search

GPUSPH / gpusph

The world's first CUDA implementation of Weakly-Compressible Smoothed Particle Hydrodynamics
161 stars 67 forks source link

DEMExample does not work with SPS viscosity #54

Open erikvanettinger opened 5 years ago

erikvanettinger commented 5 years ago

DEMExample does not work with SPS viscosity. Particles are leaving the domain for no known reason.

saltynexus commented 4 years ago

Did you make any progress on this? If yes, please share. If no, can elaborate? Are you only changing the viscosity?

erikvanettinger commented 4 years ago

Unfortunately I didn’t make any progress on this. Yes I was only changing the viscosity.

On 28 Jan 2020, at 07:50, saltynexus notifications@github.com wrote:

Did you make any progress on this? If yes, please share. If no, can elaborate? Are you only changing the viscosity?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/GPUSPH/gpusph/issues/54?email_source=notifications&email_token=ALP3VFY5A3CN4EF6WX2PCT3Q77BTBA5CNFSM4I2MVUIKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKCDYIQ#issuecomment-579091490, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALP3VF2VRH5GNXUY3Q7SUQLQ77BTBANCNFSM4I2MVUIA.

saltynexus commented 4 years ago

I know this doesn't fix your problem, but I've been recently implementing a DEM and found some difficulties as well. Long story short, I got a reply back from Giuseppe Bilotta over on the GPU discourse (https://gpusph.discourse.group/t/dem-in-combination-with-other-geometries/99/5) and I was informed on how the DEM is constructed

One ways is “geometrically”: it doesn’t get filled with particles, but you need the ENABLE_DEM flag for the framework.

I don't know much about the SPS viscosity or how you're trying to use it, but I found when I tried to use an incompatible boundary model with the DEM, the particles fell through as well. To fix my problem, I had to choose a supported boundary model. So yes, it could very well be that the SPS viscosity is incompatible. However, there's hope in the near future. Giuseppe mentions that the DEM should theoretically be able to be modeled using particles instead of planar geometry. I know this would permit me to use the boundary model I'm interested in and it might also permit you to use the SPS viscosity. That being said, it was unclear to me if this feature is currently available or on the TODO list.

You can read more in the link I provided above and possibly express you interest in resolving the issue as well. That is assuming we share the same problem. The more people that share a common problem, the more likely it will move up the TODO list.