Monographs in Aerospace History, Number 12




Gust Alleviation


[113] The response of airplanes to gusts has been a subject of concern to airplane designers since the earliest days of aviation. The Wright Brothers, flying in high winds at low altitude on the seashore at Kill Devil Hills, North Carolina, purposely designed their gliders and their powered airplane with low or negative dihedral to avoid lateral upsets due to side gusts. Pilots soon found the necessity of seat belts to avoid being tossed out of the seats of their machines when flying through turbulence. One of the first women pilots, Harriet Quimby, and her passenger, flying without seat belts in a Bleriot airplane over Boston, Massachusetts, in 1911, were unfortunately thrown out of their machine and fell 1500 feet to their deaths. In more recent years, gusts have been recognized as one of the sources of critical design loads on airplanes, as well as a source of fatigue loads due to repeated small loads. In addition to these design problems, many people are susceptible to airsickness when flying through rough air.

Despite the long interest of airplane designers in the effects of turbulence, very few efforts have been made to design airplanes with reduced response to gusts. Even in 1995, none of the commonly used transport airplanes or general aviation aircraft were equipped with gust-alleviation systems.

In the past, several attempts have been made by airplane designers to build airplanes with reduced response to turbulence. All of these attempts were characterized by an intuitive approach with no attempt at analysis prior to flight tests, and all were notably unsuccessful.

One of these airplanes (figure 13.1) was designed by Waldo Waterman. It had wings attached to the fuselage with skewed hinges and restrained by pneumatic struts that acted as springs. The effect of the skewed hinge was to reduce the angle of attack of the wing panels when they deflected upward, and vice versa. The response to gusts was not noticeably reduced from that of the airplane with the wings locked, probably because the dynamic response of the system was not suitable. Also, the degree of flexibility of the wings was limited because deflection of the ailerons would deflect the wings to oppose the aileron rolling moment, which resulted in reduced or reversed roll response.

The effect of wings with skewed, spring-loaded hinges is similar to the effect of bending of a swept wing. Airplanes with swept wings do have smoother rides in certain frequency ranges and suffer from reduced aileron reversal speed when compared with airplanes with unswept wings.

A similar method that has been tried inflight is to incorporate springs in the struts of a conventional strut-braced high-wing monoplane. This method may be likened to the springs used in an automobile chassis to.....



FIGURE 13.1.

FIGURE 13.1. Waterman's airplane incorporating wings attached to fuselage with skewed hinges. Pneumatic struts, shown in extended and compressed positions, balanced lift forces in flight.


....reduce bumps. This method has also proved ineffective, probably because of the slow dynamic response of the system.

Other schemes involving wing motion have been proposed from time to time, and some of them have been investigated in wind-tunnel tests or in flight. One method that has been given considerable attention is the 'free wing" concept. In this method, the wing is pivoted with respect to the fuselage about a spanwise hinge ahead of its aerodynamic center, and its angle of attack is controlled by a flap on the trailing edge. A serious disadvantage of this method is that upflap deflection must be used to trim the wing at a high-lift coefficient for landing. This flap obviously reduces the maximum lift, just the opposite from what is normally obtained with a downward-deflected landing flap. Also, the dynamic response of the wing may be too slow to provide reduction of the accelerations due to high-frequency gusts.

In England, shortly after WW 11, a large commercial airplane called the Brabazon was designed. In the design stage, a system was incorporated to reduce wing bending due to gusts by operating the ailerons symmetrically to oppose bending due to gusts. The ailerons were to be operated by a mechanical linkage connected to the wing in a way to be moved by wing bending. The system was abandoned before the airplane was flown, and the airplane never went into production. Nevertheless, the project stimu-[115] -lated interest in a flight project at the Royal Aircraft Establishment (RAE) in which a system of this type was tried in a Lancaster bomber. This system used a vane ahead of the nose as a gust detector to operate the ailerons symmetrically through a hydraulic servomechanism. The system was built with little preliminary analysis, and when the pilot engaged the system inflight for the first time, the flight in rough air seemed noticeably more bumpy than without the system. By reversing the sign of the gain constant relating aileron deflection to vane deflection, the ride was made somewhat smoother. Later, an analysis by an RAE engineer named J. Zbrozek showed the reasons for the unexpected behavior. These reasons will be mentioned later in the presentation.

Another experimental program was conducted on a C-47 airplane by the Air Force. This system was similar to that originally planned for the Brabazon. The ailerons were arranged to deflect symmetrically upward with upward wing bending, and vice versa, by means of a linkage which added a component of this deflection to that of the conventional aileron linkage. Since the wing deflection provided a large driving force, no servomechanism was required, and as a result, a system of high reliability was expected. The system suffered from the same objections as the one tested on the Lancaster. In addition, the inertia of the ailerons combined with flexibility of the operating linkage caused the aileron deflection to lag behind the wing deflection. Such a system is very conducive to flutter To avoid flutter, the ratio between the aileron deflection and wing bending had to be kept to a very low value. As a result, the system was unable to provide more than 9 percent reduction in wing bending moments, a rather small improvement.

Despite the discouraging results of these experiments, the advantages of gust alleviation remained worthwhile. As a result, a project was initiated at Langley to study this subject. These activities are described in the following section.


Background and Analysis of Gust Alleviation


I became head of the Stability and Control Section of the Flight Research Division in 1943. During the wartime years, there was little difficulty in deciding on the type of work to be done by the section. Most of the work was concerned with flying qualities or with improving the control systems of airplanes. Some of this work has been described in the preceding sections or may be further seen from my list of reports (appendix II). By 1947, jet airplanes with power control systems were being developed, and the use of automatic control to improve the stability characteristics of airplanes was a rapidly developing field. Applications of automatic control were therefore important subjects of research. The change in emphasis was recognized in 1952 when I was made head of the Guidance and Control Branch of the Flight Research Division.

In the military services, the new technology of automatic control was largely applied in the development of guided missiles. At Langley, the Pilotless Aircraft Research Division, under Gilruth, had been established to use missiles for aerodynamic research. Some of the engineers in that division, however, became interested in guided missiles. Gilruth and William N. Gardiner started to perform tests on a missile with infrared guidance that used gyroscopic stabilization and employed rocket propulsion techniques developed for aerodynamic testing. It later turned out that this missile was very similar in concept to the Sidewinder missile developed by the Navy, which was one of the first air-to-air guided missiles widely employed by the armed services.

The Deputy Director of the NACA at that time was Dr. Hugh L. Dryden, a noted scientist, who was also a lay preacher in the Methodist Church and a religious man. He did not like to see the NACA centers unnecessarily involved in military work. He issued [116] a directive that no work with military applications should be done at the NACA centers unless requested by the military services.

From the standpoint of a stability and control engineer, the study of missile guidance systems was one of the most interesting and challenging fields available at that time. Nevertheless, I had the same feelings as Dr. Dryden concerning the desire to avoid emphasis on military projects. With this area of research ruled out, I was faced with the question of what type of automatic control research with peacetime applications was of most interest. At that time, airplanes with swept wings were starting to be fitted with yaw dampers to improve the damping of lateral oscillations, particularly at high altitudes. Yaw dampers, however, did not require much research. In most cases, companies were able to hook up a rate gyroscope through an analog amplifier to the power control actuator on the rudder, and the system worked very well in damping out lateral oscillations.

After much thought, I concluded that one field in which automatic control could be applied and which had not previously been studied to any great extent, was gust alleviation. A gust-alleviation system is one that provides the airplane with a smooth ride through rough air. At that time, all transport airplanes had piston engines and flew at altitudes below the top of storm clouds. Fear of airsickness was a common problem and was a deterrent to many people who wanted to travel on commercial airlines.

I was already familiar with one study of gust alleviation that had been made. Philip Donely, who was then a branch head in the Loads Division, had formerly been designer of the gust tunnel at Langley. In this wind tunnel, a model was catapulted at flying speed through a vertical jet of air to simulate a gust in the atmosphere. Evidently while doing this work, Donely had come across a report by Rene Hirsch in France on a study of gust alleviation done as a doctoral thesis and published in 1938 (ref. 13.1). Donely brought this report to my attention. Hirsch had devised a gust-alleviation system in which the halves of the horizontal tail were attached by chordwise hinges. These surfaces were connected by pushrods to flaps on the wing. On encountering an upward gust, the tail halves would deflect up, moving the flaps up and thereby offsetting the effect of the gust. Other features of the system made the airplane insensitive to horizontal gusts and to rolling gusts, all without adversely affecting the ability of the pilot to control the airplane. These many ingenious features are too complex to discuss herein.

I was very impressed by the report for two reasons. First, Hirsch had performed an analysis to determine the relations between the tail, elevator, and flap hinge-moment characteristics; linkage ratios; and other parameters so that the flaps moved exactly the right amount to offset the effect of a gust. This analysis required consideration of so-called stability derivatives for all these components. At that time, stability derivatives were used to describe the stability characteristics of an airplane, such as, the variation of pitching moment with angle of attack mathematical symboland the variation of lift coefficient with angle of attack mathematical symbol. Prior to 1938, very few people had ever considered similar quantities for control surfaces or flaps, such as variation of flap hinge moment with angle of attack or variation of elevator hinge moment with deflection. Hirsch's analysis required a multitude of these quantities, as well as linkage ratios and other quantities describing the mechanism, all tied together by algebraic equations that were solved to obtain the desired design characteristics of the system.

The second important contribution by Hirsch was that he tested his system with a dynamic model. I have previously mentioned some of my experiments with a dynamic model. In Hirsch's case, the model was mounted in a wind tunnel so that it was free to pitch and to slide up and down on a vertical rod. The tunnel had an open throat and was equipped with a series of slats ahead of the test section, similar to a venetian blind, that could be deflected to produce an abrupt change in the [117] flow direction. When Hirsch tested his model in this artificial gust with the system locked, it immediately banged up against its stop at the top of the rod. With the system working, however, it showed only a small disturbance and settled back to stable flight.

Hirsch was so impressed by these results that he devoted most of his professional life to demonstrating the system in flight. I later corresponded with Hirsch, visited him in France, and saw his airplanes. This part of the story will be mentioned later. At the time I read the thesis, however, I simply realized that gust alleviation was a feasible idea and that with automatic controls, it might be possible to do the job more simply than Hirsch had done with his complex aeromechanical system.

I started, in 1948, to make analyses of the response of example airplanes to sinusoidal gusts and the control motions that would be required to reduce (alleviate) the response of the airplane or the accelerations that would be applied to the passengers. The method used was what has been described earlier as the frequency-response method, which had the advantage that the calculations were greatly simplified when compared with calculating response to discrete gust inputs. These early studies showed, as was known from experience, that control by the elevators alone was ineffective. Control by flaps on the wing was thought to be promising, but the analysis showed that in many cases, the flaps would produce excessive pitching response of the airplane. In general, control by a combination of flap and elevators was required. With flaps alone, successful results could be obtained only by careful attention to the pitching moments applied to the airplane by the gust and by the flaps.

A survey was also made of previous attempts at gust alleviation on full-scale airplanes. These included an airplane made by Waterman with wings pivoted on skewed hinges, a DC-3 modified by the Air Force so that the ailerons deflected symmetrically in response to wing bending, and an Avro Lancaster bomber modified by the British in which the ailerons were moved symmetrically in response to gusts sensed by a vane on the nose. All these attempts, made without any theoretical analysis, had been unsuccessful in that very little alleviation of airplane accelerations was obtained. My analysis showed why each of the attempts had failed. In general, the lack of attention to pitching moments produced by the control surfaces was responsible. In the case of the Lancaster, for example, the symmetric deflection of the ailerons in the upward direction proportional to a positive change in angle of attack produced a positive, or upward, pitching moment. A positive variation of pitching moment with angle of attack represents a decrease in longitudinal stability. This decrease in stability increased the response to low-frequency gusts, which resulted in an increase in the bumpy ride experienced by the pilots. In addition, the aileron motion reduced the damping of the wing bending oscillation, which increased the effect of structural oscillations on the sensation of the pilots. In considering the effect of the ailerons on the pitching moments, it should be realized that a symmetric upward deflection of the ailerons produces a direct effect on the pitching moments of the wing. It also changes the wing lift distribution to produce an increased downwash on the tail due to up aileron deflection, which further increases the destabilizing variation of pitching moment with angle of attack. My analysis showed that the downwash effects on the tail due to the deflection of flaps or ailerons on the wing are very important in designing a gust-alleviation system.

After making these preliminary studies, I analyzed a system in which a gust-sensing vane mounted on a boom ahead of the nose was used to operate flaps on the wing through a hydraulic servomechanism. Any gust-alleviation system working on this principle reduces the lift produced by a change in angle of attack. For complete alleviation, the lift due to angle of attack is reduced to zero. Since the pilot maneuvers the airplane by changing its angle of attack, this system would prevent the pilot from making any [118] longitudinal maneuvers. To restore this capability, the input from the control stick, normally used to move the elevators, was also fed to the flap servomechanism. With this arrangement, when the pilot moves the stick back to make a pull up, the flaps first go down to produce upward lift. Then, as the angle of attack increases in response to the elevator motion, the flaps move back to neutral and the pull up is continued with the airplane at a higher angle of attack. The result is a faster response to control motion than obtained with a conventional airplane. This type of control, in later years, has been called "direct lift control" and is advantageous for control situations requiring rapid response.

The analysis also showed that provision had to be made for avoiding excessive pitching moments due to the gusts and flap deflection. Since the flaps were moved in proportion to angle of attack, any pitching moment from the flaps contributed directly to the variation of pitching moment with angle of attack, which determines the longitudinal stability of the airplane. For satisfactory stability, the pitching moment due to angle of attack must be kept within prescribed limits. The additional contribution due to the gust-alleviation

system had the possibility of greatly exceeding these limits, which made the airplane either violently unstable or excessively stable. It was possible to solve the equations to determine the flap and elevator motion to completely offset both the lift and pitching moments applied to the airplane. This analysis showed that this objective could be attained in two ways. In one method, both the flap and elevator had to be moved in response to the gust, but the elevator motion was not in phase with the flap motion and generally had to lag behind the flap motion. In the other method, the elevator and flap moved in phase, but the downwash from the flaps on the tail had to be of opposite sign from that normally encountered. That is, down flap deflection had to produce an upwash at the horizontal tail.

Having reached this stage in the analysis, the results appeared sufficiently promising for a report on the results and a flight program to demonstrate a gust-alleviation system. A job order request was submitted about June 1948 to obtain official approval for this work. This job order is presented here to illustrate the type of request required to get a research program approved.


Title: Theoretical and Experimental Study of Means to Increase the Smoothness of Flight through Rough Air.
Est. Man Hrs: 3000 Cost: $6000
Description: A theoretical study has been made of various means to increase the smoothness of flight through rough air. This job order is to cover placing this analysis in a form suitable for a report, and preparation of a report on this study. In addition, measurements will be made in flight to verify the predicted response characteristics of airplanes, and bench tests will be made of servomechanisms intended for use with the automatic control device.
Justification: The preliminary analysis has indicated that the successful operation of a device to provide smooth flight in rough air requires careful selection of the design parameters, but that with a suitable design very promising results may be obtained. Such a device is not primarily intended to reduce stresses in the airframe due to severe abrupt gusts, but rather to reduce changes in acceleration which are primarily responsible for passenger discomfort and airsickness. The analysis indicates that current types of servomechanisms are capable of providing the desired rapid response of the controls. Inasmuch as a device of this type would be of great interest for airline operation, it is considered desirable to publish the results of the analysis and to continue this work with a view to eventual flight demonstration of such a device.

[119] Note the man-hour estimate. The value of 3000, about 1.5 man years, was just a guess based on the effort put into previous reports. The cost of $6000 was based on a standard rule of $2 per man-hour. Little effort was required for this phase of the job approval.

When the work described in this job order was undertaken, I appointed Christopher C. (Chris) Kraft, Jr. as project engineer. Kraft was then an engineer in my section with experience in flying qualities and with work using the free-fall and wing-flow methods. Later in his career, he was a flight controller during the Apollo missions and was made director of the Johnson Space Flight Center following Dr. Gilruth's retirement. Kraft and I made additional calculations to have a logical series of examples to place in the report, which was later entitled Theoretical Study of Some Methods for Increasing the Smoothness of Flight through Rough Air (ref. 13.2). The report starts with a review of the available data on the causes of airsickness. The main source of data, which is still believed to be the best available, was a series of tests made during WW II at Wesleyan University, in which subjects were tested with various wave forms of vertical acceleration in an elevator. The results showed that relatively low-frequency variations of acceleration that had periods of 1.4 seconds and greater were the most important causes of motion sickness.

Excellent control of the directional and rolling disturbances of airplanes could be obtained with conventional autopilots available at the time the report was written. These devices, however, were relatively ineffective in reducing the vertical accelerations of airplanes. The report therefore concentrated on the problems of longitudinal gust alleviation and control.

Following a section on the theoretical analysis, examples were included in the report to show the use of elevator alone, flaps alone, and a combination of flap and elevator motion to offset the effect of gusts. Next, two types of gust-sensing devices, a vane ahead of the nose and an accelerometer in the airplane, were considered. Because these systems affect the controllability of the airplane. detailed studies were made of static and dynamic longitudinal stability of airplanes incorporating these systems.

In various studies of gust-alleviation systems made by other researchers since the one described herein, the now popular approach of optimal control theory has been applied. In this method, some balance is sought between the amount of control motion required by the system and the amount of reduction of acceleration achieved. In the study made by Phillips and Kraft, however, an effort was made to achieve complete gust alleviation within the limits imposed by the assumptions of the theory. Complete alleviation was found to be possible with the vanetype sensor, but not with the accelerometer sensor. Inasmuch as complete alleviation was found to be possible with reasonable control motions, the use of optimal control theory when a vane sensor is used is really not optimal and represents an inappropriate application of this theory.

The theory used for the analysis is very similar to that taught by Professor Koppen in my courses at MIT. This theory, in turn, was based on the theory first presented by G. H. Bryan in England in 1903 and later in the textbook Stability in Aviation in 1911 (ref. 4. 1). This theory shows that the longitudinal and lateral motions of the airplane can be considered separately and assumes small disturbances so that all aerodynamic forces and moments can be assumed to vary linearly with the magnitude of the disturbance. This theory was extended to calculate the response of an airplane to gusts and reported in NACA Report No. 1 and later reports in the period 1915-1918 by Edwin B. Wilson, then a professor of Physics at MIT. To study the effect of a gust-alleviation system, it was necessary only to add to Wilson's theory the additional forces and moments caused by the airplane's control surfaces as they were moved by the gust-alleviation system. Like Wilson, I considered that the gust was constant across the span, though this subject was studied in more detail later.

[120] Two refinements were added to the theory, as presented by Wilson, that were found to be of importance for the study of gust-alleviation systems and that are believed to make the results very close to what would be obtained with an exact computer simulation such as would be possible with modern electronic computers. First, the penetration effect was considered, that is, the difference in the time of penetrating the gust by a vane ahead of the nose, the wing, and the tail. In Wilson's study, the gust was assumed to affect all parts of the airplane simultaneously. Second, the time lead or lag effects caused by gust penetration were approximated by a linearized representation to keep the equations linear. This technique had been used by Cowley and Glauert in a British report published in 1921 to improve the calculation of pitch damping of airplanes by taking into account the time for the downwash leaving the wing to reach the tail (ref. 13.3). A similar method was used in the present analysis to account for all the lead and lag effects, such as the lead of the vane in penetrating the gust, the lag in response of the servomechanism operating the flaps, and the lag of downwash from the wing and flaps in hitting the tail, in addition to the lag of the gust itself in reaching the tail. These lead and lag effects were found to be very important, particularly in affecting the damping of the short-period longitudinal motion of the alleviated airplane.

A useful advantage was found in this approach in that all the effects of the alleviation systems studied could be considered as changes to stability derivatives of the basic airplane. Most of the effects of these derivatives were known from experience, or at least had simple physical interpretations. In addition, the order of the equations was not increased over that of the basic airplane.

These equations made it possible to solve for the characteristics of the system that would produce complete gust alleviation, that is, that would produce zero response to a gust. In examining these formulas, I suddenly realized that the results had a simple physical interpretation. This interpretation is shown in figure 13.2, in which an airplane with a vanetype gust-alleviation system is shown penetrating a region in the atmosphere where there is a change in vertical gust velocity (called a step gust). First, the vane is deflected by the gust. If the servomechanism operating the flaps has a time lag equal to the time for the gust to reach the wing, then the flap moves just at the right time and the right amount to offset the lift change on the wing due to the gust. For the airplane response to be zero, however, the pitching moment about the center of gravity caused by the flap deflection must be zero. This condition is not ordinarily obtained with conventional wing flaps, but may be obtained by moving the elevators the correct amount in phase with the flaps. A little later, the tail is affected by the gust and by the downwash from the flaps. The effect of the forces on the tail can be eliminated in two ways. Either the flap downwash must be equal and opposite to the effect of the gust on the tail or the elevator must be given an additional movement to cancel the combined effect of the flap downwash and the gust.

The simple interpretation of the action of a perfect gust alleviation system has interesting ramifications. First, the influence of the feedback of angle of attack from the vane makes the alleviation system a closed-loop control system, but the feedback decreases as the system approaches the condition of complete gust alleviation. In this condition, the system behaves as an open-loop control, because there is no motion of the airplane to be sensed by the vane. Second, the consideration of lag effects is seen to be exact in the limiting case of perfect alleviation, even though these results were obtained from an approximate linearized theory. Third, the discovery of this interpretation could presumably have been made a priori, without use of any theory, but in my case, working through the theory first and examining the resulting formulas was necessary for me to realize that this interpretation existed. Many simple physical principles in the history of physics...



FIGURE 13.2.

FIGURE 13.2. Concept for complete alleviation of vertical gusts shown by the theoretical analysis. Note that the elevator moves in phase with the flaps, offsetting the flap pitching moment. Flap downwash, with opposite direction from normal, offsets the lift on the horizontal tail due to the gust.


.....have been realized only after long periods of thought and analysis by their discoverers.

The vane-type system studied has some disadvantages when adjusted to give complete alleviation. The system results in an airplane with zero lift and zero pitching moment due to angle of attack. The airplane would respond to pilot's commands as provided by the direct lift control system, but would have no inherent tendency to stabilize in a new equilibrium condition. For this reason, systems were studied that retained a small amount of longitudinal stability. Use of these systems was found to be feasible and did not seriously reduce the gust-alleviation properties of the perfect system.

The studies of the use of an accelerometer to sense the gusts showed that the gain of the system, that is, the amount of flap deflection used for a given change in acceleration, had to be limited otherwise a poorly damped short-period vertical oscillation would result. As a result, the amount of alleviation potentially available with this system was limited. On the other hand, the use of the accelerometer sensor inside the airplane avoids the problem of having a delicate vane exposed to potential damage from handling. In practice, the system with an accelerometer sensor would be much more likely to excite structural oscillations of the airplane, thereby further limiting the gain and requiring a more detailed analysis to insure the safety of the system.


Design and Test of a Gust-Alleviated Airplane


During the publication process of the report on the analysis, work was started on a....



FIGURE 13.3.

FIGURE 13.3. Two-view drawing of the Beech B-18 test airplane showing the modified control surfaces. Dimensions in inches.


....program to demonstrate gust alleviation in flight. The airplane chosen for the program was a Beech B- 18, a small twin-engine transport. The airplane was obtained from the Navy and had the Navy designation C-45. Because of the need for some major alterations to the airplane control surfaces, Kenneth Bush and Edwin C. Kilgore of the Engineering Division were called into the project to do the design work. Steve Rock of the Instrument Research Division was assigned to design the servomechanism to actuate the control surfaces.

At the time the design was started, about 1950, electronic control systems had a poor reputation for reliability. These systems used vacuum-tube amplifiers. Furthermore, the use of techniques of redundancy to improve reliability had not then been developed. For these reasons, many of the design features were governed by safety considerations. All autopilots in use at that time were designed so that the pilot could readily overpower the autopilot with his manual control system in the event of a failure. This method could not be used with the gust-alleviation system because the wing flaps, which required large....



FIGURE 13.4.

FIGURE 13.4. Nose boom and angle-of-attack vane installation on Beech B-18 test airplane.


....operating forces, were not normally connected to the pilot's control stick. As a result, the design features described in the following paragraphs were incorporated.

A drawing of the airplane as modified for the gust-alleviation project is shown in figure 13.3. A boom was built on the nose to hold the angle-of-attack vane as shown in figure 13.4. The wing flaps, which normally deflect only downward, were modified to move up and down because both up and down gusts must be counteracted by the system. The elevator was split into three sections, the two outboard segments being linked to the flaps for use with the gust-alleviation system and the inboard segment being used in the normal manner for pitch control (figure 13.5). Finally, small segments of the flaps near the fuselage were driven separately from the rest of the flap system so that they could be geared to move either in the same direction or in the opposite direction from the rest of the flaps (figure 13.6).

As stated previously, perfect gust alleviation according to the theory could have been obtained either by driving the elevator separately from the flaps and with a different phase relationship or the elevator could have been geared directly to the flaps and the downwash from the flaps altered to offset the gust at the tail. The latter method was selected for the following reasons. With a direct mechanical linkage between the flaps and the outboard elevators, the pitching moment due to flap deflection, a critical quantity for longitudinal stability, could be finely adjusted as required and would hold its setting. If a separate servomechanism and electronic amplifier had been used to operate the elevators, the gain of the amplifier might have drifted and caused the airplane to become unstable. In fact, the gain of the amplifier between the vane and the flaps often did vary in flight by amounts that could have caused violent instability if a similar.....



FIGURE 13.5. Three-quarter rear view of Beech B-18 test airplane showing elevator control split into three segments.


FIGURE 13.6.

FIGURE 13.6. Three-quarter rear view of Beech B-18 test airplane showing modified flap system with oppositely deflected inboard segment.


....amplifier had been used to operate the elevators.

The small inboard segments of the flaps were used to reverse the direction of the flap downwash at the tail, as required by the theory if the elevators moved in phase with the flaps. This method of course, reduces the flap effectiveness in producing lift. To regain sufficient flap effectiveness, the flaps and ailerons were geared together so that the flaps and ailerons deflected symmetrically for gust alleviation. These surfaces were driven by an electrical input variable-displacement pump hydraulic servomechanism, which was taken from a naval gun turret. Electrical signals from the vane and the pilot's control column were combined in a vacuum-tube amplifier and fed to the control valve of the servomechanism. For lateral control, the entire flap and aileron system was deflected asymmetrically through a separate servomechanism of the same type. The system was designed so that the airplane could be flown through its original manual control system if [125] the alleviation system failed or was switched off. In this configuration, the control wheel on the pilot's side remained connected at all times to the inboard segment of the elevator and to the ailerons. In the gust-alleviation mode, the ailerons were driven symmetrically through preloaded spring struts. This system remained connected in the manual mode, but the pilot could overpower the forces in the preloaded struts with his inputs to the control wheel. With the system in the gust-alleviation mode, the control wheel on the copilot's side was used to apply control inputs through the electronic control system. In addition, it remained connected through the mechanical linkage to the inboard portion of the elevator.

When the pilot turned the alleviation system off, the actuator driving the flaps was bypassed, and a separate hydraulic actuator with its own accumulator forcibly drove the flaps to neutral with a caliper-like linkage that could capture the flaps in any position.

Changes in angle of attack due to change in airspeed or drift of the amplifier used to operate the flaps could have caused the trim position of the flaps to vary slowly in flight. To maintain the trim position of the flaps at zero deflection over long periods, a mechanical ball-disk integrator driven by the flap linkage was used to feed an additional signal into the flap servomechanism to slowly run the flaps to the neutral position. This system had a time constant of 10 seconds, which was slow enough to avoid interference with the gust-alleviation function or the control of the airplane. To provide automatic control of the lateral and directional motion of the airplane in the gust-alleviation mode, a Sperry A-12 autopilot was connected to the aileron and rudder systems. This autopilot was the most advanced type available at the time of the tests.

Finally, the pilots landed the airplane with the wing flaps in neutral. This operation did not pose any problem on the long runways at Langley Field, but the design of a high-lift flap system that can also provide upward deflection of the flap for gust alleviation remains one of the engineering problems of such systems that no one has yet tried to solve.

The airplane was instrumented with strain gauges to measure wing shear and bending moments at two stations and tail shear and bending moment at the root. The project became a joint project with the Aircraft Loads Branch. After the tests were completed, two reports had been published by each group: an initial and a final report on the gust-alleviation characteristics and an initial and a final report on the loads (refs. 13.4, 13.5, 13.6, and 13.7).

The tests occupied a long period of time. One of the main problems encountered was finding suitable rough air. The NACA test pilots were understandably conservative in flying experimental airplanes and usually declined to do test flying in clouds or stormy weather. Clear-air turbulence occurred on occasions after passage of a cold front. These conditions were used as often as possible in making the tests, but frequently the turbulence was of low intensity. For these reasons, some of the data were not as extensive as might have been desired.

Despite these problems, results were obtained with various sets of gearings to obtain varying degrees of longitudinal stability, and cases with the inboard flaps in neutral and moving oppositely from the outboard flaps were studied. A time history comparing the airplane motions in flight through rough air with the system on and off is shown in figure 13.7. In these runs, obtained early in the test program, the inboard flap segments were locked.

The results of later tests in which the inboard flap segments moved oppositely from the rest of the flap system are shown in figure 13.8. These results are shown as power spectral densities of the normal acceleration and pitching moment plotted on log-log scales. The results show that the system was effective in reducing the response in both normal acceleration and pitching velocity at frequencies below about 2 hertz. These plots....



FIGURE 13.7.

FIGURE 13.7. Comparison of flights in light turbulence, basic airplane, and gust-alleviated airplane. Initial configuration with center-section flaps in neutral. (a) Basic airplane (top three). (b) Gust-alleviated airplane (bottom four).


....of power spectral density are the results of the usual evaluation procedure for randomly varying quantities, but they do not give a very clear comparison of the results obtained with the alleviation system on and off. The power spectra present the square of the recorded quantities, which tend to exaggerate the differences, while plotting on log-log paper tends to reduce the apparent differences. The question therefore arises as to how the data could be compared to give an impression of the effectiveness of the system more meaningful to the user. For this reason, data on the normal acceleration responses....



FIGURE 13.8.

FIGURE 13.8. Comparison of power spectral densities of normal acceleration and pitching velocity for basic airplane and alleviated airplane. Center-section flaps deflected oppositely from outboard flaps. (a) Normal acceleration (left). (b) Pitching velocity (right).


FIGURE 13.9.

FIGURE 13.9. Comparison of normal acceleration of basic airplane and alleviated airplane shown on linear scales on two different types of plots. (a) Amplitude of transfer function of mathematical symbol, or ratio of normal acceleration to gust angle of attack, as a function of frequency (left). (b)mathematical symbolof normal acceleration, or amplitude the component of normal acceleration at each frequency, as a function of frequency (right).


....are plotted in two different ways in figure 13.9 In part (a) of this figure, results are plotted on linear scales in the form of a transfer function, that is the ratio of normal acceleration to gust angle of attack for sinusoidal inputs of various frequencies. This plot is of interest to the control engineer and shows correctly the relative magnitude of the normal acceleration for the two cases at various frequencies, but does not include the variation of the actual gust forcing function with frequency. In part (b), the square root of the power spectral density of the response is plotted as a function of frequency on linear [128] scales. This plot attempts to show the actual magnitude of the normal acceleration at each frequency. This plot is believed to be more meaningful in interpreting the passengers' impression of a ride in the airplane.

As can be seen, the basic airplane has a large peak in the acceleration response at a frequency about 0.2 hertz. This low-frequency peak is typical of the response of unalleviated airplanes and occurs because of the larger amplitude of the turbulence input at low frequencies. In the case of the alleviated airplane, the response is reduced to a fairly constant, relatively low value in the frequency range between 0 and 1.6 hertz. For the degree of turbulence encountered, this reduction greatly improved the subjective impression of ride comfort. Though the results are not shown on the plots, the response at frequencies above 2 hertz with the alleviation system operating were slightly increased above those of the basic airplane. Also, in turbulence of relatively large magnitude, the pilots noted a fore-and aft oscillation caused by drag of the flaps at large deflections.

Though the results obtained were quite gratifying, the question arises as to why the performance of the system was not better, inasmuch as the theory predicted perfect gust alleviation. One of the main reasons was the nonlinear characteristics of the servomechanism used to drive the flap system. The designer of the system in the Instrument Research Division had been requested to provide a rather sharp cutoff in the response beyond 2 hertz, to avoid the possibility of exciting wing flutter. The C-45 was a very stiff airplane, with the wing primary bending mode at 8 hertz. It was intended, therefore, that the output of the servomechanism should be close to zero at a frequency of 8 hertz. It was not until after the device was installed that it was found that the cutoff in response had been obtained by rate limiting the output. This provided a steep cutoff, but the response was a function of amplitude. The result was that in a large amplitude gust, rate limiting was encountered and the alleviation was reduced just when it was needed most. A later unpublished investigation of the effects of rate limiting demonstrated these adverse effects and showed that with sufficiently severe rate limiting, a continuous sawtooth oscillation of the flap would be encountered in rough air, which resulted in response greater than that of the basic airplane.

A second problem was the nonlinear lift characteristics of the flaps. The flaps deflected about plus or minus 25 degrees, but the slope of the curve of lift versus flap deflection fell off markedly before this deflection was reached. As a result, a fixed gain between the vane and the flaps did not give a uniform value of the ratio of lift to gust angle of attack. This nonlinear response would be expected to introduce higher frequency harmonics into the response and probably accounts for the increase in response beyond 2 hertz. This and other characteristics were not known before the tests were made. I concluded that on any future project of this type, wind-tunnel tests of the airplane to determine the control characteristics would be advisable. Despite these deficiencies, enough was learned to show the feasibility of gust alleviation and to show how the results could be improved in a future attempt.


Analytical Studies of Additional Problems


The studies of gust alleviation introduced several problems that had not been considered previously in airplane design. One problem was the effect of variations in gust velocity across the wing span. The use of a single vane on the center line of the airplane to sense the gusts would work perfectly if the gust velocity were constant across the wing span, but would be less effective if different values of gust velocity were encountered at different stations along the span. This problem can be studied if the turbulence in the atmosphere is assumed to be isotropic (or for [129] this application, axisymmetric); that is, it has the same characteristics regardless of the direction in which the airplane is flying. This assumption appears reasonable for most types of turbulence. The theory of isotropic turbulence had been studied theoretically. The nature of the spectrum of turbulence, that is, the way the gust velocity varies with gust wavelength, had been determined. The gust velocity has been found experimentally to vary approximately directly with the wavelength even for wavelengths many times larger than the wing span of the largest existing airplanes. The most intense gusts, which cause the most disturbance to the airplane, have wavelengths long compared to the wing span and are therefore approximately constant across the span. Gusts with wavelengths short compared to the span have low intensity and therefore do not disturb the airplane much. The use of a single vane on the center line is therefore quite effective. Gusts of wavelength short compared to the span may cause up and down loads that average out across the span. For these gusts, the response of the vane at the center line would be too large. These gusts have such high frequency, however, that the vane response is filtered by the lag in response of the flap servomechanism. For a vane located ahead of the nose, the lag in the flap servomechanism is also beneficial in delaying the response sensed by the vane until the gust reaches the wing. Considering these factors, the effectiveness of a gust-alleviation system using a vane on the centerline may be shown to be about 98 percent as effective in axisymmetric turbulence as it would be with gusts constant across the span. This problem was studied in more detail after the flight investigation was completed (refs. 13.8 and 13.9).

An interesting optimization problem, which to my knowledge has not been solved, is to determine the optimal filter to place between the vane and the flap to obtain the greatest amount of gust alleviation, considering the distance of the vane ahead of the wing, the wing span, and the spectrum of atmospheric turbulence. This problem is only of academic importance, however, as about 98 percent alleviation was obtained by using a second-order linear filter with reasonable frequency and 0.7 critical damping (refs. 13.6 and 13.7).

Some time after completion of the tests, a summary report was given as part of a lecture series at Renssellaer Polytechnic Institute (ref. 13.10). This paper gives a more complete discussion of the subject of gust alleviation than contained herein.


Future Possibilities of Gust Alleviation


Some review of later efforts in the field of gust alleviation may be of interest, inasmuch as the development of computers and automatic control technology would permit approaches quite different from that used in the early NACA tests. It was a disappointment to me that very little effort was made by aircraft companies to incorporate provision for gust alleviation even after the development of control technology would have made it more feasible.

To my knowledge, the only airplane in service that incorporates a system performing some gust-alleviation function is the Lockheed 1011. In some later models, the wing span was increased by extending the wing tips to allow the airplane to carry greater loads. To avoid changing the wing structure to withstand greater bending moments, the ailerons were operated symmetrically by an automatic control system to reduce bending moments due to gusts.

One reason for the lack of interest in gust alleviation is that following the NACA tests, jet transports were introduced. As a result of higher wing loading, swept wings, flight at higher altitudes, and the use of weather radar to avoid storms, these airplanes were much less likely to encounter violent airplane motions that would cause airsickness. In addition, the problem of gust alleviation became more difficult because the structural flexibility of these airplanes placed their [130] structural frequencies closer to the frequency range of interest for gust-alleviation. As a result, structural response would have to be considered in designing the system. In recent years, these reasons for avoiding the use of gust-alleviation systems have become less significant. Extensive use is now made of commuter airplanes that fly at lower altitudes and frequently encounter rough air. In addition, methods have been developed to analyze the structural response and to damp out the structural modes by use of automatic control systems.


Review of Work by René Hirsch


In closing, a brief review is given of the work of Rene Hirsch, whose thesis was mentioned at the beginning of this chapter. Also, a few programs and studies applicable to gust alleviation that have occurred since the NACA program on the C-45 are reviewed.

After reviewing Hirsch's thesis, no more was heard of his activity until the early 1950's, during the course of the NACA program. At this time, a French report was discovered revealing that, following WW II, Hirsch had made additional wind-tunnel tests in the French large-scale tunnel at Chalais-Meudon on a model of a proposed airplane and had built this small twin-engine airplane incorporating his system. The airplane was envisioned as a quarter-scale model of a piston-engine transport of a class similar to the Douglas DC-6 or the Lockheed Constellation, which were the largest transports in service in that period. Correspondence was established with Hirsch and additional reports and information were obtained (ref. 13.11). Hirsch's airplane had a wing span of 27 feet and had two 100-horsepower motors. It incorporated the same system described in his thesis, in which the halves of the horizontal tail moved on chordwise hinges to operate flaps on the wings. The conventional elevators provided not only pitching moments, but moved the tail halves about their chordwise hinges to cause the flaps to move in the direction to provide direct lift control. In this way, the loss of longitudinal control due to the gust-alleviation system was overcome. Hirsch's airplane incorporated many other ingenious features, including provisions for reducing rolling moments due to rolling gusts and lift due to horizontal gusts. His design also incorporated large pneumatic servos operated by dynamic pressure to restore damping in roll and to stabilize the rate of climb or descent. The airplane had good handling qualities and appeared to have been very successful in providing a smooth ride, as shown by some time histories in rough air with the system turned on and off. After about 30 flights, the airplane ran into a ditch at the end of the runway and was damaged.

No more was heard until 1967, when another report appeared showing that the airplane had been rebuilt and equipped with two 180-horsepower motors (ref. 13.12). In this condition, the airplane made numerous additional flights with somewhat more complete instrumentation. A photograph of the airplane in flight is shown in figure 13.10. From the data obtained, the results appeared very similar to those obtained with the NACA C-45 in that the accelerations due to gusts were reduced by about 60 percent at frequencies below about 2 hertz, but were increased somewhat at higher frequencies.

All of Hirsch's work in designing and building his airplanes was done with his own funds, though some help with instrumentation was obtained from ONERA, the French equivalent of the NACA. Outside of France, Hirsch's work was little known. His reports described the work in general, but were not sufficiently detailed to give engineering data on all of the ingenious ideas and systems incorporated on his airplanes.

In 1975 during a trip to France, I visited Hirsch. He had found at that time that a twin-engine airplane was too expensive for him to operate during the oil crisis, and he had donated it to the French Air Museum. When....



FIGURE 13.10. Flight photograph of René Hirsch's first airplane as equipped with larger engines.

FIGURE 13.11.

FIGURE 13.11. René Hirsch in front of his Aerospatial Rallye light airplane equipped with gust-alleviation system.


[132] ....I saw the airplane, it stood like a little jewel amid a group of dilapidated antique airplanes in an old WW I hangar at Villaroche. It is now on display at the new French Air Museum at Le Bourget.

At that time, Hirsch was starting modification of a single-engine light plane, the Aerospatial Rallye, to incorporate his gust-alleviation system. This airplane was completed and flying when I visited France again in 1980. Hirsch is shown standing in front of his Rallye airplane in figure 13.11. This airplane was never as successful, in Hirsch's opinion, as the first one, probably because of its lower airspeed and lower frequency of response of the flap systems. In recent years (1995), Hirsch modified a third airplane, the Sobata Trinidad, with small canard surfaces ahead of the wing root to operate the flaps. The more forward position of these sensing surfaces was intended to improve the response to high-frequency gusts. Hirsch died in August 1995 at the age of 87 without having had the opportunity to test his latest design.

Hirsch's dedication to the pursuit of gust alleviation is a remarkable story in view of the general disregard of this subject by the rest of the aviation industry. Of course, the aeromechanical systems used by Hirsch have been superseded by automatic controls using computers and electro-hydraulic actuators. Hirsch readily admitted that he would prefer such systems but was unable to afford them.


Later Studies by Other Investigators


Though a number of studies of gust alleviation have been made during the years since the C-45 tests, most of them have not contributed any notable new developments. Only two are mentioned to bring the subject up to date. One is the so-called LAMS project, an acronym for Load Alleviation and Mode Stabilization, conducted at the Air Force Flight Dynamics Laboratory at Wright-Patterson

Air Force Base, Ohio, about 1968-1969 (ref. 13.13). In this project, a B-52 bomber was equipped with an electronic analog-type flight control system to operate the existing flight controls to damp out structural modes. This work is important because consideration of damping of structural modes would be required in any attempt to install a gust-alleviation system in a high-speed airplane. The report illustrates the success of modern control analysis techniques (as they existed at that time) in designing a modal damping system and in predicting the results obtained.

The second contribution of note is the analytical work of Dr. Edmund G. Rynaski, formerly of Calspan and now at EGR Associates, in designing a system to alleviate both the rigid-body motions and selected structural modes of an airplane (ref. 13.14). Rynaski's work, based on matrix analysis techniques, shows how to provide essentially open-loop control of the rigid-body modes, as was done on the C-45 airplane, as well as to provide open-loop cancellation of a selected number of structural modes. This method also makes possible improved damping of higher order structural modes by use of closed-loop control.

With the availability of digital flight computers and modern control actuators, different approaches should be considered for gust alleviation. One approach would be to operate the flaps on the wing as a function of angle of attack sensed by a vane or similar device, but to operate the elevators by a modern reliable pitch damper as part of a longitudinal command control system to control the pitching response of the airplane. Another approach would be to calculate absolute gust velocity on line by a method similar to that referred to previously in the work by Crane and Chilton in measuring gust velocity (ref. 12.5). This method requires correcting the angle of attack measured by a vane for the inertial motions of the airplane at the vane location. This signal could be used as an input into a gust-alleviation system without the need to modify the normal control or handling qualities of the airplane.