hello, i was puzzled by a problem in Fluent5.5. my question is "can density be defined as function of pressure in UDF ?" in fluent's help document,say" Some of the properties you can customize are density (as a function of temperature only)"why?

Fluid density is dependent on Temperature..... Ideal or incompressive ideal gas of material panel will follow pressure dependent density. PV=nRT..... If you have density as function of pressure, it will be better to make UDF. This can be selected at material panel too.

Thanks,

i have made udf that density was defined as the function of pressure.but when i import the udf,a lot of errors happen.so i think the error maybe results from the inner arithmetic. the follow is the udf. #include "udf.h"

DEFINE_PROPERTY(cell_density,cell,thread) {

real density;

real pressure=C_P(cell,thread);

if(pressure>=6.0e+5)

density=98.326+3.0013e-5*pressure;

else

density=-0.74708+5.54545e-5*pressure+3.25103e-10*pressure*pressure;

return density;}

I testes your UDF, and confirmed that your UDF itself was no problem. Please ask your Fluent distributor...

Do you get negative pressure values, if you calculate with roh=const? If yes, you have to limit the function to only positive values by adding something like:

if (pressure > 0.001)

...;

else

density=const_value;

By the way, how much do you expect the density to vary?

Thank you for your advice. My CFD simulation is the helium ejector. The helium refrigerant flows through the ejector at almost constant temperature about 6K. The inlet high pressure is about 18atm, the inlet low pressure is about 1atm or below. The out back pressure is about 3atm. The density change is about 4.23932kg/m^3(6K¡¢0.5atm), and 148.236kg/m^3(6K¡¢18atm).So the relation between density and pressure is important, but the relation between density and temperature can be ignored.