How Lockheed’s Skunk Works Built the World’s First Stealth Aircraft With Off-the-Shelf Parts, a Soviet Math Formula, and Absolute Secrecy

Excluded from DARPA’s classified stealth competition, an uninvited Lockheed team bankrolled their own entry — and rewrote the rules of modern air warfare with a Soviet math paper the enemy had published themselves.

In the early spring of 1975, the traditional laws of aerospace engineering were quietly dismantled behind the thick, eavesdrop-proof walls of Building 82 in Burbank, California. This windowless, austere concrete blockhouse — situated in plain view of the busy runways at Burbank Municipal Airport — served as the nerve center for Lockheed’s Advanced Development Projects division, the organization known universally throughout the aerospace and defense communities as the “Skunk Works.” Inside this facility — originally an old bomber production hangar left over from the Second World War — an unprecedented and highly unconventional dialogue took place that would forever alter the trajectory of military aviation.

Dick Scherrer, a seasoned preliminary design engineer who had spent over a decade conducting flight research programs and wind tunnel tests, posed a fundamental, paradigm-shifting question to his colleague, Denys Overholser. Overholser was a thirty-six-year-old mathematician, electrical engineer, and radar specialist whom Lockheed had lured away from Boeing specifically for his advanced computer training and expertise in radomes and optical scattering. Scherrer’s inquiry was as simple as it was technologically daunting: “How do we shape something to make it invisible to radar?”

For decades, the aerospace industry had operated under a singular, unbreakable covenant: form must follow aerodynamics. Aircraft were meticulously designed with sleek, smoothly blended curves, teardrop fuselages, and swept surfaces to minimize atmospheric drag and maximize aerodynamic lift. But Overholser, viewing the problem entirely through the lens of electromagnetic wave propagation rather than fluid dynamics, offered a response that sounded mathematically elegant but aerodynamically absurd.

“Well, it’s simple,” Overholser replied. “You just make it out of flat surfaces, and you tilt those flat surfaces over, sweeping the edges away from the radar view angle, and that way you basically cause the energy to reflect away from the radar, thus limiting the magnitude of the return.”

This counterintuitive thesis — that a high-performance aircraft should be constructed entirely from flat, two-dimensional slabs intersecting at sharp angles rather than smooth aerodynamic curves — was the genesis of “faceting.” Overholser understood that the computational power of the 1970s was vastly insufficient to accurately model how radar waves scattered off complex, continuously curving surfaces. However, calculating the electromagnetic reflection off a flat, planar surface was a mathematically solvable problem. If an airframe could be constructed entirely of flat panels angled specifically to deflect electromagnetic waves away from the originating transmitter, the aircraft’s radar cross-section (RCS) would theoretically drop to near zero.

To the seasoned aerodynamicists at Skunk Works, Overholser’s faceted concept was a monstrosity that defied every known principle of stable flight. Yet it represented a monumental pivot in military aviation. For the first time in history, the external shape of a fixed-wing aircraft would be dictated entirely by radar engineering rather than by aerospace engineering. The quest for true radar invisibility had begun in earnest — set against a backdrop of strict secrecy so extreme that even a customized team coffee mug featuring a cartoon skunk and an angular nose cone had to be locked in a secure safe between breaks, because it inadvertently revealed too much about the highly classified geometry.

Why This Moment Changed Everything

The culmination of that 1975 conversation in Burbank was unequivocally validated on the night of January 17, 1991. As the opening phase of Operation Desert Storm commenced, a fleet of black, sharply angular attack aircraft slipped completely undetected into the heavily defended airspace over Baghdad. The Iraqi capital was protected by a formidable, fully integrated air defense system featuring advanced Soviet-supplied early warning radars, surface-to-air missiles (SAMs), and radar-directed anti-aircraft artillery (AAA). Yet the Lockheed F-117 Nighthawks operated with total impunity, utilizing laser-guided GBU-27 penetrator munitions to execute surgical strikes against critical command-and-control bunkers without suffering a single loss or even a hit during the entire 43-day aerial campaign. The anti-aircraft artillery over Baghdad fired blindly into the dark, reacting only to the explosions of the bombs long after the stealth aircraft had departed the airspace, leaving the defenders to shoot at empty sky.

This flawless combat record fundamentally altered the global military balance, proving conclusively that survivability in modern combat could be achieved through electromagnetic invisibility rather than sheer speed or maneuverability. But the profound historical irony of the F-117 Nighthawk lies in its intellectual origins. The theoretical mathematical framework that allowed American engineers to defeat the Soviet Union’s most advanced radar networks was not born in a classified Pentagon laboratory. It was authored by a Soviet scientist — and published openly in Moscow nearly a decade before the Skunk Works engineers ever read it.

The Lessons That Lit the Fuse

To comprehend the desperate urgency behind the American quest for stealth technology in the mid-1970s, one must examine the catastrophic aerial attrition rates suffered by the United States and its allies during the preceding years. The global air combat environment had grown increasingly hostile and lethal due to the rapid proliferation and sophisticated integration of Soviet-designed Integrated Air Defense Systems (IADS). During the Vietnam War, United States military aircraft faced a dense, overlapping network of SA-2 Guideline surface-to-air missiles and radar-directed anti-aircraft artillery. Throughout the conflict in Southeast Asia, the United States military lost nearly 10,000 aircraft, including 3,744 fixed-wing planes and 5,607 helicopters.

Despite the extensive deployment of electronic countermeasures (ECM), specialized “Wild Weasel” defense-suppression aircraft, and dedicated radar-jamming support platforms like the EB-66, combat losses remained unsustainably high. The lethality of these Soviet systems peaked during Operation Linebacker II in December 1972, when the U.S. Air Force dispatched massive waves of B-52 Stratofortresses against the heavily defended targets of Hanoi and Haiphong. Despite complex routing and heavy electronic jamming, North Vietnamese air defense forces fired approximately 1,240 SAMs over twelve days, resulting in the loss of fifteen B-52 bombers and eleven tactical aircraft. While the B-52 loss rate was statistically less than two percent of the total sorties flown, the psychological and strategic impact of losing fifteen of the nation’s premier strategic bombers to surface-to-air missiles proved that traditional “air armada” tactics were becoming dangerously vulnerable.

Less than a year later, the 1973 Yom Kippur War confirmed that the era of uncontested Western air superiority was definitively over. When Egypt and Syria launched a surprise, coordinated offensive against Israel, the Israeli Air Force — equipped with highly advanced American-built F-4 Phantoms and A-4 Skyhawks — encountered a devastatingly effective umbrella of mobile SA-6 Gainful SAMs and ZSU-23-4 radar-guided anti-aircraft cannons. Within the first few days, the Israelis ultimately lost over 100 combat aircraft to the Egyptian and Syrian missile batteries. Israeli pilots discovered that standard American ECM jamming pods were ineffective against the continuous-wave seeker heads of the SA-6; in some instances, Israeli commanders feared the jamming signals were acting as beacons, guiding the missiles directly to the aircraft and forcing pilots to fly without their electronic protection enabled.

AIR COMBAT LOSSES THAT DEFINED THE STEALTH IMPERATIVE

These staggering losses forced defense planners at the Pentagon to reconsider the fundamental principles of air warfare. The Defense Science Board concluded that U.S. aircraft would soon face an almost insurmountable challenge penetrating Soviet air defenses in a hypothetical Central European battlefield. The solution was not to build aircraft that could fly faster or higher, nor to rely solely on active electronic jamming that could be circumvented by frequency-hopping radars, but to prevent the enemy radar from detecting the aircraft in the first place.

The physics of this challenge lay in the radar equation. Radar detection range is directly proportional to the fourth root of the target’s RCS. Because of this fourth-root relationship, achieving a meaningful operational reduction in detection range requires an exponential reduction in the aircraft’s radar signature — specifically, to reduce a hostile radar’s detection range by a factor of ten, the aircraft’s RCS must be reduced by a massive factor of 10,000, equivalent to a reduction of 40 decibels. In 1974, recognizing this exponential signature-reduction requirement, the Defense Advanced Research Projects Agency (DARPA) launched a highly classified conceptual study program spearheaded by Ken Perko, a program manager in DARPA’s Tactical Technology Office (TTO). Under the direction of TTO Director Robert Moore and Director of Air Warfare Programs Chuck Myers, the initiative was code-named “Project Harvey” — an ironic nod to an invisible rabbit in the 1950 James Stewart comedy film, symbolizing the goal of creating an aircraft that could be present but entirely unseen.

The Competition Lockheed Wasn’t Invited To

Despite DARPA’s urgent need for revolutionary aircraft design, the most famous and successful aeronautical engineering division in the world was conspicuously left off the invitation list. In late 1974, Ken Perko sent official requests for white papers to five major aerospace contractors: Northrop, McDonnell Douglas, General Dynamics, Fairchild, and Grumman. Lockheed Aircraft Corporation, and by extension its renowned Skunk Works division, was completely excluded from the Project Harvey solicitation. The rationale for this oversight was rooted in bureaucratic perception rather than technical reality: Lockheed was not viewed as an active participant in tactical fighter development, having not produced a new fighter aircraft in nearly a decade, its last major tactical success being the F-104 Starfighter developed in the 1950s.

The responses from the invited firms varied dramatically. Fairchild and Grumman ultimately declined to participate. General Dynamics submitted a proposal leaning heavily on traditional active ECM and jamming, featuring very little substantive technical content on physical signature reduction through airframe shaping. However, Northrop and McDonnell Douglas submitted compelling proposals demonstrating a solid grasp of radar cross-section mitigation, and in late 1974, DARPA awarded both companies initial $100,000 Phase 1 study contracts.

The exclusion of Lockheed was particularly ironic given the Skunk Works’ unparalleled, albeit highly classified, history with signature reduction. Under its founding genius, Clarence L. “Kelly” Johnson, the Skunk Works had already integrated rudimentary stealth features into its high-altitude reconnaissance aircraft. The CIA’s A-12 Oxcart featured radar-absorbing composite structures, iron ferrite paints that dissipated electromagnetic energy as heat, and a flattened “cobra” chine design that dramatically reduced the massive Mach 3 aircraft’s RCS to roughly that of a small Piper Cub. The D-21 supersonic reconnaissance drone utilized even more advanced radar-absorbing materials to mask its signature. However, these “Black World” programs were tightly controlled by the CIA and the National Reconnaissance Office, and DARPA leadership simply did not possess the necessary clearances to know that Lockheed already held practical flight data on radar signature reduction.

The existence of Project Harvey, however, was not heavily classified in its earliest stages. Rumors of the competition soon reached Ed Martin, the Director for Science and Engineering at the Lockheed California Company, while networking during a visit to the Pentagon and Wright-Patterson Air Force Base. Recognizing the strategic threat of being locked out of the next generation of tactical aviation, Martin immediately returned to Burbank to brief the new leadership at the Skunk Works.

The Uninvited Guest

On January 17, 1975, Benjamin Robert Rich officially succeeded Kelly Johnson as the president of the Skunk Works. Rich was a brilliant thermodynamicist who had spent decades solving complex engine inlet and heat-dissipation problems for the U-2 and the SR-71 Blackbird. But the affable engineer inherited a division facing an existential crisis: the conclusion of the Vietnam War and a broader national downturn in defense spending had left the Lockheed Corporation bleeding cash, with corporate financial losses reaching a staggering $2 billion in 1975. Securing a foothold in DARPA’s new stealth program was not merely an opportunity for technological advancement — it was a matter of absolute corporate survival.

When Martin briefed Rich on the XST competition, they immediately recognized that Lockheed’s prior experience with the A-12 and D-21 provided a massive competitive advantage. The primary obstacle was the CIA’s strict classification protocols surrounding that historical data. Rich and Martin took the problem to Kelly Johnson, who — despite entering retirement — retained immense clout within the intelligence community. Johnson personally petitioned the CIA Director’s office and successfully obtained unprecedented written permission to share the highly classified radar-cross-section test results of the A-12 Oxcart and D-21 drone programs with DARPA.

There was no money available for a third competitor — yet Rich refused to walk away

Armed with this empirical data, Ben Rich traveled to Washington, D.C., to meet with Dr. George Heilmeier, the Director of DARPA, and Ken Perko. Rich presented Lockheed’s prior achievements in low-observable technology and formally requested entry into the competition. Heilmeier — a brilliant but notoriously demanding technocrat who managed DARPA through a rigorous framework known as the “Heilmeier Catechism” — was impressed by the data but offered a blunt administrative reality: DARPA’s budget for the XST study was completely exhausted. The available funds had already been disbursed to Northrop and McDonnell Douglas; there was simply no money left for a third competitor.

Undeterred, Rich demonstrated the audacious risk-taking that would come to define his tenure as the “Father of Stealth.” Drawing on his formidable negotiation skills, he convinced Heilmeier to allow Lockheed into the competition via a $1 pro forma government contract. Lockheed would receive no federal funding for its Phase 1 design studies; instead, Rich secured a commitment from Lockheed’s corporate management to fund the entire Phase 1 XST study out of pocket — with no government reimbursement guaranteed, a significant financial gamble for a company already bleeding cash. This uninvited, unfunded entry completely altered the trajectory of the competition. Freed from standard bureaucratic reporting requirements, the Skunk Works could iterate at an accelerated, highly autonomous pace.

The Soviet Blueprint

As the Skunk Works team began conceptualizing their XST entry in the spring of 1975, they quickly encountered a severe mathematical roadblock. Evaluating the radar signature of a proposed shape required calculating how electromagnetic waves scattered off its surfaces. While the established principles of “physical optics” — based on the Huygens-Fresnel principle and Kirchhoff’s diffraction formula — could adequately approximate radar scattering off the broad faces of large objects, these formulas broke down completely at the edges. Physical optics could not account for the perturbation of the electromagnetic field — the “edge waves” — that occurred when radar energy hit a sharp corner, a wing trailing edge, or a surface discontinuity. For a stealth aircraft relying on angular geometry to deflect radar, the inability to calculate edge diffraction meant the team was essentially flying blind.

The solution arrived from an astonishing and deeply ironic source: the Soviet Union. In 1971, the U.S. Air Force Systems Command’s Foreign Technology Division at Wright-Patterson Air Force Base had routinely translated a 40-page Russian technical monograph. The paper, titled “Method of Edge Waves in the Physical Theory of Diffraction,” had been published in Moscow in 1962 by Pyotr Ya. Ufimtsev, the chief scientist at the Moscow Institute of Radio Engineering. Ufimtsev had expanded upon the 19th-century electromagnetic equations of Scottish physicist James Clerk Maxwell and German physicist Arnold Sommerfeld to create the Physical Theory of Diffraction (PTD). He proved mathematically that the radar return of an object is determined primarily by its edge configuration rather than its overall size or volume, and his equations provided a reliable method for calculating the exact trajectory and intensity of electromagnetic waves diffracting off two-dimensional and three-dimensional geometric edges.

Because Ufimtsev’s research was viewed by the Soviet military and scientific establishment as purely abstract mathematics with no practical engineering application, it was never classified. It was published openly, completely ignored by Soviet aircraft designers who believed that building an aircraft entirely out of flat, faceted panels to satisfy a radar equation was aerodynamically impossible and fundamentally useless. At the Skunk Works, however, Denys Overholser recognized the translation as the “Rosetta Stone breakthrough for stealth technology.”

Overholser and Bill Schroeder — a brilliant veteran mathematician who had mentored Overholser in optical scattering and resolved the equations for calculating reflections from triangular flat panels — realized that Ufimtsev’s PTD equations were the missing link. Because 1970s computers lacked the processing power to analyze complex, continuously curved surfaces, Ufimtsev’s formulas allowed the engineers to input flat, two-dimensional triangles into a computer and accurately predict their combined RCS, including the complex edge scattering. Working relentlessly over five punishing weeks, Overholser, Schroeder, and a small software team translated Ufimtsev’s calculus into a groundbreaking computer program named Echo 1. It was the first computational tool in history capable of analyzing a geometric shape and accurately predicting its radar signature before a single piece of metal was cut or a physical model was built.

The Hopeless Diamond

With Echo 1 operational, the Skunk Works team began iterating through various configurations. Because the software could only compute radar returns from flat, triangular panels, Overholser’s design philosophy strictly forbade the use of curves — even in the airfoil — because their radar reflectivity could not be modeled or predicted by the computer. The resulting design was a highly swept, faceted delta wing composed entirely of flat surfaces, with steeply angled fuselage panels and vertical stabilizers canted sharply inward toward the aircraft’s centerline.

When the Skunk Works aerodynamicists, led by Dick Cantrell, reviewed the preliminary drawings produced by Dick Scherrer, they were appalled. The airfoil section consisted of a series of straight lines intersecting at sharp angles, guaranteeing that airflow would separate violently from the surface, creating massive turbulence and drag. From a traditional flight dynamics perspective, the shape was inherently unstable across all three axes of flight — pitch, roll, and yaw. The aerodynamicists bestowed a derisive nickname upon it: “The Hopeless Diamond.” Even Kelly Johnson detested the faceted approach. “Our old D-21 drone has a lower radar cross-section than that goddamn diamond,” Johnson complained to Ben Rich.

To settle the debate, Lockheed constructed 1/3-scale wooden models of the Hopeless Diamond. In June 1975, one model was coated in metal foil and taken to Lockheed’s anechoic chamber at Rye Canyon, and subsequently to the McDonnell Douglas Gray Butte Radar Test Range in the Mojave Desert. The radar returns were staggeringly successful. Echo 1’s predictions matched reality perfectly: the Hopeless Diamond registered an RCS that defied belief — orders of magnitude smaller than even the already-stealthy D-21 drone. Ben Rich won a quarter from Kelly Johnson on a bet over the RCS readings. As Overholser noted regarding the validation of his faceted approach: “I went from being regarded as the village idiot to the village expert.”

Despite its radar-evading perfection, the pure diamond shape required significant modification to achieve sustainable flight. Senior Lead Aircraft Designer Kenneth Watson and the aerodynamics team refined the shape by thinning the sections outboard of the engine inlets to improve the lift-to-drag ratio, stretching the diamond into a highly swept, “notched-out delta” planform. To ensure that radar return “spikes” were pushed safely outside the frontal sector, the leading-edge sweep was set at a severe 72.5 degrees, while the trailing edge sweep was increased to 48 degrees. When Lockheed test pilot Dave Ferguson first viewed the angular, slab-sided mock-up, he skeptically asked Dick Cantrell how airframe ice encrustation might affect the aircraft’s aerodynamics. Cantrell’s dry response underscored the absolute absurdity of the shape: “Probably improve it.”

Phase 1 — The Competition

By late 1975, the DARPA XST competition had narrowed to a head-to-head battle between Lockheed and Northrop, McDonnell Douglas having fallen out of contention. Both companies were awarded $1.5 million contracts to proceed with Phase 1, which required building full-scale, radar-reflective wooden models for a decisive “pole-off” test. In February and March 1976, the competing models were transported under heavy secrecy to the U.S. Air Force RATSCAT (Radar Target Scatter) facility at the White Sands Missile Range in New Mexico.

The two engineering philosophies presented a stark contrast. Lacking Lockheed’s Echo 1 program and Ufimtsev’s diffraction calculations, the Northrop team — with radar analysis support from Hughes Aircraft Company — had pursued a blended diamond/delta shape officially designated as their XST submission, optimized heavily for frontal RCS at the expense of the side and rear quadrants. Lockheed’s faceted model, globally optimized by Echo 1, adhered strictly to DARPA’s requirement to suppress RCS across all threat angles.

When the radar operators at White Sands fired up their emitters to measure the Lockheed model mounted on a massive pylon, the signature was so infinitesimally small that the operators initially thought the model had fallen off the pole — the radar was registering a higher return from the specialized mounting pole itself than from the aircraft. To complete the testing, Lockheed was forced to hastily design and construct a new, lower-profile pole just to accurately measure their model — the Lockheed signature was that impossibly small. Even minor surface contamination perceptibly raised the RCS, requiring constant maintenance by range technicians.

In April 1976, DARPA declared Lockheed the winner of the XST competition. Northrop’s blended-body approach had produced a much higher side-hemisphere radar cross-section, but DARPA was so impressed by the team’s innovations that they urged Northrop to remain together — a decision that evolved through the Tacit Blue demonstrator into the B-2 Spirit stealth bomber. For Lockheed, the program leaped from theoretical study into physical development, classified as Top Secret/Sensitive Compartmented Information (TS/SCI) and transitioned to the Phase 2 flight demonstrator phase under the code name: Have Blue.

Have Blue: Where Theory Became Steel

In the secure confines of Building 82, Lockheed commenced construction of two 60-percent scale flyable technology demonstrators, carrying the factory serial numbers HB1001 and HB1002. The objective was to validate the low-observable signatures in actual flight conditions and prove that an aerodynamically unstable, faceted aircraft could be reliably controlled.

Because speed and budget were critical factors, Ben Rich mandated that the Have Blue airframes utilize as many “off-the-shelf” parts from existing aircraft as possible — a pragmatic strategy that allowed the Skunk Works to focus their engineering capital purely on the stealth geometry rather than reinventing standard aviation hardware.

HAVE BLUE: OFF-THE-SHELF COMPONENT SOURCES

The inclusion of the F-16’s fly-by-wire (FBW) system was arguably the most critical engineering decision of the program. Because the faceted Have Blue design completely lacked aerodynamic stability — relaxed static stability in the extreme — it was physically impossible for a human pilot to manually control the aircraft. The quad-redundant analog FBW system, operating without any mechanical backup, relied on constant, microsecond computer adjustments to manipulate the flight control surfaces and keep the aircraft airborne. In addition to radar evasion, Have Blue incorporated radical infrared signature reduction techniques: the engine tailpipes were transitioned from round ducts to a narrow, flattened 17-to-1 slot convergent nozzle, while bypass air was routed over the tailpipe to cool the aft structure — together dramatically reducing the aircraft’s infrared signature. A separate large two-position pitch-control flap on the trailing edge of the exhaust deck, called the ‘platypus,’ automatically deflected downward whenever the aircraft exceeded 13 degrees angle-of-attack, providing critical nose-down pitch authority.

Under absolute secrecy, the dismantled Have Blue prototypes were loaded into C-5A Galaxy transports in the dead of night and flown to Area 51 (Groom Lake) in the Nevada desert. On December 1, 1977, Lockheed test pilot Bill Park took HB1001 into the air for the first time. HB1001, which sported a bizarre, multi-colored camouflage pattern applied at Burbank, was used primarily to test the precarious aerodynamics of the faceted shape. Seven months later, on July 20, 1978, Air Force test pilot Lt-Col Ken Dyson flew HB1002, dedicated entirely to gathering RCS data against active ground-based and airborne tracking radars.

The flight tests confirmed the impossible: the radar cross-section in the air perfectly matched the minute signatures predicted by Echo 1. Although both Have Blue aircraft were ultimately lost in severe mechanical crashes — HB1002 crashing on July 11, 1979, following an engine fire after 52 sorties — the data collected was so staggeringly successful that the Air Force did not hesitate. Even before Have Blue had finished its testing regime, the Pentagon authorized full-scale engineering development of the operational stealth fighter under the Top Secret code name Senior Trend — the aircraft that would become the F-117 Nighthawk.

The transition to Senior Trend culminated in an operational force stationed at the covert Tonopah Test Range Airport, organized as the 4450th Tactical Group, where pilots trained exclusively at night. Equipped with the Infrared Acquisition and Designation System (IRADS) — which utilized Texas Instruments Forward Looking Infrared (FLIR) and Downward Looking Infrared (DLIR) — the F-117 was transformed into a deadly accurate platform capable of delivering 2,000-pound laser-guided GBU-27 Paveway III penetrator bombs against the most hardened targets in the world.

The Paper Moscow Never Counted

On January 17, 1991, as air raid sirens wailed across Baghdad, the true legacy of the Skunk Works’ frantic 1975 engineering sprint was unleashed. The F-117 Nighthawk — the operational descendant of the Hopeless Diamond — dismantled the Iraqi air defense grid with eerie precision. It was a ghost in the machine, rendering billions of dollars of Soviet-supplied early warning radars and SAM batteries utterly useless.

Yet the ultimate irony of this Cold War victory lay in its theoretical origins. The United States had achieved total air dominance by exploiting a mathematical formula written by Pyotr Ufimtsev, a scientist whose work was deemed so militarily insignificant by the Soviet Union that it was published openly to the world. When Ufimtsev later emigrated to the United States and learned of his inadvertent contribution, he expressed profound surprise. Moscow had possessed the blueprint for radar invisibility for a decade but failed to recognize its value because their aeronautical culture could not envision an aircraft built from flat, un-aerodynamic planes.

Lockheed’s genius was not in discovering the physics of diffraction. It was in possessing the audacity to build an airplane that looked like a broken diamond just to prove the math was right.