Problem regarding Kfric values and free-slip BC


I am a new user for DualSPHysics and I was just experimenting with some cases. I faced some issues that I was hoping that you could help me with:
  1. I was trying to simulate a free-slip boundary condition and since I could not find any way on the forum or the documentation to do so, I thought that reducing the friction coefficient at the boundary to zero should also, logically, do the trick. Would that be a fair assumption to make?
  2. As a test case, I built a rotating cylinder and filled that with fluid. In the default condition, that was working just as expected. But then, I set the Kfric value of the material at the boundary to zero, and the results were still exactly the same, even though I had expected that there would not be any major movement of the enclosed fluid.
I am very confused with regards to the same. Am I interpreting the meaning of Kfric wrong? And if that is the case, is there any alternate method to simulate a free-slip boundary condition?

Any guidance would be greatly appreciated.

Thanking you,

Yours Sincerely,

Saarang Gaggar


  • Hi

    If the movement is laminar the friction will have a very negligible effect. Native SPH interaction is friction less, so I just want to make sure that you have remembered to use/activate DEM?

    If you have not activated DEM no friction effects will be taken into account.

    Kind regards

  • Hello all!

    This is one of the major hurdles with particle methods, geometry induced frictional behavior.

    If you think about, there is no easy way to produce a completely smooth wall in a normal particle description, there will always be a 'particle to particle' normal component that is actually not normal to the surface.

    This is exacerbated in our DCDEM description, because particle radius is small. That is why this method is not advisable to describe long-lasting frictional contacts, slow dynamics etc.

    Look forward for our next release were we coupled a method that corrects this problem.

  • @RCanelas could you provide a graphic describing what you explain? I think I understand it, but would like to be sure :-) If you happen to have any it would be highly appreciated.

    Kind regards
Sign In or Register to comment.