[sabuj]

I am a newbie in ANSYS FLUENT. So, a lot of things still make no sense to me. I have been browsing various sources for past 3/4 days for the issue but to no avail.

I have a very simple 3D geometry of a rectangular channel and I am trying to use a temperature dependent density function and use a udf. The geometry has only inlet, outlet and wall. The code:

#include "udf.h"

DEFINE_PROPERTY(density,c,t)
{
real rho_w, rho_f, rho_p;
real ci=0.01;
real temp = C_T(c,t);
rho_w = 838.466+1.4*temp-.003*temp*temp;
rho_p = 750./(0.001*(temp-319.15)+1.);
rho_f = ci*rho_p+(1-ci)*rho_w;
return rho_f;
}

Now although I can interpret this code, I keep getting 'segmentation fault' error.

Some questions have raised in my mind:

1) Is my code correct? does this thread pointer (t) represent the whole fluid domain?
2) Why do I keep getting this error?
3) If I replace those temperature dependent functions with some constants, the simulation runs and does not show the error message.

Someone please help me with the code.

[Bruno Machado]

Originally Posted by sabuj

I am a newbie in ANSYS FLUENT. So, a lot of things still make no sense to me. I have been browsing various sources for past 3/4 days for the issue but to no avail.

I have a very simple 3D geometry of a rectangular channel and I am trying to use a temperature dependent density function and use a udf. The geometry has only inlet, outlet and wall. The code:

#include "udf.h"

DEFINE_PROPERTY(density,c,t)
{
real rho_w, rho_f, rho_p;
real ci=0.01;
real temp = C_T(c,t);
rho_w = 838.466+1.4*temp-.003*temp*temp;
rho_p = 750./(0.001*(temp-319.15)+1.);
rho_f = ci*rho_p+(1-ci)*rho_w;
return rho_f;
}

Now although I can interpret this code, I keep getting 'segmentation fault' error.

Some questions have raised in my mind:

1) Is my code correct? does this thread pointer (t) represent the whole fluid domain?
2) Why do I keep getting this error?
3) If I replace those temperature dependent equations with some constant values, the simulation runs and does not show the error message.

Someone please help me with the code.

did you enable energy equation?

[sabuj]

Yes, Mr. Bruno. I enabled the energy equation. I am using FLUENT 15.0

[AHF]

Originally Posted by sabuj

I am a newbie in ANSYS FLUENT. So, a lot of things still make no sense to me. I have been browsing various sources for past 3/4 days for the issue but to no avail.

I have a very simple 3D geometry of a rectangular channel and I am trying to use a temperature dependent density function and use a udf. The geometry has only inlet, outlet and wall. The code:

#include "udf.h"

DEFINE_PROPERTY(density,c,t)
{
real rho_w, rho_f, rho_p;
real ci=0.01;
real temp = C_T(c,t);
rho_w = 838.466+1.4*temp-.003*temp*temp;
rho_p = 750./(0.001*(temp-319.15)+1.);
rho_f = ci*rho_p+(1-ci)*rho_w;
return rho_f;
}

Now although I can interpret this code, I keep getting 'segmentation fault' error.

Some questions have raised in my mind:

1) Is my code correct? does this thread pointer (t) represent the whole fluid domain?
2) Why do I keep getting this error?
3) If I replace those temperature dependent functions with some constants, the simulation runs and does not show the error message.

Someone please help me with the code.

some points :
use another name for temperature instead of "temp" , like "z" or any thing else

the path that you save your udf , should not have space in directory address .
for example
//c:new folder is wrong
//c:newfolder is right

and other thing is that why you use "." in this line :
rho_p = 750./(0.001*(temp-319.15)+1.);

after 750 and after 1 ?

[D.M]

Hi Sabuj,
i just checked your UDF and both interpret and compile worked well with it and nothing appeard as an error. (my geometry was just a fluid zone)
if u got fluid and solid in your geometry i think u should some how specify your fluid zone in your udf (i mean specify t) like:

if (FLUID_THREAD_P(t))
{
....
}