[OnurKurum]

Hi all,

I am trying to set-up a simple 2D wave case using SFLOW however I am having trouble setting it up correctly. Attached is the description of the simple domain I am trying to set. Beach is on the left side and Offshore on the right side.

Essentially the trouble I am having is to be able to point the waves towards the beach.

In DiveMesh, I have tried flipping C11 and C14 between "wave generation" and "numerical beach" but no matter what I do, the generated waves always propagate in the increasing X-direction. I need them to travel in the -X direction.

Is this possible or all domains need to have the beach on the right side (max X) and wave generation on the left side (min X)?

Based on the attached figure, I have generated the mesh with:
control.txt:

B 1 5  
B 10 434148 434698 4702302 4704174 -8 6 (or in simpler terms: 0 550 0 1872 -10 6)
B 130 1
C 11 7
C 12 3 (assuming this will work with periodic boundaries I set at C22 1)
C 13 3 (assuming this will work with periodic boundaries I set at C22 1)
C 14 6
C 15 21   
C 16 3
C 22 1 (will need to have periodic boundaries applied to lateral boundaries so I used C22 1)
G 10 1   
G 15 2     
M 10 8
M 20 1

ctrl.txt is also attached - I have tried giving some angle to the waves with B 97 and B131 and they both generated perpendicular waves (still going in the +X direction). Then I removed

B97/B131

and used

B 105 61 434648 4704174

which gave me wave angles but still propagating in +X direction.


Thanks for all suggestions - I can provide additional information to help resolve the problem.
Cheers
Onur

[valgrinda]

Hi Onur,

unfortunately, SFLOW does not support periodic boundary conditions. Only CFD does that. Here, the periodicBC focus is on high-resolution LES modeling.

I have tried to run your setup and saw a couple of things:

- It looks like B 97 was still in the manual, it shouldn't have been. B 105 is the option that works for the wave direction

- I don't have the geo.dat file, so I probably ran with deeper water than intended. In any case, it looks like the water is too deep for the NH-SWE approach and the waves damp out. If you experience this including the geo.dat file, maybe FNPF becomes an option for you. Also, the mesh is probably a bit too coarse. A finer grid improves the result, but not enough.

I would say that in the end, I probably don't fully understand the case you want to run. Maybe we should go for a meeting and see whether we can incorporate your use case scenario.

[OnurKurum]

Hi Hans,

We have resolved most of the problems with some remaining issues/requests. I think Dave K has sent you an email about it. I will talk with him tomorrow re arranging a meeting to discuss more SFLOW development, particularly re breaking, bed friction, maybe neumann boundaries.

We figured (after my first post) from the source code that B97 was not implemented in the code and with B 105, the only problem was not being able to turn the waves around (i.e., beach is on left side, and waves were being generated left to right direction, I've tried many things, believe me). We got around the issue for now, no worries.

And thanks for confirming that periodic boundaries are not possible with SFLOW, the code that deals with it is now in divemesh apparently and didn't have to time to look into that.

thanks for your reply,
cheers
Onur

[va-ta-cfd]

Originally Posted by OnurKurum

We figured (after my first post) from the source code that B97 was not implemented in the code and with B 105, the only problem was not being able to turn the waves around (i.e., beach is on left side, and waves were being generated left to right direction, I've tried many things, believe me). We got around the issue for now, no worries.

Hello Onur,

Could you please explain how you have solved the problem with inclined wave direction?
I have the same problem. Actually by using B105, the waves still propagate in x direction.
I really appreciate your advice.

With Regards,
Vahid

[OnurKurum]

Hi Vahid,

I wish I could help but the way we worked around it was to use another mass source wavemaker (not currently open source). Then I stopped digging after the wave direction issue with B105.

I am sure it's an easy fix in the source code though.