Expedited constrained multi-objective aerodynamic shape optimization by means of physics-based surrogates

Koziel, Slawomir; Tesfahunegn, Yonatan A.; Leifsson, Leifur

doi:10.1016/j.apm.2016.03.020

Cited by 14 publications

(6 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…12 This type is based on a low-fidelity model having the same physics of the original system but with low accuracy. This low accuracy may be a result of using the same high-fidelity model but with coarser mesh and relaxed convergence criteria, 21,22 or as a result of using a simplified physics model by, for example, neglecting three-dimensional (3D) effects as in the present study. One of the promising techniques which uses physicsbased surrogates is space mapping (SM).…”

Section: Introductionmentioning

confidence: 94%

Blade shape optimization of an aircraft propeller using space mapping surrogates

Toman

Hassan

Owis

et al. 2019

Advances in Mechanical Engineering

View full text Add to dashboard Cite

Propeller performance greatly influences the overall efficiency of the turboprop engines. The aim of this study is to perform a propeller blade shape optimization for maximum aerodynamic efficiency with a minimal number of high-fidelity model evaluations. A physics-based surrogate approach exploiting space mapping is employed for the design process. A space mapping algorithm is utilized, for the first time in the field of propeller design, to link two of the most common propeller analysis models: the classical blade-element momentum theory to be the coarse model; and the high-fidelity computational fluid dynamics tool as the fine model. The numerical computational fluid dynamics simulations are performed using the finite-volume discretization of the Reynolds-averaged Navier–Stokes equations on an adaptive unstructured grid. The optimum design is obtained after few iterations with only 56 computationally expensive computational fluid dynamics simulations. Furthermore, an optimization method based on design of experiments and kriging response surface is used to validate the results and compare the computational efficiency of the two techniques. The results show that space mapping is more computationally efficient.

show abstract

Section: Introductionmentioning

confidence: 94%

Blade shape optimization of an aircraft propeller using space mapping surrogates

Toman

Hassan

Owis

et al. 2019

Advances in Mechanical Engineering

View full text Add to dashboard Cite

show abstract

“…Most aerodynamic design optimization methods using surrogate models rely on an iterative model refinement [440]. Such surrogate-based optimization strategies have shown to be effective in various aerodynamic shape optimization applications including single-point design [441,442], multipoint design [328,443], massively multipoint design [100], multi-objective design [93,444,445], inverse design [168,446], and robust design [313,[447][448][449]. Optimization using these methods is generally composed of two phases.…”

Section: Surrogate-based Optimizationmentioning

confidence: 99%

Machine Learning in Aerodynamic Shape Optimization

Li¹,

Martins²

2022

Preprint

View full text Add to dashboard Cite

Large volumes of experimental and simulation aerodynamic data have been rapidly advancing aerodynamic shape optimization (ASO) via machine learning (ML), whose effectiveness has been growing thanks to continued developments in deep learning. In this review, we first introduce the state of the art and the unsolved challenges in ASO. Next, we present a description of ML fundamentals and detail the ML algorithms that have succeeded in ASO. Then we review ML applications contributing to ASO from three fundamental perspectives: compact geometric design space, fast aerodynamic analysis, and efficient optimization architecture. In addition to providing a comprehensive summary of the research, we comment on the practicality and effectiveness of the developed methods. We show how cutting-edge ML approaches can benefit ASO and address challenging demands like interactive design optimization. However, practical large-scale design optimizations remain a challenge due to the costly ML training expense. A deep coupling of ML model construction with ASO prior experience and knowledge, such as taking physics into account, is recommended to train ML models effectively.

show abstract

“…As mentioned in the introduction, a fundamental bottleneck of PDE-constrained optimization of complex systems is the high cost of the accurate, high-fidelity simulations, which is especially challenging in MOO. Therefore, for the sake of computational efficiency, Multi-objective optimization of aerodynamic surfaces the MOO procedure (Koziel et al, 2016) used in this work exploits, apart from the original, high-fidelity simulation model f, its low-fidelity model counterpart c. In this work, the lowfidelity model is based on coarse-discretization CFD simulations (the detailed setup is discussed in Section 4), which allows for a fast evaluation at the cost of some accuracy degradation. A design speedup is achieved by performing most of the operations at the level of the low-fidelity model; however, high-fidelity simulations are also executed to yield a Pareto set that is sufficiently accurate.…”

Section: Optimization Algorithmmentioning

confidence: 99%

“…The MOO approach utilized in this work has been recently proposed by Koziel et al (2016). The approach integrates fast physics-based surrogate models and design space reduction techniques to approximately identify the Pareto front, and, subsequently, refines it using a limited number of computationally expensive system evaluations.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Design strategies for multi-objective optimization of aerodynamic surfaces

Amrit

Leifsson

Koziel

2017

Self Cite

View full text Add to dashboard Cite

Purpose This paper aims to investigates several design strategies to solve multi-objective aerodynamic optimization problems using high-fidelity simulations. The purpose is to find strategies which reduce the overall optimization time while still maintaining accuracy at the high-fidelity level. Design/methodology/approach Design strategies are proposed that use an algorithmic framework composed of search space reduction, fast surrogate models constructed using a combination of physics-based surrogates and kriging and global refinement of the Pareto front with co-kriging. The strategies either search the full or reduced design space with a low-fidelity model or a physics-based surrogate. Findings Numerical investigations of airfoil shapes in two-dimensional transonic flow are used to characterize and compare the strategies. The results show that searching a reduced design space produces the same Pareto front as when searching the full space. Moreover, as the reduced space is two orders of magnitude smaller (volume-wise), the number of required samples to setup the surrogates can be reduced by an order of magnitude. Consequently, the computational time is reduced from over three days to less than half a day. Originality/value The proposed design strategies are novel and holistic. The strategies render multi-objective design of aerodynamic surfaces using high-fidelity simulation data in moderately sized search spaces computationally tractable.

show abstract

Expedited constrained multi-objective aerodynamic shape optimization by means of physics-based surrogates

Cited by 14 publications

References 18 publications

Blade shape optimization of an aircraft propeller using space mapping surrogates

Blade shape optimization of an aircraft propeller using space mapping surrogates

Machine Learning in Aerodynamic Shape Optimization

Design strategies for multi-objective optimization of aerodynamic surfaces

Contact Info

Product

Resources

About