SPINDLES AND
GIMBALS!
The designers of autogyros built before the offset gimbal was invented had little option but to arrange the controls for the rotor so that the pilot's muscular input was amplified to allow him or her to cope with the fairly large forces produced by that rotor. This was usually achieved by the simple principle of having a long lever down from the control pivot- point at the rotor-head to the pilots hands and a short spindle from that pivot point up to the rotor CG.
Most modern day trainee gyro pilots have experienced the brutality of the forces developed when they mess up their rotor starting and their little 22 foot (6.7 m) rotor decides to start flapping. All they, or anyone else can do under such circumstances is struggle to push the stick hard forward and neutral until the thrashing and jolting stops. With that bit of experience under your belt, (which you make a point of avoiding thenceforward!) you may be able to guess what effort could be needed to handle a three-bladed 30 foot (9. 1 m) rotor in direct control if it chose to exert its influence to the full.
We still call our system "Direct control" but that is possibly something of a misnomer because the offset gimbal head acts as a sort of servo system to some extent. But I am racing ahead; let us go back a bit and look at the old and, in many ways, rather dangerous spindle head.
This is more or less what it says: the rotor assembly is mounted on a spindle which can rotate and which, because of the way it is mounted, can be tilted to the front and rear and to the left and right.
On his first machines, because they had evolved directly from Raoul Haffer's inexpensive and disposable Rotachutes. Igor Bensen retained the spindle head. The rotor hub-bar had a slot through it and the spindle passed up through this to allow a teeter bolt to be inserted through two adjacent pillow-blocks and through a cross-hole near the top of the spindle, (Figure 1).
The bearing in which the spindle rotated was simply a self- aligning type. This meant that, in the extreme, the spindle carried by the inner casing could be misaligned by as much as 9deg from the outer casing.(Figure 2).
In principle, that allowed the right amount of control movement on the rotor spindle. But the reality was that such bearings were never intended to take the sorts of radial and axial loads and at the maximum angles of misalignment which the gyro demanded of them. Fortunately, whilst they often had a short life and a happy one, they did not, as far as I can recall, ever actually fall apart but Goodness knows why not!
To get the control inputs to the rotor spindle, the control lever (initially an 'overhead type', extremely simple and infallible) was attached to it via a set of small bearings, (Figure 3). So as the control was moves the spindle was tilted but it was able to keep spinning quite freely.
Another small mechanical requirement was a sort of 'stop pin' which made sure that any movement of the overhead stick - to the side, for a roll input, for example, did actually result in a tilt of the spindle and not simply a planwise rotation of the overhead stick around its own little bearing system, leaving the spindle standing vertically. Personally, having seen Geof Whatley fly an overhead stick with such ease and confidence and because it lacked so much of the complication of bearings and pushrods, etc., which a conventional control column has to involve, be it rear-mounted or basemounted, I think it is a great pity that the overhead stick has quietly vanished from the scene. The biggest criticism of it is that it is the reverse of aviation convention. But after all, is not every flex-wing microlight a blatant flaunting of such convention? And would anyone dare to say that they are unflyable death-traps? There are. after all, well over 2000 of the little beasties flying in the UK alone. Without the undesirable and, since 1939, unnecessary spindle head, one does even need the two small bearings at the top of an overhead stick. Such a shame it seems to have gone for ever.
But, as always, I digress. At a glance, the spindle head looks very simple and, if one could find the right bearing arrangement, very robust. Why move away from it? Let us just take a look at its operation.
In the very basic form initially utilised on Bensen B7s and B8s, the control system amounted to a rotor spinning on a spindle, which passed through a self-aligning bearing (which acted as the control fulcrum) and, at its lower end, a long extension was attached. On the bottom end of this extension, Figure 4), were the pilot's hands operating on a T-shaped (and ergonomically very pleasant) crossbar; the twist-grip throttle was located on one of these handles, as on a motorcycle. This whole set-up amounted to a lever system with the teeter-bolt set 3.4 inches (86 mm) above the centre of the self-aligning bearing and the control handles set about 46 inches (1 168 mm) below it.
That amounts to a lever advantage in favour of the pilot of about 13.6 to 1.
In level flight, the rotor assembly might be developing about 100 lb (45 kg) of drag, trying to push the top of the spindle rearwards (Figure 5).
With a lever advantage of 13.6, the pilot only had to apply a force of 7.3lb (3.3 kg) to the base of the overhead stick and all was nicely in controlled balance. But, because of the angle at which the overhead stick operated on the spindle, this force would be slightly increased to, say, about 8lb or so.
One might liken this to holding a sheet of metal or thin board horizontally in a gentle breeze, (Figure 6), something which would not be particularly difficult to do.
Now, suppose we dive our gyro at the highest speed we can reasonably attain without being totally stupid. To do this, we shall certainly be holding the overhead stick fairly well into our stomach, keeping the rotor disc at a shallow angle of attack, (Figure 7).
Now, to pull out, we move the overhead stick forward fairly smartly. The spindle is inclined aft and the rotor disc suddenly sees a hefty angle of attack. The lift rises very markedly and so, unfortunately, does the rotor drag.
As far as the pilot is concerned, he experiences a sudden huge force, (Figure 8).
In a sense, a sudden mid-air 'flare' has occurred, just like that which we controllably introduce with an offset gimbal head, just prior to touchdown. Only this spindle-head flare can be horribly violent.
The combined forces are most likely to be far too much for the pilot to hold back, especially if he is taken unawares. The net result is that the rotor drags the spindle even further rearwards and the whole process is made even more violent. Coming back to the analogy of holding a thin sheet of metal into a breeze. If that breeze stays low and/or we keep the sheet flat, all stays well. But if that breeze suddenly becomes a gale and we raise the front of the board a little, it will instantly flap upwards and backwards over our head. There would be virtually nothing we could do to top it (Figure 9).
If you think I have exaggerated, I promise you I have not. The self-same sequence of events killed Ernie Brooks in his 'own design' gyro some decades ago. This had a spindle head and he was progressively flying faster and faster descents, fly-bys and sharper initiations of climb in order to impress a would-be investor. Finally, the spindle head won and the machine almost looped round to the inverted position before it dropped through its own rotor, so to speak, to the ground.
But let us be clear that the problem was the spindle head and not the overhead stick.
OK, so what makes the offset gimbal head so much better? In a nutshell, it is an elegantly simple device which is self- stabilising in pitch. Mr. M. Shrenck is usually credited for coming up with the concept and patenting it back in 1939. But that fails to explain why an offset gimbal is incorporated in the Cierva Rota, first built in 1935. Whoever did it and whenever they did it is not important since any patents have long since expired. What is important is that it is such a boon to our little craft. But what actually makes it self-stabilising?
Essentially what it does is to balance the amount of lift being developed against the amount of drag being simultaneously produced by using their separate moment arms to advantage, (Figure 10). Actually, in the versions which we all use, one should say "Almost balance", because a little bit of imbalance is deliberately left in the basic geometry.
.
We then add that last smidgkin of balance in the form of a ground - or inflight- adjustable spring system, which neatly allows us to cater for different pilot weights, flying speeds, etc. and trim back to 'hands off flight, (Figure 11).
The crucial pivot is the pitch-bolt located transversely through the gimbal block (Figure 12) and the important distances are the horizontal length from this bolt to the vertical line down through the centre of the rotor bearing and the vertical distance through the pitch bolt and the teeter-bolt. For practical purposes, we can regard the teeter bolt as representing the level at which the drag developed by the rotor occurs.
Looking at how the forces times distances
(= "Moments") balance out and referring to Figure 12, we will make some reasonable guesstimates to demonstrate what is happening. Let us rightly take the all-up mass of the gyro including pilot and fuel but excluding the rotor in this instance, as 550 lb (249.5 kg). We will assume that the rotor drag for the flight mode we are in, is 100 lb (45 kg).For the commonest form of offset gimbal head, based on Igor Bensen's original version, the vertical axis of the rotor bearing is typically 1 inch (25.4 mm) horizontally aft of the transverse pitch-bolt. The generated lift from the rotor acts more or less up this line through the bearing axis. So the lift supporting the mass of the machine and pilot, etc. causes a moment of 1 x 550 inch.pounds
= 550 in.lb. (6.337 m.kg) about the pitch bolt. This moment is trying to force the torque beam of the rotor-head up at the back end, over the top and forward. As drawn, we can call this an anti-clockwise moment, (Figure 12).If, as is not unusual for a standard Bensen or Bensen-derivative rotor-head, the teeter bolt is located about 5 inches (127 mm) above the horizontal line through the pitch bolt then, with a rotor drag of 100 lb, we have a moment of 5 x 100 inch.pounds 500 in.ib. (5.761 m.kg).
Note that this moment, in contrast to that produced by the lift force, is trying to tilt the torque beam of the rotor-head down at the back; as drawn, it is a clockwise moment. The two moments are clearly in direct opposition.
If things were left like that, the forward / downward (anti-clock-wise) moment would be the larger and would win; our gyro would want to dive.
But - and here's the nice bit - we can easily attach a spring to the back of the torque beam. Because the spring is angled forward to connect to an anchoring bracket on the mast, it acts at a vectored distance of 6.1 inches (155 mm) behind the transverse pitch-bolt and we set it up to pull with a force of 8.2lb (3.72 kg). In doing so, we are adding a moment of 6.1 x 8.2 inch.pounds = 50 in.Ib (0.576 m.kg). See Figure 12.
This is an added moment which is acting downwards at the back end of the torque beam (clockwise), making the latter try to tilt back and up and the gyro to climb, so we are adding it to the moment produced by the rotor drag, thus:
(5 x 100) + (6.1 x 8.2) 500 + 50 550 in.Ib Clockwise which is exactly the same as the moment developed by the lift supporting the machine, which was trying to make the gyro dive.
So we now have a rotor-head and the machine which dangles from it, in beautiful balance. it will happily fly 'hands off' in that situation. If we want to fly faster and still 'hands off', the rotor drag may increase, so we will have too much effort trying to pull the torque-beam into the 'climb' mode. But, Hey , presto! we simply relax the tension on the trim spring with our inflight pitch trimmer, and its contribution to the rearward moment is reduced and the head system comes nicely back into balance.
Similarly, if we slow the gyro down in level flight, the rotor probably develops less drag and we need to add tension to the trim spring in order to keep the aft moment on the head still in balance.
If you have not simply flicked over these pages but have bothered to read and understand what I have said, you will also have gained a clue as to why other odd things - which you have perhaps noted on your own machine -have occurred.
Suppose you flew your gyro with Bensen, Brock or Rotordyne blades and then switched to Dragon Wings. You found that you seemed to be having to pull back on the control column more than with the previous rotor. Perhaps you reset the position of the trim-spring bracket or trimmer unit?
OK, let us look at that in the context of the earlier statements. We had the rotor-head and trim springs set up so that, in level cruise, the machine would fly hands off. So the trim spring, operating at its fixed distance behind the pitch bolt, was giving just the right force x distance to add to that developed by the rotor drag and its distance above the pitch bolt, to equate to the force times distance produced by the lift from the rotor (Figure 12).
But now you have a rotor which apparently produces less drag than the other, earlier types. So what, in our guesstimated example- was a moment (due to rotor drag) of 5 inches x 100lb, (500 in. lb) now becomes perhaps, 5 inches x 75 lb, (375 in.lb). With the additional moment of 50 in.Jb produced by our trim- spring arrangement, that gives a total of (50 + 375) = 425 in.lb.
That leaves a moment of 550- 425= 125 in.Ib not balanced out. So you, the pilot, have to provide it by hauling on the control column extra hard.
But if you take some steps to get the balance restored, you can get back to the 'hands off' situation you previously enjoyed.
There are several ways which could all help to achieve this, (Figure 13), viz:
1. You could reduce the weight of the whole machine or (Heaven forbid!) slim yourself.
That would reduce the 550 in.Ib moment due to lift, so there would be less to balance out.
(But it might also mean that the rotor disc would cruise at a lower angle of attack, so might not achieve all that you expected).
2. You could fit stronger trim springs to increase the 'additional' balancing moment rearwards. Alternatively, you could move their attachment to the torque beam rearwards.
That would certainly work but, at very low rotor speeds or with the rotor stopped, you might begin to curse the hefty force you had to keep applying to the control column.
3. You could raise the location of the main rotor bearing, its housing and pillow blocks, etc.. This would increase the backward moment due to the (reduced) drag from the rotor acting at 5 plus inches and help to restore the previous set-up.
However, you should not do that without doing the necessary calculations or getting someone else to do them, because the loadings on the arrangement supported by the main rotor-bearing bolt will have increased. The PFA Engineering Department may not be too happy.
This change would also require you to do some revamping of your prerotator installation, with the gear or pulley having been raised.
4. You could move the axis of the main rotor-bearing closer to the pitch-bolt. As usually made, this dimension is one inch (25.4 mm) so you could reduce it to 0.75 inch (19.1 mm). The moment produced by the lift would instantly be dropped to 550 lb x 0.75 inch = 412.5 in.lb. This could probably be almost accommodated by the original trim system, if the rotor drag were indeed reduced to 75 lb (giving that total backward moment of 375 + 50 525 in. lb).
Apart from making sure that the new location of the hole for the ½inch diameter bolt does not make any part of it carve into the hole for the pitch bolt, this approach does require two other things: (i) A new torque beam has to be made, because the new clearance hole for the 1/2inch castle nut would otherwise intersect the old one, and (ii) any prerotator system would probably have to be relocated on the beam.
Finally, coming back to the very light and uncomplicated overhead stick: It is easy to see how simply it could be attached to the front of the gimbal head by boring a 'blind' hole down the centre of the torque beam, inserting the top of the stick and bolting it in place with a cross bolt. Mechanically trivial, ergonomically excellent but bureaucratically probably a nightmare. (Purists; please don't write to tell me that I have ignored the variation in drag applied to the fuselage and one or two other minuscule contributions. I know I have made some simplifications but they do not alter any of the truth of the foregoing explanation).
Return to the Technical page