[zx9cp]

Hi all,

I'm attempting to generate a simple linear wave in a 2D tank of length 20m and depth 8m. My mean water depth is 4m and my max free surface displacement is 0.5m giving a free surface peak height of 4.5m and trough of 3.5m.

The wave frequency is 0.5 Hz and the wave length is 3m.

I am trying to develop the wave using CEL expressions for both U and V velocities. These expressions compare well with my basic hand calcs i.e the U velocity decays exponentially as the depth increases.

I am new to this type of work and am a little unsure about a few things:

1. Do I need to specify the inlet surface displacement in the y direction (obtained from the circular particle orbits described in deep water wave theory) aswell as these velocity components?

2. I do not know how to specify the outlet condition, at this stage I am not concerned with reflection, I just want to see a wave propogate through the domain. Surely I cannot specify a pressure as I do not want to constrain this boundary at all, or do I?

Any help would be much appreciated!

[zx9cp]

I should probably also say that I have run a number of simulations with no success!!

The solver returns the following error code:

| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Floating point exception: Overflow

I have experimented with various outlet settings but am now pulling my hair out, I had initially thought this would be a straightforward thing to attempt!!

Thanks again........................

[ghorrocks]

The overflow error is a FAQ: http://www.cfd-online.com/Wiki/Ansys...do_about_it.3F

A reflective outlet condition will cause convergence problems. Something which will dissipate the wave energy will help - this can either be a modelled beach (so the waves can break on a shallow region) or a deliberately coarsened mesh (it uses the dissipation of a coarse mesh).

[zx9cp]

Hi, thanks for the advice. I have experimented with a coarsening mesh region equal in length to my region of interest. The elements remain of constant height but increase in length along the direction of propagation. This seems to be the approach discussed most often, I am still recieving the same error. I am convinced this is now down to my outlet bc.

For general free surface flow the literature seems to suggest a static pressure outlet with the pressure set to be consistent with a known water depth. Is it reasonable to adapt this to my case by specifying a static pressure outlet after my numerical beach to be consistent with my mean water level?

I though this would over constrain the model since my pressure expression will inevitably be based on VOF and therefore fluid depth.

[ghorrocks]

Read the CFX documentation about how it handles hydro static pressure for simulations with gravity - you will need to take this into account.