CFD Online Discussion Forums

Hi everyone,

I hope my post won't be off topic, here is my question :
I aml currently using Dakota 6.13 with OpenFoam 2012 in order to optimize parameters, and I am trying to build a surrogate-based optimization. The thing is I want to run several OpenFOAM in parallel to reduce the calculation time. Does anyone have already done something similar ? The OpenFOAM simulation I use runs on 6 processors and here is my dakota input file :

Code:

environment

 tabular_data



method

 id_method = 'SBGO'

 surrogate_based_global

   model_pointer = 'SURROGATE'

   method_pointer = 'SOGA'

   max_iterations = 15

   replace_points

 output verbose



method 

 id_method = 'SOGA'

 soga 

    seed = 10983

  max_function_evaluations = 30000

  #initialization_type unique_random

 output silent



model 

 id_model = 'SURROGATE'

 surrogate global

    dace_method_pointer = 'SAMPLING'

    #correction additive zeroth_order

    gaussian_process dakota



method

 id_method = 'SAMPLING'

 sampling

   samples = 50

   seed 531

   sample_type lhs

   model_pointer = 'TRUTH'



model

 id_model = 'TRUTH'

 single

   interface_pointer = 'TRUE_FN'



variables

  active uncertain

  uniform_uncertain 4 

   descriptors     'came_start' 'stiffness1'  'stiffness2' 'stiffness3'

   lower_bounds           -210          200           1200         2600

   upper_bounds           -150          400           1600         3000



interface

 id_interface = 'TRUE_FN'

 analysis_drivers = 'dakota3.sh'

   fork

    work_directory named 'workdir/run'

    directory_tag directory_save

    parameters_file 'params.in'

    results_file 'results.out'

    link_files 'dakotaParameter.orig' 'transfert.py' 'ecriture_xml_v3.py'



responses

  objective_functions=1 

   descriptors 'res' 

   no_gradients 

   no_hessians



   sense 'max'

Best regards,

Johnny