Changing materials in CaseDEM

JuliaB

Hi,

I am just starting using DualSPHysics and trying out stuff using the testcases, which are very helpful! I am just having issues when trying to adapt the materials and relativeweight in CaseDEM. I left everything else as it was and just changed the material of the floating body.

When, for example, using "lime-stone" instead of "hard-wood" the floating bodies start flying upwards right after starting the simulation (before the water even started moving, as if they are exploding) and therefore it stops after only a few steps saying "floating bodies out of boundaries" (of cource, since the floating bodies went through the upper border).

The same happens when using "Aluminium". With higher values for relativeweight the same happens, just a little slower (since the floating bodies are heavier). It also happens when using rhopbody instead of relativeweight.

Does any one of you have an idea, what the reason for this could be? Do I have to change anything else in the XML-file?

I would be really thankful for any advice!
Julia

Alex

Note that as mentioned in the guide, the correct behaviour of floating bodies is observed when solid objects are created instead of particles only in the faces!

If you still have problems, please send your XML to us (dualsphysics@gmail.com) and we will check it when we have time

Regards

JuliaB

Hi Alex,

thank you for your reply! The bodies are solid, just as in the CASEDEM. As I mentioned, I did not change anything in the XML file but the material in floatings. Thats why I was wondering, whether I'd have to consider anything else when changing the material, but it does not seem like that's the case!? I will send the XML file to you anyway, it would be great if you could take a loot at it!

Regards
Julia

edit: I just saw, that boxfill is set to solid, but setdrawmode is "full". I am running it right now but it seems to be better with setdrawmode=solid! I am just wondering, why it works in the testcase with "hard-wood", but not with any other material...!?

JuliaB

I tried also changing the line with "setdrawmode" to "solid", but that did not change anything, either. I sent you the XML file before and will add a screenshot so you can see what I mean.

I would appreciate it, if you could look into this!

lkhajenoori

Hi Alex,
I have the same problem as JuliaB. Have you found any reason or solution for this problem?
I would really appreciate any help.

Leila

JuliaB

Hi Leila,

I did not find a real solution, I guess it is not so easy to get this problem stable. However, a few things that might help:

- Using the Symplectic instead of the Verlet time step algorithm
- decreasing CFL and/or time-step
- decreasing the restitution coefficient
- placing your object a little bit over the ground in order for it to settle on its own at the beginning of the simulation (this did not work for me, though. After settling, it started bouncing anyways)

Furthermore I found that usually the behavior becomes stable (and the objects move as expected) once the water touches the object(s).

I hope this helps, good luck!
Julia

lkhajenoori

Dear Julia,

Thanks so much for your reply. I applied all your suggestions and seems the only effective option is decreasing the restitution coefficient. But since we are using DEM it does not make sense. and if we want to validate our model we should apply the real properties and another question: do you know what relative Weight is?

best wishes
Leila

moh

I was struggling with this problem, one year ago. I solved it by applying the scheme of "automatic limited compressibility".

lkhajenoori

Hi Moh,
Thank you for your reply. could you please give me some information about this scheme and how can I apply that?

best wishes
Leila

JuliaB

I would also be interested what exactly you mean by "automatic limited compressibility". Do you mean setting the Speed of sound to "auto"?

Generally, decreasing the time-step in one way or another (lower CFL number, lower dtini/dtmin, etc.) should also help stabilizing the simulation.

lkhajenoori

Hi Julia,
thanks for your reply. My model is working better with lower CFL. Can you please tell me what relative weight is?

moh

The mentioned scheme is based on SPH-VOF coupling. The DEM is solved regarding the flow field computed mainly on the VOF.

JuliaB

I am not sure what you are saying about VOF. I did not know that this method is used in DualSPHysics and could not find any information about it in the user guide!? I know about the Volume of Fluid method, just not in this program!? Could you explain this further please?