Welcome to the forums. Please post in English or French.

You are not logged in.

#1 2020-06-14 15:05:40

jhenk
Member
Registered: 2018-01-10
Posts: 6

OpenTurns Usage

Hello everybody,

I want to use OpenTurns with Salome Meca for uncertainty quantification. Unfortunately, there are no tutorials on the steps involved. I only found this short YouTube video, from which I was not able to reproduce all necessary steps:

"Easy-to-use uncertainty quantification thanks to AsterStudy and OpenTURNS"

Is there a tutorial anywhere? A simple model (say, a beam with normally distributed Young's modulus) to start with would be another possibility. I cannot speak french.

Offline

#2 2020-06-17 11:36:33

jhenk
Member
Registered: 2018-01-10
Posts: 6

Re: OpenTurns Usage

From the video the creation of the "POST_RELEVE_T" remains unclear.

Last edited by jhenk (2020-06-17 12:04:31)

Offline

#3 2020-06-17 12:04:49

jhenk
Member
Registered: 2018-01-10
Posts: 6

Re: OpenTurns Usage

I tried to run a toy example, which is possible to run in AsterStudy, but for OpenTurns it fails to evaluate deterministically with the following error:

(see attached "message" file, french)

I would appreciate, if somebody could help me out


Attachments:
message, Size: 28.1 KiB, Downloads: 361

Offline

#4 2020-06-18 18:06:18

dezsit
Member
Registered: 2012-06-27
Posts: 69
Website

Re: OpenTurns Usage

Hello,

I think the problem is, that you extracted the "extrema" of the displacement DZ with this:

table2 = POST_RELEVE_T(ACTION=_F(OPERATION=('EXTREMA', ),
                                   INTITULE='DZ',
                                   RESULTAT=linel,
                                   NOM_CHAM='DEPL',
                                   PRECISION=1.E-06,
                                   CRITERE='RELATIF',
                                   FORMAT_C='MODULE',),
                         INFO=1,)

This produce not only one DZ value (as necessary for OpenTurns) but four, and actually not simply 'DZ':
#
#--------------------------------------------------------------------------------
##ASTER 14.04.00 CONCEPT table2 CALCULE LE 18/06/2020 A 16:57:37 DE TYPE         
#TABLE_SDASTER                                                                   
INTITULE         RESU     NOM_CHAM         NUME_ORDRE   INST         EXTREMA  NOEUD    CMP      VALE       
DZ               result   DEPL                        1  0.00000E+00 MAX      N98      DZ        6.39285E-02
DZ               result   DEPL                        1  0.00000E+00 MIN      N134     DY       -1.00000E+00
DZ               result   DEPL                        1  0.00000E+00 MAXI_ABS N134     DY        1.00000E+00
DZ               result   DEPL                        1  0.00000E+00 MINI_ABS N52      DX        0.00000E+00

So, that is way a warning reported in the message file:
   ! Parametre absent de la table : DZ.   

So you should try this:

#all dz value extracted from the node group 'Group_1' or whatever into the table
table2 = POST_RELEVE_T(
  ACTION=_F(
    GROUP_NO=('Group_1', ),
    INTITULE='DZ',
    NOM_CHAM='DEPL',
    NOM_CMP=('DZ', ),
    OPERATION=('EXTRACTION', ),
    RESULTAT=result
  )
)

#then it is filtered (only maximum absolute -> MAXI_ABS) during the numpy table creation, so you have only one value at the end.
IMPR_TABLE(
  FILTRE=_F(
    CRIT_COMP='MAXI_ABS',
    NOM_PARA='DZ'
  ),
  FORMAT='NUMPY',
  NOM_PARA=('DZ', ),
  TABLE=table2,
  UNITE=8
)

I did not tried with OpenTurns, but it eliminates the warning, so should work.

Best regards,
dezsit

Offline

#5 2020-06-22 08:29:52

jhenk
Member
Registered: 2018-01-10
Posts: 6

Re: OpenTurns Usage

Thank you very much. With you help I was able to get it running!

Offline

#6 2020-07-02 10:30:00

bpaillard
Member
Registered: 2015-10-16
Posts: 28

Re: OpenTurns Usage

FEA for all has heard you !
I can't post links (!!) but just look for  :

how-to-interface-code_aster-and-open-turns-to-do-uncertainty-quantification-and-sensibility-analysis

on feaforall

Offline