Engineers and developers at Engys have implemented customised optimisation methods and employed Multi-Disciplinary Optimisation tools for solving a wide range of engineering problems since 2002. Our MDO offerings cover three main areas:
- Standard optimisation techniques
- Advanced regression analysis
- Adjoint method for topology and shape optimisation in CFD
Standard optimisation techniques
Our standard optimisation services focus around both the application and support of the Open Source tool DAKOTA and our extensive experience in the use of modeFRONTIER. Typical application areas include:
- Process integration
- Design optimisation
- Sensitivity analysis
- Uncertainty quantification
- Model calibration
- Data Mining
- Response Surfaces
- Combined methods
We have successfully applied these methods to analyse and optimise engineering designs based on a wide range of data inputs, from experiments to FEM and CFD results.
Adjoint optimisation for CFD
Engys developers have implemented a continuous adjoint flow solver using OPENFOAM® technology. The methodology has been successfully used to perform general topology and shape optimisation of real industrial applications.
The adjoint flow solutions provide surface or volume sensitivities with respect to one or multiple cost functions, such as minimising drag, minimising pressure drop or maximising flow uniformity. These sensitivities can be coupled with a standard optimisation algorithm to drive a surface-morphing tool or to penalise (remove) fluid cells from the domain. By modifying the shape or topology of the initial design or available fluid space, an optimal solution can be achieved in a few solver iterations by forcing the sensitivities towards zero.
OPENFOAM® is a registered trademark of OpenCFD Ltd. None of Engys offerings are approved or endorsed by OpenCFD Ltd.
