[woody85]

In my case i want to model the porosity as a function of temperature. my code looks like this:

#include "udf.h"

DEFINE_PROFILE{dir1, t, i}
 {
	cell_t c;
	begin_c_loop(c,t)

		real temp = C_T(c,t);
	    
		if (temp > 1300.)
			C_PROFILE(c,t,i) = 13.7 - (temp - 1300) * 0.0015 + pow(temp - 1300,2.) *0.0000075;
		else if (temp > 1500.)
			C_PROFILE(c,t,i) = 0.01;
		else
			C_PROFILE(c,t,i) = 13.7;
		
	end_c_loop(c,t)
 }
 
 DEFINE_PROFILE{dir2, t, i}
 {
	cell_t c;
	begin_c_loop(c,t)

		real temp = C_T(c,t);
	    
		if (temp > 1300.)
			C_PROFILE(c,t,i) = 13.7 - (temp - 1300) * 0.0015 + pow((temp - 1300),2.0) *0.0000075;
		else if (temp > 1500.)
			C_PROFILE(c,t,i) = 0.01;
		else
			C_PROFILE(c,t,i) = 13.7;
		
	end_c_loop(c,t)
 }

But i always get the following error:

New_UDF.c.8748.5.c: line 8: parse error.

which is a little bit confusing.

thx for your help

[macfly]

- the accolades should be replaced with parentheses
- you should not use "temp" because it shadows previous definition

Originally Posted by woody85

DEFINE_PROFILE{dir1, t, i}
{
cell_t c;
begin_c_loop(c,t)

real temp = C_T(c,t);

if (temp > 1300.)
C_PROFILE(c,t,i) = 13.7 - (temp - 1300) * 0.0015 + pow(temp - 1300,2.) *0.0000075;
else if (temp > 1500.)
C_PROFILE(c,t,i) = 0.01;
else
C_PROFILE(c,t,i) = 13.7;

end_c_loop(c,t)
}

DEFINE_PROFILE{dir2, t, i}
{
cell_t c;
begin_c_loop(c,t)

real temp = C_T(c,t);

if (temp > 1300.)
C_PROFILE(c,t,i) = 13.7 - (temp - 1300) * 0.0015 + pow((temp - 1300),2.0) *0.0000075;
else if (temp > 1500.)
C_PROFILE(c,t,i) = 0.01;
else
C_PROFILE(c,t,i) = 13.7;

end_c_loop(c,t)
}

[woody85]

thx for your help - now it seems to work !

[woody85]

Is this formulation of the UDF also valid for transient simulations ?

[macfly]

Yes it is valid.

[woody85]

Originally Posted by macfly

Yes it is valid.

Thx, i only have one problem left, i want to model more or less a simplified melting process of a porous media. The lines i posted are only a part of my code.

So far the melting is built up correctly, but when for some reason the temperatur in my cell decreases, the whole thing is reversible, which in my case is not realistic, as the molten liquid, flows to the bottom.

I want to ensure, that once my cell reached a certain temperature, the porosity cannot decrease again.

[macfly]

I don't understand why you code dir1 and dir2 since the equation/conditions are the same. Code only one DEFINE_PROFILE and load the same property for both directions.

Porosity should be between 0 and 1, values like 13.7 and the jump at 1500 look weird.

Regarding the reversibility problem, you will have to store previous timestep values in a UDMI and compare them with current timestep values. Then write some if condition that keeps the porosity constant if T current timestep < T previous timestep. See this post for an example: http://www.cfd-online.com/Forums/flu...tml#post372530

[woody85]

First of all thanks for your help.

As i mentioned this is only a part of my code, i don´t wanted to upload the whole thing here because it´s to long. In the code i also modelled the pressure drop. Actually i copied the line for the inertial resistance. This is not identical in all directions in my case at the beginning. Therefore i had to use two values. Nevertheless the code for the porosity looks exactly the same onle the values change, they have to be between 0 and 1, as allready mentoined by you.

Thanks for the help regarding the reversibility problem. I will have a closer look on this.

[woody85]

Dear MacFly, thx for your help.

The condition you suggested will not work out:

"Then write some if condition that keeps the porosity constant if T current timestep < T previous timestep"

because this condition would only work for 1 timestep. if i have temperature decrease for more timesteps, the condition would only look for the last timestep and can´t know that my material is allready melted e.g. 10 timesteps ago.

Therefor i suggested to take the porosity itself as condition. If the porisity 1 timestep ahead =1, then it should also be 1 in the next timestep. unfortunatly this doesnt work out so well yet, mainly to my lack of udf-experience.

my actual version of the UDF now looks like this:

 DEFINE_PROFILE(por, t, i)
 {
	cell_t c;
	begin_c_loop(c,t)

		real val_temp = C_T(c,t);
	    
		if (val_temp > 1500.)
			{
				C_PROFILE(c,t,i) = 1;    
				C_UDMI(c,t,1) = 1;    
			}
		else if (val_temp > 1300. && val_temp < 1500.)
			{
				C_PROFILE(c,t,i) = val_temp*0.000425 + 0.3575;
				C_UDMI(c,t,1) = val_temp*0.000425 + 0.3575;
			}
		else
			{
				C_PROFILE(c,t,i) = 0.91;
				C_UDMI(c,t,1) = 0.91;
			}	
			
		C_UDMI(c,t,0) = C_UDMI(c,t,1) /*not working */

	end_c_loop(c,t)
 }

The idea was to store the actual value of the porosity in den UDMI, and at the end i want to store the actual value of the UDMI, so that i can use the values in the next timestep (not working yet). Here i made an mistake.

The next steps would be to make an if condition, to check if the porosity in the timestep before was 1 or not, if the value was 1, then overwrite the actual value of the cell with the value 1.

[ghost82]

Originally Posted by woody85

if i have temperature decrease for more timesteps, the condition would only look for the last timestep and can´t know that my material is allready melted e.g. 10 timesteps ago.

If you need to know if material was already melted what about using a boolean variable?