Boundary Layer in SPH

Sandra

Hello,

Is it possible to calculate the shear stress near a wall or bed boundary condition using the hydrodynamic field? More specifically, is there a wall function applied in DualSPHysics? The velocity at the boundary is zero so it should gradually go to the freestream velocity; how is this done in DualSPHysics, does it use a wall function?

Thank you,

Sandra

gfourtakas

Dear Sandra,

Thank you for your post. Boundary Conditions (BCs) in SPH are an

open research field (please see SPH grand Challenges https://www.spheric-sph.org/grand-challenges).

 In DualSPHysics, we use two well-known boundary conditions (DBC and mDBC) to impose a no slip wall boundary condition at the wall surface. These general types of BCs impose a no penetration condition (and no-slip) either by density (DBC) or by extrapolating the field variables (mDBC) on ghost nodes on the wall.

 I assume you would like to know if we impose a wall function (log-law or otherwise). Since for most applications the wall boundary layer is quite thin and mostly unresolved by the SPH nodes (particles) we choose not to impose such a wall function and directly use a no-slip BCs as explained above. A wall function is possible, but care must be taken for the sharp velocity gradients near the wall boundary and of course use a sufficient number of particles.

 I hope this helps,

George

Sandra

Hello George,

Thank you for your detailed response.

If I want to calculate shear stress on a wall or solid boundary, would it simply be the eddy viscosity multiplied by the velocity gradient? Is it possible to extract these from the model near the wall?

For instance, the pressure force on an object can be conveniently calculated using SPH, however, it's not clear to me how to calculate the drag force on the object.

For context, I will be using the Laminar+SPS viscosity treatment.

Thank you,

Sandra