Helicopter aerodynamics

In future issues of W3MH I will be concentrating on aspects of helicopter aerodynamics and I would very much appreciate your suggestions for topics of interest. I guess that, unlike many of the other contributors to this magazine, I am one of the back-room guys who is not too well known so for this first issue and at the risk of boring you all I thought I would spend some time explaining my background, how I got involved in model helicopters and what it is about them that fires my interest. As for my background, I was until recently a lecturer in a university physics department. My research interest was in atmospheric physics and teaching mainly electronics and computing. I am now running my own design company, CSM, whose main product at the moment is the Northern Helicopter Products (NHP) heli/aero simulator, so if you are looking for an unbiased view of flight simulators (Ed: or latterly, Piezo Gyros...) don't ask me.

Colin did much of his early learning on an MFA Sport 500. The crossed sticks for learning have long since been discarded... (ClickPic for larger view)

My interest in radio control dates back to the late '50's. I was about 5 years old when my father built an R/C model boat (a Veron Marlin for those who remember them). This he fitted with ED valve (thermionic) single channel gear modified to give a single proportional channel by variable mark-space ratio modulation. The modulator was, believe it or not, mechanical! The mark/space ratio may have hovered around 50/50 but the work/fail ratio of this gear was closer to 1/99. The unreliability killed my father's interest in R/C but I was totally hooked. Given the performance and reliability of early R/C gear I suspect anyone suggesting radio control for model helicopters at this time would have been instantly locked up. For me the progress towards better R/C gear was not that fast, being a mixture of home build and second-hand. My first R/C gear of about '65 vintage still used a valve at the transmitter end. Come to think of it, my first multi-channel set, a 10 channel reed outfit, still used a valve in the transmitter output stage. I never rated the reliability of these early R/C sets and confined it to boats while flying control-liners and free-flight.

Involvement in models came to a halt during my undergraduate years but was revived when, as a post-grad. student, I found myself working with Ian Stromberg, a keen R/C flier. I still have the slope soarer (a Soarcerer) that I built then, complete with its home built Micron 27MHz radio. The servo amplifiers in this were a work of art and used discrete transistors which you had to file down a bit to get in the available space. Helicopters didn't even figure in my thinking at all until around 1987 when by chance I read an article in Radio Control Models and Electronics on the cyclic controls of helicopters. I had, up to that stage, what seems to be the typical fixed-wing fliers attitude towards helicopters - they're expensive, all you can do with them is hover, they're as interesting as watching paint dry, and I don't want anything to do with them. My attitude didn't change over night but I did find myself reading the helicopter articles in the model magazines rather than skipping them. The final spur to venture into helicopters came in 1991 when the University turned down my promotion. I felt I really would have to work much harder - on my hobbies! Within days I had an MFA Sport 500 collective, a set of Futaba Challenger Heli radio, a stack of Model Helicopter World back issues, and some R/C helicopter books. Luckily, Ian Stromberg decided to go along with my insanity and bought an identical set of gear so we could learn together. The reading matter made us realise just what a challenge we had let ourselves in for. The more we read the more daunting the flying seemed to be. The consensus seemed to be that it would cost a fortune in broken bits before we could even hover.

A lot of us learnt to hover in the back garden, there's not far to travel to repair it! (ClickPic for larger view)

A simulator seemed like a good idea. I had a 486 PC but, at the time, the only simulator available in the UK was not available for the PC. Buying a simulator and a machine to run it on was outside my budget (remember, I hadn't got the promotion!) In 1986 I had used a Sinclair QL to numerically model propellers for optimising electric flight, but the speed at which this program ran (walked?) made me wonder if there was a realistic chance of a proper numerical model of a helicopter running in real time even on a 486. Some quick 'noddy' programs showed just how much processors had progressed from the 68008 of the QL. It looked as if the sort of sums needed could indeed be done with enough time left over for the graphics, and so the idea of writing my own model helicopter simulator was born. Although I had a fair understanding of fixed wing aerodynamics, all I knew of helicopters was what I could guess from the propeller theory used earlier. I got a copy of Rotary-Wing Aerodynamics by Stepniewski and Keys (ISBN 0 486 64647 5) to bone up on the subject. Now the interest became an obsession. The more I read the more fascinating the whole subject became. A couple of months of frantic reading and programming followed and resulted in my first and very user aggressive simulator. Definitely for private use only! I used this for about 40 hours before daring to try the real thing. Having set up the helicopter on a test stand it took me one fairly unproductive flying session to get used to the different feel of a real transmitter after using the simulator with a couple of IBM joysticks, but on the second session I got the thing into a reasonable tail-in hover and held it till the tank was almost out. I guess anyone who has learned to fly a heli will know the sense of achievement at that moment. There are just so many manoeuvres a heli can do that you can keep on getting a buzz from doing something new with no chance of ever running out of challenges.

(ClickPic for larger view)

Just in case you are thinking of trying R/C helicopters here are my list of tips for beginners, though I bet every W3MH contributor will give you a different one.

• Resign yourself to the expense. If your starting with no existing gear at all you need £700 minimum to get up and running. Crashes average £70 a time. Its cheaper to learn Indian club juggling with (full) Champagne bottles.
• Read all you can. I found 'Learning to fly Radio Control Helicopters' (ISBN 1-85486-025-9) and 'Setting up Radio Control Helicopters' (ISBN 0-85242-975-4) both by Dave Day to be very helpful.
• Learn with a friend. Having someone at the same level as yourself to bounce ideas off is really great. Experts can be very depressing.
• Use a test stand. This allows you to set up the helicopter (especially the engine). You can't go through life getting the model shop/ local expert to set up your heli every time you bend it so you might as well start doing the job from the start. My test stand is a 'WorkMate' and some rope.
• Get a training undercarriage. The usual 'crossed sticks' type (shown in photo fitted to my Sport 500) will allow you to get away with murder. Shorten sticks progressively to wean yourself off the training gear - 'cold turkey' removal is too much of a shock.
• Get a collective pitch helicopter. Although they are cheaper, fixed-pitch machines are much harder to fly than collective pitch ones. Some of the fixed-pitch machines are great fun but I have seen a world-class flier have real difficulty doing a circuit with one.
• Get a gyro. Although it is possible to fly without a tail rotor gyro, it is a lot harder. The pioneers who learned to fly fixed pitch and gyroless have my full admiration.

Simulators

Now, given my financial interest, you wouldn't expect me to miss simulators off this list. Just a few hours on a sim can save a lot of time on the flying field. If you can, borrow a friend's. That way you get to drink their coffee and burn their electricity. For me, improving my understanding of why helicopters behave the way they do is almost as important as improving my flying. By way of a short introduction to heli dynamics for those who have never given much thought to how helis fly, I have shown in Figure 1 just some of the forces acting on a helicopter in the hover.

Next time I'll look at the way the main rotor generates lift and the power needed to drive it.

Part 2 (Originally published November 1995)

I know last time I promised to talk about how the main rotor generates lift but our editor has had a number of requests for articles aimed at the beginner. He suggested that I do 'something' about setting up linkages so here it is.

OK, do you need to read this? Try this simple test.

1. If your answer was 130 degrees you might want to go and read Bruce Naylor's article instead.

2. If your answer was 'in the bottom right hand corner' you may feel more at home with the Baywatch home page.

3. If your answer was anything else and you are a heli beginner (and especially if you haven't a computer radio) read on!

First let's look at the geometry of the linkages themselves. I know this sounds boring but a few minutes spent understanding this can save a lot of wasted time in setting up a helicopter. This is because the linkage geometry usually introduces non-linearities in the controls, some of which we can use and some we can well do without.

In this case the angular deflection of the control accurately matches that of the servo however, any other arrangement introduces non-linearity into the control movement.

In figure 2 we have a control arm that is longer than the servo arm. If course, the angular movement of the control is now less than that of the servo but the sensitivity of the control movement falls off near the extremes of movement.

 

Conversely, a control arm shorter than the servo arm moves through a bigger angle than the servo with an increasing sensitivity towards the ends of the servo travel (see figure 3).

Even if we stick to having the servo and control arms the same length we can get non-linear control movements. Here in figure 4 the shorter-than-ideal push-rod causes a bigger movement of the control arm when it is being pulled towards the servo than when it is being pushed away. The reverse is true with an over-length pushrod. Now we could go through all the combinations of long and short control arms with long and short push-rods but starting from these diagrams I guess you can probably see what these combinations are going to do.

Now, if you look at the linkages on a real model helicopter things look much worse than these simplified diagrams. For a start, the servo and control arms may well rotate in different planes making it harder to see what's going on. Usually the control gets passed through several linkages before it gets where its going.

As a first rule, with the servo in the middle of its travel get the push-rods at right angles to the arms like this

	Cyclic control linkages. The roll servo is mounted vertically at bottom left and fore-aft cyclic servo is mounted horizontally (picture centre). Note the pushrods are at right-angles to servo arms and bellcranks. (all pictures hotlinked)
Sometimes compromises in the linkage geometry have to be made. Ideally this bellcrank should be right-angled.
	Here the bellcrank angle has been dictated by the need to clear the sideframe. Unfortunately, to get the pushrod to the servo at the right angle to this bellcrank it needs to be attached directly to the output shaft without any arm at all. This give perfect linearity but leaves the pitch range woefully short.

There's no excuse for not getting the servo arm pointing in the right direction. The servo manufacturers have provided a really fine adjustment. Taking Futaba as an example, they use a 25 tooth spline on the output shaft of their servos. Lets assume we are using a six arm star on the servo. If you move the star round four points on the spline you move it 4/25 of a turn or 57.6 degrees. Since the arms are 60 degrees apart the effective movement of the star is just 60-57.4 or 2.4 degrees. This is only about 3% of the total servo movement.

In selecting the servo arm length the required control throw is of course the primary factor. However, where you have a choice, use the longest arm length you can. This means that the forces on the push-rods, ball links, servo output shaft bearings, and servo mounts are all minimised. By using long arms you also minimise the effect of any slop in the bearings or links and any flexing of the helicopter frame or servo mounts. So even on the throttle linkage use the outermost hole on the throttle arm. The collective pitch control probably generates the highest servo loads and so long arms are most important in this linkage. The longer the arms the softer the servo mountings can be and the less grief the servos will have to put up with. However, the effect of backlash in the servo gears is not changed by arm length.

How you go about setting up your linkages will depend quite a bit on the facilities you have on your radio. Computer sets certainly have facilities that make life easier but a computer set should not be seen as a substitute for making a decent job of setting up the linkages though there are quite a few flyers I've seen who seem to treat them this way. I think it is far better to get your linkages set up well first, with the transmitter at its default settings, and then use its facilities to fine tune things.

If you have a good computer set then you have such facilities as 'travel volume' by which the throws (either side of centre) can be independently adjusted for each servo. This allows you to fine-tune the total travel of each control without fiddling about with servo arm lengths. Because it can be separately set for each direction of throw it can also be used to compensate for unequal travel such as seen in figure 4. 'Exponential' is a facility which allows the sensitivity of the controls to vary between the mid stick position and full travel. This allows controls to be made fairly insensitive to stick movements near the middle while retaining full control authority at full stick. 'Exponential' can also be used to overcome the sort of non-linearity in linkage seen in figure 2.

Next time I'll (hopefully) look at setting up the collective pitch range.

Colin Mill

(Originally published October/November 1995)

Copyright Colin Mill and Lance Electronic Publishing 1997