[ASDzxc]

Hi there,
I'm doing a fluid-solid coupling heat transfer problem. Since I've just started using Fluent, I've created a very simple test case (no radiation) to figure out what Fluent can do (attachement):
1.png
Based on the description of udf DEFINE_HEAT_FLUX, it modifies the diffusive heat flux coefficients using equation:
qid = cid[0] + cid[1]*C_T(c0,t0) - cid[2]*F_T(f,t) - cid[3]*pow(F_T(f,t),4)
so I wrote the udf:
DEFINE_HEAT_FLUX(heat_flux, f, t, c0, t0, cid, cir)
{
real h=2000.;
cid[1] = h;
cid[2] = h;
cid[0] = 0.;
cid[3] = 0.;
C_UDMI(c0,t0,1)=h;
}
Then I got questions about the result, it seems really strange for me.
1. I tried to visualize C_UDMI(c0,t0,1) in the result. Instead of showing 2000 in some near wall cells, the max of the contour is somewhere around 1500 or 1000, not 2000, why?
2. What exactly this udf modified? In the figure that I uploaded,
h(T_f-T_interface)=-k_wall*(T_interface-T_wall)/d2
where h is the coefficient assigned using udf, k_wall is the wall conductivity.
is this correct?
Thanks for your help.

[vinerm]

This function could have many uses, such as, modeling a phenomenon that is not being resolved. The objective is to modify the relation between heat flux and temperature. And it is not supposed to be used for CHT problems. If you expect the system to behave as it should for a solid-fluid pair, then do not use DEFINE_HEAT_FLUX function.

[ASDzxc]

Originally Posted by vinerm

This function could have many uses, such as, modeling a phenomenon that is not being resolved. The objective is to modify the relation between heat flux and temperature. And it is not supposed to be used for CHT problems. If you expect the system to behave as it should for a solid-fluid pair, then do not use DEFINE_HEAT_FLUX function.

Hi Vinerm,
Thanks for you reply.
Could you please give me an example of "a phenomenon that is not being resolved"? and how does it "modify the relation between heat flux and temperature"?
I tried to model the system as a normal solid-fluid pair (hot fluid passes through a cylinder, and we would like to predict the temperature on the outer cylinder wall surface), the prediction does not match the experiments.
regards,

[vinerm]

To model normal solid-fluid pair, just do the simulation without any UDF. Ensure that the solid has at least four cells across its thickness and there are not pyramids at the fluid-solid interface.

Unresolved phenomena could be of any nature, say, thermal energy transfer by ions or by magnetic lines of force that are not being resolved, but are affecting the thermal distribution for a specified flux. Then, user can modify the coefficients in such a manner so as to include these extra effects.

[ASDzxc]

Originally Posted by vinerm

To model normal solid-fluid pair, just do the simulation without any UDF. Ensure that the solid has at least four cells across its thickness and there are not pyramids at the fluid-solid interface.

Unresolved phenomena could be of any nature, say, thermal energy transfer by ions or by magnetic lines of force that are not being resolved, but are affecting the thermal distribution for a specified flux. Then, user can modify the coefficients in such a manner so as to include these extra effects.

Thanks Vinerm, your reply is really helpful to me.
Now I understand DEFINE_HEAT_FLUX seems not be a good option for normal CHT problems.
In fact, modelling this problem (hot fluid passes through a cylinder, and predict the temperature profile on the outer cylinder wall surface as a function on time) is the first step of my project. It can be regarded as a first step to get familiar with fluent. The final objective is to change that pure fluid region to a porous medium region (the other stuff rest the same). I know in Fluent, we can model the thermal behaviour inside a porous medium, but when it comes to coupling the porous media with a solid wall. Does it the same as fluid-solid CHT problem? Do you think DEFINE_HEAT_FLUX can be useful in this situation?

[vinerm]

I don't have your complete project description, however, looking at what you described, you will not require DEFINE_HEAT_FLUX. Fluent can directly work with heat transfer between fluid medium and its solid part in porous region as well as between fluid-porous-solid domains.

[ASDzxc]

Originally Posted by vinerm

I don't have your complete project description, however, looking at what you described, you will not require DEFINE_HEAT_FLUX. Fluent can directly work with heat transfer between fluid medium and its solid part in porous region as well as between fluid-porous-solid domains.

Thank you for your reply!

[sachin tom]

I am trying to implement the RPI wall boiling model as a user-defined function in Ansys Fluent. Is the Define_HEAT_FLUX macro need to be used for it? How to give the liquid side coefficients and phase change coefficients separately with this macro.

Any help on this will be appreciable.