On June 5, 1948, test pilot Glen Edwards lifted the second prototype Northrop YB-49 off the runway at Muroc Army Air Field. Ninety minutes later, the aircraft disintegrated in the Mojave Desert, killing all five crew. The base was renamed Edwards Air Force Base in his honor. The flying wing program had four more years to live, and then Jack Northrop's masterpiece would be cut up with acetylene torches while he watched.

The YB-49 was a 172-foot-wingspan, eight-engine jet bomber with no fuselage and no tail. Just a wing. Northrop had been chasing this configuration since 1929, convinced that eliminating parasitic drag from fuselages and empennages was the path to true aerodynamic efficiency. The math was unassailable: a pure flying wing carries no structure that isn't generating lift. The XB-35 piston version flew in 1946; the YB-49 jet conversion flew in 1947. It could carry 10,000 pounds of bombs 2,000 miles. Its radar cross-section, accidentally, was so small that ground controllers repeatedly lost it on approach to Andrews Field in 1949 — a fact buried in test reports for thirty years until B-2 engineers rediscovered it.

Then came the cancellation. The official reasons were technical: yaw instability without a vertical tail made it a poor bombing platform with 1948 analog autopilots, and the bomb bay was too short for the Mark III nuclear weapon. The unofficial reason, according to Jack Northrop's 1979 deathbed interview recorded by Clete Roberts, was a meeting with Air Force Secretary Stuart Symington in 1948 where Symington allegedly demanded Northrop merge with Convair. Northrop refused. The contract went to the Convair B-36 instead. Symington became Convair's president shortly after leaving government.

Eleven flight-worthy airframes were ordered scrapped in 1953. Not mothballed. Scrapped. Jack Northrop, banned from his own factory's wing program, was reportedly told the destruction was to prevent "further development by parties who didn't deserve it." He didn't see a flying wing fly again until 1980, when classified B-2 program officials brought him into a SCIF and showed him the design. He died ten months later.

Why it works now:

• Fly-by-wire: The yaw instability that doomed the YB-49 is trivially solved by digital flight controls running at 100+ Hz. The B-2 proved this in 1989. A clean-sheet civilian flying wing — say, a 250-passenger transport — would have 30% lower fuel burn than a conventional tube-and-wing per Boeing's own X-48B blended-wing-body tests (2007-2012).
• Composite manufacturing: The YB-49's aluminum structure couldn't handle the bending loads of a true span-loaded wing. Carbon fiber laid up with automated fiber placement makes the 1948 structural problem disappear.
• Cabin pressurization: The cylindrical fuselage exists largely because cylinders handle pressure differentials elegantly. Modern multi-bubble composite pressure vessels (proven on the 787) make non-cylindrical pressurized cabins manufacturable.
• Boundary layer ingestion: Engines buried in the wing root, sucking in slow boundary-layer air, recover 4-8% additional efficiency — exactly the YB-49's engine placement.

NASA and Boeing's X-48 program demonstrated all of this between 2007 and 2013. The Aurora D8 "double bubble" took some of it forward. JetZero received $235 million in 2023 to build a full-scale BWB demonstrator by 2027. Eighty years after Glen Edwards died, the configuration is finally getting its industrial chance — but the lost decades were a choice, not a technical inevitability.