On Noise Control in Turbomachinery Using an Automated Multidisciplinary Design Optimization System

Abstract

In this work, we review our recent efforts to develop and apply an expanding database of aerodynamic and aeroacoustic prediction technologies for exploring new conceptual designs of propulsion system turbomachinery components optimized for high-efficiency performance with minimum noise radiation. In this context, we first discuss construction of our automated, distributed, industry-like multi-disciplinary design optimization (MDO) environment used in all the studies. The system was developed on the basis of commercially available optimization modules, and involves a user-friendly interface that provides an easy link to user-supplied response analysis modules. We address various issues in the automated optimization procedure with focus on turbomachinery design, including proper geometry parameterization, algorithms selection, and transparent interconnections between different elements of the optimization process. In a benchmark study testing the performance of the system in application to aero/acoustic optimization, we consider a problem of optimal blade design to minimize fan noise, a dominant source of sound radiation both in high-speed fan applications (such as high-bypass-ratio turbofans, propellers of turboprop and IC engines in general aviation, and helicopter rotors) and low-speed ones (including applications in automotive, computer, air-conditioning and other industries). Two approaches are investigated, with the first relying on commercial CFD software coupled with an unstructured mesh generator, and the second employing a panel-based aerodynamic code integrated with an integral acoustic solver. Success of various optimization algorithms (including gradient-based and evolutionary) in finding global minima of the objective function for a noise metric in both unconstrained and constrained optimization processes is examined.