Monday 1 January 2001

How to set-up a racing glider

Even pilots who have been racing for years will take many sessions in differing lift conditions to set a model up to their satisfaction.
Although a good distributor will supply set-up suggestions with their models most of the measurements will be subject to personal taste. It is also worthy of note that although two moulded designs should behave identically this is not always the case. In practice a sensitive pilot (one who is not afraid to cry?) may detect minor differences in lay-up (changing torsional rigidity) and weight distribution. Any differences should be small and only recognisable at extremes but serve to prove the point that there is no substitute for taking the time to set each model up on its own merits to your style of flying.
Also it's fair to say that your own stele is likely to develop, so the way you set up a model 12 months ago may not be the way you'd set it up if you started again.
There are two vital mixes that can be optimised and are not necessarily solely personal taste. Snapflap and differential. These will be looked at in detail shortly. First though a few other thoughts.
For racing do not use aileron/rudder mixing. As a matter of principle I do not believe this coupling should be used by anyone with two thumbs but for racing there is a legitimate reason not to use it. If you have coupled aileron and rudder you are only achieving what correct differential should be doing in the turns. The downside to the coupling is that if you have good speed on the straights and need to make an aileron adjustment you will also be putting rudder in, which can start an oscillation that will scrub off speed. The effect will be minor but they all add up. It would be possible to have a very fine coupling that would do little damage but why bother when you can simply use the correct differential in the first place?

What we are talking about here is programming your TX so that the ailerons are making the plane track correctly around a turn without the fuselage pointing high or low. When done correctly significant drag reductions can be achieved both in the turn and in reducing any subsequent control inputs required.
Flying different styles - reversals, energy management, bank and yank - may not need significant differences in differential but subtle changes can make the pilot feel more at home.
Set the model up with the aileron response that suits you. Once done perform a crosswind roll without using any rudder input. Your airspeed should be reasonable but not so fast as to make anything axial. When I'm doing this I leave the differential menu open on my TX so that I can adjust the settings without landing. Multiplex TX's are a doddle in this respect as you can use the digi-adjuster. You may have to be more safety conscious with other radios i.e. don't blame me!
Once you have adjusted your differential to achieve the perfect axial roll at all but the slowest of speeds you have a place to start from. I usually save these settings for the TX's aerobatics mode. Switch your TX to racing mode and dial in maybe 10% more differential than was used to attain the axial roll. On a good day with the wind square on to the hill start flying an imaginary F3F course. After you pass the pilots position and are around 15 metres from the base, lift the nose slightly and roll 3/4 to inverted watching the line of the fuselage the whole time. If the fuselage seems to sit nose high and tail low during the turn increase the differential. If the glider seems to be thrown more towards the hill with the tail high and nose pointing into the hill try easing off on the differential a little.
The following two photos are designed to illustrate the general point. Even if a plane has what I would consider less than optimum differential settings, it is still relatively simple to adjust your flying style to suite the differential. What I always aim for though is to make it simple for myself so that I can fly the plane how I want to not how the set-up dictates I need to fly it.

This Cobra is the best example I could find of not enough differential. The nose is comparatively high and the tail correspondingly low. Of course there could have been many other factors that left the plane on this trajectory but I'm trying to illustrate the point.

This Sting is in a much better position to complete the run without the need for unwanted control inputs. The nose is low and the tail is high.

The Cobra above could have achieved the same result with more differential.
After a session in good lift you should have the plane set up with an aerobatic mode of perfect axial rolls and a racing mode where the fuselage tracks around the turn totally in line with the direction that the plane is travelling in.
I also use a slightly different setting for lighter conditions using a little more differential.
The speed of aileron response is largely a matter of personal taste although it is worth bearing in mind that in turbulent air you may need enough response to level the wings rapidly.
With regard to the all important turns you need to strike a balance between having too little movement making you start the turn early thereby missing out on the best lift and having so much that you can turn very late but are creating a lot of drag and end up twitching about the sky. Also bear in find that shaky thumbs can get to anyone, so less can be more.

Snapflap is the mixing of the flaps with the elevator. Normally the mixing will involve up elevator producing an amount of down flap and vice versa. I say normally because if certain flight characterises are required (such as damped pitch response) there is nothing to stop you experimenting with the direct opposite of what I will be discussing.
From the mixing we hope to gain tighter, faster turns and better inverted performance. In the case of 4 servo wings, whether or not the ailerons will be mixed 100% with the flaps (i.e. the whole trailing edge droops equally) is down to the model. My experiences have been that the more cambered the section the less outboard aileron you want acting as flap. The reason is that to have the outboard surfaces moving fractionally less than the inboard ones mean that you are less likely to induce a tip stall and associated flick.
Setting up planes with the optimum amount of snapflap is simply a case of evaluating a number of different settings, although it is complicated slightly as conditions and ballast can also have an effect. As a general rule the more flap you have mixed with the elevator the tighter the model will turn, up to the point that you are adding too much drag. This becomes apparent when the plane will still turn in a small radius but will scrub off lots of speed. You are looking for the compromise between the tightness of the turn and the speed retention throughout the turn. A few years ago I started using the phrase 'grip' which people seem to empathise with. A good set up will give you good grip, like an F1 car glued to the road. Too much snap flap and you'll be like a rally car drifting around the corners.
The factors governing this are numerous. Wing section, wing loading, moment arms, tail area/thickness/movement, wind direction, airspeed, flap areas, flying style, aspect ratio, CG etc. So go  and have a play.
As a cheat you may want to start at 1.5 mm down flap with full up elevator on an RG15 60" model and around 3 mm on a 3 m RG15 model.
The up flap mix with down elevator is far less critical. Generally for aerobatics you are only looking at a few mm to de-camber the section and for racing many people, myself included, use none at all.
It is very tempting to use computer TXs to show that you can waggle everything but it is worth bearing in mind that it is probably better to have too little elevator/flap mix than too much.
One final thought is don't be afraid to experiment. The aileron side of things is pretty straightforward the real potential gains are when you pull in the up elevator to turn the plane so watch, think and tinker.
Some areas worth experimenting with on the set-up front, try putting your snapflap on a curve so that all the flap comes in at, say, 1/2 the available up elevator movement. Also play with elevator curves, if you like a really response elevator but that can make a model a bit flicky use negative exponential i.e. front load the elevator movement around neutral but have it fall off at the extremes in case you get a bit excited.
Try different set-ups for different lift conditions. Consider more snapflap for liftier days when ballast is on board and less when the lift is marginal and/or the wind is not square to the hill.
One of the things I enjoy about slope racing is the analytical side of getting the best from a plane in varied conditions and setting a model up to do this is a skill, the same as flying the model in different conditions is. Like all skills of any merit it takes time, experience and sometimes help to feel confident in what you are doing.

By the way you'll have noticed I've steered clear of the CG debate, go fly and find it yourself!

Mounting wing servos

As moulded models are becoming more and more commonplace so are the associated questions they pose. One of the most frequently heard is "what is the best way to mount the wing servos?" The technique used by myself and many others is very simple but a little explanation of the way it evolved is probably warranted.

Since servos first shrank enough to fit in wings builders have experimented with different installation techniques. There are many and varied methods used today, some naturally being better than others.

One factor affecting the way your servos are mounted is the servo gear material used. The best plastic gears in the world when put in our less than friendly wing environment are liable to resent the abuse once in a while. On the ground this can be a desirable weak link and save the rest of the airframe from further damage.

For this reason I would advocate the use of screws and blocks in a manner appropriate to the make of servo you are using. In using plastic geared servos you are accepting that the gears are probably going to require replacing at some point, whether in sport flying or in competition the easier you can make it for yourself the better. However, as anyone who has ever seen a wing servo come adrift in flight will tell you, although ease of access is nice, security is most definitely paramount.

Metal geared servos pose different problems. If you have purchased wisely the likelihood is that there will be no requirement to deliberately remove them from their mountings, as they should outlive the airframe. Of course life isn’t as simple as that and problems can arise both mechanically and through some ham fisted stick twiddler not tucking their crows away on cue (if anyone tells you they’ve never done it don’t believe them). This means that an amount of serviceability has to be built into the system, but hopefully not as much as with plastic gears.

I am full of admiration for builders who craft clever, well-made screw mounts, however I wouldn’t recommend it for metal-geared servos. When your protruding flap gets a whack something will have to give, if it isn’t going to be your gears the chances are it’ll be that clever mounting system you laboured over for many an evening. Someone will probably have come up with a clever system with a weak link but is it going to be worthwhile timewise when compared with the simple method that follows?

Firstly, yet often overlooked, is the need in most cases to strengthen the wing skin to which the servos will be attached. This means that if a servo is knocked off it won’t take your lovely wing skin with it, secondly it will reduce any skin flexing that can be a cause of flutter. A square of medium weight glass, carbon or Kevlar cloth as large as you can get to adhere properly applied with wing skinning epoxy to the area where the servo will be glued will do the trick nicely. Whilst on the subject it is often prudent to spend a few minutes shaping vertical grained balsa ribs to bridge the upper and lower skins of a moulded wing thus spreading the mounting loads further. These can be worked in through the servo access holes prior to the servo fitting, but after the aforementioned skin reinforcement, and secured in place with some five minute epoxy or cyano. Some would regard this skin bridging as overkill and they might be right but for the sake of half an hour’s work why take the chance?

What you have now is a strong and sturdy base to which the servo can be glued. If you feel the need to protect your servo cases simply wrap a few layers of tape around it. Masking tape works well but be sure to sand the gluing surface to remove any release agent. Personally I don't use any tape.

Mix up a paste of top quality five-minute epoxy and micro balloons and apply to the appropriate surface of the servo. What we want is for the excess glue to be pushed away from the servo arm and creep out at the other three sides. This is simply achieved by holding the servo by the servo arm and applying that end to the wing skin first, as the rest of the servo is pushed down the excess glue will be pushed out away from the servo arm. A wise precaution is to have a few cotton wool buds on standby to get any unwanted glue away from the moving parts, you shouldn’t need them but if you fail to plan, you plan to fail!

What you are left with is firmly mounted servo that won’t let go unless really clouted. When it does let go all that will break is the glue joint leaving you with a pre-formed epoxy/microballoons "cup" to receive the servo again with a light smear of epoxy and no loss of trim.

Of course when mounting servos in this or any other manner you wouldn’t forget the old golden rules of double checking your linkage and servo arm geometry, would you? Another cause of premature swearing is forgetting to centre your trims, sub trims, brake settings and other gadgetry before screwing the servo arm in place and securing the servo. Most of us have done this at least once and experienced the less than blissful spectacle of plugging the servo into the RX and watching in horror as the servo arm despatches itself to centre in a position that you hadn’t catered for. I have seen someone fiddling with a Christmas cracker screwdriver who’s head had been bent at ninety degrees to adjust the servo arm in a wing mounted servo. Whilst it didn’t look the most elegant of procedures and seemed to require a lot of verbal encouragement it did eventually work and was better than ripping the servo out, worth bearing in mind in an emergency.

I know a few people who use silicon sealant to mount servos and there can be no doubt that it does work but be warned. If you need to replace gears or the servo gets knocked off nothing is going to stick it back again except more silicon sealant. Not practical for a slopeside job.


Update... below is very true and very worth a read but... the latest range of magnetic switches are just fantastic and well worth a look.

Your average RC switch, used them for years, occasionally they get a bit intermittent, sometimes black wire corrosion gets right in there.

When my first Viking had rather a fast touchdown (immense crash) the switch went a bit intermittent, so I took it apart...
Even before getting a deep down to look at the state of the contacts, have a gander at the standard of the soldering, and look at the pinching where the wire entered the case.

I've now replaced three switches and the soldering on this one was actually the neatest, scarey...
This is the Maplins switch I've started replacing my switches with.
Don't know what the Amp rating is at the voltages we use, but in practice I've had no problems drawing quite a few Amps through it.
Mmmm, lots of lovely contacts...
Wire across the two sets of contacts for each position. Thread the tinned wire through the holes so that you get some mechanical strength as well as the soldering.

Make sure your battery input connection is the centre one, which means that in one position the left lead will be live and in the other position the right lead will be live (one to the Rx and one for charging).
Tin, twist and solder the negative leads together and slip a bit of heat shrink over them.
Once done, pot the lot in epoxy for strength, insulation and durability.

This method works for me. If you have a better method let me know and I'll share it here.

Boring legal bit because we live in a bizarre society where cack-handed, knuckle draggers can sue you because of their own ineptitude... Anyway, I can in no way be held responsible if you try this and it goes horribly wrong and you damage anyone, anything or yourself.