1、-1 NASA TECHNICAL NOTE FLIGHT EVALUATION OF * GROUND EFFECT ON SEVERAL LOW-ASPECT-RATIO AIRPLANES by Paul A. Baker, Willium G. Schweikhard, and William R. Young Flight Research Center Edwurds, Cub$ 93523 NATIONAL AERONAUTICS AND SPACE ADMINISTRATION WASHINGTON, D. C. OCTOBER 1970 R / i however, for
2、the larger and heavier XB-70 air- planes, usable data were obtained with winds as high as 11 knots. The Air Force Flight Test Center (AFFTC) tracking facility provides optimum data when an airplane is near the midpoint on the runway. It was found that a glide- slope indicator light (fig. 5) aided th
3、e pilot in establishing the initial conditions of con- stant angle of attack and steady sink rate. The indicator was used as a reference from which to initiate the airplanes descent so that it would be in the proximity of the ground near the midpoint of the runway. In the F5D-1 airplanes and the F-1
4、04A airplane, the angle-of-attack display was placed just inside the windshield directly in front of the pilot. pilot to determine his relationship to the ground without interrupting his concentration on angle of attack. Because of lack of space in the windshield area, the normal loca- tion for the
5、angle-of-attack display was used in the XB-70 airplanes. This location enabled the DATA ACQUISITION AND ANALYSIS Aircraft Instrumentation Each of the airplanes was instrumented to record angle of attack and control- surface deflection. In addition, the XB-70 aircraft were instrumented to record powe
6、r- lever position. The accuracies and ranges of the sensors installed in each airplane are listed in table 2. Also listed are the ranges and resolutions of the cockpit angle- of-attack displays. Time correlation was attained by using a tone switch mounted on each of the pilots control sticks or colu
7、mns. When on, the switch transmitted a 1000 -cycle-per-second tone over the UHF communication channel which was received by the tracking facility. Tracking Facility The AFFTC Takeoff and Landing Facility provided the external tracking required for the program. The Facility maintains two Askania cine
8、theodolite stations, one near each end of the main runway, as shown in figure 6. function of time were obtained from the Facility during each low approach. Wind speed, wind direction, and air temperature were also recorded by the Facility. tion is described in detail in reference 7. Precision positi
9、on data as a The installa- Method of Analysis In addition to the assumptions of constant angle of attack and power setting, the constant-angle-of-attack technique assumes a shallow flight path (y 2 3“). While 4 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from
10、 IHS-,-,-approaching the runway, the pilot establishes the initial conditions. The angle of attack corresponds to a particular speed and the power setting to a particular sink rate. These steady-state conditions are disturbed by ground effect, causing the aircraft to change both speed and sink rate.
11、 normalized lift coefficient is given by the following equation: The relationship between these accelerations and the XO sin yo + cos yo ,I- (1) ) -($os -yo - q -=40(hcosy-sytcosy ACL qg g g LO C Data Reduction The Askania cinetheodolite camera system began tracking the test airplanes descent when i
12、t was approximately 200 feet (61 meters) above the ground. airplane was still out of ground effect, initial sink rates and approach speeds could be determined. The parameters provided by the tracking facility and atmospheric pres - sures obtained from the Edwards Air Force Base weather station were
13、used as inputs to a computer program which calculated the aircrafts vertical and horizontal position and dynamic pressure each one-fourth second during a run. Because the The time history in figure 7 of an approach made in an XB-70 airplane is typical of the data that can be obtained with the consta
14、nt-angle -of-attack-approach technique. As required, the throttle angle was absolutely constant. slightly, but the increase in lift due to ground proximity still caused the airplane to flare. attack. deflection was required to maintain angle of attack. (3.1 m/sec) in true velocity is qualitative evi
15、dence of increased drag resulting from the nearness of the ground; however, conclusive quantitative analysis of these data in terms of drag increments was not possible. The angle of attack decreased Corrections were applied for minor deviations from the reference angle of Ground effect also changed
16、the pitching moment. The 3“ to 4“ change in elevon The eventual decrease of 10 ft/sec The position data calculated by the computer were reduced to obtain the vertical Two methods were used to obtain h, x, One method was to plot the altitude versus time and then fit a smooth curve The slopes of the r
17、esulting curve represented the rate of sink at each and horizontal velocities and accelerations. h, and G. to the points. time interval. Similarly, plotting the horizontal position versus time and taking slopes provided horizontal velocity. Repeating the procedure by plotting rate of sink and hori-
18、zontal velocity versus time produced the required vertical and horizontal accelerations. The quantities h and 2 were also obtained from the following relationships: 5 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-The rate of sink and horizontal vel
19、ocity were obtained in the manner described in the preceding paragraph. The slopes from a curve of altitude versus rate of sink repre- sent the quantity Similarly, the slopes of the curve of altitude versus hori- zontal velocity represent dh Once the accelerations were obtained, they were used in eq
20、uation (1) to calculatethe normalized increase in lift coefficient. Obtaining the vertical and horizontal accelerations by either of these methods was Consequently, several computer smoothing routines were tried to determine tedious. if they could fit the plotted curves, but none was able to do a sa
21、tisfactory job. were overly influenced by stray points and calculated extraneous accelerations. They PRECISION Accuracy of the Position Measurement The accuracy of the position data from the tracking facility was evaluated (ref. 7) Further, the velocities and accelerations obtained from to be ztl. 5
22、 feet (h0.5 meter). the hand reduction methods are accurate within 50.3 ft/sec (+O. 1 m/sec) and however, significant effects were not encountered until the airplane descended to lower heights. measurable pitching-moment changes. These lift increments generally preceded any The pilots in the program
23、 made several qualitative observations about the extent of ground effect encountered. There were consistent comments on the strong flare and float characteristics of the F-104A airplane, whereas the XB-70 airplanes were noted to become more stable laterally. On the other hand, the F5D-1 airplanes di
24、d not “float“ as much as the XB-70 and F-104A airplanes. float characteristics can be related to the airplanes initial sink rate. rate as low as 4.3 ft/sec (1.3 m/sec) caused the F5D-1 airplanes to touch down; where- as, on one approach the XB-70 airplanes had an initial sink rate of 7.3 ft/sec (2.2
25、 m/ sec) but did not touch down. due to ground effect reduced sink rates of 20 ft/sec (6.1 m/sec) to zero and caused the airplane to stabilize a few feet above the runway for the rest of the approach. Quantitatively, the flare and An initial sink Similarly, the increase in the lift of the F-104 airp
26、lane A mathematical analysis of the drag change due to ground effect is presented in appendix A; however, attempts to quantitatively measure the change in drag produced inconsistent results. The problem is believed to lie with the relatively small magni- tudes of the accelerations that must be measu
27、red. among the inadvertent inputs associated with flying the airplane. Although the change in drag was not measured, some deductions were made by observing the change in true airspeed during the low approaches. As an approximation, a reduction in true airspeed may be interpreted as an increase in dr
28、ag, and an increase in true airspeed, as a reduction in drag. On the basis of the general reduction in speed (tables 4 to 7), it appears that the drag generally increased as the airplane encountered ground effect during the constant -angle -of -attack approaches. niques for extracting drag data due
29、to ground effect need further research. These accelerations are easily lost Analytical methods and test tech- F5D-1 Airplane Modified With an Ogee Wing The effect of ground proximity on the lift and pitching moment of the modified As can be seen, the pitching moment due to F5D-1 airplane is shown in
30、 figure 8. 7 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-ground effect is small (less than I“ change in elevon position). Consequently, the trimmed and untrimmed lift increments are approximately equal at 24 percent at touch- down. A comparison o
31、f flight results obtained by two different flight-test techniques is shown in figure 9. In the constant-altitude-flyby tests of reference 1, measurements were assigned altitude bands rather than discrete heights; consequently, in this figure, the horizontal line is at the center of the original alti
32、tude band, and its length covers the spread of the data within that band. The vertical line at each end of the horizontzl line symbolizes the width of the band. The individual measurements within each band are indicated by the tick marks on the horizontal line. These data are superimposed on the con
33、stant -angle-of -attack-approach data from figure 8. Although the lift- coefficient data show general agreement at the lower heights, the constant -angle -of - attack data are more consistent than the constant-altitude -flyby data. The disparity between the two sets of data increases with increasing
34、 airplane height above the ground. Also, it should be noted that the constant -angle -of-attack data were obtained during 5 runs, whereas the constant -altitude -flyby data required 44 runs. Figure 10 is a comparison of wind-tunnel data and theoretical predictions with data from a 10“ constant-angle
35、-of-attack approach. predicted data from reference 4 are in close agreement in both trend and magnitude; however, extrapolating the wind-tunnel results indicates a lift increment near the ground substantially greater than either the flight or reference 4 results. hand, it should be noted that the wi
36、nd-tunnel data go to zero below one wing span, but the flight and analytical results show lift increments above one wing span. It seems reasonable that the wind-tunnel incremental-lift results go to zero at lower heights be- cause of the negative effect produced by the ceiling of the tunnel. This is
37、 especially true of the full-scale tests in which the model is nearly in the center of the tunnel at the one-half span, and, as expected, the ground effect measured in the tunnel is zero. As can be seen, the flight data and the On the other Flight, wind-tunnel, and theoretical ground-effect data for
38、 the modified F5D-1 are summarized in figure 11 as a function of angle of attack at 0.30 wing span. There is some correlation of trends but little correlation of magnitudes between the various sources of data. The NASA Ames Research Centers constant -altitude -flyby and wind- tunnel data and the dat
39、a of reference 4 indicate a slightly decreasing incremental lift coefficient with increasing angle of attack. On the other hand, the constant-angle-of- attack flight data and the NASA Langley Research Center and Lockheed wind-tunnel data indicate increasing incremental lift coefficient with increasi
40、ng angle of attack. All the wind-tunnel pitching-moment data show the same increasing trend with angle of attack; however, both sets of flight results indicate a relatively constant trim change with angle of attack. Basic F5D-1 Airplane Figure 12 shows the effect of ground proximity on the lift coef
41、ficient and elevon deflection for the basic F5D-1 airplane. The results are similar to those for the modified F5D-1 airplane, in that there is little change in pitching moment due to enter- ing ground effect, as indicated by the small change in elevon position (0“ to 2.4“). Con- sequently, the trimm
42、ed and untrimmed increases in lift coefficient are nearly the same, increasing to approximately 14 percent at touchdown. The difference between 8 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-this lift increment and the 24-percent value for touchdo
43、wn of the modified F5D-1 air- plane (fig. 8) shows the influence of the planform modifications. In figure 13, a typical midrange angle-of-attack approach for the basic F5D-1 air- plane is compared with corresponding reference 4 data and wind-tunnel results. The increase in lift coefficient predicted
44、 by reference 4 agrees well with the flight results; however, the wind-tunnel measurement is considerably higher than either the pre- diction or the flight data. The trim changes shown by the reference 4 and wind-tunnel results were both substantially greater than the flight -measured change. In fig
45、ure 14 the flight, wind-tunnel, and reference 4 results indicate similar trends in change in lift coefficient with angle of attack but differ with regard to the magnitude of the change. Both the wind-tunnel and reference 4 data predict greater lift in- crements in ground effect than were measured in
46、 flight. and wind-tunnel data suggest a greater trim change than was measured in flight. The flight-measured trim changes also indicate no functional dependence on angle of attack, whereas the predicted and wind-tunnel data indicate an increasing trim change with increasing angle of attack. Similarl
47、y, the reference 4 XB-70 Airplanes The effect of the ground on the lift and pitching moment of the XB-70 airplanes is shown in figure 15. Because the only difference between the two aircraft is a positive 5“ dihedral of the wing on the XB -70 -2 airplane, the results from approaches made by both air
48、craft are plotted in the same figure. change and incremental lift coefficient increase as the airplane approaches the ground. At touchdown the trimmed and untrimmed lift coefficients increase to approximately 18 percent and 24 percent, respectively. compared with that for the F5D-1 airplane, probabl
49、y because the XB-70 elevons are less effective than those of the F5D-1 airplane. Significant changes begin below one wing span and increase to 3“ or 4“ change in elevon deflection at touchdown. Again the results show that the trim The elevon increment of the XB-70 is large It may be noted that these results do not necessarily agree with the XB-70 results in references 3 and 10. not specifically flown to obtain ground-effect data. It appeared that the data used in t
