[B.Hamada]

Hi all.

how to express the exponential function in a UDF on FLUENT 15.0, I used the epression exp (value) as UDF following shows:
but it does not work.

#include "udf.h" 
DEFINE_SPECIFIC_HEAT(my_user_cp, T, Tref, h, yi) 
{ 
real cp;  
real Tmelt=299.6; 
real cps=1413; 
real cpl=1797; 
real lf=232000; 

if (T <= Tmelt) 
{ 
cp = cps+lf*exp(-0.5*(T-Tmelt)*(T-Tmelt)); 
} 
else /*(T > Tmelt)no need if statement, "else" is enough*/
{ 
cp = cpl+lf*exp(-0.5*(T-Tmelt)*(T-Tmelt)); 
} 
*h = cp*(T-Tref); 
return cp; 
}

but it does not work, so how I express it please
THank you.

[pakk]

"it does not work." is not very useful. To help you, more information would be better.

Does it compile?
Do you get errors?
Do you get warnings?
Do you get results at all?
What do you get for results, what do you expect?

[B.Hamada]

Originally Posted by pakk

"it does not work." is not very useful. To help you, more information would be better.

Does it compile?
Do you get errors?
Do you get warnings?
Do you get results at all?
What do you get for results, what do you expect?

Thank you for you help.

So.
I compiled the UDF.
No errors?
no warnings?
I have not achieved the desired results by that UDF
For the results, it did not work for variable specific heat.

[upeksa]

Originally Posted by B.Hamada

Thank you for you help.

So.
I compiled the UDF.
No errors?
no warnings?
I have not achieved the desired results by that UDF
For the results, it did not work for variable specific heat.

I knew you had no errors at compilation since I did that code and I checked it some weeks ago

You are not the first one to give this approach to simulate multiphase flow, check this post (also answered by me, by the way)

http://www.cfd-online.com/Forums/flu...em-cp-udf.html

I bet you have a similar problem as related in that post.

Cp variable is quite odd in Fluent: it has its particular UDF instead a "classic" DEFINE_PROPERTY macro. That because the solver integrates Cp and get enthalpy, although it is no that trivial indeed.

I think that your (only) solution is to use the multiphase models of Fluent (VOF, Euler, etc), and if you eventually need an UDF, you should use a DEFINE_SOURCE and DEFINE_MASS_TRANSFER. You will have many posts in this forum dealing with this

If you want to give a last try to your approach, you could drop the energy relaxation factor. It is set as 1 as default, try 0.8 or lower.

Cheers.

[B.Hamada]

Originally Posted by upeksa

I knew you had no errors at compilation since I did that code and I checked it some weeks ago

You are not the first one to give this approach to simulate multiphase flow, check this post (also answered by me, by the way)

http://www.cfd-online.com/Forums/flu...em-cp-udf.html

I bet you have a similar problem as related in that post.

Cp variable is quite odd in Fluent: it has its particular UDF instead a "classic" DEFINE_PROPERTY macro. That because the solver integrates Cp and get enthalpy, although it is no that trivial indeed.

I think that your (only) solution is to use the multiphase models of Fluent (VOF, Euler, etc), and if you eventually need an UDF, you should use a DEFINE_SOURCE and DEFINE_MASS_TRANSFER. You will have many posts in this forum dealing with this

If you want to give a last try to your approach, you could drop the energy relaxation factor. It is set as 1 as default, try 0.8 or lower.

Cheers.

Why are no errors in the UDI, and Fluent does not use it in calculations,