Supersonic Flying Qualities Experience Using the SR-71

NASA Technical Memorandum 4800

Supersonic Flying Qualities

Experience Using

the SR-71

Timothy H. Cox Dryden Flight Research Center Edwards, California Dante Jackson Analytical Services and Material, Inc.

Edwards, California National Aeronautics and Space Administration Office of Management Scientific and Technical Information Program SUPERSONIC FLYING QUALITIES EXPERIENCE USING THE SR-71 Timothy H. Cox NASA Dryden Flight Research Center Edwards, California Dante Jackson AS&M Edwards, California CH Cooper-Harper rating Abstract acceleration due to gravity g Approximately 25 years ago NASA Dryden Hight instantaneous vertical speed indicator Research Center, Edwards, California, initiated the IVSI evaluation of supersonic handling qualities issues using KEAS knots equivalent airspeed the XB-70 and the YF-12. Comparison of pilot lift curve slope of the aircraft, ibf comments and ratings with some of the classical Let National Aeronautics and Space handling qualities criteria for transport aircraft provided NASA Administration information on the usefulness of these criteria and insight into supersonic flying qualities issues. A second research s LaPlace operator study has recently been completed which again stability augmentation system addressed supersonic flying qualities issues through SAS evaluations of the SR-71 in flight at Mach 3. Additional 1 To2 high-frequency pitch attitude zero, rad/sec insight into supersonic flying qualities issues was the change in a parameter obtained through pilot ratings and comments. A These ratings were compared with existing military 0 pitch attitude, deg specifications and proposed criteria for the High Speed pitch attitude command, deg Civil Transport. This paper investigates the disparity 0c bandwidth time delay parameter, sec between pilot comments and the Neal/Smith criteria xp through a modification of the technique using vertical phase angle, deg speed at the pilot station. The paper specifically t_ phase angle at twice the phase crossover addresses the pilot ability to control flightpath and pitch t_2tots0.

attitude in supersonic flight and pilot displays typical of frequency, rad/sec supersonic maneuvering.

gain-limited pitch attitude bandwidth O')bwgain parameter, rad/sec Nomenclature phase-limited pitch attitude bandwidth fd'JbWphas¢ parameter, rad/sec ADI attitude director indicator flightpath bandwidth parameter, rad/sec bw bandwidth O)bw. t pitch attitude bandwidth parameter, rad/sec CAP control anticipation parameter ¢Obw0 short period frequency, rad/sec c.g. center of gravity tOsp phase crossover frequency, rad/sec (0180o Aerospace Engineer. Introduction ' Aerospace Engineer.

Copyright © 1997 by the American Institute of Aeronautics and Good handling qualities are essential for aircraft Astronautics, Inc, No copyright is asserted in the United States under Title 17, U.S. Code. The U.S. Government has a royalty-free license to performance and can be predicted during the design exercise all rights under the copyright claimed herein for Governmental process by analytic means. Handling qualities criteria purposes. All other rights are reserved by the copyright owner.

used to evaluate aircraft designs are defined by empirical American Institute of Aeronautics and Astronautics includes pilot comments and ratings of the SR-71 flown data obtained from previous aircraft. A problem exists, with a well-defined vertical altitude plane change however, when an aircraft flies in a new flight regime.

maneuver. Data from this study were used to assess the Airplanes cruising at speeds greater than Mach 2 or applicability of handling qualities criteria, especially beyond are examples of this problem.

those used in the design of the High Speed Civil Transport program. 7 Flying qualities criteria do not address some of the unique characteristics of high-speed flight because they A comparison of the pilot's ability to directly control are based primarily on subsonic data. For example, many of the current criteria assume that good flightpath flightpath and pitch attitude in supersonic f/ight and an evaluation of the vertical speed display are specifically response follows from good pitch attitude response; addressed in this report. The Neal/Smith criteria, 8 however for the high-speed case, this assumption may not be valid. As Mach number increases, the lift curve bandwidth criteria on pitch attitude, 8 and bandwidth criteria on flightpath 9 are used to evaluate pilot control slope of an aircraft (Lct) decreases proportionally, of flightpath and pitch attitude. Throughout the analysis, thereby increasing the lag between flightpath and pitch results of these criteria are compared with the piloted attitude response. If this lag characteristic is too large, evaluations. In all cases, criteria for category C flight the pilot's ability to control flightpath will be impaired.

were assumed to be applicable to the high-speed tasks.

Another unique characteristic of high-speed flight is the decrease in pitch attitude change to achieve a rate of Aircraft Description climb. This characteristic, due primarily to the increased velocity, implies that as the speed increases the pilot The SR-71 aircraft (fig. 1) is a twin engine, delta-wing must maintain precise control of pitch attitude to airplane designed to cruise at a speed of Mach 3.2 and to establish the desired altitude response. Unless accurate altitudes above 80,000 ft. The SR-71 aircraft is powered pitch attitude or rate-of-climb information is fed back to by two Pratt & Whitney (West Palm Beach, Florida) J-58 the pilot, this characteristic could potentially cause afterburner engines with axisymmetric, variable- the pilot to overcontrol the aircraft. This increased geometry, mixed compression inlets. Centerbody spikes sensitivity to pitch attitude control in high-speed flight is and bypass doors located on the forward part of the not as prevalent in subsonic flight. Nevertheless, flying nacelle are automatically modulated to control the qualities criteria of pitch attitude are based on subsonic oblique and normal shock positioning that is associated data.

with flying at high supersonic speeds. Data gathered in this report occurred with doors and inlets in this Approximately 25 years ago NASA Dryden Hight automatic configuration.

Research Center researchers gained insight into these issues by applying MIL-STD-8785B criteria I to YF-12 The majority of the cockpit contains conventional and XB-70 data. The longitudinal tasks for the YF-12 instrumentation. Some of the main cockpit instruments and XB-70 transport class aircraft, in up and away flight, used during this evaluation include a pressure-driven included precise flightpath tracking without gross instantaneous vertical speed indicator (IVSI) and a triple maneuvering, which was considered category C flight.

display indicator that shows altitude, equivalent The researchers showed for both the YF-122,3'4 and the airspeed, and Mach number in a digital format. The IVSI XB-702,5 positive correlation between category C is a circular gauge with a needle indicating vertical criterion on the control anticipation parameter (CAP) speed to a resolution of I00 ft/min. The resolutions of and pilot comments and ratings. The researchers also the triple display indicator parameters are 50 ft, 1 kn, indicated the requirements for short period damping may and Mach 0.01, respectively. Because a lag in the be relaxed although these results were not considered response of the IVSI exists at high altitude, a horizontal conclusive. 4.5 Researchers also gained insight into needle displaying inertial vertical speed located on the unique issues of supersonic cruise flight, such as the attitude director indicator (ADI) provides a reference for usefulness of an inertial vertical speed display 2.3 and climb and descent rates. This gives SR-71 pilots a more the problems associated with unstable long period precise and reliable vertical speed indicator than the IVSI.

dynamics. 2,6 Recently a second study was completed which extends Wing trailing-edge elevons are used symmetrically as the research of supersonic flying qualities issues elevators and differentially as ailerons to provide conducted by the XB-70 and YF-12 programs through longitudinal and lateral control, while twin all-movable evaluations of the SR-71 aircraft in flight at Mach 3. Data vertical tails supply directional control. The pilot American Institute of Aeronautics and Astronautics EC94 42883-04 Figure 1. SR-71 aircraft.

controls consist of a conventional stick for pitch and roll includes attitude hold (in pitch, roll, or both), Mach inputs and rudder pedals for yaw inputs. The SR-71 number hold, and knots equivalent airspeed (KEAS) aircraft has a conventional response with angle-of-attack hold. Normal aircraft maneuvering is executed by pitch and normal acceleration changes commanded by the and roll attitude inputs through thumbwheels, while acceleration and deceleration to and from Mach 3 are pitch stick. The controls are irreversible and fully powered by two independent 3000 lb/in 2 hydraulic performed with KEAS hold engaged on the autopilot.

systems that operate actuating cylinders at each control The autopilot mode is used routinely in the climb, surface.

cruise, and descent portions of the flights. However for the handling qualities evaluations, all autopilot modes The SR-71 flight control system provides a stability were disengaged, and maneuvers were performed manually with the stick.

augmentation system (SAS) to increase damping about all three axes. This is accomplished with conventional feedback of roll rate, pitch rate, and yaw rate. In addition, Maneuver Description and Pilot Evaluations lateral acceleration is used in the yaw axis to reduce the severity of engine unstarts. Three maneuvers were flown at Mach 3 to evaluate the handling qualities characteristics of the SR-71 aircraft: a An autopilot is available to reduce the workload steady level turn, an ascending turn, and a vertical plane involved in flying the SR-71 aircraft. The autopilot altitude change. These maneuvers were considered to be American Institute of Aeronautics and Astronautics

typical of maneuvers to be flown by a large supersonic

Pilot transport ah'craft. Reference 7 describes each maneuver [] A 9 -- _ e and documents pilot comments and ratings. For the Level 3 purpose of this paper, only the description and evaluation of the vertical plane altitude change is considered.

[] Two pilots flew the maneuver a total of five times. Pilot ratings and comments evaluating each maneuver using Cooper41arper predefined adequate and desired performance margins <_] Level 2 were collected immediately after performing each task and during postflight briefings.

<I] Maneuver Description Level 1 The vertical plane altitude change at constant KEAS

I I I

entailed a wing level pullup to capture a 2000-ft altitude 1 IVSI Inerthllly derived increment at a climb rate of 1000 ft/min. Once the target vertical speed altitude was established, it was to be held for an 970372 additional 10 see. Constant airspeed was to be Figure 2. Pilot ratings for the vertical plane altitude maintained throughout the maneuver. The pilots change maneuver.

evaluated this maneuver with two variations of feedback displays: one using the IVSI and the other using the inertial vertical speed on the ADI.

establish. As long as enough range of throttle motion in afterburner was available, the airspeed was easily Adequate margins for the maneuver were +300 ft maintained as well. The only difficulty arose when deviation from target altitude and + 10 KEAS deviation searching to establish rate of climb and when leading the from target airspeed. Desired margins for this maneuver aircraft as it approached the target altitude. This minor were +100 ft deviation from target altitude and compensation required in establishing flightpath added +5 KEAS deviation from target airspeed.

to the basic concentration necessary to fly this airplane and warranted CH = 3 (level 1).

Summary of Pilot Evaluations Pilot Display of Vertical Speed Pilots' comments using the IVSI and the inertially derived vertical speed are summarized below for the This section presents insight on the pilot display of vertical plane altitude change. Figure 2 presents the venical speed necessary for typical maneuvering in high- Cooper-Harper (CH) ratings for pilots A and B.

speed flight. As was mentioned earlier, the vertical plane altitude change maneuver was flown using two types of IVSI vertical speed indicators as feedback to the pilot: a pressure-driven one (IVSI) and an inertially derived one.

Flying the maneuver with the IVSI as a vertical speed indicator made establishing and maintaining the desired Figure 3 compares the altitude time histories of two rate of climb very difficult. This is caused by sluggish initial response and an excessive delay between the stick IVSl input and a reaction in the IVSI gauge. The altitude 500 f_min 7 _ Inertial change was able to be performed but not without high / concentration and some loss of performance. The ability A altitude of the pilot to hold airspeed was decreased because of the from initial 1000 excursions in rates of climb. These problems warranted condition, , ft ratings of CH = 5-7 (levels 2-3).

50 1 0O 150 20O Inertially Derived Vertical Speed Time, llec 970373 Flying the maneuver using the inertial derived vertical Figure 3. Comparison of the vertical plane altitude speed for feedback made achieving desired performance change maneuver using the IVSI and an inertiaily relatively easy. The desired rate of climb was easy to derived vertical speed indicator.

American Institute of Aeronautics and Astronautics a very long lag, and the pilot cannot use it as a feedback vertical plane altitude change maneuvers. One maneuver variable."

was flown using the IVSI and the other was flown with the inertially derived vertical speed. This figure plots the increment in altitude from the initial altitude for each of In contrast, the ability of the pilot to establish the 1000 ft/min rate of climb was greatly improved when the the maneuvers to provide a common reference point for maneuver was flown with the inertially derived vertical comparison.

speed (fig. 3). Although the pilot misjudged the target altitude by 500 ft, he was able to establish the desired It is clear from observing the maneuver flown with the altitude precisely. Pilot comments indicated that this IVSI that the pilot was unable to establish the correct rate maneuver was "to some extent a nonevent. You establish of climb (1000 ft/min). Instead, the pilot overshot rate of the 1000 ft/min and then sort of nod off. You establish a climb by a factor of 3 and then overcompensated to new trim and then be alert enough to see when you want 500 ft/min. By reducing the rate of climb to 500 ft/min, to start to level off. I held 75,500 right on at the level off.

the pilot was able to hit the target altitude within the This is easy to do." The pilot rating for this maneuver desired performance of 100 ft. The pilot comment on this was much improved also, CH = 3 (level 1). The maneuver was "rate of climb [is] all over the place.

conclusion is that to accurately control flightpath the Could not keep rate of climb anywhere near where I pilot must have an inertially derived vertical speed wanted it." The corresponding pilot rating was CH = 5.

indication.

The difficulty in performing the maneuver derived from the lags in the pressure-driven IVSI. The static pressure Pilot Control of Pitch Attitude and Flightpath decreases which are associated with high altitude make changes in altitude difficult to measure. Thus, the pilot tends to overdrive rate of climb. Figure 4 shows the This section presents insight into whether it is more relative amounts of lag contributed by the IVSI and the appropriate for the pilot to control pitch attitude or flightpath for maneuvering typical of high-speed flight.

flightpath response lag, l/T02, as a function of altitude. 2 At an altitude of 70,000 ft, nearly 12 sec of lag Pilots typically control flightpath by commanding a pitch is present, one-half of which is caused by the IVSI. One attitude change and then waiting for the desired pilot concluded that the "vertical velocity indication has flightpath to develop. This technique applies to normal, Altimeter lag Tlmelag, 8ec FIIghtpath response lag 50,000 55,000 60,000 65,000 70,000 75,000 80,000 Altitude, ft 97o374 Figure 4. Relative time lag comparison of flightpath response and altimeter. 2 American Institute of Aeronautics and Astronautics P.ilr, h.l_algi_i_a subsonic flight when the pilot uses pitch attitude as the feedback parameter either through visual cues obtained To analyze pitch attitude control, the pitch bandwidth by looking out the canopy window or through the ADI.

criterion as defined in MIL-STD-17978 was evaluated For conventional aircraft in subsonic flight, this with SR-71 and reproduced YF- 12 data. The gain-limited technique is generally successful. On the other hand, how does this apply to high-speed flight, where the bandwidth, defined as the frequency at the magnitude increase in lag of the flightpath response may make this which is 6 dB above the magnitude at the phase crossover frequency, and phase limited bandwidth, technique unusable? In such a case, direct pilot control of flightpath may be more appropriate. To directly control defined as the frequency where a 45 ° phase margin exists, were calculated from a pitch attitude from stick flightpath, the pilot must use either flightpath or vertical deflection frequency response. The lesser of the two speed as the feedback parameter.

frequencies was considered the bandwidth frequency.

This criterion places limits on the bandwidth frequency Several analysis techniques are used to evaluate the as a function of the time delay, which is estimated from pilots' ability to control pitch attitude and flightpath: the phase at twice the phase crossover frequency: pitch bandwidth analysis, pitch bandwidth as a function of flightpath bandwidth analysis, frequency response analysis, and Neal/Smith analysis. To evaluate the analysis techniques, the results from criteria are (1) compared with the ratings and comments from the maneuvers using the inertially derived vertical speed indication. The calculation of the bandwidth frequency for SR-71 and YF-12 data with the SAS turned on proved straightforward. However, the YF-12 data included four Reproduced YF-12 data were used to supplement the test points where the SAS was turned off. A typical SR-71 handling qualities data for some of the analysis.

example of the pitch attitude bandwidth calculation for Enough information existed in the reference 3 report to extract the test conditions of the YF-12 data. Because of low damped, SAS turned off, YF-12 data is shown in figure 5. The calculation of the phase bandwidth value, the similarity of the YF-12 and SR-71 aircraft in the the frequency where the phase is -135 °, is longitudinal axis, a flight-validated SR-71 linear simulation was used to model the YF- 12 data. straightforward. However, the calculation of the gain (Llbw,gain = 1.8-_ 20-- (_ow,gain = 0.8-_ 0 6 dB Magnitude, - 20 -- _- (ObW, gain = 0.19 dB _40 I I I I II II I I I I llll -60 0 h _- Ol 80 ° -lOO f 0 = - 135 _ = - 180 - 20o =14 I (_bw, phase I" -_ I IIII t I I I I I L l llll -300.1 .2 .3 .4 3 4 5 6 7 8 910 .5 .6.7.8.9 14) 2 Frequency,rad/aec 970375 Figure 5. Example pitch attitude bandwidth calculation typical of a low damped YF- 12 test point.

American Institute of Aeronautics and Astronautics qualities, because the gain margin is very sensitive to bandwidth value is less obvious. Applying the definition slight changes in phase. 8 The results of the analysis using of the gain-limited bandwidth to the data in figure 5 the lowest gain bandwidth (represented by 0.19 rad/sec, results in three possible gain bandwidth values: 1.8, 0.8, fig. 5) and the phase bandwidth are compared in the and 0.19 rad/sec. Note that a slightly increased phase following discussion.

crossover frequency, (o180o, would prevent this phenomenon from occurring. If the value of 1.8 rad/sec Figure 6 shows the pitch bandwidth results. The pilot were considered the gain bandwidth value, then the ratings for each point are plotted next to the point. The phase bandwidth value would be compared to the criterion since it is less than 1.8 rad/sec. However, if analysis indicates that good pilot control of pitch attitude is possible with the SAS turned on. The level 1 pilot either the 0.8 or 0.19 rad/sec values are the appropriate ratings of these two points correlate with the analysis.

gain bandwidth value, then these values would be When the SAS is turned off and the damping decreases, compared to the criterion because they are less than the pitch bandwidth decreases. In addition, pilot control of phase bandwidth value.

pitch attitude deteriorates. The amount of deterioration in the predicted pilot control of pitch attitude depends on This phenomenon results because of the large, shelf- whether the gain or phase-limited bandwidths are used in like characteristic created by the significant difference the analysis. Use of the gain-limited bandwidths between 1 To2 and the short period frequency as well drastically reduces the pitch bandwidth to around 0.25, as the low short period damping. The low short period near the level 3 border. The pilot ratings, which are damping characteristic produces the three possible gain mainly level 2, appear to correlate better with the phase- bandwidth values. Standard procedure for the limited bandwidth values. Although the reason behind application of this criterion would be to choose the lesser this phenomenon is unknown, it may be that ignoring the value, 0.19 rad/sec, as the gain bandwidth. The rationale two lower frequency gain-limited bandwidths is is that if the pilot tries to close the loop at the higher gain appropriate for this type of transfer function.

bandwidth values a tendency to oscillate will occur because of the lightly damped peak. The large, shelf-like Although the pitch bandwidth analysis is supported characteristic produces a wide separation between the with pilot evaluations and indicates that good pitch three gain bandwidth values. The existence of a large attitude control is possible, it may be a misleading result.

shelf is generally an indication of poor handling .35 0 SAS off - phase bandwidth 0 SAS off - gain bandwidth X SAS on .3O .25 Level 3 .2O "Cp, aa¢ .15 .10 .O5 __ 3.0 3.8 _ Level 1

I I I I I

.5 1.0 1.5 2.0 2.5 3.0 CObwo, rad/sec 970376 Figure 6. MIL-STD-1797 pitch bandwidth criteria for category C flight.

American Institute of Aeronautics and Astronautics bandwidth exists, it may not necessarily relate to good Because the vertical speed indicator is the primary flying qualities if the pilot were to control pitch attitude.

feedback display, the pilot ratings and comments appear driven more by the flightpath response than the pitch Pitch Bandwidth as a Function of Flightpath Bandwidth attitude response. Thus, it is uncertain whether the &0a!xm bandwidth analysis, which assumes pilot feedback of pitch attitude, accurately reflects the pilot evaluations.

Because pilot comments and ratings appear linked to The level 2 pilot evaluations of the SAS turned off data flightpath response, applying an additional criterion on could reflect degraded flightpath response and not flightpath bandwidth may be insightful. A criterion on degraded pitch attitude response. In this case, the pilot flightpath bandwidth as a function of pitch attitude may not even care about pitch attitude response, and the bandwidth was evaluated against SR-71 and reproduced correlation of the pilot ratings with the phase-limited YF-12 data (fig. 7). Phase bandwidth values of pitch pitch attitude bandwidth may be coincidental.

bandwidth are used for the SAS turned off YF-12 data.

The pitch attitude bandwidth analysis for the SAS The SAS turned on data falls well within the level 1 turned on data in figure 6 may also be misleading region for both flightpath and pitch attitude bandwidth.

because of the existence of the large shelf in the pitch This analysis indicates that good pilot control of pitch attitude frequency response. Reference 8 documents the attitude and flightpath control is possible. However, the potential of a frequency response which has a large shelf addition of the SAS turned off YF-12 data reduces the and a gain-margin-limited bandwidth frequency to pitch attitude bandwidth and increases the flightpath provide bandwidth frequency estimates that give bandwidth enough to be near the border between levels 2 optimistic predictions of handling qualities. Although and 3. Because pilot ratings appear more dependent on the SAS turned on bandwidth frequency estimates are flightpath than on pitch attitude, the increase in flightpath phase margin limited, the gain-margin-limited bandwidth could be a factor which degrades the SAS frequencies are very close to the phase-margin-limited turned off evaluations although the pilot ratings are frequency, and both fall right on the edge of the shelf.

better than what would be predicted by this criterion.

Small changes in gain could provide for large reductions Still, it is unclear whether the degradation results from in phase margin, providing misleading predictions of the increase in flightpath bandwidth or the decrease in pitch attitude control. Thus, although good pitch attitude pitch attitude bandwidth.

1.4 o s,so,

30 xs,so°

1.2 4.3 1.0

Z/ x .).4 3.0

.8 jj ....

J \\

Level 2 .4 Level 3 .2

I I I I i I

.5 1.0 1.5 2.0 2.5 3.0 O)bw O, rad/sec 970377 Figure 7. Criteria on pitch bandwidth vs. flightpath bandwidth.

American Institute of Aeronautics and Astronautics the SAS turned on. These frequency responses were Closer scrutiny of the flightpath bandwidth data shows some problems, however. This criterion sets generated with an SR-71 linear model that was validated requirements on the amount of lag between pitch attitude with flight data. A large shelf exists in the pitch attitude and flightpath response. The flightpath bandwidth values frequency response which is normally associated with for the SAS turned off data in figure 7 are above the bad pitch attitude control. The large shelf exists because level 1 region. This implies that the flightpath response of the significant separation in 1/Toz (which causes the lags the pitch attitude, not by too much as is expected, lag between flightpath and pitch attitude) and short but by too little. This result is counter intuitive because period frequency (tOsp). However, the large shelf is physically the lag between pitch attitude and flightpath actually the cause of good flightpath control. For the increases with Mach number. If flightpath bandwidth frequency range where the large shelf exists, the were the cause of the degradation in pilot ratings flightpath frequency response is characteristic of an between the SAS turned on and turned off data, the SAS integrator, or l/s. The 1/s characteristic is typical of turned off data would be expected to be below the level aircraft with good flying qualities. Thus, pilot control of 1 flightpath bandwidth limit (at 0.6), not above. Thus, flightpath should be better than pitch attitude control.

pilot evaluations are inconclusive in supporting these Other types of control systems could remove the shelf - criteria.

improving pitch attitude control, such as pitch rate command systems. However, such systems would cause One possible explanation for this data is that the upper flightpath control to deteriorate.

limit on flightpath bandwidth was set based on the need to separate the pitch attitude and flightpath responses for Neal/Smith Analysis precision approach and landing. In this flight phase, both pitch attitude and flightpath response need to be The Neal/Smith criterion typically involves closing the controlled by the pilot. For the data in figure 7, the pilot loop around a pitch-attitude-to-stick-deflection transfer ratings appear to be based only on flightpath control. The function and a lead-in-lag compensator. This procedure separation requirement between flightpath and pitch is modified by the addition of pure time delay, to meet attitude response may not be as important for these specific closed-loop characteristics. The characteristics conditions. The pilot simply may not care as much about the smaller pitch attitude response in cruise flight as of the closed-loop frequency response are defined as compared to landing.

-90 ° of phase at the bandwidth frequency and no less than -3 dB of droop (fig. 9). The bandwidth frequency Frequency Response Analysis represents the piloting task which is being conducted and is generally chosen based on fight phase. Criteria are Additional insight on the issue of whether it is more established based on the lead required of the appropriate to control flightpath or pitch attitude is compensator to meet the characteristics and the provided through figure 8. This figure presents a maximum amplitude, or resonant peak, of the frequency comparison of the magnitude of the pitch attitude and response of the closed-loop system; for example, the flightpath from stick defection frequency responses with compensator and airplane.

m I i

i I

p Magnitude, 0 ...... O) i dB Flightpath

t l l llllll i i i ii lll L il it

.02 .04 .06 .08.10 .2 .4 .6 .8 1.0 2 4 6 8 10 Frequency, racVsec 970378 Figure 8. Comparison of pitch attitude and flightpath from stick position frequency response.

American Institute of Aeronautics and Astronautics

F

I

IO_cl, dB -3 -10

I

Frequency, rad/sec Bandwidth frequency -9O deg - 180 970379 Figure 9. Illustration of application of Neal/Smith criteria.

To examine the issue of whether pilot control of pitch maneuver when using vertical speed instead of pitch attitude or flightpath is more appropriate, the Neal/Smith attitude. Therefore, the Neal/Smith analysis supports criteria were applied to SR-71 flight data at Mach 3.0 using direct pilot control of flightpath as opposed to using frequency responses of pitch attitude from stick direct control of pitch attitude.

deflection and vertical speed from stick deflection. The vertical speed data were obtained by integrating the Pilot evaluations of the SR-71 aircraft compare normal acceleration at the pilot station. (The inertially favorably with the Neal/Smith analysis using vertical derived vertical speed was calculated by the navigation speed in figure 10. The pilot ratings and comments of the system, which is located just aft of the cockpit.) The data SR-71 aircraft noted that leading the aircraft was presented here use 0.3 sec as the time delay of the required to acquire the target altitude during the vertical compensator. Three bandwidths ranging from 1.5 to plane altitude change. One pilot commented that "a great 2.5 rad/sec were analyzed with the pitch attitude deal of lead is required [to acquire the target altitude] in frequency response to evaluate the results for increases terms of time." However, the lead is not significant in the demands of the task. Bandwidth from 1.0 to enough to reduce the ratings from level 1 to level 2 for 2.0 were analyzed for the vertical speed frequency the inenially derived vertical speed data (fig. 2). The response. Neal/Smith analysis using the vertical speed at the pilot station predicts level 1 flying qualities up to 1.5 rad/sec Figure 10 presents the results comparing the two (category C bandwidth requirement) while requiring frequency responses. For the compensator using pitch pilot lead. Direct pilot control of vertical speed could attitude, significant amounts of pilot lag are required, result in large, objectionable pitch rate overshoots.

driving the flying qualities into the level 2 region. Level 1 ratings from bandwidths of 1.0 to 1.5 result with the Figure 11 presents a step input into a validated batch compensator using vertical speed, which was consistent simulation at Mach 3 to illustrate the pitch rate overshoot with pilot comments. In addition, the slope of that occurs when a small vertical speed increment is degradation (Aresonant peak/Apilot lead) as bandwidth made. No pilot comments indicating an abrupt initial increases using vertical speed is much less than when response were noted. Thus, pilot ratings and comments using pitch attitude. This indicates that the flying appear to be based on the vertical speed response, while qualities are less sensitive to increasing demands of the the pitch attitude response goes unnoticed. Although the American Institute of Aeronautics and Astronautics 0 Vertical speed m 4 to 2.5 Level 3 X Pitch attitude

/

:/

Resonant X 2.0 Category C bandwidth peek, dB frequency _ 4/ - X 1.5 2-- o ---_4.s 1.0 Level 1 --2 -- _\\\

I J I I_1 I

-4 - 20 0 20 40 60 80 Pilot load, deg 970380 Figure 10. Neal/Smith results using pitch attitude and vertical speed from stick position frequency responses as a function of bandwidth frequency.

.5 Pitch rate, deg/lmc

I I

--.5 Vertical speed, 20 Wee c 10

1 I I I

0 1 2 3 4 5 6 7 8 9 10 Time, eec 970381 Figure 11. Comparison of pitch rate and vertical speed responses at Mach 3.

American Institute of Aeronautics and Astronautics 2.0 = bw X 0 Vertical speed at pilot station 14! X Vertical speed at e.g.

\\\\\\\\\" 10-- 8-- Resonant pes_ 6 Level 2 dB 2 0 bw 2-- 1,0 = bw _,, \\ -2 m Level I

1 I I

--4' -40 -20 D 20 4D 60 80 Pilot lead, (leg 970382 Figure 12. Neal/Smith analysis of the vertical speed at the center of gravity and the vertical speed at the pilot station to stick deflection frequency response.

pitch rate overshoots are relatively large, the absolute addressed in this study. The following results are magnitude of the pitch attitude changes are small at high reported: speeds, so the overshoot characteristics are not a factor in The pilots' ability to perform the maneuvers was pilot opinion. 1.

strongly influenced by the information displayed to the pilot. Feedback of vertical speed based on an The NeaYSmith analysis can also be used to illustrate inertial reference frame was critical to the pilots' the affect of the location of the inertial vertical speed ability to perform accurate flightpath control. Pilot measurement. Figure 12 shows the results for the inertial ratings and comments of maneuvers flown with a system in its present location near the cockpit and at the pressure-driven vertical speed indicator showed center of gravity. The cockpit location data provide lead that performance was significantly degraded.

due to the pitch acceleration, thereby reducing the 2.

amount of lead required by the pilot. The cockpit Analysis indicates that direct pilot control of location data also reduce the sensitivity of the response flightpath for a conventional airplane in supersonic for increased demands of the task.

flight has significant advantages over pilot control of pitch attitude. A modification of the Neal/Smith Conclusions analysis using the vertical speed at the pilot station as feedback for pilot control produced level 1 results for a bandwidth up to 1.5 rad/sec. Pilot Pilot comments and ratings of the SR-71 aircraft flown ratings and comments are consistent with the at Mach 3 were collected using a vertical plane altitude level 1 Neal/Smith analysis using vertical speed at change and compared to established handling qualities the pilot station. In comparison, the analysis using criteria. The objective of this study was to extend some pitch attitude feedback to the pilot produced level 2 of the supersonic flying qualities research conducted results for the same bandwidths. In addition, the 25 years ago with the XB-70 and YF-12 programs using large shelf in the pitch attitude from stick deflection these new data. Analysis of pilot displays of vertical frequency response, typically associated with poor speed critical to performing the maneuvers and flying qualities, supported the Neal/Smith analysis comparisons of the pilots' ability to directly control pitch using pitch attitude frequency response.

attitude and flightpath were among the issues specifically American Institute of Aeronautics and Astronautics . Aircraft Safety and Operating Problems Conference, Application of criteria on pitch attitude and flightpath bandwidth was inconclusive in the vol. 2, NASA SP-271, May 1971, pp. 25-38.

evaluation of pitch attitude and flightpath control.

3 Berry, Donald T., "A Summary of YF-12 Handling Although the pitch bandwidth criterion showed Qualities," YF-12 Experiments Symposium, NASA good pitch attitude control was possible and CP-2054, Sept. 1978, pp. 31-57.

correlated with pilot ratings, this criterion may not be sufficient to predict good supersonic flying 4 Berry, Donald T., Donald L. Mallick, and Glenn B.

qualifies. Ratings were based on flightpath control Gilyard, "Handling Qualities Aspects of the NASA using vertical speed indications to the pii :_t.

YF-12 Experience," Proceedings of SCAR Conference, NASA CP-001, Nov. 1976, pp. 193-213.

.

Application of the flightpath bandwidth ctiterion showed that good flightpath control using pitch 5 Berry, Donald T. and B. G. Powers, "Flying Qualities attitude was possible. However, the flightpath of a Large, Supersonic Aircraft in Cruise and Landing bandwidth analysis was inconsistent with degraded Approach," AIAA 70-566, May 1970.

high-speed SAS turned off ratings and comments.

The SAS turned off flightpath bandwidth data fall 6 Berry, Donald T. and G. Gilyard, "Airframe/Propulsion above the minimum level 1 value in a region where System Interactions_An Important Factor in flying qualities limits are defined by poor Supersonic Aircraft Flight Control," AIAA 73-831, Aug.

consonance between pitch attitude and flightpath in 1973.

the approach and landing phase. This factor may 7 Cox, Timothy H. and D. Jackson, Evaluation of High not be as influential for the high-speed evaluations Speed Civil Transport Handling Qualities Criteria with because the pilots used direct feedback of vertical Supersonic Flight Data, NASA TM-4791, Apr. 1997.

speed in cruise.

8 "Flying Qualities of Piloted Vehicles," Military References Specifications MIL-STD-1797, Mar. 1987.

9 Hoh, Roger H., "Recommendations for Approach 1 "Flying Qualities of Piloted Airplanes:' Military and Landing Flying Qualities," Systems Technology, Specifications MIL-STD-8785B, Aug. 1969.

Inc., Hawthorne, California, Working Paper 2631-1, 2 Berry, Donald T., "Some Handling Qualities Problems Dec. 1988.

of Supersonic Cruise Aircraft" Proceedings from NASA American Institute of Aeronautics and Astronautics

REPORT DOCUMENTATION PAGE Form Approved

OMB NO. 0704-0188

Publicrlcx)rtmgburden for thiscollectionof informetcPn is estimatedto average 1 hourper response, inctuding the time forreviewing instructions, searching existingdata sources,gathering and maintaining the date needed, and completingand rmnewing the collectionof information. Send comments regardingthis burden estimate or any other aspect o_this collection of information, including suggestions lor reducingthisburden, to Washington HasOcluarters Services.Directorate for InformationOperations and RS _orts, 1215 JeffersonDavisHighway,Suite 1204, Arlington, VA 22202-430_, and to the Officeof Managementend Budget. PaperworkReduction Propect (0704-0188}, Washington,De 20503.

3. REPORTTYPE AND DATES COVERED 1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE Technical Memorandum August 1997 5. FUNDING NUMBERS 4.TITLE AND SUBTITLE Supersonic Flying Qualities Experience Using the SR-71 WU 529 50 24 s. _(s) Timothy H. Cox and Dante Jackson 8. PERFORMING ORGANIZATION 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) REPORT NUMBER NASA Dryden Flight Research Center P.O. Box 273 H-2178 Edwards, California 93523-0273 9. SPONSORING/MONITORING AGENCY NAME(B) ANDADDRESS(ES) 10. SPONSORING/MONITORING AGENCY REPORT NUMBER National Aeronautics and Space Administration NASA TM-4800 Washington, DC 20546-0001 11. SUPPLEMENTARY NOTES Presented as AIAA 97-3654 at the AIAA Atmospheric Right Mechanics Conference, New Orleans, Louisiana, August 11-13, 1997.

Timothy H. Cox, Dryden Flight Research Center, Edwards, California; Dante Jackson, Analytical Services and Material, Inc., Edwards, California.

1211. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE Unclassified--Unlimited Subject Category 08 13. ABSTRACT (Maximum 200 words) Approximately 25 years ago NASA Dryden Flight Research Center, Edwards, California, initiated the evaluation of supersonic handling qualities issues using the XB-70 and the YF-12. Comparison of pilot comments and ratings with some of the classical handling qualities criteria for transport aircraft provided information on the usefulness of these criteria and insight into supersonic flying qualities issues. A second research study has recently been completed which again addressed supersonic flying qualities issues through evaluations of the SR-71 in flight at Mach 3. Additional insight into supersonic flying qualities issues was obtained through pilot ratings and comments. These ratings were compared with existing military specifications and proposed criteria for the High Speed Civil Transport. This paper investigates the disparity between pilot comments and the Neal/Smith criteria through a modification of the technique using vertical speed at the pilot station. The paper specifically addresses the pilot ability to control flightpath and pitch attitude in supersonic flight and pilot displays typical of supersonic maneuvering.

14. SUBJECT TERMS 15. NUMBER OF PAGES Bandwidth criteria, Flightpath bandwidth, Flying qualities, High Speed Civil 16. PRICE CODE Transport, Neal/Smith criteria, SR-71 aircraft, Supersonic flying qualities.

A03 17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20.LIMITATION OFABSTRACT OF REPORT OF THIS PAGE OF ABSTRACT Unclassified Unclassified Unclassified Unlimited NSN 7540-01-280-5500 Avai/ab/e from the NASA Center for AeroSpace Informatlon, 800 E/kridge Landing Road, Standard Form 298 (Rev. 2-89) Linthicum Heights, MD 21090; (301)621-0390 Pre_,_,d by ANSi StdZ39-18 298-I02