Passive Design for Reduction of High-Speed Impulsive Rotor Noise James D. Baeder Associate Profe,fessor Centerfor Rotorcruji Fdrcarior and Research Depart~nent of Aerospace Engineerir~g Universiry of Maa,gla~ld, College Park, MD 20742 The high-speed impulsive noise from non-lifting rotors is examined in hover at large tip Mach numbers using an Eulerl Navier- Stokes solver. A parametric study is performed to systematically investigate the nonlinear effects of sweep, taper and thinning on high-speed impulsive noise in order to determine the key parameters for the design of a low noise rotor. An untwisted UH- 1H rotor hlade forms the baseline blade for this shldy. Forward sweep is found to be more beneficial for noise rednctinn than rearward sweep, effectively delaying delocalization. At lower tip speeds taper and thinning are more effective for reducing the in-olane noise. However. at high tio soeeds ohasine effects become more imoortant and sween hecomes more effective at re- - .. - ducing high-speed impulsive noise. Combinations of sweep, taper and thinning are investigated. Furthermore, a dogleg plan- form, with backward sweep inboard of the forward swept tip, is included in order to place the aerodynamic center near the root quarter chord. In particular the Forward ~eroacou~tically Swept Thin and Tapered W A S P ) hlade is seen to delay delo- calization well past a tip Mach number of 0.95 while providing at least 10 dB of noise reduction from Ule baseline untwisted UH-IH over the entire tip Mach number range investigated. The numerical results form a rich numerical data base for inves- tigating the ability of simpler methods for predicting high-speed impulsive noise for more complex planforms. Introduction Noisecan severely restrict rotorcraft usage in both civilian and military operations. When it occurs, impulsive noise is unquestionably the loudest and the most annoying source of noise. It is annoying because the ear is panicularly sensitive to pressure changes that occur over a very shon pe- riod of time. There are lwo rypes of impulsive noise: bladovonex interac- tion noise (BVI), and high-speed impulsive noise (HSI). BVI noise is very difficult to model due to the complex unsteady, three-dimensional flow- field including wakes. This paper discusses only HSI noise, its causes and the use of passive design for HSI noise reduction. High-speed impulsive noise is caused by compressibility effects. As the advancing tip Mach number of the rotor increases, the flowfield be- comes nonlinear with the formation of supersonic pockets on the rotor blade. If the tip Mach number is large enough, the phenomenon known as delocalization may occur, wherein the supersonic pocket on the blade ex- tends out to the far-field beyond the rotor. In this case the noise becomes much more impulsive in nature and focused in a narrow region (Ref. I). Fortunately, the influence of lift on HSI noise appears to be secondary (Ref. 2). Thus, one does not need to beconcemed with trying to accurately model the details of the wake, currently one of the most difficult areas in rotorcraft aeroacoustic predictions. Presented at the American Helicopter Socicly 52nd Annual Forum, Wushingion, D.C., June 4-6. 1996. Copyright O 1996 by the American Helicapler Society. Inc. All rights reserved. Mmuscripl received Aug. 1996; accepted May 1998. HSI noise usually occurs in fonvard night. In such conditions, and par- ticularly for non-conventional blade tip shapes, it has been shown (Ref. 3) that unsteady aerodynamics is playing an imponant role in the develop- ment of transonic effects which govern noise radiation and drag. However, accurately predicting these unsteady effects on noise requires more com- putational effon than calculating the hover case. For this reason, and based on (Ref. 21, it is assumed that noise charac- teristics can be in a preliminary approach, approximated in hover Under this assumption, in both cases a similar relationship exists for peak acous- tic pressure as a function o l tip Mach number (Ref. 2). Because of un- steady effects in fonvard flight, HSI noise is usually first studied in hover. Methods to study HSI noise in hover include: (1) the Lighthill (Ref. 4) acoustic analogy approach of Ffowcs Williams and Hawkins (Ref. 5). (2) the stationary Kirchhoffformulation (Ref. 6), (3) the rotating Kirchhoff formulation (Ref. 7). (4) a nonlinear Kirchhoff Cormulation of Isom and Purcell (Refs. 6-a), and (5) a purely computational fluid dynamics (CFD) approach (Refs. 9-12), The first four methods rely on experimental data or numerical calculations from CFD to provide input. Furthermore, if CFD is used as input. the cornputalional domain must include all of the significant nonlinearreeions of the flowfield. Since this means that the CFD must ex- - tend to at least half a rotor radius past the tip of the rotor in the case of de- localization, one might as well use a pure CFD method in order to inves- tigate the effect of various parameters on HSI noise reduction. Detailed experimental measurements to determine the effects of plan- form modifications on llS1 noise hove been somewhat lirnited due to the cost of manufacturing and testing various hlade geometries. Very few sys- tenlatic tests have been conducted to determine the key parameters affect-