Backing It Both Ways
by Trevor Hewson
First published as an April Fool spoof in the March 1991 edition of Sloping Off - the Newsletter of Christchurch and District MFC
Ever since the Black Diamond confounded the sceptics by actually flying, I've been looking for a new experimental project. Every time I thought of something, it seemed I picked up a magazine and found that someone, usually the Germans, had already done it. So if it was to be a truly original project, it was going to have to be pretty mad! As you may know 1 have been playing around with canards for the last year and the "Hey, look, it's flying backwards" comments suddenly got me thinking would it be possible to build a model that really would fly backwards - as well as forwards! Before you send for the little green van, perhaps I had better clarify that Just a little bit. Clearly, the idea that you could literally launch the model in either direction is not on - if this could be achieved, you could change direction in flight - and think what new aerobatic manouevres that would open up!
So how does the model decide which way to fly? The answer of course is the position of its cg relative to its centre of lift. If we leave the flying surfaces alone we will therefore, at the very least, have to move the cg in order to switch between conventional and canard flights. The next problem I faced was that of yaw stability or, to put it in basic terms, which end does the fin go! Now, Iain Rose has on occasion tried to persuade me that, with enough sweep back on the wing, one can dispense with the fin altogether. Whilst this seems attractive at first, there is a fundamental problem - what does sweepback mean on a reversible model!
So we now have two concessions to common sense - we move the cg and move the fin to the other end when we want to fly the other way. The next problem concerns the tailplane/foreplane, or stabiliser. Clearly this was going to be a flat plate, but how can I fit the elevator on the trailing edge for both directions of travel? I considered the simple option of turning the stabiliser round when it changed from a tailplane to foreplane but this seemed against the spirit of the project and in any case creates linkage problems because of the change in hinge line position.
The answer of course was to use an all moving stabiliser. However, it was some time before I realised that even this arrangement presents some difficulty. Clearly the only sensible place for the pivot is at mid chord, so whichever way we are flying, the aerodynamic forces will be acting in front of the pivot point - a potentially dangerous situation. I decided that I would just have to live with this risk, and make sure the linkage was short and slop-free. This determined the position of the elevator servo.
Now I came to the real problem, the wing. Whilst a flat plate will do for the stabiliser and might even suffice for a 'brute force' power model, for a glider, I had to do better. Clearly the first question to be addressed is why are trailing and leading edges so different - and do they really have to be different? Trailing edges are usually sharp so that the air flowing over the top and bottom surfaces of the wing can re-combine as smoothly as possible. However we have all seen beginners' models with very poorly shaped trailing edges. and they almost always fly, albeit with some loss of efficiency. By the same token, the job of the leading edge is to separate the airflow and one might expect them to be sharp too. The reason that they are in fact rounded is to enable the airflow to be split cleanly over a range of angles of attack. If we were content to fly at one speed only, the leading edge could indeed be very sharp - but the stall characteristic of the wing would be lethal!
From these musings I decided that a working compromise was possible and my first attempt at a reversible aerofoil is shown in the sketch. You will see that I have opted for a cambered aerofoil to enable the wing to generate lift at zero geometric angle of attack. If a symmetric section were used the model would have had to fly in a very nose-up (or tail-down!) attitude.
At this point I really thought I had the problem licked - and then I thought about the ailerons! My first idea was to go the 'all-moving' route again and use a wing twist system as on Sailplanes International's Axle. However, the servo loads of such arrangements can be a problem and I would have the same central pivot problem as with the stabiliser which would be sure to overload the servos. After briefly contemplating fitting ailerons to the trailing and leading edges I took the easy way out - and went for rudder/elevator control! This of course means that the rudder servo, mounted amidships, must drive snakes in both directions, in much the same way as operating a steerable nosewheel in a power model.
At last I had a complete design and construction could get underway. I kept the building simple, using foam wings and a simple balsa box fuselage. Some care was needed to facilitate the relocatable fin and cg (ie nicad!), but apart from this, it has all been remarkably straightforward so far. I don't usually build to deadlines, but for this model the date of the maiden flight had been laid down in advance - and I am just about on schedule to finish it by the end of the month - so watch this space!
P.S. To read what this article eventually led to, have a look at Each Way Bet