[juchan]

Hi,

I am a beginner with ANSYS.
i just want to solve simple rectangular shape(5mm*5mm) 2d model.
Water_liq is located from the bottom of the rectangle to 2mm height and water_vap is from 2mm to 5mm.
Bottom wall is heated so evaporation occurs

I am using a UDF function to define saturation temperature according to C_VOF ( for evaporation on the interface, not on the heated wall ).
when I try to do initialization, following error message shown and fluent program automatically shut down:


// 999999: mpt_accept: error: accept failed: Invalid argument
// 999999: mpt_accept: error: accept failedMPI Application rank 0 exited before MPI_Finalize() with status 2: Invalid argument

following is my udf code;;
------------------------------------------
#include "udf.h"

DEFINE_PROPERTY(saturation_temp,c,t)
{
real sat_t;

if (C_VOF(c,t) == 1||C_VOF(c,t)== 0)
sat_t = 373;

else
sat_t = 353;

return sat_t;
}

---------------------
using above code, i just want to give cells different saturation temperature according to the C_VOF value of each cell.

if udf works well, only interface cell's saturation temp. would be 353
But, i can't solve a problem because i do not have any idea to resolve an error.....

Please give me some tips why it doesn't work.
Thank you

[AlexanderZ]

code seems to be correct, check your model.
Decrease timestep size

[juchan]

To. AlexanderZ
Thanks for your opinion
i tried but it didn't work....

maybe there would be other solution,,,

[AlexanderZ]

if you get this error right after initialization, may be the problem is in C_VOF macro, it may not be defined before iterations, so you may try to pre-define saturation temperature for that case

#include "udf.h"

DEFINE_PROPERTY(saturation_temp,c,t)
{
real sat_t;
sat_t = 353;
if (C_VOF(c,t) == 1||C_VOF(c,t)== 0)
sat_t = 373;

else
sat_t = 353;

return sat_t;
}

[juchan]

Thanks! AlexanderZ i understand what you advised

maybe i should find the other way to resolve error because fluent program still closed accidently in spite of your advice....

[AlexanderZ]

try constant value, as a first step

#include "udf.h"

DEFINE_PROPERTY(saturation_temp,c,t)
{
real sat_t;
sat_t = 353;
return sat_t;
}

if it doesn't work, it means the problem is in your settings

[juchan]

Thanks AlexanderZ

i try to have a solution after your help but i couldn't....
when i take constant saturation_temp value, there is no error
i think just c_vof makes error

if you know any other way to settle this problem, please help one more....
then it will be very important information to me

Thanks again!!

[AlexanderZ]

you've mentioned

for evaporation on the interface, not on the heated wall

how do you apply your UDF to interface?
is it the only code you are using?

show latest version of code you are using, explain all settings, show error log

[juchan]

My geometry is simple 5mm*5mm square (60*60 no-biased mesh)

using patch From 0mm to 1.7mm height water-liquid(secondary phase)
from 1.7 to top water - vapor (primary phase)

i could not change my code because there is no way to find the reason of error.

#include "udf.h"
#include "sg_mphase.h"

DEFINE_PROPERTY(saturation_temp,c,t)
{
real sat_t;
sat_t = 353;

if ( C_VOF(c,t) >0 && C_VOF(c,t) < 1 )
sat_t = 373;
else
sat_t = 353;

return sat_t;

}

This code is what i applied for changing the saturation temperature of interface
(there is no more code that i write. above one is all)

general - solver - pressure-based, absolute, transient, planar, gravity Y(-9.81)
model - vof -explicit, courant number 0.25 / energy on / viscous - laminar

i use lee model and i want to apply udf code that i posted for changing the saturation temperature of the interface

all zone's initial temperature is 350k

boundary condition is like this
bottom line is heated wall(360k)
vertical lines(left,right) are wall(no heated)
top line is out flow

when i press the 'initialize' button after apply udf code,
i could see the message like this then fluent ends all of a sudden

999999: mpt_accept: errorrr: accept failed: No errror
MPI Application rank 0 exited before MPI_Finalize() with status 2
The f1 process could not be started

i will wait for your answer
Thanks AlexanderZ!

[AlexanderZ]

try this

#include "udf.h"
#include "sg_mphase.h"

DEFINE_PROPERTY(saturation_temp,c,t)
{
real sat_t;
Thread *ts;
ts = THREAD_SUB_THREAD(t,1);
sat_t = 353;

if ( C_VOF(c,ts) >0 && C_VOF(c,ts) < 1 )
sat_t = 373;
else
sat_t = 353;

return sat_t;

}

for more information:
Ansys fluent customization manual -> THREAD_SUB_THREAD

I'm not sure whether it help or not

[juchan]

AlexanderZ!! Thank you SSSSSo much
Error that makes me crazy for a huge long time disappears because of the modifications you've proposed

finally i get the great inner peach!!!

although bubbles generated from the heated bottom surface makes me a little confused, now i got the license to find the answer(udf) because there is no more errorr!!

i think maybe changing this part is the key

if ( C_VOF(c,ts) >0 && C_VOF(c,ts) < 1 )
sat_t = 353;
else
sat_t = 373;

Thanks one more time!!