[HPE]

Hi,

- Could you please elaborate what went wrong and its relation to the post in the link?

Many thanks.

[Tibo99]

Hi @lebc,

I'm not in the position at the moment to decide if I'm allowed to share the work I've done (sorry about that), but l found what I needed in this paper:

https://onlinelibrary.wiley.com/doi/....1002/fld.2709


The gradiant at the top for U and espilon that I use as a BC look something like this:

for epsilon: grad[faceI] = -pow(ustar_[faceI],5)*kappa_/pow(nut[faceCellI],2)

for U: grad[faceI] = dir_*pow(ustar_[faceI],2)/nut[faceCellI]

I did a 1D simulation first with the Inlet/Outlet to be set as 'cyclic' and the BackAndFront as a 'symmetryPlane'. The goal is to having a profile of U,k,epsilon and nut by using the top and bottom BC and use this results to map the 2D domain for initial conditions.

I didn't validate my results though, so, I can't guarantee the right results, but we could probably comparing our results maybe?

Hope this help!

Regards,

[lebc]

@HPE,

I'm using the version of OF from openfoam.org, so I had to adapt a little bit the files for the proposed case. Additionally, I did not compile everything that is proposed in the link you provided, so this could also be the issue. I'm not saying it is wrong, but that I couldn't make it work as I expected.

@René,

No problem, thank you for the tips! My case is simple, I haven't modified the source code, but if you want I can share it with you. I have just accepted the difference in results and used the standard approach from OF for my work.

[Tibo99]

@lebc,

My pleasure!

I didn't change the source code, but just created new BC at the top for U and epsilon in the way I mentioned before in my last post but keep the k-e model from OF as is.

In conjunction with these BC, I used the 'nutkAtmRoughWallFunction' at the bottom. And 'zeroGradient' at the top and bottom for 'k' since the profile is suppose to be uniform.

One thing I know is, if you try to achieve matching U profile at the Outlet with the Inlet, you'll struggle if you don't use these kind of BC. Obviously, there is other options in the literature, but these must match all together.

You can share it if you want. Its up to you.

Good work and stay safe!

Regards,