The
following paper was one of seven published in the 1993 BMFA Free-Flight
Forum Report.
SOME THOUGHTS ON
F1A
Mike Fantham
I thought it might be useful to review ways of improving model performance; my thoughts are applicable to the class I fly, F1A, but there is the usual scope for some cross-pollination.
1. Obvious areas of endeavour
Since a glider spends most of its flight time gliding down and duration is the aim, we must never forget the basic governing equation. Leaving thermal assistance aside for a moment, the duration is given by-.-
DURATION = LAUNCH HEIGHT ÷ SINK RATE
These are the only two basic variables and they give us two areas for exploration if the duration is to improve:-
Increase launch height OR Reduce descent rate.
Increased launch height, with a fixed line length, requires that excess kinetic energy at release is converted to potential energy before the stable glide starts - more of this later.
In 'still air', a reduced descent rate will be produced if we reduce the surface loading or improve the aerodynamics.
Surface loading is effectively fixed for an F1A but not for some other classes. Aerodynamic improvements will come from increasing aspect ratio to reduce induced drag, cleaning up the airframe to reduce other drag effects or improving aerofoils to get a better overall aircraft ‘duration factor’ (C L 1.5 /C D ). I'll leave you to try various sections and higher aspect ratios, but remember that long thin wings have lower Reynolds numbers and can give reduced aerodynamic efficiency. Their other problems are structural - you get increased root bending per area at the same time as a reduced wing thickness to resist the loading.
Don’t forget turbulation - this can cause great improvements in model performance both in sink rate and stall recovery.
The other way to reduce sink rate, or even cause it to be negative, is to put the model in a thermal! You need to make it stay there too, so warps and fin size must be balanced. My experience is that too much fin on a right turning model will make it turn too tightly in lift. If the fin is too small, the model will not be sensitive to thermals and will not centre. Warps also help a model to be sensitive to lift - the model will tend to turn towards the wash-in in an up gust. Warps are tricky - they depend on your wing section - but I think enough warp to give twice as much right rudder offset for glide as you have left rudder on tow is about right. (These deflections are for a model that turns right!). In the straight tow days I used an amount of warp which tended to give equal offsets - left for tow and right for glide.
My new model will have a smooth twist across the whole span with superimposed equal wash-out in each tip. The twist is about 0.72 degrees across the span with one degree wash-out in each tip. This has been calculated to compensate for the slower wing on the inside of the turn. Interestingly, the twist looks equivalent to the value discovered by trial and error over the years.
2. Other areas of endeavour
In most circumstances good contest results depend heavily on reliability and general preparedness and not just on the absolute performance of the models.
Aspects which deserve your attention are:-
The details of the models
Systems - if it don't work 'til you poke it - fix it
Consistent assembly - mark the bits - read the marks
Durability - good design - weather resistance - wet or hot
The supporting equipment - bike - bins - Biotrack
Yourself - keep in shapeI have written a paper covering these ‘peripheral' matters before- see the 1985 Free-Flight Experts’ Forum, 'Ways of Winning Glider Contests'.
3. lncreased launch height
Given a fixed towline length and a ban on energy storage or other power the only way to increase launch height is by ensuring a supply of kinetic energy at the moment of release - ie speed.
3.1 Potential Energy
Height gives potential energy, PE, the glider's 'power'. PE = Mass x gravity x height
This is limited by the release height to something less than about:-
0.41 x 9.81 x 52 Kg m²/s² = 209 Kg m²/s²
In practice, we release at more like 47.5 m, giving 191 Kg m²/s²
(See the Kochkarev/Makarov paper 10/91 Free Flight News or 1991 NFFS SYMPO)
3.2 Kinetic Energy
Speed gives kinetic energy - KE - this can also be provided during the launch.
KE = ½ x Mass x (Velocity)²
To get more kinetic energy, we can increase speed or mass but the latter detracts from the glide!
For a 26 m/s launch (58 mph! -see Kochkarev/Makarov again) and a 4.75 m/s glide, this gives:-
KE at launch ½ x 0.41 x 26² = 138.6 Kg m²/s²
KE in glide ½. x 0.41 x 4.75² = 4.6 Kg m²/s²
KE convertible to height 138.6 - 4.6 = 134 Kg m²/s²Note that this compares with 191 Kg m²/s² PE, giving a potential height gain without losses of:- 134/19l x 47.5 m = 33.3 m, a total height of 47.3 + 33.3 = 80.8 m !!
Unfortunately this is limited by aerodynamic and other losses to about 15 m gain, (Kochkarev/Makarov)
PE contribution + KE contribution = 47.5 + 15 = 62.5 rn
3.3 Getting the best from your KE
To ensure the maximum height gain we need to optimise KE to PE conversion. ']'his means a well trimmed zoom or bunt.
'I'o bunt or not to bunt - that is the question. Whether ’tis better in the calm ...
I am not convinced that a bunt is the way to go for 'normal' conditions. If a bunt flyer wants to launch a model into the air that he or she has just been sampling on circle tow, a carefully judged manoeuvre is required to avoid launching out of the front of the thermal. The zoom model tends to end up gliding in the same patch of air as the tow circles making for an easy accurate launch. However, to get a good consistent zoom, you have to have the right warps, dihedral and fin area balance.
My current feeling is to reserve bunting for big models (2.2 m +) and 'still air' flying. The bunt should be an easier way of trimming models with a large glide circle and I suspect that there may be a very slight advantage if ultimate performance. is required - as in 'still air' fly-offs. Of course, there is a risk involved in bunting in that there are more mechanical (or electro-mechanical) things to go wrong - but on the positive side, the technique is relatively tolerant of variations in launch speed and release angle (see Kochkarev/ Makarov).
There is a compromise which will give the ease of trimming of a bunt with the accuracy of a zoom. This is a combined zoom and bunt launch with only a soft bunt input to kill any tendency to stall that a more open zoom might normally carry.
3.4 Structural considerations
To accelerate a model on the line you need a high tension. You cannot tow a flexible model at high speed - either the wing will wash out reducing the line tension or the wing will wash in causing excess drag and reducing the speed attainable. Either way you could also get flutter either during or after release. High line tensions cause high loads on the wing as well - fortunately stiff wings are usually strong too.
Please don't believe that ‘normal’ models are not strong enough to be bunt launched. The fact is that most 'normal' models would gain little from a bunt because they cannot generate much line tension and consequently don't accelerate much when you pull them. You are pulling as hard as you can now aren't you? - if not you are not getting the best from your model. If you make a stiffer model, it will tow faster for the same running speed and you should get better launches.
4.0 Still Air Flying - a Still Air League?
‘So what' I hear the sceptics say, the weather is so rarely good enough in the UK to need a really good 'still air' time - why bother?
With notable exceptions, the average model flown in the UK is a thermal-riding machine. A good model of this type will perform well in the vast majority of domestic contests and will be happy in windy conditions. There is not much incentive to develop that special 'dead air' model that may be necessary to win the occasional calm air contest. Many flyers say that you need one set of models to qualify for the team and a different set to win the World or European Championships. I'm not sure if that is literally true, but they have a point. So how do we encourage the development of better 'dead air' performance in the UK? It seems to me that the first step is to find out where we are now by measuring what we've got.
One way of doing this is by establishing a league table of UK model performance in 'still air'. The aim of the league will be to promote the development of models with a higher ‘still air' performance and to get people out flying early in the morning! I realise that there will almost always be some vertical movement in the air, but the aim here is to minimise the effects of these variations by making a series of flights ‘at random’ in fairly rapid succession.
The table entries will show the flyer’s name and the specific model used to set the time.
To establish the 'still air' time for a model: -
- a series of flights must be made within a given time period
- each flight must glide down to the ground
- the longest flight must not be more than say 125% of the shortest flight
As an example, here is a series of flights that I made:-
Mike Fantham, model 14, 18/5/89, 06:40 - 07:15, Croydon Airport, sunny, very calm
221, 241, 196, 211, 208, 221, 211
Sorted times:-
241 221 221 211 211 208 196
Average...... 215.6 sec.
Max/min = 1.23We will need to refine the precise ‘rules’ on the basis of experience, but as an initial suggestion I think we should use 5 flights in 30 minutes for F1A. The concept can apply to other classes too:-
Class F1B F1C F1G F1H F1J
Period (mins) for 5 flights 45 60 35 25 45Class OR OG OP ETC...
Period (mins) for 5 flights 90 30 60These periods are offered for guidance, I don't want to exclude sets of flights that don't quite meet the criteria if they are otherwise 'valid'.
I would welcome thoughts and suggestions around the topic of measuring (and improving) still air performance. Peter King has already suggested that, once the glide time is established for a model, time to descend on D/T could be used as a means of comparing climb heights. Whilst this may not be returning a true measure of overall flight time, it certainly is a means of comparing different props, motors, climb trims etc. without the problems of a full length glide.
Clearly the opportunities for recording ‘still air times' in the way suggested will be quite rare, especially in the UK and with the classes that have higher durations. The air will vary from day to day and it would be more reliable to compare models directly in the same air. This will take some organising. You could fly two models together, using the same reference model at each session - Tony Young used this technique to remove any day-to-day air variation effects when he was assessing performance. Ideally different flyers could compare their models in the 'same air' - a real comparison.
I realise that there are practical difficulties, but if you record a series of consecutive dead air times under suitable circumstances, even if they don't exactly meet the criteria above, send them to me giving the following details:
Name, address, specific model identification, class.
Date, first launch and last landing times, weather details, location.
Flight times in the same order as the flights.
Any comments, e.g. 'third flight had poor bunt'.Send your information to Mike Fantham.
Either post to the BMFA Office, or e-mail admin@bmfa.org
I will collate the inputs and send the data to Free Flight News from time to time.
back to: BMFA Free Flight Forum Reports | Electronic Timers
© British Model Flying Association.
Updated: 22 May, 1998
Free Flight Forum pages created and maintained by: Model net.
sales@modelnet.co.uk