CS 4.0b - Output control strange behavior

Questions and remarks about code_saturne usage
Forum rules
Please read the forum usage recommendations before posting.
Post Reply
Pablo
Posts: 49
Joined: Thu Sep 04, 2014 11:31 am

CS 4.0b - Output control strange behavior

Post by Pablo »

Hello everyone:

I am on the first stages of the CS-4.0b testing and I have realized of an strage behaviour on the results output control for steady-state simulations: No matter how do I change the output control frequency, (either by "n" steps or true-time scale), it is always saved as one result per second in the result file.

I have checked the generated "xml" file and the paragraph seems to be right, so, for example:

<analysis_control>
<output>
<listing_printing_frequency>10</listing_printing_frequency>
<mesh id="-1" label="Fluid domain" type="cells">
<all_variables status="on"/>
<location>all[]</location>
<writer id="-1"/>
</mesh>
<mesh id="-2" label="Boundary" type="boundary_faces">
<all_variables status="on"/>
<location>all[]</location>
<writer id="-1"/>
</mesh>
<probe_format choice="DAT"/>
<probe_recording_frequency>1</probe_recording_frequency>
<writer id="-1" label="results">
<directory name="postprocessing"/>
<format name="ensight" options="binary"/>
<frequency period="time_step">5</frequency>
<output_at_end status="on"/>
<time_dependency choice="fixed_mesh"/>
</writer>
</output>
<profiles/>
<steady_management status="on">
<iterations>10</iterations>
<relaxation_coefficient>0.7</relaxation_coefficient>
<zero_iteration status="off"/>
</steady_management>
<time_averages/>
<time_parameters>
<iterations>200</iterations>
<max_courant_num>1</max_courant_num>
<max_fourier_num>10</max_fourier_num>
<property label="CourantNb" name="courant_number"/>
<property label="FourierNb" name="fourier_number"/>
<property label="LocalTime" name="local_time_step"/>
<time_passing>2</time_passing>
<time_step_max_factor>1000</time_step_max_factor>
<time_step_min_factor>0.1</time_step_min_factor>
<time_step_ref>0.05</time_step_ref>
<time_step_var>0.1</time_step_var>
<zero_time_step status="off"/>
</time_parameters>
</analysis_control>


...as it is defined a steady-state simulation with 200 iterations of 0.05 seconds/iteration ("total time" 10 seconds), and having defined an output control of the results of a record every 5 iterations, it would be expected to have 40 results records (one result every 0.05 seconds), but there is only 10 result records in the result file.

No matter if I change the output control recording to every 0.05 second per output register, the result file recording do not change: one per "physical" time.

Where could the mistake be? :?

Thanks
======================================================================================================
EDIT:
It also happens for transient-simulations.
Attachments
RTEST00.XML
(6.94 KiB) Downloaded 258 times
Yvan Fournier
Posts: 4208
Joined: Mon Feb 20, 2012 3:25 pm

Re: CS 4.0b - Output control strange behavior

Post by Yvan Fournier »

Hello,

This is strange; I just ran a similar test on my side, and did not encounter this issue (I ran 40 steps for a steady calculation, output every 5 time steps, and obtained 8 output records).

Could you also post your listing file, as it has a summary of writer parameters. ?

Regards,

Yvan
Pablo
Posts: 49
Joined: Thu Sep 04, 2014 11:31 am

Re: CS 4.0b - Output control strange behavior

Post by Pablo »

Ok, I have re-run a similar steady-state simulation than yours so that:

- 80 iterations
- 0,1 secs/it
- Writer output every 5 iterations (expected 40 results)

...and I have achieved 8 results for 8 seconds of simulation :?

Attached XML, log file and mesh files attached
Attachments
2,5DMESH.tar.gz
(497.82 KiB) Downloaded 266 times
listing.txt
(97.47 KiB) Downloaded 228 times
RTEST00.XML
(6.67 KiB) Downloaded 278 times
Yvan Fournier
Posts: 4208
Joined: Mon Feb 20, 2012 3:25 pm

Re: CS 4.0b - Output control strange behavior

Post by Yvan Fournier »

Hello,

Do you still have the issue with 200 time steps ? (maybe the problem was due to some partial update) ? If it still fails, could you post the listing for the case that fails ?

Regards,

Yvan
Pablo
Posts: 49
Joined: Thu Sep 04, 2014 11:31 am

Re: CS 4.0b - Output control strange behavior

Post by Pablo »

Hello:

I attach a case which fails: the XML and the listing file generated.
According to the simulation definition:

Steady-state simulation:
·····Constant time-step= 400 iterations
·····Reference time step= 0.05 s

Output control:
·····Log frecuency: Every "n" time-steps= 20
····Writer Frecuency: Every "n" time-steps= 10

According to what I suppose (and I've confirmed in previous simulations in previous CS releases), this configuration should simulate 20 "seconds" of result behaviour with one log every 20 time-steps in the listing and 40 states to represent (1 result every 0.25 seconds) in the result file, but there only are 20 states in the result file ( 1 every second ).


I hope this information helps.

Thank you, as always.
Attachments
RTEST00.XML
(6.64 KiB) Downloaded 260 times
listing.txt
(241.89 KiB) Downloaded 275 times
Yvan Fournier
Posts: 4208
Joined: Mon Feb 20, 2012 3:25 pm

Re: CS 4.0b - Output control strange behavior

Post by Yvan Fournier »

Hello,

Im am not able to reproduce your bug: I built an xml file based on yours, on a simple cavity, adapting only the boundary conditions to my mesh, and obtained 40 time steps, not 20...

Do you have any user subroutines ?

Regards,

Yvan
Pablo
Posts: 49
Joined: Thu Sep 04, 2014 11:31 am

Re: CS 4.0b - Output control strange behavior

Post by Pablo »

Hello,

Sorry for the delay, but I have been busy these days.

I have not any subroutine loaded at all, it is just a smple simulation generated as I realize that something failed with the pulverized coal combustion tutorial in CS4.0b.

This issue happens to me in my personal laptop (Xubuntu 14.0.4, with the 3.3 Saturne pre-requisites updated at July 2014), which, prior erasing of the previous release of CS, I installed the CS4.0b.
Anyway, and I do not know how, I have installed the CS-4.0b in the office's workstation (as I have re-installed the SO Xubuntu 14.0.4 and all the saturne pre-requisites updated at today) and now the simulations behaves as usual, without any trouble with the results recording in the result file. (even with the coal-combustion tutorial, as far as I have checked by now)

The only difference between both machines is the pre-requisites versions, so I am going to compare them and re-install CS4.0b in my personal laptop in order to check if something is not working as it should.


I will keep you informed.


Kind regards,
Yvan Fournier
Posts: 4208
Joined: Mon Feb 20, 2012 3:25 pm

Re: CS 4.0b - Output control strange behavior

Post by Yvan Fournier »

Hello Pablo,

I'll probably be uploading a new 4.0 beta later today, with some fixes relative to combustion, so you may want to wait a bit before reinstalling...

Regards,

Yvan
Pablo
Posts: 49
Joined: Thu Sep 04, 2014 11:31 am

Re: CS 4.0b - Output control strange behavior

Post by Pablo »

Well, although a little late, this problem was caused by the post-processing program (Paraview). Its reinstalling fixes the problem.
Post Reply