Sign In

SIMPRO

TUT subproject 1

Structural optimization in automated product design

Demand

Throughout the history of industrial optimisation, the main focus in a majority of the research projects has been the search for a global optimum using various strategies. This massive research has resulted in many fine-tuned and robust optimisation algorithms for different types of problems. In some problem categories, even the finding of the global optimum can be guaranteed.

Automated product design has many advantages:

  1. routine work such as finite element model generation can be avoided,
  2. structural optimisation is carried out systematically and
  3. the designer can handle multiple design cases simultaneously.

Especially in cases where the product is not limited to certain standard types and depends on the customer’s requirements, the automated design process is extremely efficient. The research questions of this SIMPRO subproject were:

  • How to automate product design process using optimisation methods?
  • How to connect black box simulations and mathematical optimisation algorithms?

Solution

In the SIMPRO project, we chose to use surrogate-model based optimisation method due to its generic nature. The platform (design software called SimPro) was used to formulate the optimisation problem while the simulations could be carried out by using some black box solver and the generated optimisation subproblems are solved using the best available optimisation algorithm libraries.

Solves problems with:

  • several objectives
  • real-, integer- and discrete-valued variables
  • discrete variables that have dependencies

Features:

  • Interface to ANSYS
  • Supports high-performance computing (HPC)
  • Interface to Techila system (HPC)
  • Several efficient optimisation algorithms
  • Task queue
  • Informative graphs and comprehensive reports

Benefits and use cases

The SimPro platform was a vital part of the project so that all the existing optimisation algorithms could be adopted to our use. Adding the Techila HPC feature in the platform made it competitive even against commercial products.

The platform was adopted successfully to the design cases of guide rails and telescopic booms.

 

 

 

Subproject deliverables

  1. Taskinen, L.-P.
    Hissin johteen rakenteen optimointi / Optimization of a elevator guide rail
    Msater's thesis, Tampere University of Technology, 2014
    Link: http://dspace.cc.tut.fi/dpub/bitstream/handle/123456789/22767/Taskinen.pdf?sequence=1
  2. Laakkonen, P.
    Laskentatyökalun kehittäminen kantavien rakenteiden optimointiin / Development of a computational tool for structural optimization
    Msater's thesis, Tampere University of Technology, 2013
    Link: https://dspace.cc.tut.fi/dpub/bitstream/handle/123456789/22177/Laakkonen%20Petri.pdf?sequence=3&isAllowed=y
  3. Lehtinen, J. & Pajunen, S.
    A perfect fit – Researchers develop an automated process for optimizing marine structural components
    ANSYS Advantage, vol. 7, no. 2, 2013
    Link: http://resource.ansys.com/staticassets/ANSYS/staticassets/resourcelibrary/article/AA-V7-I2-A-Perfect-Fit.pdf
  4. Pajunen, S. & Heinonen, O.
    Automatic design of marine structures by using successive response surface method
    Structural and Multidisciplinary Optimization, May 2014, Volume 49, Issue 5, pp 863–871
    DOI: 10.1007/s00158-013-1013-7
    Link: http://link.springer.com/article/10.1007%2Fs00158-013-1013-7
  5. Pajunen, S. et al.
    Surrogate-model based method and software for practical design optimization problems
    To be published
  6. Laakkonen, P.
    Structural optimization in automated product design
    Poster, SIMPRO-SCarFace Joint Seminar 2, Espoo August 25–26, 2015

Contacts