Handling

Save to de.icio.us

OVER the past two years I have covered various aerobatic manouevres and we have now exhausted most of them. So, our plan for the next few months is to write more about individual aerobatic aircraft, commenting on the good and not-so-good points of each.

These articles are not full flight test reports. They will concentrate on particular aspects of performance and handling as they relate to aerobatic flying, from a training viewpoint and also for competition and display flying.

These articles will have quite a lot to say about ‘handling’ and ‘performance’. To make meaningful comments on these attributes it is important to say first exactly what I mean by ‘handling’, and what constitutes some method of discussing and comparing ‘performance’.
So the aim of this month’s article is to lay down a more analytical framework for the future discussion of handling qualities.
Next month I will deal with performance issues.

HANDLING
MANY magazine flight tests make comments along the lines of ‘good’ or ‘nicely balanced’, or perhaps even ‘horrible’ handling, but such comments are all subjective and often anecdotal. Not all authors will agree on what constitutes ‘good’ handling. Aerobatics, however, is all about pure flying: the ability of the pilot to move the aircraft with precision about all three axes.

The aircraft I will be discussing all have manually operated controls with no artificial ‘feel’, no hydraulic-powered assistance and no computerised interference between hand and control surface.

In such simple aircraft, the most simple analysis of the ‘handling’ of the machine consists of only two parameters. How much effort you have to make to move the control and how quickly the aircraft responds to the control input.

There are but three controls, elevator, aileron and rudder, operated by significantly different muscles or muscle groups.

For optimum aerobatic handling qualities, these three controls should have different force profiles that I will detail later. Each aircraft type, too, has different relationship between the mass of the fuselage and of the wing, and the way the mass of each is distributed along the appropriate axis also varies.

These things affect the response of the aircraft to the control inputs, controlling the ease of starting and stopping as well as the time of reaction to a sudden input.

PITCH LOADS
When you fl y any aircraft, the first thing you find out about is its elevator feel and response. In an aerobatic aircraft, I want to be able to start a loop quickly and accurately, getting up to my desired G level without undue effort and with very precise aileron control to make sure I pull straight. How much effort do I need to make to get from level flight at +1g to looping at +4g or even +6g?

The aircraft designer, of course, has decided the answer to this question by the layout of the control system. Mind you, the designer’s aim may have had other targets or motivations than those which I might consider to be the ideal for aerobatics.

For example, the Robin 2160 and the Extra 300 are both twoseat certificated aircraft. In theory at least, they both comply with the same EASA/JAR design requirements, but the feel of their two elevator controls is different, to say the least. Much less pulling force is required to get +6g out of the Extra than is needed to achieve +4g in the Robin (see Figure 1). Why so much difference to reach, in each instance, about two-thirds of the normal G limit?

My guess is that the designer of the Robin was trying to make the aircraft ‘safer’ by making it next to impossible to exceed the 6g Flight Manual limit load. This design philosophy would argue that safety is ‘built in’ rather than achieved by appropriate pilot training. Thank goodness all designers do not share this philosophy.

As an aerobatic flying instructor, I often fly four or five instructional sorties in a day. I can do this in a Pitts or an Extra without getting unduly tired or losing accuracy. I don’t think I could do the same under the Robin designer’s safety regime!

In general, accuracy of control movement decreases as the effort required increases. Try knocking in a panel pin with a sledgehammer! So my preference is for lighter elevator controls, even though, at the extremely light end, such controls demand extensive amounts of both practice and relaxation.

PITCH RESPONSE
In competition flying there is a requirement to fl y loops that look round. This means that from flying straight and level, I need to get the aircraft loaded up to, say, 4g as quickly as I can. With a relatively light elevator, I can make this control input quickly and accurately. What then varies, however, is how quickly the aircraft actually changes attitude. This depends on the distribution of weight in the fuselage and on the position of the Centre of Gravity (CG) with respect to the Mean Aerodynamic Chord (MAC) of the wing.

Figure 2 shows the concept of MAC, along with two different CG positions. With the CG at 20%, the aircraft will be quite nose heavy, pitch stability will be high and response to rapid control inputs will be relatively slow. At 30%, the aircraft will be very lively and will respond much more quickly.

In this latter situation, though, the pilot must be careful not to overshoot and pull too much (see Figure 3). Overshooting is more likely if the control force is high than if it is light – back to the panel pin and the sledgehammer again.

Some Flight Manuals allow computation of CG in the form of a percentage of the MAC. With these aircraft it is relatively easy to predict their pitch response, based on this example of 20% and 30%. Other manuals, however, only produce a calculation in terms of the CG being a certain distance ‘aft of the datum’, and you as the operator are required to ensure that this is within the envelope.

Of course, because designers have varying design philosophies, the aft end of the permissible CG range is not always the same percentage of MAC.

Designer A might allow the aircraft to be loaded to 30% (CAP 232 for example), while designer B may be more conservative and have the most aft allowable position at 27%. But you will not know what these percentage limits are if you calculate in ‘inches aft of the datum’. Thus the only way truly to compare likely pitch responses of two aircraft is to have the CG calculated as this percentage of MAC.

AILERON LOADS
For similar reasons to those given above for the elevator, aircraft designers also create aileron controls with immensely different load characteristics.

Of course, there is a safety argument with aileron control forces, just as there is for elevator loads. Ailerons should not be fully deflected at Indicated Air Speeds in excess of Manoeuvre Speed. So a designer who believes pilots are unable or unlikely to exercise this discipline designs ailerons that are so heavy that the average pilot will be totally unable to make such a deflection at such a speed.

You must decide for yourself whether you think this a sound design philosophy and choose your aircraft accordingly.

I would just say that accuracy in flight path when rolling any aircraft depends on very precise elevator control at (probably) full aileron deflection. Such precision is much more readily achieved if the aileron control force is relatively small. The other problem with heavy ailerons is that it takes more time to move the control to full deflection. Thus it takes more time to get from zero to maximum roll rate.

Consider two aircraft with ostensibly the same roll rate, but with noticeably different aileron control forces. The aircraft with lighter ailerons will reach maximum roll rate sooner and thus will actually complete a half or a full 360° roll quicker than one with heavy ailerons (see Figure 4). In a half-roll, it may not even be possible to reach maximum roll rate before the half-roll is complete.

With more traditional aileron design, even in specialist aerobatic aircraft like many Pitts, the ailerons continue to get heavier the further you displace them. The loading curve is pretty much linear.

The very latest designs adopt a different approach. The smart designer assumes that if you actually move the aileron past a certain deflection, usually somewhere about halfway, you actually want to have all of it as quick as you can.

By adjusting aileron profile and hinge point, the ailerons can be made non-linear, such that after a certain point no more force is required to deflect them further. This enables full deflection, and hence max roll rate, further cutting down the time needed to complete a fraction of a roll. Such designs are especially useful in aircraft that perform many four-point rolls in their careers! This is most easily exploited in modern carbon-winged machines such as the Edge 540.

A further refinement is to design a noticeable break-out force into the aileron system. This prevents hugely powerful ailerons from being ‘sloppy’ in the middle which would give rise to wobbles in level flight and at the end of very fast rolls. Figure 5 shows the benefi cial characteristics of non-linear ailerons. Point A indicates the break-out force while Point B shows the point at which the ailerons deflect further without more eff ort. Designers have to be careful that the ailerons do not actually get lighter (red line) beyond Point B, as this leads to ‘snatching’ and loss of precise control at higher deflections.

AILERON RESPONSE
Roll acceleration is not just a matter of aileron design. It is also a function of wing span and wing mass distribution, both of which contribute to inertia.

Just as heavy aileron control forces will result in longer times for rolls, so will a heavy or especially long wing. At opposite ends of the wing inertia spectrum, consider the Slingsby T67 and the Pitts S1S. I don’t think a diagram is really needed here!

Aileron roll rates are invariably quoted in sales brochures as maximum rates and this inevitably means roll rate at full defl ection at Manoeuvre Speed. While two aircraft might appear ostensibly similar in such brochure fi gures, the real question is whether they will still roll the same as each other at lower air speeds.

Of course, this is not always the case. Some aircraft lose roll rate more quickly as speed drops off , and this is a noticeable detriment. An aircraft’s ability to maintain an adequate, if not sparkling, roll rate at lower air speeds seems to be dependent on the amount of the wing that is actually given over to the moving surface. This argument usually favours aircraft with short wing span and deep chord ailerons, such as the G202 or even the diminutive Jurca Tempête.

Lastly, aileron response will vary depending on the position that the wing is mounted on the fuselage. High-winged aircraft, like the Decathlon and the Cessna Aerobat are stabilised in roll by being suspended under the wing. This somewhat inhibits their roll response, even to maximum control deflection, especially at lower speeds.

Conversely, during the second half of a level roll, these aircraft are destabilised in roll because they are noticeably ‘low winged’ when inverted. Thus the roll rate always seems to speed up in the second half of a roll. Now you know why. Biplanes and mid-winged planes, where the centre of roll is close to the CG, tend to have the most consistent roll characteristics for reasons that should now be more obvious.

RUDDER HANDLING
Legs are stronger than arms. Thus rudder loads are designed to be heavier than stick forces. For aerobatics, again, lighter is generally better, up to a limit.

Maximum control deflection of the rudder is a common thing in spinning, but here entry speeds are low and control forces usually quite reasonable. For flick rolls, however, which might well be initiated at nearly double the spin entry speed, a heavy rudder can cause problems. Not only does it make repetitive practice very tiring but it also delays the ability to reach full deflection and hence to demonstrate a quick, clean wing drop.

There are various ways available to the designer to make rudder forces light, should he wish. These include horn balances, servo tabs and also the benefit of longer pedal travel.

Most GA pilots will be flying aircraft whose rudder pedals are hinged at the bottom and swing through a small arc to get full deflection. The tendency is to use ankle movement rather more than leg movement, but other design solutions are out there.

The Giles G202 and CAP 232 both have rudder controls that employ some sort of longitudinal sliding mechanism that actually means the whole leg can be brought to bear easily on the problem. The Sukhoi 26, 29 and 31 have large-radius, topmounted, hinged rudder pedals that, again, demand a lot of forward and aft leg movement. All these aircraft employ large rudders to good effect, made possible by affording the pilot the larger mechanical advantage in the linkage that derives from more leg movement.

YAW RESPONSE
In most aircraft, at low speed, it is quite easy to impose full rudder deflection. If you use less than this amount, though, how does your aircraft respond? Just as an understanding of pitch stability and response is important for aerobatics, so too is a knowledge of what contributes to rudder effectiveness.

When you yaw the aeroplane, you are effectively asking the fuselage to act like a wing and generate lift. As a result of its low aspect ratio, it also generates a lot of drag. Hence we think more of the rudder as a drag inducer than we do of the fuselage as a lift creator. But we would be well advised to think the other way around when looking at the handling of an aerobatic aircraft.

By design, most aircraft are more stable in yaw than in pitch. We pull the stick back, fully expecting the aircraft to respond by looping. Why then do all aircraft not perform ‘flat’ loops when we apply the rudder?
Well, some do, albeit more slowly in terms of turn rate than in pitch rate. Most don’t. The majority of aircraft are reluctant to turn flat. They respond initially with yaw when rudder is applied but then the vertical stabiliser has its way and they yaw no more, despite further rudder deflection.

In the days of the Wright Brothers and Bleriot, however, aircraft were able to turn fl at. Indeed, turns then were made with the rudder while the bank angle was kept close to zero. How else would you turn an aircraft with a 30-foot wingspan that only flew 10 feet off the ground? Certainly not bank and yank.

When we make CG calculations, they are always with reference to the lift characteristics of the wing. But it would be an equally valid and especially interesting exercise to derive a CG envelope based on the lifting characteristics of the fuselage and its centre of pressure in knife-edge flight.

In most cases, I am sure we would find the CG is effectively further forward in yaw than it is in pitch. In terms of yaw responsiveness, the more lively aircraft tend to be those with short fuselages and more fl at keel surface areas forward of the CG. Pitts with bungee undercarriage and CAP 232 are both good aircraft in this respect. Many others are disappointing.

Next time you are up, try flat turning your aeroplane. Add some rudder to turn and keep the wings level with opposite aileron (wing warping if you are in the Bleriot!). See how it responds at, say, 1.5 times normal 1g stalling speed. Can you get a steady turn? What is the turn rate?

CONCLUSION
Any analysis of the handling characteristics of an aeroplane demands a systematic approach based on an understanding of the underlying aerodynamics.

A few simple tests will soon give an idea of the control forces and response characteristics of any aircraft and thus an understanding of its likely strong and weak points aerobatically. Designers of aircraft intended for aerobatics should not approach the subject thinking that they can make it perfectly safe. They cannot. Aerobatic safety comes from good instruction, sound understanding, lots of practice and, above all, self discipline.
Be safe and enjoy your flying.

Send To A Friend