Learning Forum

[Edmond Lam]

Hi all, I am simulating a laminar multiphase flow in Fluent using the VOF model. However, when I am trying to extract forces acting on a no-slip wall parallel to the xz plane, the value from Fluent force report using the direction vector (1,0,0) is different from the one reported by a surface integral of X-wall shear stress on the same surface. In fact, in the Fluent force report, the viscous force component along the direction vector (0,1,0) is non-zero, which does not make sense for a wall parallel to the xz plane. In comparison, the surface integral of Y-wall shear stress on the same surface results in a value almost equal to zero (E-27). Are there any ideas? Thanks.

[SRP]

Hi,

Will you kindly send a screenshot of the geometry, force report results, and surface integral of the shear stress in the relevant direction?

Thank you

Saurabh

[Edmond Lam]

Hi Saurabh,

Thank you for your reply.

The geometry is very simple, only a box. The surface I am interested in is the bottom surface, which is parallel to the xz plane.

Here are the results from the force reports and surface integrals:

 

As you can see, the forces are very different.

This mismatch seems to occur when on this surface there are 2 phases.

[Amine Ben Hadj Ali]

Can you please include the report of the force vector and not the dot product of the force vector with the direction.

[Edmond Lam]

Here it is, thanks:

 

As mentioned, this mismatch seems to occur only when there are 2 phases in contact with the wall. (I have 2 phases in total for the case.)

I have wall adhesion turned on for the wall and a specfied contact angle. The surface tension is modelled using CSF. I am wondering if using these models adds additional terms to the force calculation on the wall in Fluent.

 

Thank you very much.

[Amine Ben Hadj Ali]

Surface tension force should is not part of the force report.

For a VOF case with two phases I am not able to reproduce your issue. Is your problem reproducable on a simple dummy case?

[Edmond Lam]

Thank you for your reply.

I ran two simple dummy cases for 10 time steps after patching a liquid droplet with a velocity in -x.

The boundary conditions for the 2 cases are exactly the same.

There is no gravity, and the wall that I am measuring forces is at the bottom with wall adhesion enabled, surface tension modelled with CSF.

1) a liquid droplet moving in the -x direction in its vapour:

Here, with only one phase (vapour) in contact with the bottom wall, integral of x-wall shear stress is exactly the same as the force report in (1 0 0). The integral of static pressure is also exactly the same as the pressure component in force report in (0 1 0). However, although the integral of y-wall shear stress is almost zero, the viscous component in force report in (0 1 0) is non-zero, or at least much larger.

2) a liquid droplet moving in the -x direction in its vapour, but also touching the bottom wall:

Here, with both phases in contact with the bottom wall, the integral of x-wall shear stress no longer matches the force report in (1 0 0). The integral of static pressure also no longer matches the pressure component in force report in (0 1 0). And similarly, although the integral of y-wall shear stress is still almost zero, the viscous component in force report in (0 1 0) is still much larger.

[Amine Ben Hadj Ali]

Can you check if there is a mismatch if you disable the surface tension force?

 

[Edmond Lam]

The mismatch is gone when wall adhesion or surface tension modelling is disabled. But the integral of y-wall shear stress (E-27, almost zero) is still different from the viscous component of the force report in the (0 1 0) direction (non-zero).