There is a photograph taken at the B-2's public rollout at Air Force Plant 42 in Palmdale, California, on November 22, 1988. The aircraft is sitting on the ramp, and the crowd is standing about 300 feet away — far enough that nobody can see it clearly. The Air Force positioned the crowd on purpose. The trailing edges of the B-2's wing were still classified, and a photograph showing the saw-tooth exhaust nozzle configuration or the exact shape of the flying wing's rear surface would have been a security violation.

This is the B-2 Spirit's story in miniature: a program so tightly held that even its public debut was classified down to the angle at which reporters were allowed to stand.

The B-2 entered service in 1997 and costs approximately $2.1 billion per aircraft in 1997 dollars — the most expensive military aircraft ever built. Twenty-one were produced. All remain in service with the 509th Bomb Wing at Whiteman Air Force Base, Missouri. Each one can carry 40,000 pounds of weapons, fly 6,000 miles unrefueled, and approach a defended target at an altitude and speed its radar cross-section makes practical.

The Flying Wing Problem

Northrop had been building flying wings since the 1940s. Jack Northrop's obsession with the aerodynamic elegance of a wing-only aircraft — no fuselage, no tail, all lift and no parasitic drag — produced the XB-35 and YB-49 bomber programs of the late 1940s. Both were cancelled. The YB-49, a jet-powered flying wing, was unstable in the yaw axis in ways that 1940s flight control systems could not reliably correct. One crashed on a test flight in 1948, killing the crew. The Air Force chose the Convair B-36 instead, and Northrop's flying wings went into storage.

By the early 1980s, two things had changed. Digital fly-by-wire flight control systems — the same technology that made Have Blue controllable — could now manage the stability problems that had killed the YB-49. And the Air Force's SENIOR CJ program (later designated ATB, for Advanced Technology Bomber) was looking for a long-range penetrating bomber with low observable characteristics. Northrop, now run by Thomas Jones and with a team that included veterans of the earlier flying wing programs, proposed combining the two revolutions: a flying wing with a radar cross-section small enough to penetrate defended Soviet airspace without a fighter escort.

Northrop won the ATB competition against Lockheed in October 1981. The contract was classified. The program was classified. The existence of the ATB program itself was acknowledged only in 1988 — seven years into development — and even then only because budget documents being reviewed by Congress had become impossible to fully redact.

Engineering the Invisible Wing

The B-2's low-observable design differs fundamentally from the F-117's approach. Where the F-117 used faceted flat surfaces to deflect radar energy, the B-2 uses carefully shaped curved surfaces that minimize radar return through a combination of geometry and radar-absorbent materials. The curved approach produces superior aerodynamic performance — the B-2 is a far better-flying aircraft than the F-117 — at the cost of substantially more complex manufacturing, since every curve must be precisely maintained to preserve its radar cross-section contribution.

The aircraft's radar cross-section is classified, but has been described in open-source literature as approximately comparable to a large bird — similar to the F-117 but achieved with an aircraft that is vastly larger (172-foot wingspan versus 43-foot wingspan). The design includes screened engine inlets to prevent radar energy from reflecting off the engine faces, a saw-tooth trailing edge configuration on the wing that scatters radar returns, and a cockpit designed to minimize the radar return from the crew compartment.

The B-2 has no vertical tail surfaces of any kind — directional control is provided entirely by differentially deflected control surfaces on the trailing edge of the wing and by split drag rudders on the wingtips. This configuration, combined with the aircraft's physical size and flying wing geometry, creates handling characteristics that the flight control computer manages continuously. Pilots describe it as exceptionally stable at cruise — the computer does the work.

"The computer flies the airplane," said Colonel Avery Thompson, one of the B-2's early operational pilots, in a 1998 Air Force interview. "Your job is to tell the computer what you want to do. It figures out how to do it."

The Cost and the Program's Near-Death

The B-2 program's cost trajectory became politically untenable in the late 1980s as the Cold War wound down. The original acquisition plan called for 132 aircraft. By 1990, with the Soviet Union visibly deteriorating, the Bush administration cut the program to 75 aircraft. In 1991, as the Soviet Union dissolved, it was cut again to 21 aircraft — the minimum the Air Force argued was operationally viable for a two-theater nuclear deterrent.

At 21 aircraft amortizing development costs of approximately $23 billion plus production costs, each B-2 cost roughly $2.1 billion. This figure — accurately calculated but somewhat misleading, since recurring production costs per additional aircraft would have been substantially lower — became the political anchor around which every subsequent debate about the program's value was organized. Members of Congress seeking to cut defense spending returned to the $2.1 billion figure repeatedly throughout the 1990s.

The Air Force's argument for retaining the program rested on capability: the B-2 could penetrate defended airspace that would cost fighter aircraft and conventional bombers prohibitive losses. During Operation ALLIED FORCE over Yugoslavia in 1999 — the B-2's combat debut — six Spirit aircraft flew roundtrip missions from Whiteman AFB in Missouri to Yugoslavia and back, each mission covering roughly 30 hours and requiring multiple air refuelings. They delivered more than 600 JDAM precision-guided bombs across 33 targets. None were detected by Yugoslav air defenses.

Stealth's Limits and the Maintenance Reality

The B-2's low-observable characteristics require intensive maintenance. The radar-absorbent coating on its surfaces is sensitive to moisture, temperature extremes, and physical abrasion. Each aircraft requires a climate-controlled hangar — one of the reasons Whiteman AFB, with its moderate continental climate, was selected as the primary basing location. A deployment to Diego Garcia in the Indian Ocean during Operation ENDURING FREEDOM required the construction of temporary climate-controlled shelters at significant cost.

The maintenance burden per flight hour is substantially higher than conventional bomber aircraft. The Air Force has not publicly disclosed exact figures, but analyses based on available budget data suggest the B-2's cost per flying hour is several times that of the B-52, which remains the Air Force's most cost-effective nuclear and conventional bomber on a per-hour basis.

This tradeoff — extraordinary capability at extraordinary cost and maintenance burden — defines the B-2's operational position. It exists because there are targets that no other aircraft can reach without unacceptable losses. Whether those targets will remain relevant, and whether the B-21 Raider (its classified successor, now in early production) represents a more sustainable path to the same capability, are questions the Air Force is living through in real time.

Northrop's flying wing flew. Jack Northrop, who died in 1981, reportedly saw classified drawings of what his company was building before he died. Whether he was told what it was for is not recorded.