[AVRR]

Hi all, I am new in Fluent and my problem is the next.

I have to simulate a tank with walls and a presure outlet on the top. The problem is in the pressure outlet. I have to define a constant heat flux there. Is that possible? I have been talking with some people about this and many of them advice me to do it using UDF, but I do not know if I have to do it by DEFINE_SOURCE or DEFINE_PROFILE.

Anything you can tell me it will be helpful for me.

Thanks.

[farhanjaved]

Hi,

Make use of DEFINE_PROFILE.

Use this UDF. Put the below code in notepad and save file in .c format and then interpret in in ansys fluent.


#include "udf.h"
DEFINE_PROFILE(wallheatgenerate,thread,i)
{
real source = 0.001; // Put the value of heatflux that you would like to apply
face_t f;
begin_f_loop(f,thread)
F_PROFILE(f,thread,i) = source;
end_f_loop(f,thread)
}

[tatiana33]

Hello everybody,

Did you solve the problem with the pressure outlet? If so, what did you do?
I have a quick question, a UDF with DEFINE_PROFILE can be used in a pressure-outlet? or this is exclusive for a wall?

I will really appreciate your comments.

Thanks in advance!

[obscureed]

Hi Tatiana33,


From the page for DEFINE_PROFILE in the UDF/Customisation manual, you should be able to see that this UDF can work for some user-defined parameters of a pressure outlet. However, DEFINE_PROFILE only ever replaces a constant value that you might type into a parameter at a boundary condition. There is no input parameter for heat transfer rate at a pressure outlet, so you cannot define that via DEFINE_PROFILE UDF. The heat transfer that occurs at a pressure outlet is deduced from the flow and the conduction in the model, based on temperatures and temperature gradients next to the outlet.


I hope this helps. Good luck!
Ed

[tatiana33]

Ed,

Thank you so much for your comments.
Yes, you are right. As there is no input parameter for heat transfer rate at the pressure outlet B.C., I cannot define it through a UDF.
Do you know any other alternative to attach this condition (zero heat flux) at the outlet? A user-defined scalar would be useful?

Thanks in advance,

Tatiana

[obscureed]

Hi Tatiana33,

This is an unusual thing to ask for, and there is no simple way. In fact, it would be a strange thing to happen -- what do you actually want the model to do if hot fluid leaves at the outlet?

You could separate out some cells near the outlet, by making them into a different cell zone, and add a heat source to those cells. For example, if hot fluid is leaving at the outlet, you could note the excess energy that is escaping and define the heat source to return an equal amount to the cells. But then the fluid leaving at the outlet will be hotter, so the source term will increase. This is probably a positive feedback loop (that is, a bad idea -- it will get hotter and hotter until something breaks or some other effect kicks in). Maybe you could create a bigger source zone (for example, spreading out the re-injected heat), even encompassing the whole model -- sometimes this might help, but often it will not.

If you can explain the motivation for your question, it might be possible to see a better way.

Good luck,
Ed

[tatiana33]

Good morning,

Sorry for my late reply. I have been trying to improve my CFD model during the last month with not much success.
I am developing a 2-D depressurization model for pipelines transporting pure CO2. The model is transient/unsteady and my domain is a rectangle that represents the inside of the pipeline. I am using ANSYS Fluent. I am currently facing convergence issues when using the Peng-Robinson EoS. Apparently, the residual with the worst behavior is the radial velocity (y-velocity), which is particularly unstable close to the upper wall of the pipeline. I have tried different meshes with different y+ values, as well as different turbulence models. Unfortunately, I have not been able to run the model for more than 24 time steps (Deltat:10^-7s).
A summary of my model settings is next:
- Density based axysimmetric.
- Turbulence models: I have tried standard k-epsilon, realizable k-epsilon and k-omega SST.
- Energy: on
- Boundary conditions: inlet-wall, upper wall, axisymmetric, pressure-outlet.
- Implicit formulation.

Thanks in advance. I will really appreciate any input or suggestion on this matter.
Note: at the very beginning, I was blaming the positive sign of the heat flux at the outlet as the reason for the divergence. Then, I realized this sign was just a consequence of the difference in temperature between the inside and the outside.