[souria]

Hey everyone,

I'm simulating sediment transport using DPM. So, I tried an UDF to follow the particules. I built and hooked the udf, no problem. But, when I try to track the particules (means when I run the DPM) I got this message :

" DPM Iteration ....
MPI Application rank 0 exited before MPI_Finalize() with status -1
The fl process could not be started. "

Please, does anyone know where it comes from ?? !!

Thanks for your help

Souria

[Harry321]

Could you show your UDF?

[souria]

Originally Posted by Harry321

Could you show your UDF?

Here is the UDF :

/************************************************** ************************************/
/* UDF for computing cell co-ordinates */
/************************************************** ************************************/

#include "udf.h"
#include "sg.h"
#include "prop.h"
#include "dpm.h"
#include "surf.h"
#define WALL_ID 5

DEFINE_DPM_SCALAR_UPDATE(particle_coords, c, ct, initialize, p)
{
real A[ND_ND];
int n;
face_t f;
Thread *ft;
c_face_loop(c, ct, n)
{
f=C_FACE(c, ct,n);
ft=C_FACE_THREAD(c, ct, n);
if (NNULLP(ft))
{
if (THREAD_ID(ft)== WALL_ID)
{
p->user[0] = p->state.pos[0];
p->user[1] = p->state.pos[1];
p->user[2] = p->state.pos[2];

}
}

}
}



DEFINE_DPM_OUTPUT(Particle_coords_output, header, fp, p, thread, plane)
{
char name[10];
if (header)
{
if (NNULLP(thread))
par_fprintf(fp,"(%s %d)\n", THREAD_HEAD(thread)-> dpm_summary.sort_file_name,5);
else
par_fprintf(fp,"(%s %d)\n", plane->sort_file_name,5);

par_fprintf(fp,"(%10s %10s %10s)\n","X", "Y", "Z");
}
else
{
sprintf(name,"%s:%d", p->injection->name,p->part_id);
sprintf(fp,"(%10.6g %10.6g %10.6g)\n", p->user[0], p->user[1], p->user[2]);
}
}


thanks for your help

[Harry321]

I don't know:/

[souria]

Originally Posted by Harry321

I don't know:/

thanks anyway

[pakk]

I have my doubts with

sprintf(fp,"(%10.6g %10.6g %10.6g)\n", p->user[0], p->user[1], p->user[2]);

I think you'd rather use:

par_fprintf(fp,"(%10.6g %10.6g %10.6g)\n", p->user[0], p->user[1], p->user[2]);

For the rest, it might be that you have to add some compiler directives (#if !RP_HOST) to the DEFINE_DPM_OUTPUT macro, because the host does not know about some variables.

[souria]

Hi Pakk,

I change it but it gives the same error ! Excuse me, but what didi you mean by (#if !RP_HOST). PS. I'm not an expert in UDF, I just begun ...
Thanks for your help

Originally Posted by pakk

I have my doubts with

sprintf(fp,"(%10.6g %10.6g %10.6g)\n", p->user[0], p->user[1], p->user[2]);

I think you'd rather use:

par_fprintf(fp,"(%10.6g %10.6g %10.6g)\n", p->user[0], p->user[1], p->user[2]);

For the rest, it might be that you have to add some compiler directives (#if !RP_HOST) to the DEFINE_DPM_OUTPUT macro, because the host does not know about some variables.

[pakk]

The Fluent help has a "UDF manual", chapter 7.5.1 explains the compiler directives that I referred to.

[souria]

Originally Posted by pakk

The Fluent help has a "UDF manual", chapter 7.5.1 explains the compiler directives that I referred to.

Thank you, I will check this...

[Saidul]

I compiled the the same UDF with two warnings first time (Image attached) and no warning when I compiled it for second time. Then run the simulations without any error but can't find the output file . The fluent console shows that some particles are trapped.

Can anyone help regarding this problem.

Thanks in advance.
Saidul

[pakk]

Did you select "write to file" in the appropriate place, and remember which name you gave the file?

[Saidul]

Originally Posted by pakk

Did you select "write to file" in the appropriate place, and remember which name you gave the file?

Yes, I did everything with no results. I think the problem is with code. We have to check the print command. When I changed the print command
{
char name[100];
if (header)
{
if (NNULLP(thread))
cxprintf(fp,"(%s %d)\n", THREAD_HEAD(thread)-> dpm_summary.sort_file_name,8);
else
cxprintf(fp,"(%s %d)\n", plane->sort_file_name,8);

cxprintf(fp,"(%10s %10s %10s)\n","X", "Y", "Z");
}
else
{
sprintf(name,"%s:%d", p->injection->name,p->part_id);
cxprintf(fp,"(%10.6g %10.6g %10.6g)\n", p->user[0], p->user[1], p->user[2]);
}
}
and run the simulation, it showed a fatal error as " Error: received a fatal signal (Segmentation fault)" and stopped the simulation.

I don't know what is happening ....

saeed

[fishball]

From the error message, it seems the paraellel Fluent process could not be strated. Have you try to run in Serial Model?

Originally Posted by souria

Hey everyone,

I'm simulating sediment transport using DPM. So, I tried an UDF to follow the particules. I built and hooked the udf, no problem. But, when I try to track the particules (means when I run the DPM) I got this message :

" DPM Iteration ....
MPI Application rank 0 exited before MPI_Finalize() with status -1
The fl process could not be started. "

Please, does anyone know where it comes from ?? !!

Thanks for your help

Souria

[souria]

I fixed the problem.
It comes from parallel, so I parallelized the UDF. I just added :

#if !RP_NODE /*in the begining of the UDF*/

#endif /*in the end of the UDF*/


Hope it could help others...

Souria

Originally Posted by fishball

From the error message, it seems the paraellel Fluent process could not be strated. Have you try to run in Serial Model?

[fishball]

Originally Posted by souria

I fixed the problem.
It comes from parallel, so I parallelized the UDF. I just added :

#if !RP_NODE /*in the begining of the UDF*/

#endif /*in the end of the UDF*/


Hope it could help others...

Souria

Cool, Thanks

[Saidul]

Hello,
I compile the following UDF with a warning.
# Generating udf_names.c because of makefile dpm_modification.obj
udf_names.c
udf_names.c(9) : warning C4113: 'void (*)()' differs in parameter lists from 'void (*)(void)'
udf_names.c(10) : warning C4113: 'void (*)()' differs in parameter lists from 'void (*)(void)'
# Linking libudf.dll because of user_nt.udf udf_names.obj dpm_modification.obj
Microsoft (R) Incremental Linker Version 10.00.30319.01
Copyright (C) Microsoft Corporation. All rights reserved.

How can I fix it.....


/************************************************** ************************************/
/* UDF for computing cell co-ordinates */
/************************************************** ************************************/
#include "udf.h"
#include "sg.h"
#include "prop.h"
#include "dpm.h"
#include "surf.h"
#define WALL_ID 8 (line 9)


DEFINE_DPM_SCALAR_UPDATE(particle_coords, c, ct, initialize, p)
{
real A[ND_ND];
int n;
face_t f;
Thread *ft;
c_face_loop(c, ct, n)
{
ft=C_FACE_THREAD(c, ct, n);
if (THREAD_ID(ft)==WALL_ID)
{
p->user[0] = p->state.pos[0];
p->user[1] = p->state.pos[1];
p->user[2] = p->state.pos[2];
}
}
}
DEFINE_DPM_OUTPUT(Particle_coords_output, header, fp, p, thread, plane)
{
char name[1000];
if (header)
{
if (NNULLP(thread))
par_fprintf_head(fp,"(%s %d)\n", THREAD_HEAD(thread)->dpm_summary.sort_file_name,3);
else
par_fprintf_head(fp,"(%s %d)\n", plane->sort_file_name,3);
par_fprintf_head(fp,"\n %10s %10s %10s","X", "Y", "Z");
}
else
{
sprintf(name,"%s:%d", p->injection->name,p->part_id);
par_fprintf(fp,"%10.6g %10.6g %10.6g\n", p->user[0], p->user[1], p->user[2]);
}
}

[pakk]

Look more carefully: the warning that Fluent is giving, points to the file "udf_names.c", not to your file.

It is a file that Fluent makes while (preparing to) compile your code.

I get these warnings all the time, since migrating to Fluent 15.0. It is not a mistake in your code, but a mistake in Fluent code. You can ignore it since it does no harm; if you want to get rid of the code then contact the Fluent developers.

[Saidul]

Yes, thanks a lot. I got the point.

Originally Posted by pakk

Look more carefully: the warning that Fluent is giving, points to the file "udf_names.c", not to your file.

It is a file that Fluent makes while (preparing to) compile your code.

I get these warnings all the time, since migrating to Fluent 15.0. It is not a mistake in your code, but a mistake in Fluent code. You can ignore it since it does no harm; if you want to get rid of the code then contact the Fluent developers.