CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Starting transient calculation from steady solution via Journal-file (https://www.cfd-online.com/Forums/fluent/239159-starting-transient-calculation-steady-solution-via-journal-file.html)

farbflash_ October 23, 2021 06:59

Starting transient calculation from steady solution via Journal-file
 
Hello,

I'm trying to run a transient simulation with a steady start-solution on our Workmachine. Therefor I have to start the simulation via a Journalfile:

Quote:

; initialising the system
/solve/init/init

;import case- and data-file
/file/read-case trans-setting.cas.h5
/file/read-data steady-solution.dat.h5
/file/auto-save/data-frequency 1

;set calc-settings
/solve/set/time-step 0.00025
/solve/dual-time-iterate 1000 20

;save the results
/file/write-case-data trans-calc

;exit fluent
exit
The problem is, that if i look into the evolution of the single timesteps, the calculation clearly doesn't start from the steady-solution. But If I do it manually on my own computer (importing the stady-data and the trans-case) the simulation is doing just fine. Maybe there is something wrong with my journal-script? I hope somebody can help me out.
Many thanks in advance!

Lukas

LuckyTran October 23, 2021 13:02

The first line confuses me: /solve/init/init
You are initializing a case that hasn't been read in yet. Even though it makes no sense and will throw an error message, it won't prevent you from going further.

The rest of the journal seems okay but it's quite possible that there are prompts that might come up that you do not answer. What does your output look like? You might want to look in there to see what Fluent is actually doing.

By manually testing it on your own computer, I presume you are using the GUI to do this. You should test the journal "manually" on your own computer by typing in the text in the journal file into the TUI window.

farbflash_ October 24, 2021 04:07

Quote:

Originally Posted by LuckyTran (Post 814899)
The first line confuses me: /solve/init/init
You are initializing a case that hasn't been read in yet. Even though it makes no sense and will throw an error message, it won't prevent you from going further.


Sorry for my question, but where do I should put this line instead? After loading the cas-file or can I just keep it away?


Quote:

Originally Posted by LuckyTran (Post 814899)
The rest of the journal seems okay but it's quite possible that there are prompts that might come up that you do not answer. What does your output look like? You might want to look in there to see what Fluent is actually doing.

By manually testing it on your own computer, I presume you are using the GUI to do this. You should test the journal "manually" on your own computer by typing in the text in the journal file into the TUI window.

I will give it a try, thank you for your reply so far! :-)
I will report back if I figured out the problem.

LuckyTran October 24, 2021 05:00

Well if you want to solve from a previous solution then you should never initialize at all. Like I said, that line just plain confuses me. It shouldn't be anywhere. But if it was somewhere it shouldn't be before you read in your .dat. But since you read in the .dat afterward anyway, it wouldn't matter either that it is (wrongly) where it is. So if your simulation is not using what is in the .dat that you read, it would indeed be mysterious... This of course leads me to believe that your .dat contains a different solution than what you think it contains.

farbflash_ October 24, 2021 06:19

Quote:

Originally Posted by LuckyTran (Post 814924)
Well if you want to solve from a previous solution then you should never initialize at all. Like I said, that line just plain confuses me. It shouldn't be anywhere. But if it was somewhere it shouldn't be before you read in your .dat. But since you read in the .dat afterward anyway, it wouldn't matter either that it is (wrongly) where it is. So if your simulation is not using what is in the .dat that you read, it would indeed be mysterious... This of course leads me to believe that your .dat contains a different solution than what you think it contains.


Thanks again.
I solved the problem, it was, like you said, a small request from fluent which I didn't noticed until you gave me the tip. I had to confirm, that I want to overwrite the existing data.

For all troubling with the same issue, I solved it with this little code adaption:
Quote:

/file/read-data steady-solution.dat.h5
o


All times are GMT -4. The time now is 07:44.