[User12]

Hello,

I have some experience using fluent, but I am quite new at UDFs. I am trying to set a volumetric heat source which is dependent on the mass flow rate. I think I know how to implement a UDF for a volumetric heat source, but my problem begins when I want to calculate the mass flow rate in each centroid and use this value in the definition of the heat source.

Can anybody give me a clue?

Thank you very much!!

[pakk]

With "mass flow rate" you mean density * velocity, right?

Do you have access to the Fluent manual? That gives info on how to access these variables in a UDF. (C_R(c,t) for density, and so on.)

Can you show how far you already got with the rest of the UDF? That gives a better idea on how much and what kind of help you need.

[User12]

Thanks for the answer.

Yes, it would proportional to velocity and density, but I think that in the case of defining it this way, the cell area should also be included. I explain myself.

My need is to define a heat source, q (W/m3), which is proportional to the flow rate, m (kg/s). Then, the easiest way to define it would be:

q (W/m3)= C (J/kg) * m (kg/s) * 1/Vcell (1/m3)

C is a constant and Vcell is the cell volume.

However, the only way I know to read the mass flow rate is through F_FLUX(f,t), which is a face variable. When I use this parameter in the macro DEFINE_SOURCE, i get a segmentation fault error, which I think is due to that this macro only deals with cell variables (not face variables).

Then, I tried to define the mass flow rate using density, velocity and area.

q (W/m3)=C (J/kg) * v (m/s) * rho (kg/m3) * A (m2)

I am stuck on this point, since I have two problems. 1) I don't know how to define the area. I think this would be again a face variable. 2) I can't see which is the area I need. I mean, which is the area that defines the amount of fluid inside the cell? Is it the area of a specific face? Is it the area normal to the velocity vector?

Here I paste the UDF I have been writing.



#include "udf.h"
#include "mem.h"

real U;
real V;
real VEL;
real RHO;

DEFINE_SOURCE(heat_source,thread,dS,eqn)
{
real source;

cell_t c;
Thread *t;

U=pow(C_U(c,t),2);
V=pow(C_V(c,t),2);
VEL=pow(U+V,0.5);
RHO=C_R(c,t);

source=20000*VEL*RHO*...; AREA IS MISSING


return source;
}

20000 is an arbitrary value for C. VEL is velocity magnitude, calculated through the squared radial and axial velocity components (axisymmetric simulation).

Regards

[pakk]

My need is to define a heat source, q (W/m3), which is proportional to the flow rate, m (kg/s).

From a physical point of view, I don't understand this. Flow rate is always through some surface, and heat source is local. I think there you already have your 'missing surface'.
Suppose in your model the flow rate would be 1 kg/s, and the corresponding heat source would be 1 W/m3. If you then would look at half of your model (such that it splits the flow 50/50), the flow rate would be 0.5 kg/s, so the heat source in that half model would be 0.5 W/m3, but that does not make any sense: the heat generation per unit volume should not depend on how big the volume is you are looking at.

I don't know what you need to do, so I don't know what to change it to. Below are two options that make sense from a physical point of view, but that does not mean they describe your situation.

• Total heat generated (W) can be proportional to the total mass in the system (kg). Then the heat source (W/m3) is proportional to the density (kg/m3).
  
• Total heat generated (W) can be proportional to the magnitude of the linear momentum of your fluid (kg m/s). Then the heat source (W/m3) is proportional to density multiplied by the magnitude of the velocity (kg/m2.s).

[User12]

Thanks again for answering.

From your reasoning I could change my point of view and use a different approach for my problem.