compressibleInterFoam incl. viscous dissipation heating
I am working on a modification of compressibleInterFoam, more specifically, I am trying to implement the term for viscous dissipation into TEqn.H in order to take viscous dissipation heating of highly viscous fluids (i.e. oil) into account. I managed to extend the source code of the TEqn.H file according to this thread: http://www.cfd-online.com/Forums/ope...n-heating.html
The results so far:
T_aw = 305.4 K | standard compressibleInterFoam
T_aw = 328.7 K | modified compressibleInterFoam
Somehow, it seems to be working, however, it's strange that T_aw of the standard solver is almost 3 K higher than T_jet.
Theoretically, it should be the same, right? The wall and outlet patches all have zeroGradient BCs.
Another thing which makes me second-guessing my above mentioned implementation of the viscous terms, is a CFX simulation of the same problem.
With viscous dissipation heating, it yields a adiabatic wall temperature of T_aw = 305.9 K which is significantly lower than the result of the modified solver.
My question is quite simple: Did I miss something during the implementation of the viscous term?
I tend to think that my definition of muEff is not completely right.
AFAIK, it's muEff = mu + mut, which should hold since I'm using the kOmegaSST model.
But I think, I made a mistake regarding the multiphase character of the case.
If anyone has experience with this, your help will be greatly appreciated :)
Did you get any good result ? Because i need to do the same.
Did you take a look at this: http://www.cfd-online.com/Forums/ope...tml#post428926
By the way, you use muEff from turbulence->muEff(), but this is multiphase:
this should be something like mixture.muEff instead ?
I'm not able to access my final version of the modified solver or my data, however, the results were pretty good compared to the CFX solution I used as reference. Maybe, with a bit of fine tuning, they would be extremely good ;)
Let get back to you as soon as I can access my stuff again (I am moving at the moment :D)
Did you find your final version of your modified solver ?
sorry for the delay, setting up everything took a bit longer than expected :)
In fact, I used the code from my first post and did a re-run of the exact case. The only thing I did was implementing a ramp for my mass flow rate.
The modified solver now yields a T_aw = 306.7K, compared to T_aw = 305.9K from the CFX run. For the time being, it's sufficiently exact, but your point regarding the muEff is completely valid. I'm gonna check up on that, when I find the time ;)
|All times are GMT -4. The time now is 10:58.|