The Daedalus Project

The Daedalus Project

Pilot Frank Scarabino got his breath back after the prize winning flight of Monarch on May 11th 1984, the Monarch made a few more flights and was then displayed in the Museum of Science at Boston, Mass. Then they all wondered "What next ? " Some of them had helped to make the last desperate attempt to get BURD to leave the ground. Many had been involved with both Chrysalis and Monarch or with all three machines. BURD could no more fly than could a caterpillar. They had been through the "Chrysalis" stage and on to the prize winning butterfly stage of the Monarch. What could follow that ? The next stage above emulating real living creatures must surely be nothing short of emulation of the gods, or the mythical heroes of ancient legend who consorted with them. The archetype of mythical fliers were the father and son team of Icarus and Daedalus. Could they possibly have crossed the Aegean Sea then ? Could it be done now ? For a year, they asked this question. Then they decided to find out for  real.

PRELIMINARY ANALYSES

May 1984 to April 1985 By the end of May, five of the Monarch team, Juan Cruz, Mark Drela, Steve Finberg, John Langford and Barbara Langford were studying charts of the Aegean and starting to calculate what sort of plane would be needed. They concluded that if the flight could be made, it would not be done in the same sort of frantic rush as was used to get Monarch into the air, and it would not be done using the simple technology of MIT's three previous machines. Maybe, it was felt, it could not be done at all, and therefore it was proposed that a year be devoted to a feasibility study. In February 1985 the Smithsonian Institution gave the group some initial backing on the basis of this proposal.

PHASE 1 April 1985 to April 1986 Feasibility Study The official start of the Daedalus project with students at MIT making a preliminary design. The only major difference between this first design and the eventual aeroplanes was the propeller position which was shown at this stage as behind the pod. The classical Greek texts were studied in an attempt to pinpoint the route of Daedalus and Icarus. A trip to Crete set up weather stations and made contact with authorities. Efforts were made to get further sponsorship. In the group's study of the likely power available, it was realised that up till then all experiments on the power that a human can output had not recorded the weight of the person producing it. Clearly a choice of pilot will be based, amongst other factors, on both power-output and weight.

 STANDARD FOR MEASURING POWER REQUIREMENT

Drela showed that if an airframe has half the weight of the pilot, then a reasonable approximation, within the range of practicality, is that the power needed is directly proportional to the weight of the pilot. Therefore the parameter most relevant to pilot physical fitness is power-output per weight. Drela chose to quantify this in watts per Kg. This value could be expressed in ft lb/sec per lb weight. This reduces to ft/sec (implying 1g conditions). One could then get a feel for the magnitude of this parameter by comparing it with, for instance, how fast one can run up stairs. This is readily measured in ft/sec and is also equal to the person's power-output per weight. Watt-units become relevant if there is electrical-energy-storage. As a result of the Daedalus project, watts/Kg has become the standard measure of HPA power-requirement. In April 1986 the group decided that a crossing of the Aegean was possible, but that technology would need to be developed, and that a prototype aeroplane would need to be built to determine the form of the final plane.

 PHASE 2 April 1986 to April 1987

Michelob Light Eagle With sponsorship from Anheuser Busch Inc, makers of "Michelob Light", construction started on the Eagle. An Eagle is one of the trade marks of this company.

NEW AEROFOIL

Mark Drela had written the program XFOIL which enables the design of aerofoils and which can accurately predict the performance at any Reynold's Number. It is actually superior to a wind-tunnel test of a proposed section, because tunnels suffer from turbulence and interference from the tunnel walls and model supports. A series of sections was produced to suit the varying Reynolds numbers along the wing due to wing-taper.

WING PROFILE

A computer-driven foam-cutter was built, based on a hot-wire stretched between the pen-holders of two plotters. This was used not only to cut the ribs, but to cut the nose skinning panels from solid blocks. The wing sections never existed on paper, either as lists of numbers or as drawings; the information came out of the design program onto disc, and this guided the cutter. The spar of the MLE (Michelob Light Eagle) was three tubes aligned vertically. This scheme was abandoned for the final Daedalus design where the more usual single tube was used. Construction of Light Eagle took 15,000 hours work by 18 members of MIT.

PILOTS SELECTION, TRAINING & TESTING

Because of the sheer distance to be travelled on the cross-Aegean flight, pilots had to be selected on athletic stamina, based on a carefully set up test which measured heart-rate at 70% maximum oxygen uptake during prolonged exercise. Being athletes rather than pilots meant that they needed to be trained to fly. Many other projects, faced with a plane with a high power-requirement have asked the experienced pilot to step out, and asked a cyclist to step in - with disastrous consequences. On the Daedalus project the quality of the flying-training, and the piloting-ability acquired by those that flew the three aircraft,(MLE, Daedalus87 and Daedalus88), is evident in that in all the flights made there were only two crashes, both in local weather conditions against which no amount of experience would have availed. Like most HPA, both Eagle and Daedalus are single-seaters, and a  flight simulator was built to familiarize pilots. Other than on the simulator which had the same seating as the aircraft, the pilots kept physically fit by training on upright unfaired bicycles.

DRINK

A special drink was developed by Ethan Nadel to refresh the pilots during flight. Nadel calculated how much fluid, glycogen and sodium would need to be replaced per hour, and found that no available drink would provide this. His own concoction was tested by the pilots on the ergometer and all who tried it could keep up a sufficient pedalling rate for six hours, except one who gave up after four hours because he had been sitting on the uncomfortable seat of the rig long enough, not because his legs were tired. It must be mentioned that this was done in the spirit of engineering in order to find a satisfactory solution to a problem, not as pure research, since no comparative tests were done with other drinks.

JANUARY 1987 TEST FLIGHTS

The Eagle first flew in October 1986 at Hanscomb Field. The wingspan was then increased to 114 ft (34.7 m) and the first series of Michelob Light Eagle test-flights were done at NASA Dryden during January 1987. On January 22nd with Glenn Tremml at the controls the distance record previously held by the Gossamer Albatross was reset at 36.5 miles (58.7 Km). The World Distance Record for a female pilot was set at 10 miles (16.1 Km) by Lois McCallin, in 37 minutes 38 seconds. McCallin's flight also established Duration-Female and Closed-Course-Distance and Closed-Course-Distance-Female records recognised by the FAI.

But the main purpose of these January 1987 flights was to gather data to guide the design of Daedalus itself.

 LAMINAR FLOW

To discover whether the flow on the wing surface was laminar, the team used the following technique. Immediately prior to flight, the wing surface was painted with a powder/liquid mixture. As the liquid dried off, the type of flow could be detected by the streaks that remained. This was photographed during or immediately after flight. On the Eagle it was seen that the flow was indeed laminar in all the areas expected, including at some points on the lower surface where the Melinex had been deflected up to touch the carbon fibre spar, but not at the wing-tips. This turbulence at the wing tips was attributed to the coarseness of the texture of the foam skin-panels on the relatively sharp leading edge. Accordingly, a denser foam,(pink-coloured), was used for Daedalus.

ATHLETIC LIMITS

The pilot's heart-rate and breathing-rate were recorded during these flights and results confirmed those obtained on the training-rigs and the simulator.

REQUIRED POWER

Attempts were made to determine how much power was needed to fly Light Eagle. These were not very conclusive. It is notoriously difficult to use gliding angle as a measure, as any slight upgust will affect the results. The other method used was to fit a strain gauge to the propeller shaft, and use readings from this in conjunction with recordings from the airspeed indicator and altimeter to determine power-required. Altimeter readings were necessary so that the power absorbed by any height increase during the test-period could be subtracted to obtain the net power for level flight. This system gave consistent results. Surprisingly perhaps, it even gave fairly consistent results when the mean height was varied between 10 ft and 25 ft (3 m and 7.6m). The best measure of required-power still seems to be the subjective estimate of pilots, backed up by the records of their heart-beats during prolonged flights after the effect of take-off on heart-rate has subsided. This can be quantified as a power value from the records of each pilot's performance on a ground-rig on the same day. One snag with this system was discovered later, when on any pilot's first Daedalus flight, there was an increase in heart-rate which was due to the excitement of flying the new plane.

AUTO-PILOT

An attempt was made to make an autopilot, as had originally been envisaged for Daedalus. This did not prove possible because neither the artificial horizon display nor the sensors ever showed the degree of reliability necessary for autopilot operation, possibly because of a buildup of an electrostatic charge on the aircraft as it flew. This meant that the crossing of the Aegean would need to be made in daylight when there is more wind and turbulence.

PHASE 3 April 1987 to April 1988

DAEDALUS

The Daedalus airframe was considerably lighter than Eagle because it was designed to a lower load-factor, 1.75g instead of 3g, and because a higher grade of carbon was used for the spar. There were no ailerons on Daedalus, whereas on Eagle the tip 12 ft (3.7 m) had rotated. This more than offset the slightly heavier wing-skinning foam used. Parts for two Daedalus airframes were made simultaneously.

The United Technologies Corporation backed the project with $500,000, and there were many other sponsors.

Daedalus test flight before going to Greece.Pictue by Steve Finburg

   
AERO-ELASTICITY FLIGHT TESTS

A separate series of flight-tests was made with Light Eagle from December 1987 to March 1988. This research was aided by J.E.Murray of NASA-Dryden because one of the purposes of these tests was to gather information which could be useful in the design of a high altitude drone, which would operate at between 55,000 and 70,000 ft  (16,750  and  21,330 m) altitude. The flights were also invaluable pilot training for all the five Daedalus pilots Erik Schmidt, Frank Scioscia, Glenn Tremml, Greg Zack and Kanellos Kanelloppoulos. A video recorder and a pair of mirrors were fitted to Light Eagle so that both wing-tips and the tail could be seen in one picture. The airframe was fitted with accelerometers, rate gyros and strain gauges to measure aircraft motion and wing and tail-boom bending. The computer that polled these sources 20 times a second weighed 2 lb (0.9 Kg) and had a memory of half a megabyte. This system devised by the Daedalus team was recognised by NASA as providing good quality data. Each set of data was processed on a NASA mainframe the same day to give useful results for the next day's flying. The Eagle has a long thin flexible tail-boom; on a plane like this, it is fairly obvious that if you waggle the elevator up and down fast enough, you won't move the plane, you'll just flex the boom. But that is just stating one of the problems that the Daedalus team set out to solve. The natural frequencies of the pedalling-cadence, the structural flexure and the aircraft flight dynamics were known to be of the same order and would therefore be expected to interact. Therefore both the aircraft stability and the response to any control input could not be calculated using standard theory. Analysis of the stresses observed during tests showed that a drone with a similar structure to the Eagle could well survive the turbulence at the anticipated operating altitude. With regard to the involved flight dynamics question, the report on these tests found that the influence of aircraft flexibility and unsteady aerodynamics was greater than originally expected and concludes "The flight data collected so far may prove adequate for model validation, although special software will need to be written to handle the larger sets of equations of motion" (Zerweckh, Flotow & Murray, 1988).

Of direct relevance to the project was a study of the response of the aircraft to rudder only, to simulate Daedalus. It was observed that the aircraft moved into a turn satisfactorily, albeit with a 12 degree angle of sideslip. This was seen as confirming the decision to not have ailerons on the final plane, although at least one of the group considers that the lack of ailerons did play a part in the April 1988 accident. The first flight of Daedalus 87 was in January 1988. Flight trials were satisfactory until a loss of control into spiral divergence resulted in a crash in February 1988. This was attributed to insufficient rudder movement because the control cables were stretching, and to insufficient dihedral.

The flight simulator proved useful in investigating how much more dihedral would be necessary, and it was seen as fortunate that the faults became evident before the plane was out over the sea. Construction work on the second airframe, Daedalus 88 was stepped up, incorporating the modifications, and Daedalus 87 was repaired.

On 26th March 1988, the Hellenic Air Force transported all three aircraft to their Heraklion base on Crete. This conveniently has a runway pointing straight out to sea. Greek Navy ships stood by to escort Daedalus, the Greek National Tourist Organisation provided meals and lodgings to the team, and the Hellenic Industrial Development Bank helped to sponsor the epic flight.

THE CROSSING

23rd April 1988 It had been estimated that there were maybe 3 or 4 possible days a year when the weather would be suitable, and there had already been two of these just before arrival! The pilots were on a rota system. Each morning one would sit in Daedalus and wait to be told by Steven Bussolari that the weather wasn't quite good enough today. Even if it was calm at Heraklion at that moment, reports from team members on the destination island of Santorini or from a yacht offshore or from Greek meteorologists would indicate unsuitable weather later on.

It is the 23rd April, the turn of Greek national cycling champion Kanellos Kanellopoulos to be in the pilot's seat. The weather initially is ideal, with a 3 mph (1.3 m/s) tailwind. Flight controller Steve Bussolari decides that the flight is on.

Team leader John Langford announces that "The research phase of the project is over"

The Daedalus flight starts. The Heraklion runway ends at the top of a cliff, so Kanellos suddenly finds himself 130 ft (40 m) up, but glides down to join the waiting escort boats. The speed of Daedalus, 15.4 mph (24.8 Kph) is more than that of Gossamer Albatross on the English Channel crossing (12.1 mph (19.5 Kph). It has been so designed in order to make the distance in a short enough time for there to be a reasonable chance of the weather holding. The inflatables with their outboards are making a lot of wake just to keep pace, as photographs later show. The course had originally been planned as leaving from the Western tip of Crete, but this would involve too much flying close to coastlines with their associated turbulence.  On the course chosen there is only the small island of Dia to detour around until the coast of Santorini is reached. Dia is passed after 26 minutes and all is well, but soon after this the altimeter fails. Bryan Allen had to fly the last five miles to France without altimeter. Kanellos has another sixty miles to fly without one. The morning is still cool and Kanellos is not drinking at the prescribed rate. Yes, the rate was prescribed because otherwise pilots tend to ignore drinking until dehydration started to impair performance. Suddenly a cargo vessel is seen sailing towards the flight path, but a Greek Navy boat heads this off. Kanellos pedals smoothly on, making Daedalus 88 the first HPA to lose sight of land, the first HPA to cover such a distance and then after 2 hours 48 minutes the new duration record holder. The tops of the mountains of Santorini are seen first and then through the clear Greek air could be seen the whole island. Beautiful as it looks, the weather report radioed from the beach is not so cheering - an 8 mph (3.6 m/s) wind, and in order to land into this wind, the plane will need to turn almost back on its current flight path. Thermals from the hot beach are causing turbulence. At 10.58 am, during this turn for a final approach parallel with the beach a gust lifts the plane and brings the groundspeed to virtually nil. A second gust from the side catches the rudder and snaps the tailboom. This momentarily causes full up-elevator deflection. The starboard wing is overloaded and snaps. Kanellos and the Daedalus 88 fuselage fall into the sea 20 feet (6m) from shore. He rips through the Mylar unharmed and with plenty of energy left to swim ashore, the whole trip is done under his own power. The local people are ecstatic, as are Kanellos and the rest of the team, and as indeed are HPA enthuthiasts the world over when we hear the news and see the television pictures later that day. Journalists on the beach are baffled as to why the team should be pleased when their aeroplane is lying there, smashed up. The team are pleased because they have accomplished what they have been planning for four years. Furthermore, because the project was a pursuit of the ultimate, the fact that the plane just completed the course before breaking up was proof that it was designed just right for its intended purpose.

The flight clearly was a success and must be recognised as such; however it is double-counting to claim both that it was a "success because Kanellos wasn't exhausted" and a "success because the plane only just made it".

RESULTS

The two national flags of Greece and the United States had flown side by side by the temporary hangar at Heraklion, showing that this was an endeavour supported by individuals and organisations, both private and public from both nations.

As mentioned above, the technology of Daedalus indicates the feasibility of high-altitude winged craft.

The Daedalus project led to a large number of technical reports, covering the wide range of topics studied. Some are mentioned in the bibliography including (Dorsey 1990), (McIntyre 1988) and (Langford August 1989). The last two authors accompanied the epic flight. Langford was project manager and his August 1989 publication lists many other references.

But, of course, for us HPA enthusiasts, or indeed for those with soul enough to appreciate the realization of a classical myth, the flight needs no such apology.  Sufficient that it happened