CO8-2m Flying Wing 

by Glyn Fonteneau and Dave Camp

Introduction:

An introduction to set the scene - we have both had a long term interest in flying wings and have planned to do a design for flat field use for several years now. However there have always been other projects on the go, plus a degree of uncertainty over sections, planform etc. A lot of text book work had narrowed down the various parameters, but we still had a few unknowns which held us back from committing to time and materials. A few years ago Dave bought a Multiplex ‘Cortina’ (3.6m wing) and although rather dated (a discontinued line by the time he bought it) it did give us an idea of what a ‘proper’ wing could do on the flat.

The trigger for the design described here was some information picked up on the Internet. We found the web site of Andy MacDonald, (web site no longer active unfortunately) which was largely devoted to wings, with a lot of information, in particular on the work of Hans-Jurgen Unverferth in Germany. The description of his designs (in particular the CO series) showed that there was real potential, but initially we held back, this was still going to be a big jump. However Hans-Jurgen had designed a hand launch wing (CO7hlg) which was similar in design to his larger designs. At this point Dave took the plunge and produced a version using an all balsa sheeted built up construction as opposed to the indicated foam/glass, in an attempt to keep weight to a minimum. The end result has a really good performance which we think surprised those who have seen it in action. Glyn had also made the Herk Stokely designed wing, which after some adjustments and trimming also showed good performance, but we didn't think the layout had as much potential as the CO type designs.

This was the turning point in pushing us to hold any other projects and to make a real effort for a serious wing.

Size was the one remaining query - for out and out performance something around the 3.3 m mark would suit us, but we were influenced by a desire to have something quick to build with lowish material outlay, and linked to this is our club 2m evening event we run in the summer. Neither of us had a decent 2m model, so this set the main parameter. So onto some details of the design, which we have named CO8-2m.

Glyn was the driving force here, but rather than reinvent the wheel the obvious thing was to take an existing design and adjust it to suit our main criteria - span 2m. The last design described by Hans Jurgen was his C08; 2 versions were produced by him - one at 2.6m, the other at 2.7m. These were F3B orientated designs and although we were going to be using our 2m designs in a mainly thermal role we both agreed that this would be the layout to work to. The decision was taken to take the 2.8m design and to reduce the span, but keep the chord the same. This would enable us to use the CompuFoil  foam cutting templates that Hans-Jurgen had made available. If you look at the attached drawing you may see that the design consists of 3 panels, the section remains constant across the span, but the wing twist is stepped. This was a bit of an unknown for us, but given the pedigree we had no intention of questioning it, after all textbooks indicate this is the preferred layout, even if it complicates construction.

Blue foam cores, as mentioned above templates taken from web site. However we have noticed a discrepancy between template twist values and the values quoted on the original CO8 drawing, the templates have around 1.5 degrees less twist and a slightly different distribution. Note there is 1degree anhedral built in. The wing section is RS004a (see R Sielemann's web site) 

We varied our construction details to help our learning process, e.g. a full depth 1/32 ply spar was used, but we varied the sweep and positioning of it. Glyn used steel joiners, Dave used carbon for the main one, steel for the secondary one. Note the wing is in 3 pieces, and for our convenience the joint is at 90 degrees to the chord line, rather than in line with the airflow. Having finished our designs we discovered that someone in America (John Dvorak) has made a standard CO8 using the same approach to this joint. Glyn corresponded with John to exchange experiences and construction ideas.

1.125" wide carbon uniweb spar caps (from Aerospace Composites products) are used, again we varied the lay-up between models; carbon uniweb was used for the trailing edge as well. The wings were vac bagged with epoxy/glass, with the cloth laid at 45degrees. Using painted Mylar's in the process allowed some experiments with paints and possible schemes for future projects.

Both gliders have Volz Wing- Star servos for the elevons and new JR 331's for the flaps. Winglets are a feature of these designs; we used the same sections as the wing, rather than the flat plates specified for the original CO8.

Fuselage Construction:

The fuselages were both made using the lost foam method. This was our first attempt at this technique. It was quite straightforward, although we had to use more glass than we originally expected in order to get sufficient rigidity. The fuselages are quite slim, roughly the size of a medium sized receiver, and we were somewhat in the dark as to the length and proportions of the fuselage. We thought nose weight would be required, so we ended up with around 5" projecting from the leading edge to carry the battery - you will read soon that this was wrong! We both used slip off nose cones, taking the construction method from R/C Soaring Digest March 1998 article. The fuselages weigh about 3 oz each empty.

Again these were vacuumed with epoxy/glass. The receiver aerial runs in a plastic snake projecting from the rear.

We made use of the Panknin twist spreadsheet to determine the CofG; there was slight sweep variation between the two, so the CG came out at 10.9 and 10.5 ins respectively. For safety we both set the CofG over 1 inch forward of the calculated position to start with.

Both wings have come at very similar weights of around 2 lbs.

Hand launches showed that as expected up elevon was required to trim out the forward CG, but glide angle was quite impressive.

Confidence was high from the hand launches, and because Glyn had already installed a tow hook a bungy launch was tried with his model. At this point perhaps it should be pointed out that the launch phase of a flying wing can be ‘interesting’ - it is possible for the model to enter a ‘propelloring/catherine wheel’ state, with recovery not necessarily being down to control by the pilot! Hence a bridle type towline is often used, but we wanted to keep it simple, and are only using a single standard fuselage hook. Certainly John Dvorak (the modeller of a CO8 in USA) had reported a lot of problems.

Well the launch was dead straight to the top of the line, at this stage Glyn made no attempt to stretch the bungy whilst on the line in the strong wind. Controls were positive without being vicious.

A few more attempts were made, no vices were seen, but it was obvious that a series of trimming sessions would be required.

The next day the wind was lighter but overcast sky and very little lift, both gliders used the bungy. Now trimming started in earnest, the CG being moved back progressively by taping lead to the

rear of the fuse. The performance increased each time.

Most flights lasted about 3.5 min - it doesn’t sound much, but a 3.5m floater was only getting around 5 minutes. We were both trying launch flap to increase launch height, which it did and were also drooping

the flap during thermal turns. It was noticeable that both would thermal hands off with drooped flap. Many flights took place, each getting progressively longer, best time being around 5 minutes.

When Glyn returned home he found that he had added 4.5 oz of lead to the rear of the fuse and the CG was now at 10.5 ins. Dave found something similar; my CG was at 10". We both re-balanced the wings by moving the batteries much further back together with the RX

Below a couple of pictures of the CO8-2m wings in flight:

Both went out late in the afternoon, wind now 5 mph or less, several moulded 3.5m F3J type gliders were slowly cruising the air and staying up very well. Dave had one launch that resulted in propelloring/catherine wheel - spectacular, particularly the fact that a recovery was made and the launch continued OK (the propelloring was at around 20 ft up!) The cause was thought to be too much flap on launch, plus a nose up throw. The launch phase is always going to be the tricky part of a high performance wing, winching would be better, but given the restrictions of the Working Mans League event we will stick with the bungy.

Further improvements in performance were seen, with the CG being moved back further on each model. We were surprised that they would stay up so well considering the speed they have to fly, even with thermal flap set they were travelling faster than the F3J machines.

Glyn had a 12-min flight and Dave a 20 min; we think this shook up the other fliers because Dave only came down to explore the aerobatics capabilities. Climbing (just!) inverted turns were possible in the weak lift. Coupled aileron/flap improves roll rate, very little flap req. (2 mm down) used. The main handicap is visibility - lack of it! A 2m wing with the small cross section of these designs is not easy to see - whilst circling away it is Ok, but when the time comes to straighten up and head back - oops - where did the model go! This was the only reason Glyn had a 'short' flight; both models were performing as well as each other.

Apart from the one incident of propelloring we have not experienced any of the problems John Dvorak had, like stall spinning, having to use a bridle or using vortex generators on the tips. Landing is however a problem at this time; certainly if attempting anything like a spot landing, we have both set up large down flap and down elevator (to combat the pitch up from flap) and are starting to get some good results. As John Dvorak found they land fast, hence the large down flap deflection required. A few weeks ago we had a mixed session on a Sunday morning. Glyn had moved his CG too far aft - launches were Ok, but the model would spin quite easily. As a result performance suffered. Dave managed to get into good lift and work it back; his problem was getting out of the lift before the model disappeared! It is noticeable that a lot of ground is covered - a bit like a moulded F3B machine - it’s no good trying to hang around over the field at low speed, you have to go and search for good air. It will be interesting to see how they work in the evening against the ‘Gentle Ladies’ and the like.

With the current CofG settings we have lost some of the hands off thermalling ability, but the roll rate is better - you pays your money… We accept the launch limitations now, certainly a crosswind element makes life interesting.

As these have gone so well we are considering either a proper Co8 or a bit larger say 3.3m or 3.5m perhaps with a bit more chord? Potential wing fences are being explored to assist in the launch phase; the flap settings are critical to getting a good launch. A larger design might well incorporate 3 moving surfaces on each wing panel, which is the ideal according to the textbooks, although set up would be even more critical.

Since writing the above report we have both experimented with potential fences (see drawing). These are made of 25 thou Plastic card and are mounted in line with the start of the elevons. Do they work? We think so, they allow you to pull much more up on the line and allow you to slow the plane done in thermal turns. There is a price to pay in terms of drag but we can not detect it.

We both entered a club slope slalom event, wind speed varied between 12 and 30 mph, but mostly between 20 and 30. The idea was to launch and fly round a very tight course in the least time. We were up against F3B and F3J machines, many ballasted up. We both flew the wings with no provision for ballasting (at 2 lbs AUW). Glyn came first in class (just wings) and 5 th. overall. We think it shook up some more of the competitors, particularly as neither of us had flown the wings on the slope before.

The 3.3m wing is now being drawn up ready to start cutting foam.(see Vitesse)

These successes have spurred Glyn to start drawing up the SZD 20 Wampir (Polish full size design) at long last - work on this first started in 1987! I am seriously considering a SB13 (German modern full size design). Both will be at 1/4 scale.