added: 0011620|
|2020-10-16 17:18||henry||Note Added: 0011621|
|2020-10-16 17:32||henry||Severity||major => feature|
|2020-10-16 17:32||henry||Category||Bug => Feature|
|2020-10-16 19:26||sjohn2||Note Added: 0011622|
|2020-10-16 20:53||henry||Note Added: 0011624|
|2020-10-17 02:23||sjohn2||Note Added: 0011626|
|2020-10-17 08:28||henry||Assigned To||=> henry|
|2020-10-17 08:28||henry||Status||new => closed|
|2020-10-17 08:28||henry||Resolution||open => suspended|
|2020-10-17 08:28||henry||Note Added: 0011628|