[nikhil108]

Hallo foamers,

I am trying to simulate propane leak from a hole inside a box, which is located in a big computational domain. Whole computational domain has air, and propane starts to leak at 5g/s inside the box. For this I am using rhoReactingBuoyantFoam solver with reactions off. The problem is, after some time propane stops from leak, and there is no change in the leaked jet. It’s like, it got freezed. I tried different boundary conditions, but seems no change. If I change the position of the leak to outside the box then it seems working (not sure). With some changes in the geometry of the box like adding some more outlets, propane jet develops little bit more, but after some time it stops leaking, same as before. What might be the possible reasons for this behavior?.

This jet image attached below is after 4 sec of simulation, after that it stopped over there, the simulation is running, also there is change in velocity field but no change in propane. And the integrated propane flux values (in paraview) also stuck at 4 sec. The other jet image is, when the leak is located outside the box. The box, boundaries are walls. And this box is in a big computational domain, the propane is supposed to leak inside the box and flow out after some time. But that’s not happening . The case files are attached below. I am stuck here, any suggestions ?

case files:
https://www.dropbox.com/s/cjf6oy2bznl5ni0/case.7z?dl=0

[Tobermory]

Sounds like the propane field is not being solved for - check the log file: is it actually doing any solver iterations for propane, or is it skipping it because the residual is already below its tolerance? I am not sure how OF normalises the residuals, but any changes in jet mass flux could be lost in a large domain when normalised by the whole domain mass. If so, you could try tightening up the propane solver residuals to force it to solve.

[nikhil108]

Originally Posted by Tobermory

Sounds like the propane field is not being solved for - check the log file: is it actually doing any solver iterations for propane, or is it skipping it because the residual is already below its tolerance? I am not sure how OF normalises the residuals, but any changes in jet mass flux could be lost in a large domain when normalised by the whole domain mass. If so, you could try tightening up the propane solver residuals to force it to solve.

Thanks for the suggestion, i will tighten up the residuals, and let you know if it works.

[nikhil108]

Hallo Tobermory,

yes you are right, the residulas are not tight enough. simple and easy to miss. Its working now. Thanks a lot

cheers,
Nikhil.

[Tobermory]

No problem - easy to miss. There are plenty of bear traps set by OF ... some of which I have had to climb out of already, including that one!