A numerical tool for optimal airfoil design has been described. The design tool uses a fully viscous Navier-Stokes solver to obtain the aerodynamic characteristics of airfoils. The solver design tool is used to compare several optimization algorithms by solving a minimum drag airfoil subject to a minimum lift requirement using three different optimisation algorithms: the modified method of feasible directions (MMFD), the sequential linear programming (SLP) method, and the sequential quadratic programming (SQP) method. The SQP method outperformed the SLP and MMFD methods in computational efficiency. Therefore, of the three existing methods, the SQP method is the method recommended. Furthermore, the paper highlights that it could be possible to increase the efficiency of the SQP algorithm by substituting the line search with a moving limits methodology similar to the one used in the SLP algorithm. This would reduce the number of function evaluations in the SQP algorithm. The SLP method is also considered to be a potential candidate for shape optimization if an efficient method to compute the gradients is available. When using the SLP method, special care must be taken when selecting the initial move limits. Finally, the MMFD method is not recommended because it requires a large number of internal function evaluations that cannot be parallelized, making this method highly computationally inefficient.
Numerical evaluation of optimization algorithms for low-Reynolds-number aerodynamic shape optimization
Numerische Bewertung unterschiedlicher Algorithmen zur Optimierung von Tragflügelprofilen bei niedrigen Reynolds-Zahlen
AIAA Journal ; 43 , 10 ; 2262-2267
2005
6 Seiten, 7 Bilder, 8 Tabellen, 25 Quellen
Article (Journal)
English
Missile Aerodynamic Shape Optimization Using Genetic Algorithms
British Library Conference Proceedings | 1999
|