chtmultiregion: pipe in a wall
Even if there are many threads about cht, I haven't fixed my problem.
I've got a wall (concrete) on half a circle (radius = 40m). There is a sheathed cable (steel) through it, on the length.
I have to blow hot air (313K) in the sheath, and study the heating of the wall (initial temperature = 263K).
I use the chtMultiRegionSimpleFoam solver.
I've made the mesh with Gambit, converted it using 'fluent3DMeshToFoam', and then splitting it by the three regions (air, cable, concrete), with 'splitMeshRegions -cellZones -overwrite'.
The attached picture shows a section of the domain
I've used the multiRegionHeater tutorial to set up my case: files needed, BC's, fvSchemes, fvSolutions...
1) on the diretMappedWalls, temperature is set as compressible::tubulentTemperatureCoupledBaffle. Wall functions are also 'compressible::' for k and epsilon. How may I do in order to run an incompressible case?
2) I gave correct values (I hope so :)) to rho, Cp and K for both concrete (2500; 880; 0.92) and steel (7867; 502.48; 16.27). However, the wall is heated by the fluid faster than the cable. I don't understand why.
I ran the case on Fluent, and I haven't this problem. I certainly mistaken in my BCs, but I don't know where.
3) I've seen the wallHeatFlux utility was only for combustion cases. Therefore I've tried the wallHeatFluxRho found on a thread. But I can't make it work:
The second picture shows the temperature at the inlet, after 5000it, running the case with the files in the attached archive (unfortunately without prompting for a log file).
Thanks in advance for your help
running in parallel
I've almost fixed my problem on the wallHeatFlux utility thanks to this thread
By now, I'd like to run my case using 3 CPUs. I've adapted the script given in the multiRegionHeater tutorial.
However, I've got this error:
I use the Scotch's method.
Has somebody got any idea?
new solver ?
In order to simplify, I'd like to know if exists a way to solve this kind of problem with incompressible and steady-state conditions.
I've already tried the chtMultiRegionFoam with the "ddt" set up as "steadyState", but with no success. And conversely it seems impossible to use the chtMultiRegionSimpleFoam with incompressible case.
Do I need to (test my weak C++ abilities and) code a new solver ?
Do you recall how did you fix 3).
The wallHeatFluxRho utility works only in fluid domain. In order to use it you have to set up a fake case with every data from your fluid (ie constant, system and times directories).
For more details, check this thread.
Hope my answer is clear enough.
Any progress or new advice?
I would like to evaluate heat flow or heat flux on walls in chtmultiregionfoam or chtmultiregionsimplefoam. Has there been anyone who has developed a way to do this, other than the approach mentioned. The only approach that i can find in these threads is to take the results and divide them into mutiple folder sets, and run the utitilty on each region as if it was an individual foam problem.
|All times are GMT -4. The time now is 20:18.|