[yanghai]

hello everyone,

i want to set the temperature profile to a wall, the value is include the pressure of the wall, the equation is :T=100+ln(p). the below is my udf, in my case, the flow have pressure drop along the wall. so the temperature value of the wall is different, but the result is that the pressure drop is very small ,it is about 0.01Pa. so the temperature of the wall is constant. so where is the problem?Thanks anybody. thanks thanks thanks!!!


#include "udf.h"
#include <math.h>
real p_eva;
DEFINE_EXECUTE_AT_END(eva_pressure)
{
real p_op = RP_Get_Real("operating-pressure");
int ID=11;
Thread*thread;
Domain*domain;
face_t f;
domain=Get_Domain(1);
thread=Lookup_Thread(domain,ID);

begin_f_loop(f,thread)
{
p_eva = F_P(f,thread)+p_op;
}
end_f_loop(f,thread)
printf("p_eva is %g\n",p_eva);
fflush(stdout);
}

DEFINE_PROFILE(eva_temperature,t,i)
{
face_t f;

begin_f_loop(f,t)
{
F_PROFILE(f,t,i)=100+log(p_eva);
}
end_f_loop(f,t)
}

[pakk]

Your p_eva is the pressure of the last face that Fluent happens to visit... I don't think it is what you want.

If you want the temperature to be a function of the local pressure:

#include "udf.h"
#include <math.h>
DEFINE_PROFILE(eva_temperature,t,i)
 {
	  face_t f;

	  begin_f_loop(f,t)
	   {
		F_PROFILE(f,t,i)=100+log(F_P(f,t)+RP_Get_Real("operating-pressure"));
	   }
	  end_f_loop(f,t)
 }

[yanghai]

Originally Posted by pakk

Your p_eva is the pressure of the last face that Fluent happens to visit... I don't think it is what you want.

If you want the temperature to be a function of the local pressure:

#include "udf.h"
#include <math.h>
DEFINE_PROFILE(eva_temperature,t,i)
 {
	  face_t f;

	  begin_f_loop(f,t)
	   {
		F_PROFILE(f,t,i)=100+log(F_P(f,t)+RP_Get_Real("operating-pressure"));
	   }
	  end_f_loop(f,t)
 }

thanks for your reply, i have tried the way you said. it will have a Error: received a fatal signal (Segmentation fault). if i change the profile of temperature like :F_PROFILE(f,t,i)=100 that don't include the F_P(f,t), it is OK. i think that maybe the profile shouldn't have some variable just like F_P(f,t). is it true? do you know this ? thank you so much.

[pakk]

The problem is that F_P(c,t) does not exist yet when you initialize, because (apparently) the temperature is calculated before the pressure.

The quick solution: first initialize (with whatever pressure you want), then attach this UDF.

[yanghai]

Originally Posted by pakk

The problem is that F_P(c,t) does not exist yet when you initialize, because (apparently) the temperature is calculated before the pressure.

The quick solution: first initialize (with whatever pressure you want), then attach this UDF.

you mean click the Initialization with the pressure i want firstly,and then to compiled the udf?? i have tried this , it still have the same problem. do you mean that?
what's more, in fluent, shouldn't the pressure be solved before the temperature?
thanks for your reply, thank you very much.