Learning Forum

[soloviev]

Hello,

I was running a VOF to DPM model on Fluent 2020 R2 and writing a lump transcript file to track when particles were generated. This was going well until I had to transfer and continue the simulation on either 2021 R2 or 2022 R1 (due to an HPC OS update). Now when the file continues writing the time for each new particle line is 0, and it continues to be 0 as long as it continues to run. If I switch back to 2020R2 on a smaller workstation, the file continues with time normally. I tried starting a new lump transcript file (new name) in the new Fluent version and it just has a time of 0 for the entire new file. Is there a possible solution for this?

Thanks,

Alex

[Karthik Remella]

Hello Have you tried setting up your case from scratch in the latest version? If so, do you see this issue there?
Karthik

[soloviev]

Hello Karthik I have not tested this. I can try testing this, but the model we were using these files for had already been running for 2 weeks prior to this issue, so we were hoping not to have to re-run it.
I will see if the issue persists though if I start the case fresh in 2022R1.

Thanks Alex

[soloviev]

Hello again Karthik I started the model from the beginning and the time in the file is still 0. I guess maybe this is an issue with the 2022 R1 release.
Thanks Alex

[Amine Ben Hadj Ali]

I tested that and a transcript file has been written. Are you able to see from Fluent transcript that particles have been transitioned from VOF World to the Discrete World?

[Amine Ben Hadj Ali]

Are you using 22R1 or 21R1? (title bit misleading).

[soloviev]

Yes, I see the transcript file and it has lines or the particles being transitioned, but the time is still 0 for all particles when I do this in 22R1.

Thanks Alex

[soloviev]

Hello again I am trying to write a UDF to fix this issue (export particle variable data via a UDF text file). I looked through the documentation for UDFs, specifically for DPM, and didn't see wha the macro would be to specify or write particle generation time (when particle is transitioned from VOF to DPM). Is this a possibility?
Thanks Alex

[Amine Ben Hadj Ali]

Back to the problem: Is it possible to add more details? What is "zero" Can you add screenshot of the Lump Transcript file? I will need to run a test again.
In the DEFINE_DPM_OUTPUT Documentation you might see the Macro for the time of birth.

[Amine Ben Hadj Ali]

From quick verification: flow-time is written but time is zero: particle residence time is zero as the particles are born at that flow time.

[soloviev]

Hi Dr Amine I've attached a screenshot of the file below. The residence time (column 't' is not 0, but it is the same for all particles in the file. The 'time' column (the time of generation) is 0.
I will look at DEFINE_DPM_OUTPUT macros, thank you!

Alex

[soloviev]

Hello again
I managed to write a UDF for exporting the particle info, including the generation time (time of birth):
/*** DPM UDF to write particle positions to file ******/
#include "udf.h"
#include "dpm.h"
#include "math.h"
DEFINE_DPM_OUTPUT(particle_info,header,fp, p,t,plane)
{
if(header)
{
par_fprintf_head(fp,"X-Position Y-Position Z-Position X-Velocity Y-Velocity Z-Velocity Time Generation Time\n ");
}
if(NULLP(p))
return;
par_fprintf(fp,"%d %" int64_fmt " %e %e %e %e %e %e %e %e \n" P_INJ_ID(P_INJECTION(p)),p->part_id,P_POS(p)[0],P_POS(p)[1],P_POS(p)[2],P_VEL(p)[0],P_VEL(p)[1],P_VEL(p)[2],P_TIME(p),p->time_of_birth);

}

The UDF works without errors, compiles fine, and writes the file with the proper headers and information. The issue I'm having is that there are way too many lines in the particle file. The Fluent console says there are 5319 particles, meanwhile the written particle file has 928911 lines of particle info. Is there a way to fix this?

Thanks Alex

[Amine Ben Hadj Ali]

The output you are getting is correct and that is similar to what I get. The column with "time" is zero the one with "flow time" is not zero. That "time" does not not correspond to time of birth by the way. The time of birth in the transcript will correspond to the column with flow time!

[soloviev]

Thank you! I compared and the old files had the exact same data in flow time and time columns. When I loaded old case/data files into 2022 R1 it works with the time continuing to write in the flow-time column.

[Amine Ben Hadj Ali]

Yes we corrected that behavior as it is odd to have particle residence time equal to its birth time!