The Avrocar

The Rig view is the machine in 3D, at true scale, with a person standing beside it. It is 3D because the question about this aircraft is a distance: the gap between the rim of an eighteen-foot disc and the ground. Raise it and watch the cushion dim, the augmentation fall and the disc break into hubcapping past about two feet. Push it past the ceiling and there is no drama, just a machine settling because the lift is not there. A staff on the left is ticked at eighteen inches, two feet and three feet, which is the entire operating world of this aeroplane.

The Ledger view is lift against height, and it is where the aircraft is explained. The amber curve is the machine: L(h) = 3,150 + 1,038 / h pounds, anchored on two published weights and nothing else. The dashed curve above it is standard thin-jet theory, the cushion the configuration was chosen for, which it never got. The red dashed line across the bottom is 3,150 lb, the lift with no ground under it. The weights are drawn across all three, including the 150 lb bracket between the empty airframe and that red line, and a side panel prices the two losses: 1.14× through the turborotor, 0.24 of the theoretical cushion.

The Mistake view tests the file's 1954 memorandum. On the left, angular size: an 18 ft disc against the full moon and a thumb at arm's length, with a range slider, because the memo's claim is about appearance. On the right, an altitude ladder from this machine's three-foot ceiling up through the design goal, Frost's projection, and four reported altitudes this site already carries on its own instruments. Between them, two cards at equal size: what the memorandum said, and what the machine that class produced actually did.

The File view is the paper trail: a 1952 Canadian research contract, the 1953 leak, the Scientific Advisory Board's briefing at Omaha and its November 1954 report, the confusion memorandum, Doolittle's forwarding letter, the January 1955 paragraph that was cut, the KC-97 case the committee could not explain, CIRVIS, rollout, first flight, two trips to the Ames tunnel, cancellation, and two museum floors. It closes on two verdicts at equal size: it was real and it flew, and it never left the cushion.

The station has priced a lot of saucers. INST-32, Electrogravitics, put LaViolette's two competing thrust models on one pressure dial; the warp and Casimir benches costed negative energy. What all of them share is a claim about physics and no machine. This file contains the opposite: an actual machine. The briefing "The saucers they were building themselves" covers DOW-UAP-D095, which holds the Scientific Advisory Board's review of Avro's Project Y2, a memorandum arguing that circular aircraft could be mistaken for UFOs, a KC-97 case the Air Force recorded itself unable to explain, and the CIRVIS reporting programme. That line produced the VZ-9 Avrocar: two airframes, test pilots, a NASA wind tunnel, and both survivors on museum floors today.

It earns its place because it separates two things that usually arrive fused. The first is the engineering: a real saucer really flew, and everything it could do lived inside about a metre of ground, for reasons you can compute exactly from two published weights. The second is that 1954 memorandum, which is the most reasonable-sounding sentence in the whole saucer literature: observers unfamiliar with the technology may mistake circular-planform VTOL aircraft for UFOs. This instrument puts both on the bench and declines to merge them for you. An 18 ft disc really does match the full moon at 600 m; and the machine that class produced flew from two airfields, never above three feet, for 93.5 hours in total, twelve years after the scare it gets invoked to explain began.

One cushion law, two published weights and one fitted constant run the entire instrument, and the panel tags every number measured, reported, modelled or read.

p_c = J(1+cos θ)/h · L(h) = T + η·A·J(1+cos θ)/h = 3,150 + 1,038/h lb · h_max(W) = 1,038/(W − 3,150) m

The Avrocar is a hovercraft that wanted to be an aeroplane. The turborotor's efflux leaves as a curtain around the rim, and the curtain's momentum traps a pad of raised pressure under the disc. Thin-jet theory makes cushion pressure the curtain's momentum flux per unit length times (1 + cos θ), divided by the gap. Lift is therefore a flat thrust term plus a term in 1/h, and the standard augmentation for a circular planform is 1 + D/4h. This is textbook ground-effect machine theory, applied to an aircraft that was sold as a fighter.

Two published weights fix the whole curve. Maximum lift out of ground effect is 3,150 lb, which is the flat term with no ambiguity: no ground, no cushion. The machine held its 4,285 lb zero fuel weight at a ceiling of about three feet, which fixes the 1/h term. The result is one line, L(h) = 3,150 + 1,038/h pounds, and its rearrangement gives the ceiling for any all-up weight directly: h = 1,038/(W − 3,150) metres.

The one fitted number is the finding. Textbook augmentation at three feet is 2.5×, about 7,900 lb. The machine made 4,285. So it realised 0.24 of the cushion the theory owed it, and that constant is the only free parameter on the bench. Read it beside the other multiplier, 1.14× from a lift fan that should have returned twice its input thrust, and the engineering story is complete without a word of opinion. The technical record names the cause in one clause: excessive losses in the ducting system, which became apparent and were never resolved.

The ceiling has a pole, and the pole is the obituary. As the all-up weight falls toward 3,150 lb the ceiling runs away to infinity, and above it the ceiling collapses fast. The empty airframe is 3,000 lb. That leaves 150 lb for the pilot, the fuel and the payload combined, which is less than a pilot. Out of ground effect there is no cushion to improve, so no amount of development inside that budget moves a number of that shape.

Instability is reported, not solved. The critical height near two feet, the loss of control above it, and the leading-edge separation and reattachment cycle in forward flight are all from the record. The amplitude ramp and the waveform you watch on the bench are drawn, and the instrument says so on its own face. Fifty-two holes drilled three feet from the centre in 1961 improved hover and moved nothing else.

The weights, dimensions, speeds, dates, flight hours and the angular-size arithmetic are the record used exactly as it gives them, and where sources disagree, on maximum takeoff weight and on engine thrust, the disagreement is printed rather than resolved. The cushion law is textbook and is labelled as modelled; the single constant fitted to it is labelled as the finding. What the 1954 memorandum is entitled to explain is a reading, so the bench hangs the memorandum and the machine as two cards at equal size, and adds no third.

A gap, a weight, a speed and a range: four sliders, and between them the whole career of the only flying saucer that ever existed.

• The gap at the rim, 0 to 1.6 m. The distance the programme lived in. Cushion pressure goes as 1/h, so this slider is not a camera control, it is the aircraft. Eighteen inches is where it was flown; two feet is where it started to come apart; three feet is where the lift ran out at zero fuel weight.
• All-up weight, 3,000 to 7,000 lb. Empty at 3,000, zero fuel at 4,285, the two design transition weights at 5,650 and 6,970. Drag it and watch the ceiling move on the ledger. Below 3,150 lb the ceiling goes to infinity, which is the cleanest way to see that the whole problem is 150 pounds wide.
• Forward speed, 0 to 40 knots. Best free flight was 30 knots in 1960 and 20 at the final evaluation in June 1961, against a design goal of 300 mph. Add speed and the pitch cycle takes over, which is the second way the same cushion failed.
• Slant range, 50 m to 30 km. For the Mistake view. This is the memo's claim reduced to geometry: how big does an 18 ft disc look, against the moon and against a thumb, and at what range does it stop being a disc at all.
• Three flight cards. Hover as flown, forward at the best speed ever achieved, and the attempt to climb out. The third one is not a manoeuvre, it is a wall.
• Layers. The cushion, whose brightness is the share of the weight the ground is carrying; the curtain; the downwash; the scale figure and the ticked staff; and the six-foot ditch it crossed at eighteen inches, which is the part of the story usually left out.

Six claims about the only flying saucer that was ever actually built, with who established each and where it lands today. Two are settled, two are still argued, one does not survive contact with the flight log, and one is the kind of claim no bench can close.

1. Start at zero and lift it to eighteen inches. Watch the cushion come up and the augmentation chip climb past 1.7×. That is the aircraft working, and it is working entirely because the ground is there.
2. Now go to two feet. The disc starts to wander in pitch and roll. This is hubcapping, and its onset height is from the record. Nothing on the machine changed; only the distance to the ground did.
3. Push to three feet at zero fuel weight. The state chip flips to SINKING and the lift number drops under the weight. There is no failure event and no bang. The aeroplane simply runs out of ground.
4. On the ledger, drag the weight down to 3,000 lb. The ceiling marker runs off the right-hand edge, and the 150 lb bracket is the only thing between the empty airframe and the flat line. That bracket is the programme.
5. Compare the two curves. The dashed line is what standard theory owed this configuration; the amber line is what it got. At three feet that is 7,876 lb against 4,285. Then read the side panel: 1.14× through the fan, 0.24 of the cushion.
6. In the Mistake view, set the range to 600 m. The disc is exactly moon-sized: the memo's premise is real. Now drag out to 19 km and watch it stop being a disc at all. Somewhere between those two numbers is the entire honest version of the confusion argument.
7. Read the ladder from the bottom. Three feet, then eighteen inches below it, then nothing until the design goal at ten thousand. Every rung above is a figure this site carries on another instrument. Click one and it will tell you which.
8. End at the file, on the January 1955 card. The office's best efforts were constantly directed to discouraging, minimising and playing down UFOs, and the article's final paragraph, which carried a UFO tie-in, was to be eliminated. Their saucer could be printed. Everyone else's could not be mentioned beside it.

The honest line between what is measured, what the record reports, what is modelled here, and what is a reading:

In one line: the weights, dimensions, speeds, dates, flight hours and angular sizes are the record plus arithmetic, auditable on the bench, with the two source disagreements printed rather than smoothed; hubcapping's onset and mechanism, the design weights and the design goals are quoted and tagged REPORTED; the thin-jet cushion law is MODELLED, and the single constant fitted to it, 0.24 of the textbook cushion, is the instrument's finding rather than its input; and what a 1954 memorandum about a class of aircraft is entitled to explain is a reading, for which this bench hangs two cards at equal size and adds no third. Raise the gap, load it up, and decide for yourself.

• DOW-UAP-D095, "Joint U.S.-Canadian Aviation Projects and UFO Sighting Reports, 1954-1955", PURSUE Release 04, U.S. Department of War, hosted at war.gov/ufo. The file carries the Scientific Advisory Board cover on Avro Project Y2 (November 1954), the Doolittle forwarding letter of 10 December 1954, the 14 December memorandum drawing attention to its generally negative findings on technical feasibility, the 13 January 1955 memorandum for record on the Air Intelligence Digest article, the 1954 memorandum on mistaking circular-planform VTOL aircraft for UFOs, the 11 July 1955 evaluation of the KC-97 Newfoundland incident, and the 3 November 1955 memorandum on CIRVIS under AFR 200-3 and the JANAP 146 series.
• Signals from the Periphery, Release 04 Briefing 5, "The saucers they were building themselves", the site's own summary of that file, which this instrument is built from.
• Lindenbaum, B. & Blake, W., "The VZ-9 Avrocar", the American Helicopter Society technical account of the programme. Source of the two figures this bench is anchored on: maximum lift attained out of ground effect 3,150 lb, zero fuel weight 4,285 lb, therefore incapable of hover out of ground effect; and of the finding that excessive losses in the ducting system became apparent and were never resolved. Also the source for three Continental J-69 turbojets at 920 lb thrust each and the 5 ft, 124-blade turborotor.
• The technical review "Avrocar: a real flying saucer" (Advances in Aircraft and Spacecraft Science 4:4, 2017; arXiv:1507.06916): programme span 1954-1961 at about US$10 million, design transition weights of 6,970 lb in ground effect and 5,650 lb out of it, 18.5 hours on the first airframe and about 75 on the second, 30 knots reached in April 1960 and 20 at the final evaluation in June 1961, 100 knots by the modified model in the Ames tunnel, a 90 degree turn in 5 seconds left and 11 right, and the fifty-two stabilisation holes drilled radially three feet from the centre.
• Avro Canada VZ-9 Avrocar reference summaries for the dimensions and dates: 18 ft diameter, 3 ft 6 in height, 254 sq ft planform, 3,000 lb empty, first tethered flight 29 September 1959, first free flight 12 November 1959, testing halted 5 December 1959, NASA Ames 40 ft by 80 ft wind tunnel tests April 1960 and April 1961, funding ended March 1961, cancellation December 1961, serials 58-7055 and 59-4975, and the estimated-versus-actual performance table (300 mph against 35, 10,000 ft against 3, 995 miles against 79). These summaries list a maximum takeoff weight of 5,560 lb and 660 lbf per engine, both of which disagree with the technical sources above; the disagreement is printed on the instrument rather than resolved silently.
• Contract and programme history: Canadian Defence Research Board contract 1952 at CAN$400,000; the press leak of 11 February 1953; USAF contract 1955 at US$750,000; Avro private venture funding of $2.5 million in 1956; additional Air Force funding from March 1957 under Project Y-2 / Weapons System 606A; a joint-services contract of $2 million and about $700,000 more; the second airframe contracted March 1959 at $1.77 million; rollout from the Malton plant in May 1959, three months after the CF-105 Arrow cancellation. John Frost joined Avro Canada in June 1947 from de Havilland, where he had worked on the Vampire, the Hornet and the Swallow.
• The hubcapping description as the programme and its historians give it: rapid, unpredictable oscillation in pitch and roll from an unstable ground cushion above a critical height of about two feet, with flight above about three feet impossible; and in forward flight, separation over the leading edge dropping the pressure under the front of the disc, a nose-down pitch, reattachment, a pressure spike and a nose-up pitch, repeating as a limit cycle. Ames wind tunnel work also indicated the configuration would be unstable and uncontrollable at speed.
• Ground-effect theory: the thin annular-jet momentum curtain of Chaplin (1957) and Stanton-Jones (1961), in which cushion pressure is the jet momentum flux per unit length times (1 + cos θ) divided by the height, giving the standard augmentation 1 + (1 + cos θ)·D/4h for a circular planform. The bench uses it as the reference curve and fits one constant, the realised fraction, to the two published weights.
• Museum disposition: 58-7055 acquired by the National Museum of the United States Air Force in November 2007 and displayed from June 2008; 59-4975 at the U.S. Army Transportation Museum, Fort Eustis, Virginia, after a showing at Man and His World in Montreal in 1968; a replica at the Royal Aviation Museum of Western Canada.
• Cross-referenced altitudes on the ladder are the figures this site already carries on its own instruments: the GoFast trigonometric solution near 13,300 ft (INST-27), the Gimbal jet altitude of 25,000 ft (INST-28), the consensus reconstruction of Mantell's climb to about 30,000 ft (INST-36), and the reported 80,000 ft start of the 2004 Nimitz descent (INST-24).