GOSSAMER CONDOR.

Until now, all human powered aircraft had followed the glider and the bicycle. There had been nothing really new since the 1935 Mufli of Haesller and Villinger. Designers had added to this by copying wheel drive from the bicycle, which meant unnecesary weight once airborne. HPA had flown in straight lines, but the Mufli had flown in a straight line. Every time the straight line got a little longer we got excited, but the 2290 yards (2094 m) of the Stork compared with the 779 yards (712 m) of Mufli? In forty years one expects more progress than that. And no HPA could repeatably demonstrate controlled sustainable turns. There must be a better way.

As this account shows, the quantum leap in HPA progress made by the Gossamer series occurred through a combination of novel ideas and thorough analysis, as well as iteration through flight testing and modifications.

In 1976 experienced glider-pilot and aerodynamicist Dr Paul MacCready decided to follow the hang-glider concept. His first calculations showed that with a much larger wing area than other HPA, flying speed could be brought so low that the drag of wire-bracing would be permissible. Also, at this low speed, wheel drive for take-off would not be needed.

PASADENA VERSION

MacCready decided that as a first step he would make the simplest possible machine to just fly and prove the principle. This would look like no other aeroplane, in fact it would look like no other thing at all. It consisted of 12 ft (3.66 m) lengths of 2 inch (50.4 mm) diameter aluminium tube and a network of wires. If it didn't work, MacCready didn't want to be laughed at; if it did work he wanted to keep it secret. So if anyone asked what "that" was he replied "This is a work-of-art. It portrays the meddling of government in the affairs of citizens". This first test aircraft flew just once in the car park of the Pasadena Rose Bowl, California, near where it had been made by Jack Lambie.

MOJAVE VERSION

The first real aeroplane had a single surface Mylar covered aerofoil, the aluminium spar tube at the leading edge with a second tube near the trailing edge, no pilot fairing, but otherwise the general configuration of the later versions. First flight was at midday on 26th December 1976 in the Mojave desert. Pilot was one of Paul MacCready`s sons. They had been taking turns to get their breath back on the initial runs and hops and stated later that "Who is to say what is a real flight, so who can say who made the first one". Best that day was a 40 second flight, already exceeding the durations of many previous HPA. The plane was named Gossamer Condor to draw attention to the condor, an endangered species of bird native to the locality. This aeroplane had a wing-span of 96 ft (29m) and an area of 1100 sq ft (102.2 m²). Longer flights were made and enough experience gained to show that the plane could not be made to turn. Also it was unstable, but the motions were so slow that all pilots could cope with this. Initially, the canard stabilizer only worked as an elevator to control pitch. Later it was made to bank, so that it would " pull the plane round by the nose ".

The first 40 second flight of the Mohave version of the Gossamer Condor, Dec 26, 1976, Parker MacCready pilot. RAeS collection

SHAFTER VERSION

An improved version flew at Shafter airport, California in March 1977. Shafter had calmer weather at this time of year, and nearby lived Vern and Maude Oldershaw, Sam Duran, Bryan Allen and Bill Richardson. The team were able to draw on Dr Chester Kyle's experience with human-powered road vehicles. A special aerofoil was designed by Dr Peter Lissaman. This has a low pitching moment and moderate concavities. It has been used on several subsequent planes built by others, ( see Table of Types). At this stage it would seem that the plane was never quite the same shape on any two flights as a variety of lateral control surfaces were tried out. Spoilers were tried, then a forward rudder, then an aft rudder. The banking canard stabilizer on its own was insufficient. Before long, the Gossamer Condor's total time in the air had exceeded that of all other HPA put together. Minor crashes were frequent; if the spar broke the pilot would jump out, so as to land feet-first and to reduce the weight on the plane. If anything broke it was rebuilt better. The Condor was evolving from the initial contraption into a sophisticated flying machine. While the team were still searching for a method of making the plane turn controllably, it was so severely damaged as to require a rebuild. They promptly started thinking about what improvements to incorporate in the next version.

Final configuration of the Shafter version of the Gossamer Condor. Bryan Allen piloting, which made the Kremer Prize flight August 21, 1977. Copyright Don Monroe

WASHOUT

Many wings are built with a twist such that at each wing-tip the angle of attack is the same as at the other tip, but different from the angle of attack at the centre of the wing. This is known as wash-out or wash-in.

While discussing wash-out for the new Gossamer design to replace the damaged machine, someone said as a joke that if the wing had wash-out on one side and wash-in on the other then at least it would turn, which is what they couldn`t make it do any other way. Then the serious practicable application of this bizarre idea dawned on others present. The new version incorporated a lever in the cockpit which when operated pulled on the rigging wires. This action twisted each wing. The system worked as hoped for in flight. The pilot retained the use of the banking stabilizer for small corrections and the wing-twist was used both to initiate and to hold a steady turn. This and all other HPA until then had done little more than straight flights, unsuccessfully trying all sorts of lateral-control-devices except, until then, this wing-warping which, in fact, was the system which the Wright brothers had used. But a bright idea is not enough: MacCready then had to calculate how much twist would be required. With Bryan Allen as pilot, an attempt was soon made on the Kremer figure of eight course with a prize of £50,000 to be won. On the first attempt the wake from a passing crop-duster plane put the aeroplane out of action. On Bryan's next attempt on the 23rd August 1977 the course was completed. The first Kremer Prize, that had been known as 'the' Kremer Prize’ for eighteen years had been won. The Gossamer Condor was now flown for the fun of it by all members of the team and others including Professor Geoffrey Lilley of Southampton, England.

KREMER CROSS CHANNEL COMPETITION

Henry Kremer's response to the first figure-eight flight was promptly to propose a new challenge, and the rules for the Channel crossing were drafted by a sub-committee chaired by Ron Moulton. Also there was to be a new figure-eight competition, restricted to entrants from any country other than that of the first winner. As mentioned above, after the first flights of SUMPAC and Puffin, it was generally thought that it would be a short time before the figure eight was flown. This was not what happened. When a competition was announced with the course being the English Channel, then it was generally thought that it would be a long time before this course was flown, if it ever was. "22 miles?", "Impossible !". Again, this is not what happened.

GOSSAMER ALBATROSS

The project started in October 1977 as a development of the Condor. New materials would lighten the weight. The technique of producing carbon-fibre tubes was developed. Although carbon-fibre had been used for some time, and composite tube structures were used for instance for fishing-rods, this was the first time that carbon was used on tubes of the proportion that we now think of as HPA spar proportions. Kevlar, another material taken for granted today, was used on an HPA first on the Albatross.

The wing surface profile was improved by having closer spaced ribs. The wing area was reduced to suit a higher speed. First flight of the Albatross was at Shafter in July 1978.

HPF>HPA

Human powered flight entails more than just a human powered aircraft. MacCready was realising that the organisational problems involved were such that sponsorship would be needed. In March 1979 the Du Pont Company, makers of most of the materials of the plane, agreed to be chief sponsor. The publicity generated by MacCready's ingenious use of their new material Kevlar resulted in millions of pounds of extra business for the firm.

Organisational tasks included :- transport to England of the plane and two back-up planes, test-flying these over land, obtaining boats, studying navigation, providing communications and making arrangements for the team during the trip which might last several months waiting for the right weather. The aeroplane was refined but was still not indicating the capability of crossing the English Channel, as Bryan Allen's best duration was 18 minutes. It was considered that inefficiency in the propeller was holding them back. With help from the Chrysalis team (see below), Gossamer Albatross soon had a new propeller, and in April 1979 Allen's duration improved from 18 minutes to 69 minutes, and he was not exhausted. He landed only to let Kirk Giboney have a flight. At this time, the intention was that the next stage would consist of some practice flights over water, but in a way that at the time seemed to the team to be magic, with the aid of test pilots and Royal Air Force authorities whom they had never met, transport to England had been arranged for the very next day in a Royal Air Force Hercules. So the overwater testing was skipped. It was some weeks before the weather was right but on the 12th June 1979 at 5.51 am the Gossamer Albatross with Bryan Allen aboard headed out to sea, with the many escort boats. Part way across, the plane met a headwind. It had been anticipated that the first attempt might not get all the way, hence the spare aeroplanes, but Allen kept pedalling. Although France can be seen from the top of the Dover cliffs, it cannot be seen from the beach. This means that for almost half the trip Bryan Allen was pedalling apparently to nowhere, but following MacCready in the pilot boat. Then comes the moment when it can be realised that that blur on the horizon is the continent, which gradually gets firmer.

The headwind was getting stronger and when barely three miles from Cap Gris Nez , Allen felt he could go on no longer and it was decided to abort the flight. The first move was made in the pre-planned procedure, namely climb to the height of a fishing-rod ready to be hooked up for a tow from a fast rubber dinghy. Surprisingly, he found that at this greater height he was able to proceed, so he never made the hook-up and flew all the way under his own power, arriving at 8.40 am, feeling that he could not have gone another 100 yards. But the Channel had been crossed, and the prize won by a superhuman effort, the equivalent of a marathon race in the air.

Back in England to collect the reserve aircraft, MacCready kindly invited all those of us known to be interested in HPF in England to view these aeroplanes; and he then related the background to the achievement. (See Gossamer Odyssey).

Gossamer Albatross flying the English Channel on 12^th June 1979. RAeS collection

CHRYSALIS

The project started when some MIT students tried to fly BURD II (see above) with the assistance of model aircraft engines in November 1978. This resulted in a smashed BURD, but the experience sparked new ideas, and Harold Youngren, Bob Parks, John Langford, and Hyong Bang decided in early December 1978 to design and build a new machine.

This had to be flying by May 1979 when these final-year students were due to graduate. Other objectives were, a maximum 75 ft (22.9 m) wingspan so as to fit the hangar, and a plane able to be flown by anyone and easy to repair.

These objectives led logically to the choice of a wire-braced biplane. The usual theoretical optimisation process was performed to finalise dimensions, and to decide the control system an 1/8th scale flying-scale-model was made. In its final control configuration, with developed control techniques, it was possible to fly this radio-controlled model in turns as tight as 2 1/2 spans in radius. As a result of these model tests it was decided to adopt conventional tail controls and wing-warping for the full-size plane. A test-section of wing was made and resulted in several important changes in design and constructional details. The team developed a method of anchoring the bracing wire 36% more efficient than that used on the Gossamer Condor. The Czerwinski method of joining the aluminium tubes was used.

The project was officially recognised by MIT and funded. More students, including Mark Drela, joined the team to help with construction. Professor E. Eugene Larrabee's minimum induced loss propeller design process was used.

In March 1979, Larrabee met Paul MacCready who suspected that the propeller on the Gossamer Albatross could use some improvement. Guppy Youngren designed a propeller for Albatross using Larrabee's method and sent the plans to MacCready.

A few weeks later MacCready was able to help the Chrysalis project by supplying the Mylar, an essential item which the team could not obtain elsewhere.

Anyone involved in a project today, can often benefit by co-operating with other groups, and the two problems mentioned here can be more easily solved. Other enthusiasts can put you in touch with suppliers of Mylar (Melinex in Europe), or other materials, and with a propeller design program.

Construction of Chrysalis took 3 times the originally estimated number of person-hours, but the plane was built in 91 days, and when nearly complete, proceeded for 72 hours round the clock.

Chrysalis made two flights on the 5th June 1979, the first day it had left the hangar. Within 3 weeks it was making 120 degree turns. One major modification was the addition of a fairing at the aft of the pod to lower wing junction. See interference drag in SUMPAC above.

Chrysalis provided recreational flying for 44 pilots, some inexperienced, and one very experienced: Bryan Allen, just back from flying to France. He found Chrysalis to require less power than Condor for cruise although more for climb and particularly that it was a more stable aeroplane.

Chrysalis in 1979

MiLan

There were no new aircraft at Nihon University from March 1978 to December 1981, when the MiLan'81 first flew. This was a totally new configuration, an almost rectangular high wing to which were attached twin tail booms with the propeller between them just behind the pod. A large number of bracing wires from the top of a kingpost to the wing and from the wing to near the small wheel. The pilot was upright. This aircraft flew 645 yards. A similar plane, the MiLan'82, flew on the 16th Oct 1982. Cruising speed was 12 mph (5.4 m/s). The twin booms were far enough apart for a 14 ft (4.3m) propeller.(Continue with Swift)

PHOENIX

Frederick To, an architect, made a film about HPF called "The Last Challenge" in 1974. He started building Solar One with the help and advice of David Williams (see SUMPAC) in 1977. In 1978, Solar One became the world's first solar-powered-aircraft. In this same year he began considering inflatable HPAs, seeing the advantage of portability and less vulnerability. He acquired the Reluctant Phoenix, and followed several of Perkins' techniques, including the elevon hinge system, in the design of his own inflatable, the Phoenix.

Wanting to use Melinex, rather than traditional balloon fabric, To first had to develop a system for bonding this material (F.To Aerospace June 1985), enabling the "T" joints to be made. Built like a flying airbed, there are a quarter of a mile (400 metres) of such joints between the skin and the vertical webs. F.To chose a rectangular configuration, to maintain constant skin tension and Reynolds number along the span, a flying wing configuration, because otherwise the inherent simplicity of rigging and transportation would be lost, winglets to enhance ground effect, and a span of 100 ft (30.5 m) with a 16.7 ft (5.1 m) chord. He estimated a weight of 85 lb (38.5 Kg) and power requirement of ¹/₃rd HP. He realised that cable operated controls would be out of the question because he anticipated the wing stretching 9 inches (230 mm) when inflated. Operation was by model aircraft servos, 4 servos to each elevon.

Work on construction of this largest ever HPA wing took place in a suburban house in a room 12 ft by 26 ft. (3.7 m by 8 m)The workbench was just longer than the chord. The material was fed across spanwise as it was worked on.

Wind-tunnel tests at Imperial College, London showed that the winglets reduced induced drag by a quarter. They also showed that the profile drag would be ten times more than for a smooth shape of the same section, whereas To had assumed a factor of only 1.25 in his power-requirement estimate. The model's profile replicated the series of flats and curves to be expected on the inflated shape. By this time, construction of the actual wing was advanced. Frank Irving of Imperial College sugested a series of hoops, with the outer skin above them, and these were used on the plane. Performance indicated their effectiveness.

Eventually inflation could be done in 20 minutes and deflation 30, but the first few times it took much longer than this. Phoenix could be carried deflated on a car-roof-rack. Fred To said the most awkward thing to transport was the propeller. On 28th March 1982, it was planned that Ian Parker would make the first test-flight inside the 600 ft (180 m) warehouse where it had been rigged. However Ian Parker thought better of this, and took it outside and took off. The weight was 105 lb (48 Kg), but there was 200 lb (91 Kg) of air inside the wing which had to be accelerated during the take-off run. (How to tell the difference between weight and mass ? Try to do a take-off run in an inflatable HPA.) The pressure of this air was small enough that the density above atmospheric air was insignificant, but because of its enormous volume, the mass was nearly double that of the airframe. Control was adequate and flights were made by all group-members, but no great distances were covered. In May 1982, Barry Jacobson took over the Phoenix, intending to reduce its thickness to make it more manageable.

Phoenix flying in Docklands, London in 1982. RAeS collection

KREMER FIGURE-EIGHT COMPETITION ( continued )

A prize of £ 10,000 was still available in 1980 for this course. The first figure-eight had been flown by Bryan Allen in the Gossamer Condor, in the USA. The rules stated that an entrant for the competition must be from a different country. The competition was to close at the end of June 1984. Three aircraft raced to beat this date, and each other.

HVS

The HVS was designed by Hutter, Villinger and Schule. Franz Villinger had been responsible for the Mufli 48 years previously. Wolfgang Hutter had designed gliders and Wilhelm Schule had constructed airframes. With the experience of the designers, and with backing from industry a neat and sophisticated-looking plane resulted. However it would appear that this was another disappointing low-wing aeroplane. The longest distance flown being little more than that of the Mufli. Achievement of the HVS has been to operate in high winds; speeds greater than 21 mph (9.4 m/s) (the plane's cruising speed) have been quoted. (Aero Modeller, Aug 1983).

ADJUSTMENTS would have been difficult to make on the HVS. The "V" tail precludes adjustment of vertical or horizontal tail area; and the canopy fits pilots of only one height. On this plane, the pilot operated non-rotating pedals to drive an adjustable pitch propeller. The drive mechanism was complex, entailing cable runs to the pylon where there are chains to contra rotating flywheels on the prop-shaft. With a cantilever wing of 54 ft span and a carbon-fibre structure, the HVS first flew on 12th March 1983.

COMPETITIONS

The HVS was designed and built before the setting of the Kremer World Speed Competition, yet its cruise speed of 21 mph was just in the range to make it a contender had it been more successful.

PELARGOS 2

In 1980, Max Horlacher the head of a small company which worked with composites, hoped to build a plane which his son, then eleven, could fly and which could be entered for the Kremer figure eight competition. His first plane, Pelargos 1, did not take off, and although he blames this partly on the fact that his son had put on too much weight, he decided to call in assistance for his next design. Fritz Dubs, an aerodynamicist from Zurich who collaborated on the II pointed out that the aerofoil section must be changed to be suitable for the very low speed at which they intended to fly the new plane. The firm of Reichhold Chemie AG supported the project by supplying the synthetic materials used. It is noteworthy that the wing-ribs and other secondary structural items on this aircraft were made from carbon, compared with the more usual and cheaper foam sheet construction. Other unusual structural features were a rigid strut and a light rear-spar. This rear-spar, as well as stiffening the wing generally was also made to serve as a place on which to make the joint between the sheets of Mylar, the greatest available width being six feet, and the chord of the wing over eight feet. It might seem that the wing could have been covered by using the width spanwise, and forming the joints on ribs; but if "tensilised" Mylar was used then this would explain their procedure. Tensilised Mylar shrinks more in one direction than the other when warmed. If correctly oriented, it will provide a better approximation to the rib shape between the ribs.

The propeller was 12 feet in diameter, one of the largest used. For ease of construction, a rectangular shape was chosen for almost everything else. This aeroplane was one of the first to be easily transportable in sections. The constructors benefitted from Horlacher's business experience, and the lessons learned on the plane were, in turn, useful to the business, as well as an advertisement for them and for Reichhold. Once the Pelargos 2 was constructed, it was put on display at a trade show in March 1983, but did not make its first flight until seven months later, because of difficulties in finding a satisfactory transmission system. A Cardan type universal joint was one of the options considered for this.

On January 20th 1984 Horlacher wrote to the RAeS to enter for the Kremer figure eight competition. Max Horlacher knew that there was competition both from the HVS and from Rochelt's Musculair. He wrote "I do not grudge the German team its success up to now. Both the aeroplanes are distinctive. One could be curious which will be successful".

The Pelargos 2 first flew in December 1983. On a later flight it covered a distance of 1100 yards (1000 m).

 MUSCULAIR I was built by Holger Rochelt and Gunter Rochelt, and designed by Gunter Rochelt, Ing. Ernst Schoberl and Dr.Ing. Heinz Eder. Gunter Rochelt had previously built a successful and radical solar-powered-aircraft. This human-powered-aeroplane was not built with the intention of winning a Kremer Prize. It was built with the intention of winning two Kremer Prizes. But it didn't. It won three of them ! Although there are many difficulties in the design of an HPA, one of the ways in which it is easier than for most other classes of aeroplane is that one has generally needed to only consider a narrow speed range. But Ernst Schoberl in his optimisation process aimed to find a shape which would be suitable both for the figure-eight course and for the speed course (see next page), and it would be piloted by Holger Rochelt who was of only average athletic ability. As can be seen from the table, the wing area of 173 sq ft (16.5 m²) was one of the smallest used on an HPA, and contrasts with the 760 sq ft (70.6 m²) of the Condor, the only other plane to fly a figure eight. A conventional configuration was chosen, with the propeller behind the tail, the drive shaft being the length of the tail boom. This propeller had previously been used on the solar-powered-plane. Spar construction was by laying-up cloth onto a rectangular block of foam: firstly diagonal weave all round, then spanwise strips top and bottom. Ribs were foam sheet, and the wing was skinned at the nose area, then covered in Melinex. The torsional stiffness deriving from the box-spar and the Melinex covering was found to be insufficient and it was found necessary to involve the "C" shape nose skinning in the task of load-carrying. This was done by adding diagonally-braced carbon-fibre rovings and effectively changing the "C" to a "D", thus forming a closed box. The plane was built in three months and made its first flight, as yet without cockpit fairing, at the end of May 1984. Within two weeks, on 18th June 1984, Holger Rochelt made a figure-eight flight. One Kremer prize for the Rochelts, the second ever figure-eight in HPF.

Musculair I story continues later