[lukasf]

Hi,


now I want to post process Fluent data acquired by data sampling for time statistics with ParaView.


When I read the case into Fluent/CFX Post everything is fine.


However, using ParaView I have the problem that the Mean and RMS values do not make sense. E.g. The Temperature ranges from 200 to 400K in my simulation. The available cell arrays are e.g.
Temperature, Temperature_M1, Temperature_M2, Temperature_Mean, and Temperature_RMS.


I think that M1 and M2 are the fields needed for the second order time integration. Those values range from 200 to 400K which is good.
Looking at the Temperatur_Mean scalars there are values around 8-9 . The Temperature_RMS has a value of 1000-4000. One would expect a Temperature_mean in the range from 200 to 400K and RMS values of maybe 0-50K. Just from looking at the Mean field it looks like an average field but the values are somehow messed up.



Does anyone know how to handle this?

[lukasf]

I ended up using the ensight-gold format as well.


My summary of what I am doing:


I have a script called Journal_Ensight_Script which starts fluent in the shell and reads a journal (.jou) file. In this .jou file there are the command to export the data in ensight-gold format. Afterwards the file is renamed to .case. Note, that all " symbols have to vanish in the .case file so that Paraview has no error:



Content of the script:

# run Fluent
/home/software/ansys_inc/v201/fluent/bin/fluent 3ddp -g -i Journal_Ensight.jou
mv encase_files/encase_format.encas encase_files/encase_format.case

Content of the .jou file:

file read-case /home/C12_FS/C12_FS.cas
file read-data /home/C12_FS/C12_FS_result-2-20000.dat

file/export/ensight-gold
/home/C12_FS/encase_files/encase_format

mean-x-velocity
mean-y-velocity
mean-z-velocity
rmse-x-velocity
rmse-y-velocity
rmse-z-velocity
mean-temperature
rmse-temperature
q
yes



yes

exit ok

All enters are importat because those are answers in the TUI command line of Fluent. Moreover, adjust the fields in which you are interested in! "q" tells Fluent that I do not want to enter any more fields. It has nothing to do with the q-criterion.

[aroma]

Originally Posted by Moerten

Hey together,


by starting my masters thesis in the last few weeks I ran into the same Problems between Fluent and Paraview, utilizing the EnSight data format, that you guy's mentioned previously.


Long story short, I came up with a sulution. At least for anyone that uses the Fluent 2020 R2 release:


Your Ensight .encas-file looks after the export somewhat like the example beneath. Paraview seems to have Problems with the ""-marks, which hold certain filenames, and doesn't support the whole SCRIPTS-section of the file. If you remove the problematic bits of the file (marked in red) and save it as .case-file, you should be good to go importing your solution data in Paraview. However, be careful to no leave a blank line at the end of the textfile. This, for some reason, sends Paraview in a inf-loop while reading the data.

FORMAT
type:  ensight gold
GEOMETRY
model:  2  2   "generic_meshfile.geo"
VARIABLE
scalar per element:  1  1  total_pressure  "generic_solution_variable.scl1"
vector per element:  1  1  velocity  "velocity_solution.vel"
TIME
time set: 1 Model
number of steps: 2
time values:  2.59587e-09  1.86688e-08 
time set: 2 Model
number of steps: 1
time values: 2.59587e-09
FILE
file set: 1
number of steps:2
file set: 2
number of steps: 1
SCRIPTS
metadata: "generic_metadata.xml"

Best regards
Moerten

Worked perfectly for me. Thanks

[lukasf]

Originally Posted by lukasf

I ended up using the ensight-gold format as well.


My summary of what I am doing:


I have a script called Journal_Ensight_Script which starts fluent in the shell and reads a journal (.jou) file. In this .jou file there are the command to export the data in ensight-gold format. Afterwards the file is renamed to .case. Note, that all " symbols have to vanish in the .case file so that Paraview has no error:



Content of the script:

# run Fluent
/home/software/ansys_inc/v201/fluent/bin/fluent 3ddp -g -i Journal_Ensight.jou
mv encase_files/encase_format.encas encase_files/encase_format.case

Content of the .jou file:

file read-case /home/C12_FS/C12_FS.cas
file read-data /home/C12_FS/C12_FS_result-2-20000.dat

file/export/ensight-gold
/home/C12_FS/encase_files/encase_format

mean-x-velocity
mean-y-velocity
mean-z-velocity
rmse-x-velocity
rmse-y-velocity
rmse-z-velocity
mean-temperature
rmse-temperature
q
yes



yes

exit ok

All enters are importat because those are answers in the TUI command line of Fluent. Moreover, adjust the fields in which you are interested in! "q" tells Fluent that I do not want to enter any more fields. It has nothing to do with the q-criterion.

I have adjusted the script to automatically create the postprocessing folder, run the postprocessing, rename the file extension from .cas to .case and to remove the symbol (") from the case file so that ParaView can work with the file.

mkdir encase_files
# run Fluent
/home/software/ansys_inc/v211/fluent/bin/fluent 3ddp -g -i Journal_Ensight_UNSTEADY.jou
# rename file so that paraview knows the file extension
mv encase_files/encase_format.encas encase_files/encase_format.case
# delete symbols from encase file which prevent paraview to read the file
sed 's/"/ /' encase_files/encase_format.case > encase_files/remove_strings_encase
sed 's/"/ /' encase_files/remove_strings_encase > encase_files/encase_format.case