[tomasfreitasrocha]

Hello everyone,

I am trying to implement a real-gas model for a CO2 ejector. My properties are interpolated from lookup tables, and this is working well for thermal conductivity, specific heat and viscosity. As for density, the simulation converges when I use an ideal-gas model; however, when I switch to my UDF, the simulation does not converge. Even if I write a UDF for density that replicates the ideal gas model, I have problems.

I would appreciate any help.

Thank you.

[AlexanderZ]

well, if you have problems with UDF, you should show it, at least
show the error log as well, show compilation log

[tomasfreitasrocha]

Yes, of course.

Here it is:

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

DEFINE_PROPERTY(property_density,c,t)
{
float density;

density = C_P(c,t)/(69.29*C_T(c,t));

return density;
}

The molecular weight of my fluid is 120 kg/kmol.

When I use this UDF, I get negative pressure values.

[pakk]

C_P is gauge pressure, you have to add the reference pressure.

[tomasfreitasrocha]

Thank you for your reply.

I have set my operating pressure at 0. Also, I get no compilation or loading errors! All the other UDFs work.

[pakk]

If you initialize it at e.g. 5 Pa, 300 K, don't iterate, and plot density, does it make sense?

(I don't know what is causing your problem, but this might give some information to help solve the problem.)

[tomasfreitasrocha]

Yes it does! Which adds to my confusion...