1957 - 0484

2
486 ;  FLIGHT 1 2  April  1957 A u t o p i l o t t o F l i g h t c o n t r o l S y s t e m  As far as concentration of panel instruments was concerned, the first and probably most important group to be tackled was that giving heading information. Various forms of gyro-magnetic compass were developed i n the 1930 s to combine t he magnetic comp ass a nd directional gyro. The R.A.F. distant-reading com- pass originally conc eived by A ir Ministry and S. G. Brown, Ltd., in 1934, was put into full-scale production for bombers and was, in fact, a standard instrument throughout World War 2 in heavier Service aircra ft. In America, Sperry and Pioneer were working on the same problem and examples of the result of their efforts were the Sperry slaved gyro-magnetic compass in 1937 and the Pioneer gyro-fluxgate compass which w as w idely used in U.S. military aircraft during World War 2. These were the forerunners of the present day Sperry Gyrosyn, Bendix, Kelvin Hughes and other gyro-magnetic compasses which are now almost universally used in all but the smalle st and simplest types o f aircraft. The gyro-magnetic compass wa s a step in the right direction, but the number of instruments required continued to grow as all forms of radio aids, including Lorenz, Standard Beam Approach, homing and, for military purposes, Gee and the other position- fixing radio aids came to be added to the range and scop e o f instrument information required at various times during a flight. From the purely military point of view, during World War 2, bombing aids and weapon sighting instruments were added, and these still further aggravate d the problem. Towards the end o f the war it became obviou s that, for accurate bombing, much close r integration of aircraft-control and weapon-sighting was needed; and this laid the foundation for t he modern electronic automatic pilots which are now almost a standard fitment in heav y transport as well as bomber aircraft. The combination of the American Norden bomb-sight and Minneapolis-Honeywell electric autopilot or the Sperry integrated bombing and autopilot system were wide ly used in Liberators and Fortr esses. Work along similar lines for the British Services was carried out under the guidance of R.A.E. and based on the Mk 14 bombsight, D.R. comp ass and Mk 8 autopilot. Thu s, for bombers, a sta ge had bee n reached by the end of World War 2 where there was a large degree of integration of gyro-magnetic compass, autopilot and bomb-aiming equipment. At this stage it is o f advantage to digress for a while from the main theme of integration and concentrate on the development of autopilots, i.e., of the basic automatic Control equipment. This equipment can be divided largely into two general cate- gories : firstly, that which assists the pilot to fly his aircraft more eff ici en tly in the face of stability limitations in the basic air- frame; and, second ly, that which relieves fatigue duri ng flying for comparatively long periods. Int o the second ca tegory, o f course, fal l all those civil autopilots of the pre-war era and, to a less er extent, the military equipments of the same and rather later years. These were essentially displacement-type autopilots which, with references about all three axes, w orked through all three control channels. The majority had fa cilities f or the superimposition of manual control so that normal flight manoeuvres such as turns, climbs, and descents could be executed by overriding the outopilot. But for military purposes—and it was here that the greater autopilot development effort was applied—the initial emphasis wa s rather differe nt. For example, the original Mk 1 autopilot, developed by R.A.E., was intended primarily as a stabilizing control for bombing. It was nominally a two-a xis control for rudder and elevators only, used a single conventional displacement gyro, and corrected ya w an d pitch disturbances. In fact, an indirect control about the roll axis was achieved by the girnballing eff ec t of the inclined gimb al-axis of the gyro. Later versions, f rom the Mk 2 up to the Mk 7, utilized a separate aileron-control gyro and servo syste m in addition to this ru dde r/ elevator control. Other forms of early military autopilots, such as the Pollock Brown and some German Siemens and Askania types, were two- or single-axis system s principally concerned with the steering and pitch func- tions.  But once the value o f the stabilizing effect in these two axe s was appreciated there came a tendency to adopt a control on the thir d axis, either by means o f an additional separate gyro-referen ce and servo-channel or by some form of combination of all three channels. Further development of airframe design, bringing improvement of aerodynamic stability, particularly about the yaw axis, tended to reduce the importance of the rudder-control function, thus allow- ing the design of autopilots for control through ailerons and elevators only. A notable example o f this predominantly aileron/ elevator type of control was the R.A.E. Mk 8 autopilot widely used in R.A.F. bomber aircraft during World War 2. Thi s was a dis- placement-type autopilot utilizing a single, gravity-controlled, inclined-axis gyro to provide pitch an d roll reference. A short-ter m yaw reference was provided indi rectly by gimballing eff ect and the long-t erm heading reference was supplied by signals f rom the The Curtiss F Boat  in  which Lawrence Sperry first installed  a  gyropilot. It  is  here seen  i n  France the  Seine in  1914 D.R. compass. Control was applied by pneumatic servos to the aileron and elevator channels only. Complete manual override was possible for climbs, descents, and t urns; and a novel means of  jinki ng or taking avoiding action, was incorporated. As aircraft development continued it becam e apparent t hat t he aerodynamic characteristics or shor t-term stability o f the airfram e would not be suf fic ien t to damp out t he inherent tendency of the purely displacement-type of automatic control to hunt about the mean flight-path, although various artificial methods were adopted in the follow-up or feed-back function of autopilots to reduce this hunt ing to a minimum. Thu s a rate function was introduced and (at least in the earlier stag es) added to the displacemen t function about one or more axes to provide a dam ping factor in the auto- matic control system. In som e case s the secon d derivative, i.e., an acceleration term, was added; but until comparatively recently, with one or two notable exceptions, acceleration terms were not considered to be of m ajor importance. Before rate terms were incorporated, most autopilots had em- ployed vacuum-driven gyros and either pneumatic or hydraulic signallin g and actuation. But the Polloc k Brown equipment was sign ifican t in that it was entirely hydraulic i n operation, except for an electrical sy stem for applying manual override. The inclusion of rate, and particularly the mixing o f rate with displaceme nt terms, was much easier to achieve in electrical signalling systems; and for this and other possibly more important reasons electrical operation was introduced into autopilots. One of the most elab- orate early electrical designs was a German three-axis control, the joint product of Askania, Patin and Siemens, which, quite early in the war, incorporated displacement, rate and acceleration terms. The displacement and rate references were obtained from electric- ally driven displacement and rate gyros, and the acceleration term (about the pitch axis o nly) fro m a fore-and-aft linear accelerometer. The various pic k-o ff signals in each control channel were suitably mixed in a galvanometer type of movement and the output was used to control a rotary electrical servo coup led to the aircraft controls. This equipment was especially interesting as it was one of the first examples of an autopilot in which control was by the rate  of application of control, i.e., a servo speed was made pro- portional to the control signals and no follow-up system was required. In America, the Minneapolis-Honeywell electric autopilot was another early example of the new technique of combining rate and displacement. But here just two d isplacement gyros were used, one v ertical and one directional, the rate function being obtained fo r the yaw plane by an ingenious viscosity throttl e/pot enti ometer device which, operated by the directional gyro, supplied yaw dis- placemen t signals . Rate and displacement control was applied only in the yaw plane, the other two planes being controlled solely according to displacement. Control signals were electronically amp lified to operate the electric servos coupled to all three control surfac es. The i nteraction between the three control surf aces w as particularly carefully studied in order to achieve proper co-ordina- tion during disturbances or controlled manoeuvres. In America , too, the Sperry A-5 b egan a lon g line o f electronic autopilots, developed initially fo r precise aircraft control to pro- vide a stable bombing platform. In this equipment electrical differentia tion of displacement signals w as used to provide bot h rate and acceleration terms and the electrical and electronic cir- cuitry enabled a number of additional features to be incorporated, such as automatic airspeed control, height lock and, later on, radio approach couplers. By this time (shortly after World War 2) the heavier mili tary bomber and transport and civil airliner requirements had become very similar. But the development story may be inter rupted here by recording one particular autopilot which caused a substantial proportion of  Flight  reader s acute disco mfort towards the end o f World War 2, namely, that in the V.I. This was a rudder -and-

Upload: heiri

Post on 28-Feb-2018

214 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: 1957 - 0484

7/25/2019 1957 - 0484

http://slidepdf.com/reader/full/1957-0484 1/1

486 ;  FLIGHT 12

 April

  1957

A u t o p i l o t t o F l i g h t c o n t r o l S y s t e m  

the long Empire air routes for Imperial Airways in the late 1930s.

As far as concentration of panel instruments was concerned, the

first and probably most important group to be tackled was that

giving heading information. Various forms of gyro-magnetic

compass were developed in the 1930s to combine the magnetic

compass and directional gyro. The R.A.F. distant-reading com-

pass originally conceived by Air Ministry and S. G. Brown, Ltd.,

in 1934, was put into full-scale production for bombers and was,

in fact, a standard instrument throughout World War 2 in heavier

Service aircraft. In America, Sperry and Pioneer were working on

the same problem and examples of the result of their efforts were

the Sperry slaved gyro-magnetic compass in 1937 and the

Pioneer gyro-fluxgate compass which was widely used in U.S.

military aircraft during World War 2. These were the forerunners

of the present day Sperry Gyrosyn, Bendix, Kelvin Hughes and

other gyro-magnetic compasses which are now almost universally

used in all but the smallest and simplest types of aircraft.

The gyro-magnetic compass was a step in the right direction,

but the number of instruments required continued to grow as all

forms of radio aids, including Lorenz, Standard Beam Approach,

homing and, for military purposes, Gee and the other position-

fixing radio aids came to be added to the range and scope of

instrument information required at various times during a flight.

From the purely military point of view, during World War 2,

bombing aids and weapon sighting instruments were added, and

these still further aggravated the problem. Towards the end of the

war it became obvious that, for accurate bombing, much closer

integration of aircraft-control and weapon-sighting was needed;

and this laid the foundation for the modern electronic automatic

pilots which are now almost a standard fitment in heavy transport

as well as bomber aircraft. The combination of the American

Norden bomb-sight and Minneapolis-Honeywell electric autopilot

or the Sperry integrated bombing and autopilot system were

widely used in Liberators and Fortresses. Work along similar lines

for the British Services was carried out under the guidance of

R.A.E. and based on the Mk 14 bombsight, D.R. compass and

Mk 8 autopilot. Thus, for bombers, a stage had been reached by

the end of World War 2 where there was a large degree of

integration of gyro-magnetic compass, autopilot and bomb-aiming

equipment.

At this stage it is of advantage to digress for a while from the

main theme of integration and concentrate on the development of

autopilots, i.e., of the basic automatic Control equipment.

This equipment can be divided largely into two general cate-

gories : firstly, that which assists the pilot to fly his aircraft more

efficiently in the face of stability limitations in the basic air-

frame; and, secondly, that which relieves fatigue during flying for

comparatively long periods. Into the second category, of course,

fall all those civil autopilots of the pre-war era and, to a lesser

extent, the military equipments of the same and rather later years.

These were essentially displacement-type autopilots which, with

references about all three axes, worked through all three control

channels. The majority had facilities for the superimposition

of manual control so that normal flight manoeuvres such as

turns, climbs, and descents could be executed by overriding the

outopilot.

But for military purposes—and it was here that the greater

autopilot development effort was applied—the initial emphasis

was rather different. For example, the original Mk 1 autopilot,

developed by R.A.E., was intended primarily as a stabilizing

control for bombing. It was nominally a two-axis control for

rudder and elevators only, used a single conventional displacement

gyro, and corrected yaw and pitch disturbances. In fact, an indirect

control about the roll axis was achieved by the girnballing effect

of the inclined gimbal-axis of the gyro. Later versions, from the

Mk 2 up to the Mk 7, utilized a separate aileron-control gyro and

servo system in addition to this rudder/ elevator control. Other

forms of early military autopilots, such as the Pollock Brown and

some German Siemens and Askania types, were two- or single-axis

systems principally concerned with the steering and pitch func-

tions.

  But once the value of the stabilizing effect in these two axes

was appreciated there came a tendency to adopt a control on the

third axis, either by means of an additional separate gyro-reference

and servo-channel or by some form of combination of all three

channels.

Further development of airframe design, bringing improvement

of aerodynamic stability, particularly about the yaw axis, tended to

reduce the importance of the rudder-control function, thus allow-

ing the design of autopilots for control through ailerons and

elevators only. A notable example of this predominantly aileron/

elevator type of control was the R.A.E. Mk 8 autopilot widely used

in R.A.F. bomber aircraft during World War 2. This was a dis-

placement-type autopilot utilizing a single, gravity-controlled,

inclined-axis gyro to provide pitch and roll reference. A short-term

yaw reference was provided indirectly by gimballing effect and

the long-term heading reference was supplied by signals from the

The Curtiss F Boat

 in

 which Lawrence Sperry first installed

 a

 gyropilot.

It  is here seen in France

the Seine

in 1914

D.R. compass. Control was applied by pneumatic servos to the

aileron and elevator channels only. Complete manual override was

possible for climbs, descents, and turns; and a novel means of

  jinking or taking avoiding action, was incorporated.

As aircraft development continued it became apparent that the

aerodynamic characteristics or short-term stability of the airframe

would not be sufficient to damp out the inherent tendency of the

purely displacement-type of automatic control to hunt about the

mean flight-path, although various artificial methods were adopted

in the follow-up or feed-back function of autopilots to reduce this

hunting to a minimum. Thus a rate function was introduced and

(at least in the earlier stages) added to the displacement function

about one or more axes to provide a damping factor in the auto-

matic control system. In some cases the second derivative, i.e., an

acceleration term, was added; but until comparatively recently,

with one or two notable exceptions, acceleration terms were not

considered to be of major importance.

Before rate terms were incorporated, most autopilots had em-

ployed vacuum-driven gyros and either pneumatic or hydraulic

signalling and actuation. But the Pollock Brown equipment was

significant in that it was entirely hydraulic in operation, except for

an electrical system for applying manual override. The inclusion

of rate, and particularly the mixing of rate with displacement

terms, was much easier to achieve in electrical signalling systems;

and for this and other possibly more important reasons electrical

operation was introduced into autopilots. One of the most elab-

orate early electrical designs was a German three-axis control, the

joint product of Askania, Patin and Siemens, which, quite early

in the war, incorporated displacement, rate and acceleration terms.

The displacement and rate references were obtained from electric-

ally driven displacement and rate gyros, and the acceleration term

(about the pitch axis only) from a fore-and-aft linear accelerometer.

The various pick-off signals in each control channel were suitably

mixed in a galvanometer type of movement and the output was

used to control a rotary electrical servo coupled to the aircraft

controls. This equipment was especially interesting as it was one

of the first examples of an autopilot in which control was by the

rate  of application of control, i.e., a servo speed was made pro-

portional to the control signals and no follow-up system was

required.

In America, the Minneapolis-Honeywell electric autopilot was

another early example of the new technique of combining rate and

displacement. But here just two displacement gyros were used,

one vertical and one directional, the rate function being obtained

for the yaw plane by an ingenious viscosity throttle/potentiometer

device which, operated by the directional gyro, supplied yaw dis-

placement signals. Rate and displacement control was applied only

in the yaw plane, the other two planes being controlled solely

according to displacement. Control signals were electronically

amplified to operate the electric servos coupled to all three control

surfaces. The interaction between the three control surfaces was

particularly carefully studied in order to achieve proper co-ordina-

tion during disturbances or controlled manoeuvres.

In America, too, the Sperry A-5 began a long line of electronic

autopilots, developed initially for precise aircraft control to pro-

vide a stable bombing platform. In this equipment electrical

differentiation of displacement signals was used to provide both

rate and acceleration terms and the electrical and electronic cir-

cuitry enabled a number of additional features to be incorporated,

such as automatic airspeed control, height lock and, later on, radio

approach couplers.

By this time (shortly after World War 2) the heavier military

bomber and transport and civil airliner requirements had become

very similar. But the development story may be interrupted here

by recording one particular autopilot which caused a substantial

proportion of Flight readers acute discomfort towards the end of

World War 2, namely, that in the V.I. This was a rudder-and-