code_saturne User's Forum

Ok, so, my next doubts are:

-|Doubt no.1 |--------------------------------------------------------------------------------------------------------------------------------------------
According to what you said, the src/lagr/lagsta.f90, I guess, copied in the DATA folder of a certain simulation would perform the statistical calculation on a certain volume (so-called "statistiques globales" in lagsta.f90, I guess) or a boundary (so-called "statistiques globales" in lagsta.f90, I guess too).
According to what I can read on the "lagsta.f90" file in line no. 391, for example, after other calculations, a command seems to express an estatistical operation so:
stativ(iel,ildck) = stativ(iel,ildck) &
+ tepa(npt,jrpoi)*tepa(npt,jrdck)*tepa(npt,jrdck)
...assigns the variance ("stativ") of a certain volumetric variable "ildck" (which meaning I can't find within the .f90 file ) of a certain particle ("iel") to the result of the operation "stativ(iel,ildck) + tepa(npt,jrpoi)*tepa(npt,jrdck)*tepa(npt,jrdck)"

is this approach correct?

-|Doubt no.2 |--------------------------------------------------------------------------------------------------------------------------------------------
As I can see in the 3.3.0 practical user's guide (page 47/203), a suplementary user volumetric statistic could be defined via "ilvu(ii)" pointer, so....

- Where is this supplementary pointer defined?
- How is this pointer linked with the .XML volume statistics Variable Name (mean value) "NOMLAG", Variable Name (variance value) "NOMLAV" and post-processing control "IACTFV"?

-|Doubt no.3 |--------------------------------------------------------------------------------------------------------------------------------------------
In the slagging tutorial, the boundary statistics variables "Part_fouled_impact_number"/"Part_fouled_mass_flux"/"Part_fouled_diam"/"Part_fouled_mass_Xck"...seems to be "supplementary" (due to these variables does not appear in the standard lagrangian calculation unless the combustion/slagging models are active)
Where can be seen the operations which results ares stored on the aforementioned boundary statistic variables?


Kind regads and thanks in advance

Hello,

1) Actually, this is an accumulation in stativ(iel,ildck), which accumulates
tepa(npt,jrpoi)*tepa(npt,jrdck)*tepa(npt,jrdck)
(statistical weight + shrinking diameter**2) at each time step.
This is probably divided in lagsta for a final result.

2) I'll need to check. If you can't find "ilvu" using grep on the code, don't trust the doc (the Lagrangian doc might be partially out of sync).
Using a standard user-defined property is probably the simplest solution here (easy to do in version 4.0 beta, see cs_user_parameters.c). The property will be postprocessed automatically. I recommend using the user property for instant values (not accumulation), then defining user time moments (in the same file) to obtain the variance (which will then also be handled automatically in the restart files).
Check the Doxygen user examples for details (they are up to date, and often nicer to user than searching through the code).

3) for boundary statistics, search for parbor in the Fortran code (or boundary_stat for updates n cs_lagr_tracking.c).

Regards,

Yvan

Hello everyone:

I have performed the pulverized coal combustion/slagging tutorial with CS-4.0b and althogh minor differences has been observed, everything seems to run adequately respecting to what it was observed in the previous CS release (3.3).

However one of the mentioned troubles is still there so that the slagging simulation fails if the .XML is not modified before running the simulation, the so called Error no.1 at the beginning of this post:

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Error 1 message:
"cs_gui_util.c:1027: Fatal error.
Invalid attribute value:
Xpath: /Code_Saturne_GUI/lagrangian/output/raw_coal_mass_fraction/@status"

Posible reason:
GUI does not fill the XML properly so that certain default values does are not have correct "on/off" values. If the XML file is edited, the lagrangian post-processing particles variables values are set to blank values instead of "OFF".

Correction:
Option 1: Manually in the XML file, fill the correct value for the parameters to "OFF" (which are the post processing values indicated in the tutorial)
Option 2: Using the GUI, select all the lagrangian post-processing particles variables ("Calculation control" --> "Lagrangian solution control" --> "Post processing for particles" --> "Variables selection") and save the XML file. (this will fill the variables to "ON"). Then un-select all the lagrangian post-processing particles variables and save again (this will fill the variables to "OFF", which is the variable state looked for in the tutorial and theoretically the default values)

After these correction, the slagging simulation runs perfectly.


Kind regards,



PS: XMLs attached.

Hello,

This is good news. I'll try to see with Cyril next week about the remaining GUI xml generation bugs.

Regards,

Yvan

Hello, I am interested in working through the coal combustion tutorials. However, I am unable to find the mesh file for getting started: Tutorial3D.cgns anywhere on my computer. I have installed both code_saturne 3.06 and 4.0-beta. That file does not seem to be in any directory associated with those installs on my computer. Could someone point me in the right direction for that file?

I'm retired now but wish to keep looking at advances in combustion modeling. I did alot of coal combustion/nox formation modeling decades ago with Brian Spalding and his PHOENICS code. We developed a restart method that looks similar to what is employed in the coal combustion slagging tutorials. I'd like to give it a whirl.

Thanks,

Tom Kosvic

Hello

Very happy to know you are interested in Code_Saturne, and welcome to this forum.

I don't have a local copy of this mesh, and this issue has appeared fo several tutorials not part of the "core" tutorials (which are versioned).

I should be able to get a copy of the mesh from the colleagues who wrote the tutorial on Monday, and I'll post it on the website, with a permanent link in the doc section.

We'll probably also upload a new version of 4.0 beta early in the week, with additional bug fixes.

Regards,

Yvan

Hello,

I have uploaded the mesh in the tutorials section, with a link just under that of the pdf file :
http://code-saturne.org/cms/sites/defau ... ial3D.cgns

I also updated the 4.0 beta version.

Regards,

Yvan

Yvan,

In clicking on the link in your response and also on the new link in the tutorial section, I am finding that my web browser (using chrome) is reading the file directly rather than allowing a download. I am seeing strange characters that look like trying to view a binary file.

I need some more assistance in how to download this file.

Thanks for the help.

Tom Kosvic

Try right-clicking on the link and selecting something like "save destiny as" or "save direction as" in the contextual menu. I don't use Chrome but that kind of option is almost universal to every browser.

tckosvic wrote:Yvan,

In clicking on the link in your response and also on the new link in the tutorial section, I am finding that my web browser (using chrome) is reading the file directly rather than allowing a download. I am seeing strange characters that look like trying to view a binary file.

I need some more assistance in how to download this file.

Thanks for the help.

Tom Kosvic

I have the Tutorial3D.cgns file downloaded per CAVT suggestions. I hadn't downloaded from web in the "save" fashion before.

Now to move on to the technical steps of this tutorial.

Thanks, Tom Kosvic