[ Openturns tutorial 2] Building manually an Openturns / Code

Let’s talk about uncertainty quantification and sensibility analysis!

There is an open-source tool for that called OpenTurns which is integrated in Salome_Meca 2019 (Linux version).

This is the tutorial number 2, in which we will talk about deeper settings:

How to export your code_aster study manually
How to prepare the files in the correct way
How to set it back up in Openturns using YACS.

Here’s the video:

A big thanks to Fernando Oleo Blanco (Irvise) for making this video and sharing his experience of Openturns and Code_Aster with the community! I trust it will help a lot of engineers to get started.

The various scripts used in the video are here:


Young = VARIABLE(NOM_PARA='Young', VALE=210e9)

Force = VARIABLE(NOM_PARA='Force', VALE=-37e3)

---------------------------
 
def _exec(Force, Young):
    from asterstudy.api import ParametricCalculation

    calc = ParametricCalculation(".",
                                 ['Force', 'Young'], [Force, Young],
                                 keep_results=True)
    calc.set_logger(print)
    calc.setup()
    # following parameters are set in the 'parametric.export' file
    # but can be overridden here:
    calc.use(version="stable",
             memory=2048,
             time="00:15:00",
             mpicpu=1,
             nodes=1,
             threads=0,
             language="fr",
             args="")
    calc.run()
    results = calc.output_values()
    if len(results) != 1:
        calc.log("ERROR: See log files:", calc.logfiles())
        raise ValueError("Calculation failed with inputs: {0}"
                         .format((Force, Young)))

    DZ = results[0]
    return DZ

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

P actions make_etude
P memory_limit 2048
P mpi_nbcpu 1
P mpi_nbnoeud 1
P ncpus 0
P server localhost
P time_limit 900
P version stable
F comm parametric.comm D  1
F libr _ExportedFromSalomeObject_0_1_2_3.med D  20
F nom inputs.pick D  0
F libr outputs.npy R  8

The first tutorial is here if you didn’t watch it: