SNIC
SUPR
SNIC SUPR
Test and Assessment of Derivative Computation Architecture in OpenMDAO and its Application for MDO
Dnr:

SNIC 2018/3-455

Type:

SNAC Medium

Principal Investigator:

Lilit Axner

Affiliation:

Kungliga Tekniska högskolan

Start Date:

2018-09-26

End Date:

2019-01-01

Primary Classification:

10105: Computational Mathematics

Webpage:

Allocation

Abstract

Topic Description: This thesis work will be carried out by the joint supervision between Airinnova & KTH-PDC . The student is supposed to • learn how to set-up an (multi-disciplinary) optimization problem using the planform OpenMDAO; • understand & compare the different derivative computation architectures in OpenMDAO and measure their scalability; • develop an automation process to set-up the derivative computation in OpenMDAO with Python scripts • If time permits, develop an automation process to transfer the computed sensitives to other optimization platform. Topic Description: This thesis work will be carried out by the joint supervision between Airinnova & KTH-PDC . The student is supposed to • learn how to set-up an (multi-disciplinary) optimization problem using the planform OpenMDAO; • understand & compare the different derivative computation architectures in OpenMDAO and measure their scalability; • develop an automation process to set-up the derivative computation in OpenMDAO with Python scripts • If time permits, develop an automation process to transfer the computed sensitives to other optimization platform. The tools/software will be used in the Project:  OpenMDAO: http://openmdao.org/  CFD tool SU2: https://su2code.github.io/