Stochastic Optimization of Aeroelastic Response (SOAR)

Period of Performance: 01/01/2003 - 12/31/2003

$100K

Phase 1 STTR

Recipient Firm

Mechanical Solutions, Inc.
11 Apollo Drive Array
Whippany, NJ 07981
Principal Investigator
Firm POC

Research Institution

Rutgers University
Engineering Building, Room B203
Piscataway, NJ 08854
Institution POC

Abstract

Modern airfoil design evaluation requires accurate analysis of both structural dynamic as well as fluid oscillatory response. The level of calculation reliability has, in principle, increased significantly, so that now the opportunity exists to optimize aeroelastic behavior in the design stage. This is made possible by new high fidelity structural and CFD tools which predict airfoil steady and transient aero loading, as well as stress and vibrational behavior. However, while single runs of the applicable COTS codes can be parallel-processed reasonably quickly, computational time becomes pacing in airfoil development due to the many iterations that must be performed in order to optimize lift/weight ratios, while ensuring that aeroelastic flutter is avoided. Automated Multi-Disciplinary Optimization (MDO) computer algorithms can address this analysis bottleneck, but attempts to date have had only limited success, due to problems with slow convergence and local minima. MSI will locate the true minimum of the multi-disciplinary objective function through a unique optimization approach making appropriate use of stochastic methods. The technique accounts for design constraints, such as component thicknesses and aspect ratios, for airfoils that are practical and cost-effective. Using MSI?s Stochastic Optimization of Airfoil Response (SOAR) methodology, time-savings of several-fold are expected in airfoil design optimization.