[kamu123]

Hello, i am panic in applying the profiles in to UDF_DEFINE SOURCE

I want to define source with "eg" depending on location x, y.

I have a text file with the information of "eg" as a series of data with x,y coordinates.

I failed to make a function like eg= f(x,y).

So I want to read directly the profile on the fluent. (profiles -> fluent)
Reading the profiles from the fluent was successful

But, desparately, i can not get any sense
applying "eg" profiles (uploaded in fluent) to UDF code


If you know some hint, please help me TT


"eg" profiles example

((eg 5 point)
(x
0.2
0.35
0.7
0.9
0.91)
(y
0.00147406
0.00147546
0.0014834
0.00148353
0.00139519)
(eg
0.8269
0.81492
0.88675
0.75573
0.83069)


UDF code shown below

=====================================
DEFINE_SOURCE(xmom_src,c,t,dS,eqn)
{
double src,x[ND_ND],source,base1,base2;

C_CENTROID(x,c,t);

{

src = C_MU_L(c,t)*eg/pk;
source = -src*C_U(c,t);
dS[eqn] = -src;
break;
}
return source;
}

[pakk]

If your scheme variable is just one number, you can read it in a UDF as:

RP_Get_Real("variable-name")

Here, you have a list of numbers, so this will not automatically work, but maybe this brings you a step closer to a solution.

[kamu123]

Thank you for your kind reply!!!

I am searching with your hint.

If you have any more information, please reply me for help. Thank you!!

[AlexanderZ]

you may define your table in udf and use interpolation to get heat source according to coordinate

this code may help you

#include "udf.h"
#include <stdlib.h>
#include <stdio.h>

#define N 10

float Temp[N] =
{
	300	301	302	303	304	305	306	307	308	309
}

float Heat[N] =
{
	100	201	302	403	504	605	706	807	908	1009
}

DEFINE_EXECUTE_ON_LOADING(report_version, libname)
{
	Set_User_Memory_Name(0,"My heatsource [W/m3]");
}

DEFINE_SOURCE(UDF_heat_source, c, t, dS, eqn)
{
	return C_UDMI(c,t,0);
}

DEFINE_ADJUST(adjust_heat_source, d)
{
	Thread *t;
	cell_t  c;
	real source,T;
	real j, Tpre, Tcur, Hpre, Hcur, A, B, M;
	
	thread_loop_c(t,d)	// loop over all cells in model
	{
		begin_c_loop(c,t)
		{
			M=N-1;
			T = C_T(c,t);
			if(T<Temp[0])
			{
				source = 0;
			}
			for(j=1;j<=M;j++)
			{
				if(T>=Temp[j-1] && T<Temp[j])
				{
					Tpre = Temp[j-1];
					Tcur = Temp[j];
					Hpre = Heat[j-1];
					Hcur = Heat[j];
					A = (Hcur - Hpre)/(Tcur - Tpre);
					B = Hcur - (Tcur * A);
					source = (A*T)+B;
					C_UDMI(c,t,0) = source;
				}
			}
		}
		end_c_loop (c,t)
	}

best regards

[kamu123]

Dear AlexanderZ

Thank you very much
i got the key information about interpolation from your reply

However, my data are actually about 120 data
and each "eg" depends on both x and y coordinates.
Even if i tried to get some fitting function with MATLAB, i failed.

So i want to apply whole 120 data points with profiles to fluent directly.

Is there any method connecting the profiles in UDF??

Thank you TTTT