Gradient-Based Single and Multi-points Aerodynamic Optimizations with the elsA Software

54th AIAA Aerospace Sciences Meeting

et al. 2016

General rightsThis document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Department of Aerospace Engineering, University of BristolThis work presents a shape parameterisation method based on multi-resolutional subdivision curves and investigates its application to aerodynamic optimisation. Subdivision curves are defined as the limit curve of a recursive application of a subdivision rule, which provides an intrinsically hierarchical set of control polygons that can be used to provide surface control at varying levels of fidelity. This is used to construct a progressive aerofoil parameterisation that allows an optimisation to be initialised with a small number of design variables, and then periodically increased in resolution through the optimisation. This brings the benefits of a low dimensional design space (high convergence rate, increased robustness, low cost finite-difference gradients) while still allowing the final results to be from a high-dimensional design space. In this work the progressive refinement technique is tested on a variety of optimisation problems. For each problem a range of 'static' (non-progressive) subdivision schemes (equivalent to cubic B-splines) are also used as a control group. For all the optimisation cases the progressive schemes perform comparably or better than the static methods, often providing a significant computational advantage, and in many cases allowing a solution to be found when the static method would otherwise finish prematurely in a local optimum.

Section: Rae2882 Viscous Drag Reduction (Vgk)supporting

confidence: 77%

Section: Rae2882 Viscous Drag Reduction (Vgk)mentioning

confidence: 99%

Progressive Subdivision Curves for Aerodynamic Shape Optimisation

Masters

Taylor

54th AIAA Aerospace Sciences Meeting

et al. 2016

“…This resulted in a significant reduction in overall computation time. It was also found to increase the chances of a similar solution being found when multiple solutions exist 59,60 . This is believed to be the main contributor to the improvement in results compared to previous work from the authors 58,61 .…”

Section: Symmetric Naca0012mentioning

confidence: 99%

“…A considerable number of papers have been published on this case 1,2,52-60 and it has been shown to be a particularly difficult problem due to a range of characteristics such as multiple local minima 58 , non-symmetric solutions 59 and hysteresis 60 . In this previous work, results ranging from 100 to 25 drag counts have been reported with the best cases produced by Lee Due to the symmetry of the problem a symmetry plane was aligned with the aerofoil chord and just half the mesh was solved.…”

Section: Symmetric Naca0012mentioning

confidence: 99%

Multilevel Subdivision Parameterization Scheme for Aerodynamic Shape Optimization

Masters

Taylor²,

et al. 2017

AIAA Journal

Subdivision curves are defined as the limit of a recursive application of a subdivision rule to an initial set of control points. This intrinsically provides a hierarchical set of control polygons that can be used to provide surface control at varying levels of fidelity. This work presents a shape parameterisation method based on this principle and investigates its application to aerodynamic optimisation. The subdivision curves are used to construct a multi-level aerofoil parameterisation that allows an optimisation to be initialised with a small number of design variables, and then be periodically increased in resolution throughout. This brings the benefits of a low fidelity optimisation (high convergence rate, increased robustness, low cost finite-difference gradients) while still allowing the final results to be from a high-dimensional design space. In this work the multi-level subdivision parameterisation is tested on a variety of optimisation problems and compared to a control group of single-level subdivision schemes. For all the optimisation cases the multi-level schemes provided robust and reliable results in contrast to the single-level methods that often experienced difficulties with large numbers of design variables. As a result of this the multi-level methods exploited the high-dimensional design spaces better and consequently produced better overall results.

“…As such, the AIAA Aerodynamic Design Optimization Discussion Group (ADODG) a was formed. As part of this, a number of inviscid and viscous aerofoil and wing benchmark optimization cases were suggested, with a number of research groups presenting results; see [13][14][15][16][17][18][19][20][21] for example.…”

Section: Introductionmentioning

confidence: 99%

Global Optimization of Wing Aerodynamic Optimization Case Exhibiting Multimodality

Poole

Allen

2018

Journal of Aircraft

An investigation into an aerodynamic optimization benchmark case that exhibits multimodality is presented using a global optimization approach. The recently suggested case 6 of the AIAA Aerodynamic Design Optimization Discussion Group involves the drag minimization of a rectangular NACA0012 wing subject to lift and root bending moment, as well as other geometric constraints. In this paper, optimization of that case using a state-of-the-art constrained population-based global optimization framework is presented.Shape control is achieved via a hierarchical application of the radial basis function domain element approach. Three different fidelity of optimization cases are performed to investigate the multimodality of various aspects of the full problem, with four independent runs of each case being performed. When considering chord optimization, the optimizer converges to two independently identifiable minima with very similar performance. When adding dihedral and sweep variation, further multimodality is introduced, though this multimodality is reasonably well defined, with minima including forward and rearward sweep, and an upward and downward winglet. Introducing thickness variables leads to a highly multimodal problem due to the coupling between the chord and thickness variables. The theoretical minimum drag for this case is 24.7 counts, which a number of the runs get close to achieving.