CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > Fidelity CFD

UQ-Simulation with Design3D

Register Blogs Community New Posts Updated Threads Search

Like Tree1Likes
  • 1 Post By dwunsch

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   December 17, 2018, 11:53
Default UQ-Simulation with Design3D
  #1
New Member
 
Tim Schraeder
Join Date: Dec 2018
Posts: 2
Rep Power: 0
tischrae is on a distinguished road
Hello everyone,
I´m new to numeca/Fine and doing a study with the UQ-module in Design3D on a turbine row with Stator/Rotor geometries. I have parametrized the Stator with autoblade and want to combine the grids with a multistage script in Design3D. The Script combines Stator and Rotor grids and works fine when exeuted manually. However running the Design3D chain, there is a problem generating the grid-file for multistage and the solver crashes.


.std File shows the following problems:
(German marked lines are saying cannot find the following Files)



----------------------------------------------------------------------------------------------


PARAMETRIC MODELER /////////////////////////////////////////////////

STEP 1 : Update model ........................................... OK

CFD MESH GENERATOR /////////////////////////////////////////////////

STEP 1 : Locate .trb file ....................................... OK

STEP 2 : Rebuild geometry .................................. SKIPPED

================================================== =====

Run AutoGrid

================================================== =====

mkdir: das Verzeichnis »/nfs/.../initialisierung_compiled_1712/initialisierung_compiled/initialisierung_compiled_Design_1/_unc/_flow_000001/_mesh« kann nicht angelegt werden: Die Datei existiert bereits

job #0 - AutoGrid - started on config #0 job #0 - AutoGrid - finished

STEP 3 : Generate mesh .......................................... OK local_quality_report_file = /nfs/.../initialisierung_compiled_1712/initialisierung_compiled/initialisierung_compiled_Design_1/_unc/_flow_000001/_mesh/initialisierung_compiled.qualityReport

STEP 4 : Check mesh quality ..................................... OK

mv: der Aufruf von stat für '/nfs/.../initialisierung_compiled_1712/initialisierung_compiled/initialisierung_compiled_Design_1/_unc/_flow_000001/_mesh/initialisierung_compiled.X_T' ist nicht möglich: Datei oder Verzeichnis nicht gefunden

mv: der Aufruf von stat für '/nfs/.../initialisierung_compiled_1712/initialisierung_compiled/initialisierung_compiled_Design_1/_unc/_flow_000001/_mesh/initialisierung_compiled.par' ist nicht möglich: Datei oder Verzeichnis nicht gefunden

mv: der Aufruf von stat für '/nfs/.../initialisierung_compiled_1712/initialisierung_compiled/initialisierung_compiled_Design_1/_unc/_flow_000001/_mesh/initialisierung_compiled.trb' ist nicht möglich: Datei oder Verzeichnis nicht gefunden

job #0 - AutoGrid - started on config #0 job #0 - AutoGrid - finished

STEP 5 : Merge mesh (multistage) ................................ OK


CFD SOLVER /////////////////////////////////////////////////////////

mkdir: das Verzeichnis »/nfs/.../initialisierung_compiled_1712/initialisierung_compiled/initialisierung_compiled_Design_1/_unc/_flow_000001« kann nicht angelegt werden: Die Datei existiert bereits

================================================== =====

Run Euranus

================================================== =====

NIFineBatch::merge_mesh, ERROR: Cannot find the mesh file: '/nfs/.../initialisierung_compiled_1712/initialisierung_compiled/initialisierung_compiled_Design_1/_unc/_flow_000001/_mesh/initialisierung_compiled.igg'

STEP 1 : Save .run file ..................................... FAILED

mkdir: das Verzeichnis »/nfs/.../initialisierung_compiled_1712/initialisierung_compiled/initialisierung_compiled_Design_1/_unc/_flow_000001« kann nicht angelegt werden: Die Datei existiert bereits

cp: reguläre Datei '/nfs/.../initialisierung_compiled_1712/initialisierung_compiled/initialisierung_compiled_Design_1/_unc/_flow_000001/_results/_pictures/logo.gif' kann nicht angelegt werden: Datei oder Verzeichnis nicht gefunden

cp: reguläre Datei '/nfs/.../initialisierung_compiled_1712/initialisierung_compiled/initialisierung_compiled_Design_1/_unc/_flow_000001/_results/_pictures/wave.gif' kann nicht angelegt werden: Datei oder Verzeichnis nicht gefunden

[error - 13:49:58] Impossible to merge the run file with the generated mesh. The running process has been stopped. pre.cjk { font-family: "Courier New", monospace; }p { margin-bottom: 0.25cm; line-height: 115%; }

---------------------------------------------------------------------------------------------------


I also get an error in "job_igg_multistage":


STARTUP_ERROR: Cannot Access Display itsm-numerik417:0.0 from itsm-numerik417

Can anyone help me where to find the problem?
Best Regards,
Tim


pre.cjk { font-family: "Courier New", monospace; }p { margin-bottom: 0.25cm; line-height: 115%; }
tischrae is offline   Reply With Quote

Old   December 22, 2018, 04:19
Default
  #2
New Member
 
abdul
Join Date: Apr 2017
Posts: 24
Rep Power: 9
eng.abdul is on a distinguished road
I have encountered similar problems and work for many weeks to figure out. This is a strange problem in Numeca, I guess. The manually meshing does not give any error but one can have an error when the same is generated through the interface of Desing3D.

I would suggest you that the template .trb should be of a high-quality mesh. pay attention to min. skewness, aspect ration and span angular deviation. and the parametric model from which you generated .trb should be of the minimum number of sections. there is no solid argument in favor of these considerations. I did not use UQ module though.
eng.abdul is offline   Reply With Quote

Old   January 7, 2019, 05:32
Default Display problem FINE/Design3D
  #3
New Member
 
Dirk Wunsch
Join Date: Jan 2019
Posts: 1
Rep Power: 0
dwunsch is on a distinguished road
Dear Tim,

let me first summarize the situation as I understand it:
  • You built a multi-stage script for AutoBlade following the instructions in the manual. This works fine.
  • Then you built a script for meshing with AutoGrid5, which you import under 'Mesh Generation'
  • When you start this script outside of FINE/Design3D it runs and generates the mesh successfully (It most likely opens the GUI of AutoGrid5)
  • When you run the same script now through FINE/Design3D you get the error message of the meshing job: "STARTUP_ERROR: Cannot Access Display itsm-numerik417:0.0 from itsm-numerik417"
It seems to me that the underlying problem is related to the environment variable DISPLAY.
  • In your case, echo $DISPLAY will return itsm-numerik417:0.0, which is the name of your machine and the number of the display. This should be equivalent to :0.0.
  • However, we have observed that in some cases a display such as itsm-numerik417:0.0, prevents the starting of GUI's. If this is the case, you cannot open, for example, xcalc neither
  • The problem comes from the environment, but impacts some of our software
What can you do to get your job running:
If you built your meshing script yourself and if it opens a GUI of AutoGrid5, in case that you run it outside FINE/Design3D, you can add a flag for batch mode to the start command of AutoGrid5[*]-real-batch -autogrid5[*]In this case no GUI will be opened[*]Please note that you might encounter a similar problem with CFView for the post-processing. Here, however, we need to open the GUI, there is no batch mode. [*]The next release version of FINE 13.2 will come with a solution for this DISPLAY problem also for CFView, where we modify the start command.[/LIST][LIST=1]

There is a possible second workaround (but not recommended for AutoGrid5, only for CFView), that sometimes works
  • From user's side it is possible to connect to the local host with ssh and display forwarding:
  • 1. Terminate running pvm
  • > killall pvmd
  • 2. Connect to the local host using ssh
  • > ssh -X user@hostname
  • 3. launch fine: display will be automatically set to localhost:10.x
  • > printenv DISPLAY
  • DISPLAY=localhost:10.0
Best regards,
Dirk
colinda1 likes this.
dwunsch is offline   Reply With Quote

Old   January 31, 2019, 09:50
Default UQ-Module Non Deterministic Results
  #4
New Member
 
Tim Schraeder
Join Date: Dec 2018
Posts: 2
Rep Power: 0
tischrae is on a distinguished road
Hello everyone,

@dwunsch: Many thanks for your reply. It really helped me to get my project running. Your solution with connecting to localhost with ssh solved the problem after some issues with the local network.

Currently the Design3D Chain is running and all tasks finish with the computation. I get all the deterministic results for the defined collocation points. However I have the problem, that the Non Deterministic Results in the Uncertainties Monitor are not displayed correctly.
In the project folder, there is a .ssd file which I think is the problem. Example:

Number_of_quantities 0
Number_of_uncertainties 2

.uq file same

It seems like there are no quantities defined for the Non Deterministic Results. A section in the User manual notes, that the selected Quantities in the Post processing page are not used in the non deterministic post processing.
How do I set up the Quantities for the non deteministic post processing? The run File doesn´t show anything.

Many Thanks,
Tim
tischrae is offline   Reply With Quote

Reply


Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Convergence Problem - Transient Simulation gemxx Main CFD Forum 0 July 15, 2018 10:36
Mapping Field Data for Mesh Regions from Another Simulation veterator OpenFOAM Pre-Processing 1 July 10, 2018 06:28
Surface Source - Fixed Temperature? robtheslob FloEFD, FloWorks & FloTHERM 18 May 12, 2017 03:28
Simulation FPEs - turbulence for transient and steady-state? DaveR OpenFOAM Running, Solving & CFD 5 March 5, 2017 16:06
setting up a simulation with multiple interactions phandy OpenFOAM Running, Solving & CFD 1 October 6, 2014 04:16


All times are GMT -4. The time now is 07:17.