Pick up the nose.pdf

plate such a project do not have any real idea of just how ..... slipstream when slowing the airplane with power on. .... they have hold of a tiger or a pussy cat.
2MB taille 2 téléchargements 314 vues
By M. B. (Molt) Taylor (EAA 14794) Box 1171 Longview, WA 98632

V^ONVERSATIONS WITH HOMEBUILDERS over the years have convinced us that many people who either are building their own lightplane or who contemplate such a project do not have any real idea of just how their plane flies. In particular, they really don't seem to

understand the limitations and principles for control of their particular plane. If they do have some idea of how things work, they still have little concept of the reasons why their favorite design may have certain flight qualities as contrasted with some other design. First, let us begin by saying that all airplanes fly

and are controlled pretty much alike. This goes for conventional designs as well as unconventional designs, taildraggers or tri-cycle gear, pusher or tractor, high wing or low wing, canards, powered planes or gliders. Name it, they are all working by basically the same

principals and fly by the same rules. No one has come up with anything basically new in a long time. There

have been refinements, and there has been recent publicity on some things, but none are basically new or revolutionary. One area where it is evident that many homebuilders are getting into trouble with their pride and joys, particularly on their first flights, has to do with the control authority of the elevators. While it is necessary to keep a lightplane running basically in a straight line and to keep it on the runway during the takeoff, few

homebuilders seem to have very much trouble in that regard. It is when they try to lift off, or "rotate" for

takeoff with a tri-geared design that things begin to happen. This is because they usually have not given the matter of how their airplane is controlled too much thought. This may sound very basic to many readers, and it is. However, we seem to see more and more takeoff accidents these days. This is particularly true of some of the so-called unconventional designs such as pusher arrangements, etc., which are now becoming

more and more accepted. No one will disagree with the big argument for the pusher arrangement for propellers. Certainly it doesn't take much imagination to see that the airplane should fly better if it doesn't have to fly in the wake of its own propeller. However, most of us have learned to fly tractor propellered aircraft and have learned to live with the peculiarities of that arrangement. Transitioning to a pusher requires a certain amount of relearning since pusher aircraft have their own set of peculiarities. Nowhere does this manifest itself any more than it does on takeoff. The reasons for

this are obvious if you think about it, but a lot of people seem to take if for granted that they can fly one since they see other people fly them, and as a result they jump into their new airplane, give it the throttle, run down the runway and over-rotate, whereupon they do the worst thing they could do and that is jerk the throttle off. We have seen several people try to fly their own

designs or some new design for the first time, and have sadly helped tow the wreckage back to the hangar, or seen them off to the hospital in the ambulance. All because they just hadn't given the takeoff thing enough thought and consideration. What happens is quite clear, but nevertheless a bit

complicated. In the first place, most pusher and unconventional designs have the engine placed to the rear of the pilot. This usually means that the center of gravity (CG) of the aircraft is as far aft as possible in the empty condition. Then, anything they put in the airplane (including people, baggage or fuel) ends up effectively mov-

ing the CG forward. This is just the reverse of conventional designs where they usually end up with the CG further and further aft the more you put in them. Thus, in the conventional design you have the most forward CG condition with minimum fuel and the lightest possible pilot. Heavier pilots, more people, more fuel or baggage results in the CG moving aft. Look at most 4 place

designs (such as a Bonanza) and you will see that the people in the rear seat are sitting almost on the trailing edge of the wing or at least on the trailing edge of the

mean aerodynamic chord (MAC). In pushers the reverse is true. Usually the designer of a pusher ends up with

his empty CG just about over the wheels. Some of them are even so tail heavy when they are empty that they

will fall back on their tails unless the designer incorporates some sort of arrangement to prevent it when everyone gets out of the airplane. These designs end up with the CG moving forward with more people aboard, or more fuel or baggage. The reason for this is simple. The designer wants the airplane to be able to take off. If the CG is too far forward of the main wheels, the designer has to incorporate a larger tail with more up elevator control surface movement in order to develop enough down force to push the tail down as the airplane rushes down the runway on the takeoff roll. This is, of course, desirable since if the airplane will not rotate by the time it is up to at least its stall speed, it will then have to accelerate even well above stall speed before it SPORT AVIATION 19

can be rotated and the nose lifted so that the wings will present enough angle of attack to the relative wind to lift the airplane. It is easy to see that in some ways this is good since it means that the more people there are in the airplane, or the heavier it may be, the faster it has to get going before it can be rotated and flown. On the other hand the "conventional" type design ends up with the CG furthest forward when it is at its lightest. Thus, the conventional airplane has to be designed with enough up elevator to rotate it when it has only a lightweight pilot aboard, and as people, fuel or baggage are loaded aboard and the airplane gets heavier, it will rotate just that much easier for takeoff. This can be dangerous since as the airplane gets up to its maximum gross weight with a full load it can sometimes be rotated much too early in the takeoff run and will stagger into the air before it really is ready to fly. This is due to the proximity of the wings to the ground (usually called ground effect) and more than one overloaded conventional design has been seen to get airborne only to fly down the runway without ever getting out of ground effect and never being able to accelerate and climb. As can be seen, there are penalties and peculiarities for either conventional or unconventional aircraft configurations. The thing that bothers us most about this problem is the fact that a lot of homebuilders seem to ignore the fact that one design doesn't fly like another. While designers try to come up with final configurations which will not have any dangerous characteristics, they still find it necessary to live with some different inherent characteristics the moment they design anything but a conventional airplane. With the pusher configuration where the propeller blast is hitting the elevators on some designs (such as our own Coots) the resulting elevator effectiveness is even greater than it is with tractor arrangements. That is, effectiveness for a given amount of elevator deflection. So, the designer must determine how many degrees the elevator control surfaces must move for a given amount of stick (or wheel pull) movement. It is easy to see that the effectiveness (that is, what happens as a result of a given elevator deflection) can vary greatly with the amount of throttle or power applied (or withdrawn). The airplane may rotate easily at full throttle, whereas with part throttle it may not rotate at all. Further, with an arrangement such as the Coot or Lake) the effectiveness of the elevator is also related to the air speed of the airplane at any given instant, just like any tractor type. Probably the biggest problem of the pusher configuration as far as elevator effect is concerned has to do with the fact that the source of the push from the propeller is quite a ways above the point of ground contact for the wheels. Thus since the wheels tend to pull back on the airplane due to their rolling resistance, etc., the propeller is tending to push the nose of the airplane down. This down force requires the elevators to have to

where the thrust has to overcome the rolling resistance and drag of the landing gear and tire contact. It is here that another peculiarity of such designs comes into play

and that is the fact that the moment the airplane leaves

the ground the wheel contact forces tending to pitch the

airplane nose down are instantly removed. The result is that the airplane will at that moment tend to pitch up. This may be ever so slight with some designs, but others have a very distinct tendency to nose up (sometimes quite sharply) the moment the wheels leave the ground. From this it is easy to see that unless the pilot is prepared for this characteristic on his first flight with a new design, he may end up with the nose pointed too far skyward for comfort and his natural tendency is to shove the wheel (or stick) forward. He may even cut the throttle. This complicates matters even further since the propeller blast is no longer hitting the elevators so that their effectiveness in holding the nose up is instantly reduced at the same time the pilot is reducing their up deflection. This often ends up with the airplane pitching nose down into the ground, whereupon the nosewheel will either fail or bounce up to pitch the nose up again. The pilot invariably gets excited about this point and tries to keep up with the alternate pitch up and pitch down, only he usually gets out of phase in his reactions and the resulting crash can put an end to a lot of dreams. There are several lessons to be learned from the above description. First, be prepared to feel out the responses of your own particular example of any design, whether it is a well-proven one or one of your own. It is obvious that the CG position with respect to the wing chord as well as the position of the wheels with respect to the wing chord are of utmost importance in determining the response of the flight controls, particularly on takeoff. While it is, of course, also a matter of where the CG is in respect to the position of the wheels, both the CG and the wheels have to be properly located with respect to the position of the wing. Further, the angle of attack of the wing with respect to the angle of attack

LANDING GEAR OR PITCH COUPLE

CENTER OF MASS A N D APPROX. DRAG CM

FORCES TO BE OVERCOME TO ROTATE

develop additional counteracting down load on the tail to overcome the nose down pitch due to the thrust of the

high propeller. This can be relieved to some extent by the designer by inclining the thrust line downward, but there are practical limits in this regard — obviously. It behooves the designer to arrange things so that the

propeller thrust line passes as nearly through the mass

center of the airplane as possible since that always gives the very desirable characteristic of having very little pitch trim change with various throttle positions or

amounts of power applied. This is good from the

standpoint of trim changes during approaches with the variation of throttle, but the designer still has to live with the fact that in pusher configurations he has a case 20 JANUARY 1979

LESS UP ELEVATOR REQUIRED

FORCES TO BE OVERCOME AFTER TAKE-OFF

(usually negative in conventional designs) of the tail has a direct bearing on how well the airplane is going to respond to a given amount of up (or down) elevator deflection. While on that point we should recognize that the actual angular travel of the elevators with respect to the amount of flight control movement has to be considered in all of its importance. We find people change the

However, the limiting of up elevator travel was the most effective part of the design considerations and, of course, this feature is now being incorporated in some of our newer homebuilts. This is particularly needed in some of the canard type machines since it is essential that the forward wing (or canard surface) be arranged so that it

homebuilt projects with absolutely no regard or knowledge of the eventual effect it will have on the flight characteristics and controllability of their airplane. Remember, the designer went to a lot of trouble, experiment, test and calculations to find out how much tail surface area his design should have, how far those surfaces should be moved, what limits they should have to their travel and how much they should move with regard to movement of the flight controls in the cockpit. When you change things you can be playing with dynamite. Particularly, if you don't know what you are doing (or don't care). This brings up another point with regard to the movement of flight control surfaces and in particular the elevators and that is the matter of limited up elevator travel which we are seeing as part of some of the newer designs. There is really nothing new about this and the

speeds somewhat above their stall speed. If the main

length of the control stick or change the leverage of the flight controls for one reason or the other in their

well-known Ercoupe is, of course, a historic example of

this sort of arrangement. In the case of the Ercoupe the

designer was particularly interested in developing a spin

proof design. He also incorporated several other design features into the airplane in addition to the limited

elevator effectiveness in order to obtain the spin proof characteristic, such as cutting away the center portion of

the elevator surfaces to reduce the effect of the propeller

slipstream when slowing the airplane with power on. The rudder surfaces were linked to the aileron control

system in such a way as to only move outward, but the rudders were also put outboard of the propeller wash to get away from power effects. The engine was also offset to take care of the assymmetrical slipstream ("P" effect).

will stall before the main wing. This has the disadvantage of also limiting the lift obtainable from the main wing so that such designs usually must be landed at

wing were permitted to stall, such designs usually would pitch up violently, and doing this close to the ground ends up with something bent.

As we have seen, elevator effectiveness is merely a matter of the angular displacement of the surfaces, the velocity of the air going over them and the area of the

surfaces. With a conventional design it is usually necessary to size the elevators and determine their deflection

so that the airplane can be flared for landing. Any more effectiveness is usually not needed except possibly in some of the acrobatic designs. Usually, this consideration is quite adequate for the takeoff condition since with the tractor type airplane the elevators are obviously going to be more effective on takeoff with the

slipstream on them than they are during landing where the engine is usually at an idle. This "flareability" situation changes when the propeller is installed closer to

the elevators in some pusher arrangements, such as our Coot (or the Lake) and other considerations complicate things when the propeller is placed even further aft as in such designs as our Mini-IMP, IMP and other designs where the propeller is entirely aft of the airplane. Such

designs must then take into consideration such matters as the takeoff problem as well as being able to flare for landing. While some people get concerned about the possibility of striking the propeller on takeoff and landing,

this is usually taken care of by merely having the landing gear long enough so that at the stall angle of the wing, the propeller just cannot touch the runway (as we do with the Mini-IMP). It should be remembered that the landing gear usually drops down somewhat from the position it has when the airplane is sitting on the ground and while it may look like the prop could touch the ground on landing or takeoff, the free position of the gear at the instant of touchdown can be far enough down so as to prevent any propeller strikes. On takeoff

if the airplane is going fast enough for the elevators to

rotate the airplane, the moment the wings are brought to sufficient angle of attack to fly they instantly lift the airplane before the propeller ever has a chance to touch the runway. However, the designer still has to have

enough up elevator effectiveness to flare for landing. This is always complicated by the fore and aft travel of TAKE-OFF AND CLIMB CONDITION

the CG, and in some conventional designs it is possible to touch the tail skid with aft CG conditions (usually fully loaded). On the other hand, some conventional designs exhibit limits in regard to being able to hold the nose off the runway when lightly loaded. The so-called unconventional design may show limits of being able to hold the nose up when fully loaded but can be landed well nose up in the light condition. This characteristic

permits the airplane to sometimes be flown with the

C.P. MOVES FWD DUE TO REFLEX '

WING REFLEXED

REDUCED DRAG "CRUISE VALUES FOR EXAMPLE ONLY

CONDITION

stick full back, and sometimes such designs can be climbed out with full up elevator if the flight controls are suitably limited since it is the full load condition that designs the limitations for up elevator. At lighter weights full up elevator still will not stall the airplane if the CG travel is sufficiently limited by proper loadings or if suitable airfoil characteristics are used by the designer so that stall angles cannot be approached with the limited elevator effectiveness designed into the aircraft. SPORT AVIATION 21

It is interesting to consider the effect of airfoil sections on elevator effectiveness. In the case of our own Mini-IMF, it was decided to initially build the airplane with a fixed horizontal tail surface. While the anhedral (inverted V) tail uses only two surfaces, it still has the effect of being both horizontal and vertical tails. We also elected to incorporate the very latest (at the time) NASA airfoil section called the GA (PC)-l. This wing section has a full span "flaperon" which serves both as aileron and flaps. In addition, these surfaces can be collectively moved up about 10 degrees for cruise. As a result the flaperons have three positions — namely, 1. Takeoff, 2. Landing (flaps) and 3. Cruise. Differential movement of the flaperons is used for the usual lateral (aileron) function and in that movement they have a 3 to 1 differential travel with the down aileron movement only one third that of the up surface travel. Further, the actual aileron movement of these surfaces is quite small (that is all that is needed). However, we quickly found that when we moved the flaperon control to the cruise position we were having to carry lots of down elevator control movement (and attendent down elevator surface position). Although we had initially designed considerable down travel into the elevator system, we found that the moment we got the airplane into any appreciable nose down condition where the speed would build up, we were having to hold full down on the elevators. Since the anhedral tail arrangement necessitates the use of two trim tabs (if you use tabs for pitch trim), because one tab would give a "rudder effect", we had initially tried to use a "bungee spring" type trim control. This proved to be completely inadequate for holding the necessary down elevator needed in cruise condition for the wing. As a result, it was necessary to devise a moveable horizontal tail surface type trim system for the MiniIMP; that arrangement is now incorporated on the prototype and is what we show in the plans for the design. The interesting thing about this is that when we went to the full moveable horizontal tail (stabilizer) trim system we found that we had an appreciable gain in cruise speed, and examination of this showed that the reason was two-fold. First, since the down load on the tail of a conventional airplane has the same effect on the amount of lift required from the main wing as if it were just that much added weight in the airplane, the main wing then has to be flown at a sufficiently high angle of attack to not only hold up the weight of the airplane but also the down load weight on the tail. This obviously results in a certain amount of induced drag (drag due to lift). Further, the negative angle of attack on the tail required to hold the tail down (necessary because the CG is ahead of the center of pressure on the wing) also adds to the induced drag. Reflexing the trailing edge of the wing up as is done with the GA (PC)-l airfoil for cruise results in the center of pressure moving forward or closer to the CG (horizontally). This reduces the amount of down load required from the horizontal tail surfaces. Thus, we found that we could cruise with very little down load (actually about one degree negative angle of attack). This change (reduction) in tail angle of

attack and the reduction in induced drag on the wing (due to its not having to also hold the tail load up) as well as the actual reduction in induced drag on the wing

due to the trailing edge being pulled up into the wake of the wing all resulted in an appreciable increase in cruise speed. The reduction in the negative angle of attack of the horizontal tail, of course, reduced the amount of down elevator required for cruise to zero. The total effect on the performance at cruise for the Mini-IMP proved to be quite worthwhile and we are very pleased with the performance of this relatively new airfoil as developed by NASA. However, we should point out that 22 JANUARY 1979

this, too, is really nothing so new and that the sailplane people have been doing this same thing with their sailplanes for years. However, the NASA people have developed the PC airfoil for powered planes whereas most sailplane airfoil sections are designed specifically for that type aircraft where they have very little landing flare and are flown somewhat differently with aircraft tow for takeoff, etc., where it is usually a problem to hold the nose down rather than lift it. It is interesting to note that in the cruise condition where the Mini-IMP has very little decalage (difference in angle of attack between the main wing and tail), the longitudinal stability of the aircraft is excellent and the ability to change the decalage angle in flight is very desireable. The reflexed (collective up) aileron (flaperon) condition is only used for flight at cruise speeds, and when the airplane is slowed down for landing the flaperon control is again moved to the takeoff and/or flap condition. In these positions it is again necessary to move the trim control into the tail down position. While this control movement could easily be interconnected so as to automatically change the trim control when the flaperon control is moved to cruise, we have elected to keep the two controls independent and have found that this, in effect, has added considerable flexability to flying the Mini-IMPs. After all, who ever saw a bird that couldn't change the angle of attack of his wings with regard to his body or couldn't change the angle of attack of the tail surfaces. So what's new? There are a lot of other things that affect the flight characteristics of any aircraft design, but most pilots find that they can quickly accommodate themselves to these if they can once get the airplane into the air where they can feel out things and get some idea of how the airplane is going to respond. The fact that some designs are extremely sensitive to control movement when others are extremely docile is ample evidence that pilots can usually learn to fly just about anything. However, they have to get them in the air to find out whether they have hold of a tiger or a pussy cat. This usually entails a bit of tiger tail tickling to start with, and anyone flying his homebuilt airplane for the first time should regard those first attempts at takeoff with just that much respect. Personally, we feel that it just makes good sense to tickle the tiger's tail with things happening just as slowly as possible. Thus.it isn't a good idea to give the airplane full throttle and go rushing down the runway on a first takeoff attempt. The initial testing of any airplane should embody a lot of gradually increasing speed runs as you feel out the controls. With pusher designs it is always good to get the airplane gradually up to speed with all wheels still on the ground, reduce throttle to idle and then (and only then) try to lift the

nose off the ground. This practice will quickly give you some idea of elevator effectiveness without having the complication and loadings of the thrust trying to push the nosewheel into the ground to start with. Once you get the feel for things with the power either off or reduced, then you can gradually approach the condition of

rotation with power continuously applied during the takeoff process. However, if things get hairy the power

should never be reduced sharply. It is far better to reduce power gradually even if the nose is (or seems to be)

too high. Remember that reducing power changes the effectiveness of the elevator rapidly, but it is usually far better to keep the nose up than it is to dive it into the ground. It is our hope that these remarks will give at least one pilot a better idea of what he is trying to do, and if we can save just one homebuilt airplane and homebuilder from having to go through the agony of rebuilding

his airplane (or himself) the story will have been worthwhile.