Multi‐point optimization of shapes and settings of high‐lift system by means of evolutionary algorithm and Navier‐Stokes equations

Abstract

Purpose

Today, the design process of high‐lift configurations in industry mainly relies on experts' knowledge, and lacks a simple exploration of the design space. Therefore, the introduction of high‐fidelity tools in an optimization chain is now envisaged. The purpose of this paper is to define and solve a realistic high‐lift design problem by the use of a constrained evolutionary algorithm, coupled to a Navier‐Stokes (RANS) solver. The complete optimization (shape and settings) of a 3‐element configuration has been carried out for landing and take‐off configurations using a sequential approach.

Design/methodology/approach

In a first step, the elements' shapes and settings of the landing configuration have been optimized simultaneously. Then, shapes have been frozen and settings have been optimized for take‐off conditions. The flow evaluation during the optimization process is made through 2.5D Navier‐Stokes computations on chimera grids. The optimization technique used is an evolutionary algorithm, with a dynamic adaptation of the covariance matrix (CMA‐ES). Geometric and aerodynamic constraints have been considered through a dynamic penalization technique of the cost function.

Findings

Solutions obtained have been analyzed and compared to the reference initial configuration. In term of cost functions improvement, 5.71 per cent drag reduction has been obtained for landing, and 2.89 per cent improvement on climb index at take‐off.

Practical implications

Compared to the global optimization process, the use of a sequential approach can be quite efficient.

Originality/value

This paper presents a first step for the introduction of recent advanced methods into a design process of high‐lift configurations in an industrial environment.

Citation

Moens, F. and Wervaecke, C. (2013), "Multi‐point optimization of shapes and settings of high‐lift system by means of evolutionary algorithm and Navier‐Stokes equations", Engineering Computations, Vol. 30 No. 4, pp. 601-622. https://doi.org/10.1108/02644401311329398