Libso - error - openfoam 11
 

Hi,

I'm using OpenFOAM 11 on Ubuntu 22.04 LTS and I'm not able to use the solver I developed. I used it safely in version 9, I made the changes to compile it successfully in version 11 but unfortunately I'm getting the following error:

controlDict:

Can someone help me? I couldn't evolve alone.

Well Guilherme, the error message is saying that it cannot find the LES model multiphaseNicenoKE; it gives you the list of available models in the list after the error message:

So, assuming that models NicenoKEqn and multiphaseNicenoKE are your own user models, then I guess the real question is why is your multiphaseNicenoKE model not being recognised? Did it compile properly? I'd start there.

Hello,

Can you go to your $FOAM_USER_LIBBIN and confirm that your libphaseMomentumTransportModel.so is there or not?

Also, which version of OpenFOAM are you using?


Bests,
Amirhossein.

Exactly,
That's the question I can't answer. I put the compilation log in the previous message. For me it was a success. I don't understand what I could have done wrong.

Check, it's there.
I'm using OpenFOAM 11.

Do one thing, try to compile it in $FOAM_LIBBIN, and see whether it works by compiling there or not,

If it worked, it seems that the PATH to $FOAM_USER_LIBBIN is not defined properly.

I'll try and get back to you as quickly as I can.

For your knowledge:


and...


file:
log:
ERRO :(


How can I make sure OF11 is looking at the correct folder ($FOAM_USER_LIBBIN)? Although that doesn't make ANY sense. It locates the folder correctly.

Something that is bothering me is the fact that kEqn was not supposed to be cited. It is not a multiphase model...

and, OF mentions the error in line 176, but honestly I didn't see anything...

Don't worry about the line 176 - the code exits via the call to FatalErrorInFunction in the following lines:

where line 176 is the one with the call to the FatalErrorInFuntion. So basically, the code cannot find your chosen LES model in the list of available models, hits end of table and calls the FatalErrorInFuntion, which exits cleanly with the warning message. So all fine there.

Your problem, as I understand it, is that your LES model multiphaseNicenoKE is not appearing in the list of available models, but kEqn is. So here's a thought - try remove your multiphaseNicenoKE model from the library and recompile, and now what is the list of available models? Does kEqn disappear? If so, then we are getting closer - your model has compiled, but is registered with the wrong name; try then checking your coding to see if you have a simple boo-boo (like forgetting to update the Typename to multiphaseNicenoKE).

I remain unsuccessful.

I reviewed all the code and nothing! It compiles and the solver don't recognize it. I even compiled OF11 from the source code, placed my solver in the LES folder (phaseCompressible/LES) and NOTHING!
It compiles without errors, but the solver don't recognize it when I try to use it. I don't know if I'll lose hope and go back to OF9. :(

I don't know if the problem is related to the way I reference the solver, through phaseCompressibleMomentumTransportModels.C
However, I can't see a way out.

Bom dia!

The linker tells you that

Are you trying to overwrite the OpenFoam installation for which you do not have the permission?

Should you copy the library file momentumTransportModels to a working directory that you can savily write to?

Hi,

The software output is below:

This output you showed was an attempt suggested by the forum.

Now, in a second moment, I compiled OF11 from the source code and redid the process of compiling my solver in the conventional way. As I said, it compiles but is not recognized by OF. In a second attempt, I took my code and threw it into the OF folder (now, inside the original software), it compiled, but it also remains unrecognized.

I understand.

Question: are you sure that the solver you use links with there library that you modify? You are using multiphaseEuler, correct? Can you provide us with the output of

and subsequently

Thx.

I am not sure,

I will do what you asked and post the result here as soon as possible.

Hi,


Nothing heappen.



log:

Was this really what I should be looking for?

1/ what solver are you currently using?

2/ where is the executable of that solver located?

3/ which library does this executable require?

Hi,

I'm using multiphaseEuler. The turbulence model I developed. I was able to use this without any problems in version 9 (it is based on the NicenoKEqn model).

I don't know if I understood the question well, but I created a folder inside $FOAM_RUN and compiled it.

I can't say specifically which ones it requires, but as I said, they are the same ones required by NicenoKEqn.

It is hard for me to judge for me from a distance what is going on.

OpenFoam works with shared libraries https://en.wikipedia.org/wiki/Shared_library

It us therefore perfectly possible that the library compiles fine, but the executable (the solver) does not pick up the changes, simply because the solver continues to pick up the old (outdated, unmodified) libraries. This (in my very limited understanding) describe the scenario above.

My above suggestion is to find the executable, and to check which libraries are used to build the executable (the solver). The unix command ldd https://en.wikipedia.org/wiki/Ldd_(Unix) does precisely this.

An alternative approach might be to write a C++ program hello-world with a hello-master function in a library file. Not sure how this helps you (in case at all)?

Keep writing us here.

Good luck.