Correlation of airloads on a two-bladed helicopter rotor

NASA Technical Memorandum 103982

Correlation of Airloads on a

Two-Bladed Helicopter Rotor

/"-

Francisco J. Hernandez and Wayne Johnson

(NASA-TM-I03982) CORRELATION OF N94-26143 AIRLOAOS ON A TWO-BLAOED HELICOPTER ROTOR (NASA) 26 p Unclas G3/02 0208994 April 1993 National Aeronautics and Space Administration

NASATechnicalMemorandum 103982

Correlation of Airloads on a

Two-Bladed Helicopter Rotor

Francisco J. Hernandez, Ames Research Center, Moffett Field, California Wayne Johnson, Johnson Aeronautics, Palo Alto, California April 1993 National Aeronautics and Space Administration Ames Research Center Moffett Field, California 94035-1000

Correlation of Airloads on a Two-Bladed Helicopter Rotor

FRANCISCO J. HERNANDEZ AND WAYNE JOHNSON* Ames Research Center speed of sound Cs Summary rotor thrust coefficient, T/pnR2(D.R) 2 CT Airloads measured on a two-bladed helicopter rotor in flight during the Ames' Tip Aerodynamic and Acoustic dimensionless blade-section lift, d(CT/ff)/dr Test are compared with calculations from a compre- L/p( D-R ) 2 c hensive helicopter analysis (CAMRAD/JA), and the L blade-section lift (force per unit span) pressures compared with calculations from a full-potential rotor code (FPR). The flight-test results cover an advance rotor-tip Mach number, D.R/c s Mtip ratio range of 0.19 to 0.38. The lowest-speed case is M characterized by the presence of significant blade-vortex blade-section Mach number, U/c s interactions. Good correlation of peak-to-peak vortex- N normal force (per unit span); number of induced loads and the corresponding pressures is blades obtained. Results of the correlation for this two-bladed

rotor are substantially similar to those for three- and four- Q

main rotor torque bladed rotors, including the tip-vortex core size for best P pressure correlation, calculation of the peak-to-peak loads on the r blade radial station, measured from retreating side, and calculation of vortex-induced loads on center of rotation inboard radial stations. The higher-speed cases are charac- terized by the presence of transonic flow on the outboard R rotor radius sections of the blade. Comparison of calculated and T rotor thrust measured airloads on the advancing side is not considered appropriate because the presence of shocks makes chord- U blade-section velocity (normal to span) wise integration of the measured data difficult. However, V good correlation of the corresponding pressures is helicopter flight speed obtained.

X chordwise distance from leading edge angle of attack Nomenclature rotor tip-path-plane angle, positive c blade chord forward

& rotor longitudinal tip-path-plane tilt

C L blade-section lift coefficient, L/1 pU2c 2 relative to shaft, positive forward C N blade-section normal force coefficient, rotor lateral tip-path-plane tilt relative to shaft, positive towards retreating side N/1 pU2c advance ratio, V/DR P Cp blade-surface pressure coefficient, air density P p/lpu2 a,S rotor solidity, Nc/zrR 1V rotor azimuth angle, measured from CQ rotor torque coefficient, downstream in direction of rotor Q / pztR 2 (DR) 2 R rotation f_ rotor rotational speed ° Johnson Aeronauucs, Palo Alto, California.

Introduction teetering rotor with a constant-chord, rectangular- planform blade. The rotor radius was 22 ft. This flight-test The accurate prediction of rotor airloads is one of the program was conducted at NASA Ames Research Center most challenging problems in the field of theoretical aero- during the early 1980s, using a set of highly instrumented dynamics and one that is far from being solved. The diffi- rotor blades to study rotor-tip aerodynamics and acoustics.

culties encountered making such predictions are caused To accommodate the additional instrumentation (ref. 4), by the highly complex nature of the rotor flowfleld, which the blades used a symmetrical airfoil with a modified OLS in forward flight includes compressibility effects on the (Operational Loads Survey) section. With this airfoil, the advancing side, dynamic stall on the retreating side, and blade chord was increased from the standard 27.0 in. to blade-vortex interactions. All of these features of the 28.625 in., and the thickness-to-chord ratio was increased aerodynamic loading must be adequately handled before from 0.09330 to 0.09677. The rotor solidity was 0.06910, an accurate calculation of rotor performance, structural and the twist was linear from root to tip with a magnitude loads, or noise can be expected. This paper examines of-10 degrees. The gross weight of the aircraft was several issues involved in rotor airloads calculations, in approximately 8000 lb.

particular focusing on blade-vortex interaction at low The set of instrumented blades was developed for the speed and the influence of transonic flow at high speed.

U. S. Army Operational Loads Survey (OLS) test. For the In order to look in detail at the factors that affect the accu- TAAT, transducers at three new radial stations were rate representation of the rotor aerodynamics, the compre- added near the tip of the blade. The complete set of pres- hensive helicopter code CAMRAD/JA (ref. 1) was used to sure transducer locations is shown in figure 2. Reference 2 model the AH-1G rotor system. The results are compared presents a comprehensive description of the measure- with measured flight-test dam on an AH-1G taken during ments with an analysis of the key phenomena.

the Tip Aerodynamic and Acoustic Test (TAAT) con- ducted at Ames Research Center (ref. 2). This is the first Flight data from the TAAT test are stored on digital tapes major correlation effort undertaken using the AH- 1G air- and are available through DATAMAP (ref. 5). A cycle loads database.

average of two rotor revolutions was used for the correla- tion with theory in this paper. DATAMAP was also used The CAMRAD/JA model was used to calculate the per- to process the data, making use of its large number of formance, dynamics, and aerodynamic behavior of the analysis tools, which include the derivation of the section rotor. The principal objective of the paper is to compare load coefficients. The airloads database consists of six the measured and calculated blade-section lift at high and forward-speed cases and one hover condition, as shown in low speed. It is useful also to examine the pressure corre- table 1. For this paper, all forward-speed cases were used, lation for a subset of the results. For this purpose, the Full with special emphasis on the low-speed case of 82 knots Potential Rotor code (FPR) was used (ref. 3), coupled and the high-speed case of 159 knots.

with CAMRAD]JA to account fully for blade motion and wake effects. The pressure calculations are also of direct interest for predictions of rotor noise. This paper investi- Analytical Methods gates the detailed pressure loading associated with the blade-section lift correlation.

CAMRAD/JA The calculations are compared with flight measurements The bIade airloading was calculated using the compre- at different advance ratios ranging from/1 = 0.19 to 0.38.

hensive helicopter code CAMRAD/JA (ref. 1). The rotor The lowest-speed case is characterized by the presence of aerodynamic problem is based on lifting-line theory, using significant blade-vortex interactions. The higher-speed steady, two-dimensional airfoil characteristics and a vor- cases are characterized by the presence of transonic flow tex wake. The rotor wake model is based on a vortex lat- on the outboard sections of the blade. This paper describes tice approximation of the wake. A small, viscous-core the flight test, the analytical methods used, the modeling radius is used for the tip vortices. A large core size is used of the AH- 1G rotor, the assumptions made, and the result- for the inboard wake elements to produce an approxima- ing correlation between theory and flight measurements.

tion for sheet elements. A model of the wake rollup is included.

Tip Aerodynamic and Acoustic Test (TAAT) The analysis separates the aerodynamic problem into inner (wing), and outer (wake) problems, which are The aircraft used during the Tip Aerodynamic and solved independently and then combined through a Acoustic Test (TAAT) was the fh'st production AH-1G Cobra built (fig. 1). The AH-1G had a two-bladed, • Table 1. AH-1G TAAT operating conditions Test points (counters) Variable 2152 2153 2154 2155 2156 2157 2370 RPM 307.2 315.0 314.7 315.2 315.5 315.9 321.0 OAT, o C 11.5 18.5 18.5 18.5 18.5 18.5 16.5 Stat. press., psia 13.65 13.60 13.45 13.30 13.24 13.18 14.75 Airspeed, KTAS 159 146 129 116 98 82 0 /.t 0.377 0.341 0.303 0.268 0.230 0.189 0.000 Gross weight, lb 8066 8000 7941 7920 7890 7870 9115 CT x 100 0.474 0.460 0.462 0.464 0.464 0.464 0.485 MR torque, in-lb 224310 184126 152109 130789 105313 93472 -- Long. flap., deg -1.13 - 1.87 -2.20 -2.38 -2.29 -2.13 -5.05 Lateral flapping, deg -I.11 --0.60 --0.51 --0.19 -0.01 0.15 -4.12 Fuselage _t, deg -3.9 -1.7 --0.5 1.4 3.4 4.0 -- Pitch attitude, deg -4.56 -2.36 -2.51 0.37 -0.16 0.89 -4.21 Long. eye. pitch, deg 11.8 10.2 8.9 7.9 6.5 5.5 1.9 Lat. eye. pitch, deg -3.6 -2.4 -2.4 -2.1 -1.8 -1.7 2.4 Collective pitch, deg 18.0 15.8 14.5 13.4 12.2 11.7 14.4 matching procedure. The outer problem consists of an more computationally efficient. These options were used incompressible vortex wake from a lifting line, with in the computations performed for this paper.

distorted geometry. The inner problem consists of FPR Code unsteady, compressible, viscous flow about an infinite- The Full-Potential Rotor code (FPR) (ref. 3) was used to aspect-ratio, yawed wing. The inner problem is split into two-dimensional, steady, compressible, viscous flow calculate the blade surface pressures. It was iteratively (airfoil tables) with empirical corrections for unsteady coupled with CAMRAD/JA to account fully for blade aerodynamics, dynamic stall, and yawed flow. A detailed motion and wake effects. This code, developed by the description of the aerodynamic analysis is given in U. S. Army at Ames Research Center, solves the reference 1.

unsteady, three-dimensional, full potential equation in conservation form. The code employs a finite-difference The rotor structural model is represented by a section scheme that is solved using the method of approximate analysis based on engineering beam theory. The equations factorization. It has been demonstrated (ref. 7) that the of motion are obtained from equilibrium of the inertial, FPR-CAMRAD,rJA analysis produces nearly the same aerodynamic, and elastic forces on the portion of the blade section lift as CAMRAD/JA, and hence is an appropriate outboard of a particular blade section. The interface tool to investigate the detailed pressure loading associated between the aerodynamics and dynamics models is with the blade-section lift correlation.

defined by the section aerodynamic forces and the section velocities.

The grid system used consists of a spanwise series of par- allel O-grids. For the computation of rotor flows, an The wake-geometry models in CAMRAD/JA include: approximate rotational coordinate velocity is assigned to uniform inflow (linear variation of inflow over the rotor each grid point. Boundary conditions consist of a disk), nonuniform inflow with a prescribed wake geome- transpiration velocity at the surface, and nonreflection at try, and nonuniform inflow with a free wake geometry. As the outer boundary.

suggested in reference 6, for/.t < 0.25 the free wake analysis is used because of the highly distorted wake that A typical grid size used for the calculations for the remains close to the rotor plane. For higher advance AH-IG Cobra consists of 80 points in the chordwise ratios, where the wake is convected downward faster, a direction, 25 in the spanwise direction, and 25 in the nor- prescribed wake analysis gives the same accuracy and is mal direction. The finite-difference grid extends in the spanwise direction approximately 4 chords inward of the

spanwise direction approximately 4chords inward ofthe

blade shifts aft 0.281 in. relative to the feathering axis at

tipand1.2 chords outward fromthe tip.The extent ofthis

r/R = 0.31. This is caused by the addition of the instru- computational fluiddynamics (CFD) region isdependent mentation sleeve to the blade.

ontheadvance ratioutilized inthe calculation. At high

The calculated blade collective- and cyclic-mode frequen-

advance ratios, thestalled region ontheretreating side of

cies are given in tables 2 and 3. These were obtained from

thedisklimitstheinboard extent oftheCFD domain. The

the flutter analysis in CAMRAD/JA with no aerodynam- outer boundary of the grid is located 5 chords from the ics included_ The boundary condition was teetering surface of the blade. Constant time steps of 0.25 deg of motion. The first four flapwise modes are identified along azimuth angle were used for the calculations. Computer with the first two lag and torsion modes.

time for a 360-deg computation was approximately 1600 CPU seconds on the NASA Ames Cray Y-MP. Additional For the calculations, the rotor was trimmed to the mea- information on the FPR code can be found in references 3 sured flight conditions, defined by helicopter weight, shaft and 7.

angle of attack, and longitudinal and lateral tip-path-plane angles relative to the shaft. These quantities are given in table 4 (ref. 2).

Model Description Here, ct, tp p is the sum of the shaft angle and the longitudi- The rotor aerodynamic tables utilized in the nal flapping relative to the shaft. In figure 4 these quanti- CAMRAD/JA model of the AH-1G were based on the ties are plotted vs. advance ratio. Notice the drastic Bell Helicopter Textron, Inc. (BHT)-developed 540 airfoil change in tip-path-plane angle for advance ratios greater tables given in reference 4. The BHT 540 airfoil tables than 0.27.

were generated from a wind-tunnel test, with standard corrections applied to maximum lift and profile drag.

In figure 5, the rotor-shaft torque coefficient is shown as a Because data were not available at high angle of attack function of advance ratio. The disagreement at high and high Mach number, NACA 0012 tables are used, with advance ratios may be caused by the presence of dynamic a smooth transition between the two.

stall. However, the calculated power for advance ratios above 0.34 follows the trend of the tip-path-plane tilt The 540 airfoil tables were modified using wind-tunnel (which was the specified trim state in the calculations), data for the OLS/TAAT airfoil from a test conducted at suggesting that measurement of the shaft angle may be in NASA Langley Research Center (ref. 8). The OLSfrAAT error. The correlation is sufficient for the present pur- airfoil was tested over a Reynolds-number range from poses, however, since the airloads calculations are not 3 x 106to 7 x 106, angles of attack from approximately very sensitive to the difference in measured and calculated --4 deg to 12 deg, and Mach numbers from 0.34 to 0.88 in propulsive force implied by figure 5.

the Langley 6- by 28-in. Transonic Tunnel. The BHT 540 airfoil tables were corrected using the Langley wind- tunnel data. The two databases were very similar, except Table 2. Collective-mode frequencies (per/rev) at high Mach numbers. Using the Langley data introduced corrections to the maximum lift coefficient, the lift curve Modes Flap Lag Torsion slope, and drag and moment coefficients at the stall con- 1st 1.109 1.631 2.656 dition. At the high Mach numbers, the Langley data were used, with a smooth transition to the NACA 0012 values. 2nd 3.107 10.709 7.656 In all cases, symmetry between positive and negative 3rd 5.227 -- m angle of attack was imposed, including cases where the 4th 8.831 ....

data did not indicate such behavior. This new airfoil table was used as part of the AH- 1G CAMRAD/JA model.

The baseline for the OLS blade structural properties was Table 3. Cyclic-mode frequencies (per/rev) taken from reference 4. Such data include blade mass, geometric twist, center of gravity offset, tension center Modes Flap Lag Torsion offset, flapwise and chordwise bending stiffness, moment 1st 1.000 1.433 2.869 of inertia, polar radius of gyration, and torsional stiffness.

2nd 2.506 10.329 8.376 Each structural property was generated in a stepwise 3rd 4.515 .....

manner for 48 segments as a function of blade radius.

Some of the blade's structural characteristics are given in 4th 7.502 .....

figure 3. For the OLSbqade, the elastic axis is assumed to coincide with the feathering axis. The quarter chord of the Table 4. Flight-test conditions V, knots /.t eZxp p,deg CT/Cr fls, deg tic, deg 82 0.189 1.24 0.0672 0.15 2.13 98 0.230 2.45 0.0671 --0.01 2.29 116 0.268 2.75 0.0672 --0.19 2.38 129 0.303 4.71 0.0669 --0.51 2.20 146 0.341 4.23 0.0666 --0.60 1.87 159 0.377 5.68 0.0686 -1.11 1.13 Blade Airloads of a significant effect of the blade dynamics on the air- loads was anticipated. To examine this effect in detail, a The rotor airloads will _ presented in the form of section parametric study was done on the first and second blade lift around the azimuth, as defined by the following torsional degrees of freedom. Figure 6 shows the effect of equation: different combinations of torsional degrees of freedom on the blade loading at r/R = 0.91 and r/R = 0.97. It is seen d(C T/S)/dr = L/p(f_R)2c (I) that the effects occur on the front part of the rotor disk, The experimental data provided the normal force, which where the blade undergoes the greatest torsional loads.

for small angle of attack (CL =--CN) and a constant chord Minor differences are seen between the different options, blade can also be written in the form: but using both first and second torsion degrees of freedom, the calculations correlate slightly better with the d(C T IS)/dr = _l MECN (2) flight measurements. This option was selected as the 2Mtip 2 baseline for all subsequent calculations.

Equation (2) was used to convert the flight data into the Effect of Tip-Vortex Core Size same form as the section lift given by the calculations in equation (1).

Because of its great effect in blade airloads, specifically As a check on the AH-1G flight-test data, the total rotor on blade-vortex interaction, the size of the tip-vortex core was varied in the calculations to establish the most lift was calculated and compared with the gross weight of the aircraft. This was accomplished by integrating the appropriate size for each flight condition. The tip-vortex blade pressures chordwise and computing the azimuthal core radius determines the maximum velocity induced by average of the section lift over the rotor disk. The thrust the vortex. Core sizes ranging from 0.015R to 0.040R (approximately 15 to 40% chord) were used and com- was then calculated by integrating the section lift over the blade radius. For the case of V = 82 knots, a rotor thrust pared with flight data at different radial stations. Figure 7 shows such a comparison for V = 82 knots at several of 7989 lb was obtained, which corresponds to an aircraft weight of 7870 lb. The difference between the thrust and radial stations on the blade. The calculations overpredict gross weight might be caused by a download on the fuse- the loading on the front part of the disk and underpredict lage created by the rotor downwash. Again, the correla- the magnitude on the forth quadrant. This overprediction occurs for all radial stations examined.

tion is sufficient for the present purposes, since the air- loads calculations are not particularly sensitive to a thrust Significant blade-vortex interaction is seen at the 90-deg- change of this magnitude.

and 270-deg-azimuth stations. The core size has a modest influence on the peak-to-peak vortex-induced loading on the advancing side, but no influence on the retreating side.

Discussion of Results: Low-Speed Cases This indicates that there is a large vertical separation of the tip vortex from the blade; hence the loads are not sen- Effect of Torsional Degrees of Freedom sitive to tip-vortex core radius. Rather, the loads are more dependent on the strength of the vortex than on its peak The AH-1G rotor blade has a fundamental pitch/torsion velocities. Peak-to-peak amplitudes are well matched on frequency below 3/rev (tables 2 and 3), so the possibility the retreating side. Peak-to-peak amplitudes are under- however) by using a larger core size when calculating the predicted for r/R = 0.864 on the advancing side. For the vortex-induced velocities at inboard collocation points on outboard stations, smaller core sizes generally give a bet- the blade. Figure 10 also shows the improved correlation ter representation of the peak-to-peak amplitudes (fig. 7(e) produced by this model, using a core size of 0.14R for is an expanded view for r/R = 0.91).

inboard stations (transitioning to 0.02R for radial stations from r/R = 0.76 to 0.88). It is therefore observed that This two-bladed rotor exhibits less of an influence of core when the vortex-induced loads are calculated using a core size on the airloads than would be seen for a three- or size that gives good correlation at the blade tip, the four-bladed rotor at the same advance ratio, since with strength of the blade-vortex interactions is significantly two blades, the tip vortex has more time to convect overpredicted for inboard stations. This phenomenon has between its creation and its interaction with the following been observed in other correlation studies as well (ref. 9).

blade. While it is therefore more difficult to deduce the core size based on the present data, it appears that a core Blade-Surface Pressure Distributions radius of about 20% chord is a good choice. This is approximately the same core radius found to be appropri- Calculated blade-surface pressure distributions for ate for three- and four-bladed rotors in reference 9. Notice V ffi 82 knots are given in figure I 1 at r/R = 0.910 for also that the vortex-induced loads tend to be overpredicted azimuths on the advancing and retreating side (near the on the retreating side (a feature also observed in refer- blade-vortex interactions). Data for the last chordwise ence 9 for three- and four-bladed rotors). Since for this pressure transducer (x/c = 0.91) were not available for this case the retreating-side blade-vortex interaction is not radial station. On the advancing side, calculations under- sensitive to core size, this discrepancy may be caused by predict the blade upper-surface pressures, especially the partial tip-vortex rollup. The tip vortex may not be pressure rise close to the leading edge. Better correlation completely rolled up by the time it reaches the following is seen on the retreating side, except at 300 deg azimuth.

blade, so the strength may be less than the value of peak This correlation is consistent with the airloads results bound circulation (as assumed in the calculation).

(fig. 7(b)), where better correlation was obtained for the In figure 8, the influence of core size is shown for retreating-side blade-vortex interaction.

V = 98 knots at r/R = 0.91, r/R = 0.955, and r/R = 0.97.

Figure 12 shows similar pressure correlations for The results are similar to those of the previous case, the r/R = 0.97 (corresponding to airloads in fig. 7(d)). Again, calculations exhibiting only modest influence of core size the calculations underpredict the blade-surface pressures on the advancing side and none on the retreating side. In throughout the rotor disk. At inboard radial stations the order to explore this lack of sensitivity to core size in surface pressure correlation was better, as evidenced in these two cases, a lower-speed case of V = 43 knots (for figure 13 for r/R = 0.60 at three azimuths on the advanc- which no flight-test data were available) was examined.

ing side (corresponding to airloads in fig. 10(b)).

Figure 9 shows this case, in which a more significant effect of core size is seen in both the advancing and retreating side of the disk. This confirms the assumption Low-Speed Airioads Correlation that, at 82 knots, the blade vortex passes the blade at a The present paper and reference 9 compare the measured large vertical distance; thus the core size has a negligible and calculated airloads for the three cases given in table 5.

effect.

All the rotors considered had rectangular planforms In all cases analyzed, small core sizes of about 0,020R (except for the trapezoidal tip cap of the H-34) and linear showed a slightly better Correlation and were used as twist (except for zero twist at the tip of the SA349/2).

baseline for the following calculations.

These three cases exhibit the following common behavior: a) Good correlation with measured peak-to-peak vortex- Inboard Blade-Vortex Interaction induced loads is obtained using a tip-vortex core radius of Figure 10 shows the airloads calculated using an inboard Table 5. Low-speed airloads correlation cases core size of 0.02R, on both outboard and inboard radial stations. Although good calculation of peak-to-peak vortex-induced loads is achieved on outboard stations Rotor N cr c/R CT/O /./ (r/R > 0.91), the oscillatory loads are significantly over- AH-IG 2 0.069 0.101 0.067 0.19 predicted on inboard stations when this core size is used.

SA349/2 3 0.064 0.067 0.065 0.14 CAMRADIJA can simulate this effect (with no implica- H-34 4 0.062 0.049 0.082 0.18 tion that the physics of the phenomenon are understood, approximately 20% chord. This val:t]e i_in the range of appears betwe_the x/c = 0.25 and 0.35 pressure trans- measured core sizes for rotor wakes, although it is proba- ducers. The baseline section lift is obtained using direct bly still somewhat too large.

interpolation between the measured pressures (fig. 16(a)).

Figures 16(b) and 16(c) show alternate interpretations of b) With a single tip-vortex core size for the wake model, the pressure change associated with the shock, which will and assuming that the strength of the tip vortex equals the produce integrated lift coefficients higher and lower than peak bound circulation of the blade when the vortex is the baseline. Figure 17 shows the measured and calculated generated, there is a tendency to overpredict the peak-to- section lift and lift coefficient, including, for several peak loads on the retreating side and underpredict on the azimuths, results of these alternate interpretations of the advancing side. This result suggests that, on the retreating flight-test data. The correlation between calculated and side, the core size is larger, or that the vortex strength is measured airloads is evidently very sensitive to such less than the peak bound circulation.

interpretations. Notice also that considering lift rather than c) Something is happening on the inboard part of the lift coefficient magnifies the differences, because it is the blade to reduce the measured vortex-induced loads.

advancing side that is of interest.

References 6 and 9 speculate that this effect is associated Another phenomenon present in the experimental data at with a smaller blade-vortex separation calculated for this flight condition is the sharp bump in section lift at interactions on the inboard part of the blade than that r/R = 0.955, seen in the second qua&ant (e.g., fig. 14(c)).

calculated for interactions at the tip of the blade.

A similar effect has been observed in flight and wind- tunnel data on other rotors. To examine this effect, figure Discussion of Results: High-Speed Cases 18 presents the surface pressures before, at, and after the peak lift. The pressures indicate that at 120 deg azimuth there is a strong shock wave present on both upper and Blade Airloads lower surfaces that rapidly decreases in magnitude at 135 deg and is non-existent at 150 deg azimuth. Because Airloads were calculated for the higher-speed cases pre- sent in the database, using the baseline parameter values of the absence of pressure tranducers on the lower surface determined for 82 knots. Figure 14 compares the mea- between 5% and 25% chord, direct interpolation of the sured and calculated airloads at 159 knots _ = 0.377). pressures (as illustrated by fig. 18(b)) produces a sharp For all the tip radial stations shown, the measured section rise in the integrated Section lift as soon as the lower- surface shock moves forward of 25% chord. Hence this lift exhibits a roughly constant value in the first quadrant of the disk before dropping to near or below zero at about and other sharp changes in the measured section lift are most likely attributable to a lack of sufficient data to 90 deg azimuth. In contrast, the aerodynamic mechanisms present in the calculation produce a continuous decrease define the movement of the shocks on the advancing side of the section lift in the first quadrant, as would be of the blade, rather than to any aerodynamic phenomenon of rotor blades in transonic flow.

expected because of roll moment balance. There is a slight bump in the calculated airloads, produced by interaction Given the difficulties associated with making a highly with the tip vortex from the preceding blade. In the analy- instrumented rotor blade, it may be impossible to achieve sis (ref. 6), the tip vortex has negative strength in this a chordwise and spanwise resolution sufficient to examine region, since the tip loading is negative. Figure 15 com- all aerodynamic phenomena. A more sophisticated pares measured and calculated airloads for speeds of 116, method of integrating the pressures to obtain section lift, 129, and 146 knots. The results are similar to those at specifically one that estimates the shock location implied 159 knots. The behavior of the measured airloads, and the by the data, would be most useful for future tests. For the resulting correlation with calculations, have been present investigation, it is concluded that there is little to observed in investigations of high-speed airloads with be gained by trying to compare the measured and other rotors (ref. 6). For this two-bladed rotor, with con- calculated section lift.

stant airfoil and low-aspect-ratio blades, significant com- pressibility effects occur even at relatively low advance ratios (fig. 15). Blade-Surface Pressure Distributions The nearly constant lift present in the plotted experimental Figures 19 and 20 compare the measured and calculated data on the advancing side may be caused by difficulties blade-surface pressure coefficients for V = 159 knots at r/R = 0.970, for several azimuths in the first and second in determining the proper shock location during data reduction. As an example, figure 16 Shows the measured quadrants, respectively. Transonic flow is seen to be pre- pressure coefficient at 75 deg azimuth, where the shock sent at 30 deg azimuth and becomes more dominant close to 90 deg. The shock strength and location are well

predicted bythe calculations. UpperZsurface pressures are

Blade-surface pressures were compared for the high-speed

underpredicted inmost cases, The phenomena evident in

case and correlated fairly well with flight data, even

the7neasurements aregenerally captured bythe calcula-

where measured and calculated lift disagreed. Shock

tions, withtheexception ofthe delay information ofthe

location and magnitude were good, but upper-surface lower-surface shock at60deg azimuth.

pressures were underpredicted in most cases. The phenomena evident in the measurements were generally captured by the calculations, with the exception of the Conclusions delay in formation of the lower-surface shock in the first quadrant.

Airloads calculations were performed for a two-bladed rotor using a comprehensive helicopter analysis and were correlated with flight-test data. Fairly good correlation of References peak-to-peak vortex-induced loads was obtained for a low-speed case. Sensitivity to tip-vortex core size at this 1. Johnson, W.: A Comprehensive Analytical Model of speed was observable but small on the advancing side, Rotorcraft Aerodynamics and Dynamics, and negligible on the retreating side. This relative insensi- Johnson Aeronautics Version. Johnson tivity to core size (compared to a four-bladed rotor at the Aeronautics, 1988.

same advance ratio) is attributed to a larger vertical sepa- 2. Cross, J. L.; and Watts, M. E.: Tip Aerodynamic and ration of the tip vortex from the blade. Therefore, the Acoustic Test. NASA RP-1179, December 1988.

peak-to-peak loading is more dependent on the vortex strength than on the core size. Nonetheless, the results of 3.

Strawn, R. C.; and Caradonna, F. X.: Conservative the correlation for this two-blade rotor were substantially Full-Potential Model for Unsteady Transonic similar to the results for three- and four-bladed rotors, Rotor Flows. AIAA J., vol. 25, no. 2, February concerning the tip-vortex core size for best correlation, 1987.

calculation of the peak-to-peak loads on the retreating .

Van Gaasbeek, J. R.: Validation of the Rotorcraft side, and calculation of vortex-induced loads on inboard Flight Simulation Program (C81) using radial stations.

Operational Loads Survey Flight Test Data.

Blade-surface pressures for the low-speed case correlated USAAVRADCOM-TR-80-D-4, July 1980.

reasonably well with flight data, particularly on the .

Philbrick, R. B.: The Data from Aeromechanics Test retreating side. This was consistent with the section lift and Analytics--Management and Analysis correlation, where the second blade-vortex interaction was Package (DATAMAP), Volume 2mSystems better matched.

Manual. USAAVRADCOM-TR-80-D-30B, Airloads were calculated for the higher-speed cases pre- December 1980.

sent in the database. All cases showed similar differences .

Johnson, W.: Wake Model for Helicopter Rotors in between the calculations and the flight measurements. In High Speed Flight. NASA CR-177507, particular, the measured section lift exhibited a roughly November 1988.

constant value in the first quadrant before dropping to near zero, whereas the calculations show a continuous 7.

Strawn, R.C.; Dessoper, A.; Miller, J.; and Jones, A.: decrease of the section lift. It was speculated that the Correlation of Puma Airloads Evaluation of behavior of the flight data was caused in part by d___cul- CFD Prediction Methods. NASA TM-102226, ties determining the shock location when integrating the August 1989.

measured pressures. It was demonstrated that sharp .

Watts, M. E.; Cross, J. L.; and Noonan, K. W.: Two- changes in the measured section lift are most likely Dimensional Aerodynamic Characteristics of the attributable to a lack of sufficient pressure transducers to OLS/TAAT Airfoil. NASA TM-89435, April define the movement of the shocks on the advancing side 1988.

of the blade. For the present investigation, it was con- cluded that there is little to be gained by trying to compare Johnson,W. Calculation of Blade-Vortex Interaction .

the measured and calculated section lift.

Airloads on Helicopter Rotors. J. Aircraft, vol. 26, no. 5, May 1989.

Figure 7. AH-1G Cobra test helicopter at NASA Ames.

3f 3_ 3E 3E 3E3( 3E X X X X X XX X X X X X XX X .2 X XX X X X X X XX X X X X X XX X "¢ .4 X X XX X X X X X X XX X X XX X X X X X XX X

"g

X XX X X X X X XX X X X X X XX X I 1.0 I I I I I .9o .4-

_I

.2 -I- ,,¢ -il- .=

4=

-i- .8 -I- I 1.0 Z I Z I I I I I .4 .1 .2 .3 .5 .6 .7 .8 .9 1.0 r/R Figure 2. All- 7G TAA T blade pressure instrumentation locations (94 upper, 94 lower transducers).

Center of gravity .15 m _ Neutral axis --- Quarter chord axis = elastic axis ! .05 • _ o • _,. ,i..... W._. ........... .. ................

location ---I --IMI l .t.. + -.05 (o.os3) II u_ -.10 I I I I I .2 .4 .6 .8 1.0 r/R Figure 3. All-7 G blade description.

]o -- O - - Longitudinal flapping -- _ -- Shaft angle of attack - @ Tip-path-plane angle of attack .2 .O .t -- .O" _'_ ".2 i -.4 /J .-O -.6 e I -.8 ..J /i -1.0 I Flight test -- -- -- Calculations °l I I I .010 I I -1.2 .o69 P tt .068 t !

i t I- t t o .... o- ..... o...

.067 .11112 O-. I t .0114I I I I I I I I I I I .O6( .15 .20 .25 .30 .35 .40 Advance ratio Figure 5. Rotor shaft torque coefficient vs. advance A ratio.

// ¢ 3 / C O " 2 I_ / ""o /

o- ,,_, /

A" -2 I I ..... I I I .15 .20 .25 .30 .35 .40 Advance ratio Figure 4. Trim quantities used in the analysis.

]!

Flight test No torsion "1 First torsion L> d.o.f., Second torsion I" calculation .... First and second torsion J .16 .12 r_" .08 "_ .04 O m u) 0 • (el r/R = 0.91 I f I I -.04 .16 .12 _- .08 "_.04 o

o

(b) r/R = 0,97 I I I !

-.04 90 180 270 360 Azimuth (deg) Figure 6. influence of torsion degrees of freedom on airioads (V = 82 knots).

.2O " .12 .16 _" .o8 g , la) r/R = 0.864 (d) r/R = 0.97 I I I I 04 I I I I 0 90 180 270 360 0 9O 180 270 360 .2O .08 , .16 tf I(e) r/1:l : 0.91 -.04 I I I I I =, I I I 0 90 180 270 360 0 45 90 135 180 Azimuth (deg) .2O .16 L.

.12 I- Flight test o o.ot s 1 0.020 k __'_ 0.030 1 _ Core (R), calculation .... 0.040 J ', (c) r/R = 0.955 I l, I I -._4 0 9O 180 270 360 Azimuth (deg) Figure 7. Effect of vortex core size on airloads (V = 82 knots).

-,---,---- Flight test 0.0IS 1 0:030 Core (R), calculation .16 .12 o_.oe i .040 Core m) 0.015 l (a) r/R = 0.91 0:02_ _> Calculation .2O ] I I I -.04 .19 0.040 J A .16 (b) r/R = 0.955

i

l_ .12 .12 I-- o _..o8

]-04

i

r/R : 0.91 I I I I -.04 "'040 I I I I 90 180 270 360 .16 (c) r/R = 0.97 Azimuth (deg) Figure 9. Effect of vortex core size on airloads .12, (V = 43 knots, p = O. I0).

.04

]o

I l I r , l I I -.04 0 90 180 270 360 Azimuth (deg) Figure 8. Effect of vortex core size on airloads (V = 98 knots).

]4 Flight test 0.14'[ -r-, 0.02_ Cu (R), inboard, calculation .2o (d) r/R 0.864 (a) r/It = 0.40 .16 .12 11% i1 (n II II

oil

I , , \7- v l I I I I .2O _)r/R=0.60 "(e) r/R = 0.91 | I\ .16 ,\ _L _, i- .12 (J

A ,f

¢ .08 I i I I I I I I I I .2O (c) r/FI = 0.75 (f) r/n = o.gs5 ,.^,_ ^_, _J_ .16 i.

/ _" .12 e- .05 \1 .04 !

I I I I I I I 90 180 270 90 180 270 360 0 Azimuth (deg) Azimuth (deg) Figure 10. Effect of inboard core size on airloads (V = 82 knots).

.20 '(g) r/R = 0.97 ......--.- Flight test .16 0.14"L _ 0.02,I" Core (R), calculation _" .12 /) \ .08 k _ / .04 I 90 180 270 36O Azimuth (deg) Figure 10. Concluded.

Flight test ..... Calculation, FPR-CAMRAD/JA 3 _ _ = 285" 1.0 .5 -Cp 0 ...... _'_'_i -1.0 °'51 I I 15 I | I f I 1.5 _= 90 ° = 270 ° 1.0 .5 -Cp 0 -Cp -1.0 -.5 l I I I I I I I I I I -1 o5 -2 _= 105" = 300 ° 1.5 IA 3 1.0 2 .5 c' I ....

-I"S I I I I I -2 -1.0 1 -1 I I I I I 0 ,2 .4 .6 .8 1.0 0 .2 .4 .6 .8 1.0 x/c x/c Figure 11. Calculated and measured blade-surface pressure distributions (V = 82 knots, r/R : O.910).

Flight test ..... Calculation, FPR-CAMRAD/JA = 90 ° 1.5 _P = 75" 1.5 .5 .S 1.0 _ 1.0 -Cp 0 -Cp 0 !

-.5 _ -.5 -1.0 -1.0

l ' r

! ! I ! I -1.5 I I ! I I .1.5 1.5 = 105 ° = 270 ° 1.0 .5 -Cp 0 -Cp !

-1 -1.0 "5 I I t I I I I I I I I -1.5 -2 3[ _=3_0 _ _ = 285 ° -Cp -1 -1 l _ .2 / ....... I I I I I -2 I ! I I I 0 .2 .4 .6 .8 1.0 0 _ .4 .6 .8 1.0 x/c Wc Figure I2. Calculated and measured blade-surface pressure distributions (V = 82 knots, r/R = 0.970).

Flight test 2.5 -- ..... Calculation, FPR-CAMRAD/JA - V = 30 ° 2.0 I I J tt 1.5 1.0 -Cp .5 -.5 -1.0 I I I I I -1.5 2.0 2.5 f _ = 90° 1.5 1.0_ -Cp -1.0 I I , ] I I -1.5 2.5 , _= 180 ° 2.0 1.5 t • 1.0 -Cp .5 ".5 -1 o0 I I I I I -1.5 0 .2 .4 .6 .8 1.0 xlc Figure 13. Calculated and measured blade-surface pressure distributions (V = 8Z knots, r/R ,= 0.60).

]9 Flight test E__ Calculation .24 " (c) r/R : 0.955 r ( a ) r _ : 0 . 8 6 4 ._ .2O .16 .12

,,\ //

.08 C _.o4 u_ \A._/ I I I I -.04 I I I I .24 "(d) r/R = 0.97 -20_ • (b) r/R = 0.91 _/ o "0 .d .12 ¢= /// 0 .08 '_ ,04 // I I I I I | I I -.04 90 180 270 360 0 00 180 270 360 Azimuth (deg) Azimuth (deg) Figure 74. Calculated and measured blade-section rift (V - 159 knots).

2O .24 V = 116 knots Flight test Calculation .20 1.5 Flight test, baseline C, I- w .12 .5 _g.oe -Cp 1.0 G -.5 ".04 I I I i (a) I I I I I -1.0 V • 129 knots 1.5 Maximum C L .16, ,, I .12 1.0 , l / .5

.o8

/ -.

\\%\ _ -Cp 0 J -.5 -,_4 I i I I (b) I | I | I -1.0 .24 V- 146 knots .2O Minimum C L 1.¢ _,,16 F- U .12 .S -Cp __._ -.S °.04 I I I I (c) I I I | . I 0 90 180 27O 360 -1.00 .2 .4 .6 .8 1.0 Azimuth (deg) x/c Figure 15. Calculated and measured blade-section Figure 16. Determination of maximum and minimum rift at different advance ratios (r/R - 0.955).

rift due to shock location (V = 759 knots, r/R - 0.970, y - 75°).

Flight test ....... Calculation, FPR-CAMRAD/JA = 120 °

l

Flight test Calculation, CAMRAD/JA .28 -1.0 Change due to shock location (flight teat) "'5 I -- -- -- Calculation, FPR-CAMRAD/JA la) ! !

.24 I I I -1.5 I I , .20 1.0 r- _= 135 ° _Z k- .16 _" .12 C o .08

-cp

_ .114 -1.0 ..Sl ( I I I I I b) .t.s ! I I I I 1.0 1.0 - ?=150 o .8 .5 -Cp °.5 -1.0 ! 1 I I I -.2 , I z = -1.5 0 90 180 270 360 0 .2 .4 .6 .8 1.0 x/¢ Azimuth (deg) Figure 18. Blade-surface pressures at section rift Figure 77. Maximum change in section rift and rift bump (V = 759 knots, r/R = 0.955).

coefficient to shock location (V = 759 knots, r/R = 0.970).

Flight test .... Calculation, FPR-CAMRAD/JA 2.5 " 1.s|,- + = _o 2.0 1.0 1.5 .5 1.0 -Cp .5 -Cp i t % -.5 -I.0 -1.0[ I I I I I -1.5 -1 5 I I I I I t.5 m 1.5 1.0 1.0 .5 .5 ¢_-7, -Cp 0 -Cp 0 • ¢__= _a,,,b I I !

-.5 -1.0 "'51 -1 .OF -1.5 I I I I I -1,5 I I I I I 1.5 1.5

: 45= / 9o0

1.0 .5 -Cp -Cp 0 "%a _- !

-*5 -1.0 -1.0

s I

I | I I I -1.5 I I I I I -1.5 4 .2 .4 .6 .8 1,0 0 .2 .4 .6 .8 1.0 x/c x/c Figure 79. Calculated and measured blade-surface pressure distributions (V- 759 knots, r/R = 0.970).

Flight teat .... Calculation, FPR.CAMRAD/JA 1.5 _= 150 ° = 120 ° 1.0 .S -Cp 0 !

-1.0 "'5 I I I I I I I I I I I -1,5 B 1.5 _ = 180" _= 135" 1.0 .5 -Cp 0 qli I -1.0 I I I I I I I I ] I -1 5 .2 .4 ,6 .8 1.0 0 .2 .4 .6 .8 1.0 X/C x/c Figure 20. Calculated and measured blade-surface pressure distributions (V = 159 knots, r/R = 0.970).

Form Approved

REPORT DOCUMENTATION PAGE oM8No. o7o..o188

Public reporting burden_or this collection of information is estimaled to average 1 hour per response, i_e_m_ for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, Including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washinglon, DC 20503.

1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED April 1993 Technical Memorandum 4. TITLE AND SUBTITLE 5. FUNDING NUMBERS Correlation of Airloads on a Two-Bladed Helicopter Rotor 505-59-36 6. AUTHOR(S) Francisco J. Hernandez and Wayne Johnson (Johnson Aeronautics, Palo Alto, California) 8. PERFORMING ORGANIZATION 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) REPORT NUMBER Ames Research Center A-93001 Moffett Field, CA 94035-1000 9. SPONSORING/MONITORING AGENCY NAME(S) ANDADDRESS(ES) 10. SPONSORING/MONITORING AGENCY REPORT NUMBER National Aeronautics and Space Administration NASA TM-103982 Washington, DC 20546-0001 11. SUPPLEMENTARY NOTES Point of Contact: Francisco J. Hernandez, Ames Research Center, MS 237-5, Moffett Field, CA 94035-1000 (415) 604-1322 12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE Unclassified-Unlimited Subject Category - 02 13. ABSTRACT (Maximum 200 words) Airloads measured on a two-bladed helicopter rotor in flight during the Ames' Tip Aerodynamic and Acoustic Test are compared with calculations from a comprehensive helicopter analysis (CAMRAD/ JA), and the pressures compared with calculations from a full-potential rotor code (FPR). The flight- test results cover an advance ratio range of 0.19 to 0.38. The lowest-speed case is characterized by the presence of significant blade-vortex interactions. Good correlation of peak-to-peak vortex-induced loads and the corresponding pressures is obtained. Results of the correlation for this two-bladed rotor are substantially similar to those for three- and four-bladed rotors, including the tip-vortex core size for best correlation, calculation of the peak-to-peak loads on the retreating side, and calculation of vortex- induced loads on inboard radial stations. The higher-speed cases are characterized by the presence of transonic flow on the outboard sections of the blade. Comparison of calculated and measured airloads on the advancing side is not considered appropriate because the presence of shocks makes chordwise integration of the measured data difficult. However, good correlation of the corresponding pressures is obtained.

15. NUMBER OF PAGES 14. SUBJECT TERMS Airloads correlations, Two-bladed rotor, Vortex-induced loads 16. PRICE CODE A03 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT 17. SECURITY CLASSIFICATION OF ABSTRACT OF REPORT OF THIS PAGE Unclassified Unclassified NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89) Prescribed by ANSI Srd Z39-1S