There is an aeroplane in the Cold War Gallery at Dayton that the museum has deliberately not restored. Its wings are off. Its nose is crushed. The paint is scorched.

The Lockheed XF-90 lost its competition, never entered production, and was officially recorded as destroyed. It was then subjected to three atomic detonations at close range, left in the Nevada desert for half a century, and is still here.

It is the only surviving example of the second design Kelly Johnson’s Skunk Works ever produced, and it survived because it was built far too heavily to be any good.

An escort fighter for the atomic age

In autumn 1945 the USAAF asked for something difficult: a jet penetration fighter to do the job the P-51 had done over Germany. It was to be supersonic, carry no fewer than six 20 mm cannon, and have a combat radius of around 900 miles.

Lockheed studied 65 configurations before settling on a swept-wing design with side intakes and a long, pointed nose. It was, by common consent, one of the best-looking aircraft of its generation.

Then Lockheed built it like a bridge.

Overbuilt on purpose

The XF-90 used 75ST aluminium rather than the standard 24ST of the era, along with heavy forgings, machined components and half-inch rivets. The result was an airframe cleared for 12 g.

A Lockheed flight-test engineer described the main I-beam as resembling a bridge girder, which was accurate and not entirely a compliment. Empty weight came out at 18,050 pounds, more than fifty per cent above its competitors and, by one comparison, some seventy per cent heavier than an F-86.

None of which would have mattered with enough thrust. There was not enough thrust.

The first aircraft flew on two non-afterburning Westinghouse J34s and frequently needed rocket-assisted take-off bottles to get airborne at all unless it was flown nearly empty. The second was fitted with afterburning J34s, the first afterburner installation on a USAF fighter, and was better without ever becoming good. The museum’s own summary is that these engines proved inadequate, and that the second aircraft had better but not great performance.

Published thrust figures for the type genuinely conflict, and no authoritative dry-and-wet pair for the XF-90A could be established. The qualitative verdict is not in doubt.

Mach 1.12, and a noise nobody recognised

It did go supersonic, in a dive, about fifteen times. The best recorded figure was Mach 1.12 on 17 May 1950, and the moment is preserved in the memory of a Lockheed flight test engineer who was standing next to Kelly Johnson at the time.

The crew on the ground had assumed the aircraft had hit the desert, because in 1950 almost nobody had heard a sonic boom and had no reason to associate the sound with anything but a crash.

Losing

The fly-off went to the McDonnell XF-88, with the North American YF-93 also competing. The arithmetic was brutal and simple: the XF-88 used the same two J34 engines and weighed about six tons, against the XF-90’s nine. Identical thrust, half again the weight.

Sources differ on whether the XF-90 placed second or third, and on whether the decision fell in August or September 1950. It hardly mattered. The Korean War redirected the USAF toward building more of what it already had, and the penetration fighter requirement was cancelled outright. Nobody won.

The XF-88 line was not entirely wasted; an enlarged derivative later became the F-101 Voodoo. The XF-90 got a different sort of afterlife.

Three atomic bombs

The first airframe went to the NACA laboratory in Cleveland and was tested to destruction, which is a reasonable end for an aircraft built to withstand 12 g.

The second was taken to Yucca Flat in Nevada and parked about half a mile from ground zero alongside a B-45, a B-29, some F-47s and B-17s, to find out what nuclear weapons do to parked aircraft.

On 15 April 1952 a one-kiloton device was dropped from a B-50 with the aircraft facing nose-in. Its wings were undamaged. It picked up dents, cracks and a buckled fire door, and repairs were estimated at around 106 hours.

On 22 April, a 31-kiloton shot, again nose-in. The nose bent. Roughly two thousand military personnel and a crowd of journalists watched.

On 1 May, a 19-kiloton shot, this time with the aircraft turned side-on to the blast. That one finished it: the tail was severed, the landing gear was torn off the wing, and the main wing structure buckled and scorched.

Period footage of the Tumbler-Snapper series is above.

What it had done, though, was survive three atomic detonations well enough to be recognisable. A Department of Energy official wrote in 1990 that the data from those tests continued to be used in aircraft design and operational deployment.

Fifty years in the desert

After a 5.3-kiloton underground blast in December 1963 the contaminated hulk was moved to Area 11 of the Nevada Test Site, where it was used for Broken Arrow exercises: training crews to recover a downed aircraft carrying a nuclear weapon.

Officially, it had been destroyed. On paper the XF-90 no longer existed.

A Lockheed engineer, Ernest Joiner, spotted what looked like an intact airframe from the air years after it had been written off. But the person who actually saved it was Robert Friedrichs, a physical scientist at the Nevada Site Office, who recognised what was sitting in the desert in the late 1980s and then spent thirteen years on historic assessment, museum negotiations, funding and contracting.

Decontamination ran from 2001 to 2002 and was not straightforward. Every rivet had to come out so that radioactive sand could be purged from the structure, plutonium residue had to be scoured off, and a colony of white-tailed antelope squirrels carrying hantavirus risk had to be dealt with. Summer work was done at night.

The engines produced the detail that lands hardest.

Fifty years after three atomic detonations, the Westinghouse J34s still spun freely by hand.

A C-5 flew the pieces to Wright-Patterson in 2003.

Why it is displayed broken

The museum chose not to restore it, and that was the right call. A polished XF-90 would be a handsome footnote about a fighter that lost a competition in 1950. The wreck is a much better object.

It sits in the Nuclear Deterrence section of the Cold War Gallery, in a diorama representing the Nevada Test Site, with its nose crushed and its wings removed.

There is a neat irony in it. The XF-90 lost because it was built far more strongly than a fighter needed to be, and every pound of that structure was a pound of performance it could not afford. That same overbuilding is the only reason there is anything left to look at.

Its rival won the contract and the lineage. Not one XF-88 survives.

Sources: Jorge and Karen Escalona, Nukes vs. Airplanes, Air & Space (Smithsonian), July 2008; National Museum of the US Air Force fact sheets; Robert Gutmann, Aviation History, March 2018; Operation Tumbler-Snapper records; Sebastien Roblin, 19FortyFive, 2021.