Bluff Body Drag

A large portion of the drag of a helicopter is due to the bluff body drag of the rotor hubs and landing gear. The main and tail rotor hubs are bluff bodies, which,

because of their rotation and function, are impossible to fit with simple streamline fairings that might keep the flow from separating. This is especially true at high speeds when the rotor is tilted nose down. Wind tunnel tests have generally led to the conclusion that streamlining individual bits and pieces without unduly blocking possible air paths between them leads to the lowest hub drag. A summary of published wind tunnel tests of both faired and unfaired rotor hubs is shown in Table 4.2. For comparison purposes, the drag values are presented for zero angle of attack and zero rpm.

The obvious way to minimize the hub drag is to keep the relative size of the hub as small as possible, as on the Vertol CH-47 and on the Hughes OH-6A. The effect of angle of attack and rotor speed on hub drag is shown in Figure 4.22 for an unfaired hub and for a faired hub. At least part of the difference between hubs is due to the lift-induced drag on the faired hub.

The drag of the rotor shaft can be estimated from the cylinder drag data of Figure 4.23. The total drag of the rotor hub and mast in close proximity to a fuselage or mast pylon may be higher than if they were isolated. This is due to separation on the fuselage or pylon triggered by the neighboring bluff body wake

TABLE 4.2 Rotor Hub Drag Data Helicopter No. of Blades Hub Frontal Area, (ft2) Hub-to- Disc Area Ratio Drag Coefficient Equivalent Flat Plate Area (ft2) Reference Unfaired Faired Unfaired Faired CH-47 3 5.0 0.0027 1.38 0.88 6.9 4.4 ■ 4.9 OH-6A 4 1.5 0.0028 1.13 0.80 1.7 1.2 4.10 UH-1B 2 5.6 0.0037 0.98 0.45 5.5 2.5 4.11 LOH wind 3 1.65 0.61 0.53 1.0 0.9 4.18 tunnel model* LOH wind 2 1.15 0.47 — 0.54 — 4.18 tunnel model1 S-58 4 7.5 0.0031 1.53 0.57b 11.5 4.3 4.19 S-65 4 16.6 0.0041 1.01 — 16.8 — 4.12 S-65 4 21. T — 0.59 — 12.9 4.12 S-65 4 33.9d — 0.22 — 7.6 4.12 Wind tunnel 1 3 0.062 1.26 — 0.078 — 4.14 model | 3 0.102 — 0.76 — 0.078 4.14 AS Twinstar 3 2.5 0.0026 1.55 3.9 4.20 AS Puma 4 5.8 0.0030 0.98 5.7 4.20 AS Dauphin 4 4.3 0.0037 1.56 6.7 4.20 aModel had no control system. ‘’Model had a boundary-layer control system. QRigid head fairing. dFloating head fairing.

“shaft = 0°

.85

I

.80

0 20 40 60 80 100

Percent rpm

FIGURE 4.22 Effect of Angle of Attack and rpm on Hub Drag

Source: Linville, “An Experimental Investigation of High-Speed Rotorcraft Drag,” USAAMRDL TR 71-46, 1971.

on the hub. Obviously, this interference drag is a function of the exact configuration, but one set of tests reported in reference 4.21 and summarized in Figure 4.24 may be taken as typical. One way of decreasing the interference drag is to suck off the low energy boundary layer, as was done on the S-58 wind tunnel model in Table 4.2. Another way of doing the same thing—though perhaps not as effectively—is to use a pylon cap, as on some Sikorsky and Yertol helicopters. This

FIGURE 4.23 Drag of Circular Cylinders

Source: Hoerner, “Fluid Dynamic Drag," published by author, 1965.

cap acts as a low aspect ratio wing whose tip vortices energize the boundary layer on the aft portion of the pylon and thus delay separation. A comprehensive survey of hub drag, along with suggestions for minimizing it, is to be found in reference 4.22.

Yet another interference drag is caused by the rotor downwash on the aft fuselage, which can induce areas of local separation. Figure 4.25 shows test results from reference 4.23 of this drag for one configuration. The interference drag increases with increasing angle of attack, apparently because the aft portion of the fuselage becomes more susceptible to separation triggered by wake turbulence as its pressure gradient becomes more and more unfavorable. The wake behind the hub constitutes a low-pressure sink that can draw the flow off of the upper fuselage, thus producing separation. Again it is obvious that this type of interference drag is highly dependent on the configuration; but in order to evaluate it accurately for a specific design, rather elaborate wind tunnel models are required with a rotor that has the correct disc loading mounted separately from the drag model, but in the correct relative position. In lieu of this, it is suggested that Figure 4.25 be used for most applications.

Nonretracting landing gears also generally produce bluff-body drag. Reference 4.2 gives the drag coefficients of several types of wheeled landing gear. Several of its examples are shown in Figure 4.26. Skid gear are combinations of tubes and struts of various shapes, and two examples are also shown in Figure 4.26. Measurements of landing gear drag on small-scale wind tunnel models may be high

Source: Keys & Wiesner, “Guidelines for Reducing Helicopter Parasite Drag,” JAHS 20-1, 1975.

because of the low test Reynolds numbers of the components. Figure 4.23 can be used to estimate this effect.