TR-3B Black Triangle UFO: The Engineering Behind the Mystery

October 4, 2026

For aerospace researcher Jarod Yates, the search for an engine that could explain the alleged TR-3B black triangle UFO led into the papers of fusion physicist Friedwardt Winterberg. Working with Tim Ventura and Winterberg’s family, Yates examined correspondence, research proposals, calculations, and handwritten notes, looking for connections between a remarkable scientific career and an alleged aircraft capable of extraordinary flight. There wasn’t a confirmed TR-3B blueprint waiting in the archive. Instead, the search revealed an intersection of UFO testimony, Cold War research, and technologies that might explain portions of the mystery. Pulsed fusion, superconducting magnets, and relativistic plasma mirrors all have scientific histories. Whether they add up to the machine witnesses describe is another question entirely.

Before the Black Triangle Had a Name

A black triangle is a description. The TR-3B is an explanation. Witnesses can report three lights, a dark underside, or an object that seems to hover without knowing its identity. Calling it a TR-3B adds a military designation and technological history. The distinction matters because the sightings aren’t uniform. By August 2004, the National Institute for Discovery Science, or NIDS, said it had accumulated almost 400 reports involving triangular, boomerang-shaped, and wedge-shaped objects—not an inventory of identical aircraft.

Some accounts describe thin, nearly equilateral triangles with lights near their corners. Others describe elongated wedges, broad chevrons, thick bodies, or arrangements of lights without a clearly visible connecting hull. The reports also vary in their lighting: three white sources, central red illumination, colored perimeter lights, or illuminated rear surfaces. Those differences could reflect viewing conditions, different objects, or mistaken interpretations. They shouldn’t disappear simply because a familiar illustration provides a more consistent-looking machine.

The combination that interests Yates is more specific than triangular geometry: apparent large size, slow movement, little sound, and occasional rapid departure. He’s particularly interested in whether changes in a central glow reveal changes in an engine’s operating state. That makes lighting and sound potential engineering clues, but only after their observational limits are understood. A bright corner isn’t necessarily a thruster, and an observer who hears nothing hasn’t measured an aircraft’s acoustic output.

Distance is the first missing specification. An object spanning 10 degrees of the sky would be approximately 17.5 meters wide at 100 meters, or 175 meters wide at a kilometer. Both produce the same apparent size. Unknown range similarly prevents angular motion from becoming a reliable speed measurement. The challenge is to preserve what witnesses actually observed while resisting the temptation to supply the dimensions and performance of the aircraft they might have seen.

Belgium, Phoenix, and the Making of a UFO Icon

The Belgian wave gave the modern triangle story an unusually substantial public record. Beginning on November 29, 1989, reports around Eupen and elsewhere attracted police, civilian investigators, and the Belgian Air Force. Retired Major General Wilfried De Brouwer, who had been chief of operations in the Air Staff, later recalled approximately 140 sightings reported during the opening evening in eastern Belgium. He described accounts of large triangular forms with powerful lights, slow movement, and little perceived noise. Yates interprets the repeated appearances as suggestive of an organized operation.

Two F-16s investigated radar contacts during the night of March 30–31, 1990, without visually identifying an unusual craft. De Brouwer later acknowledged that weather-related interference and false radar returns complicated the interpretation of extraordinary changes in speed and altitude. Another complication emerged in 2011, when the photographer behind the famous Petit-Rechain triangle image admitted that it showed a model. Those weaknesses don’t explain every Belgian observation, but they separate the continuing witness mystery from its most widely circulated supposed proofs.

Phoenix presents a different problem: several observations have been compressed into one event. On March 13, 1997, witnesses reported a moving V-shaped formation earlier in the evening, while the later, widely filmed lights around 10 p.m. became associated with military illumination flares. Accounts of the earlier formation diverge. Some observers described one enormous structure; amateur astronomer Mitch Stanley reported seeing separate aircraft through his telescope, according to the Phoenix New Times. A broad chevron, a formation of aircraft, and the familiar equilateral TR-3B aren’t interchangeable explanations.

Popular culture offered a way to connect the dots. The X-Files episode “Deep Throat,” broadcast in 1993, placed a triangular craft inside a secret military aerospace narrative. It didn’t invent the older motif, but it made the human-built-UFO interpretation memorable. Reports also outlasted the decade. In January 2000, police officers across several Illinois communities began describing unusual lights and a large object moving through the predawn sky. Their encounter brought the mystery closer to the ground—and left a record more immediate than a retrospective reconstruction.

What Witnesses Actually Saw

Around 4 a.m. on January 5, 2000, Melvern Noll stopped to check his miniature golf business in Highland, Illinois, after a delivery run. According to investigator David Marler’s March 2000 report, Noll initially noticed what looked like a bright star. Watching again, he realized it was moving. He described a large rectangular body, illuminated openings arranged on two levels, and dim red lights underneath. He compared its height to a two-story house. He drove to the police station, hoping another observer could confirm what he’d seen.

In nearby Lebanon, Officer Ed Barton responded to the resulting call. Marler recorded his account of brilliant lights resolving into an elongated triangle with white lights near its corners and a smaller flashing red light underneath. Barton stopped and switched off his patrol car’s lights and radio to listen. He reported no sound, stars disappearing behind the object, a level change of heading without banking, and a sudden increase in speed. Officer David Martin in Shiloh subsequently reported an unusually broad triangular or arrowhead-shaped object.

Officer Craig Stevens had heard the radio traffic and drove to Liederkranz Park in Millstadt. His written report places his observation around 4:28 a.m. He described slow movement at an estimated 500–1,000 feet and wrote, “I could only hear a very low decibel buzzing sound.” Stevens retrieved a Polaroid camera from his trunk, but it was cold and operated poorly in the approximately 18–20°F weather. The resulting image showed indistinct lights rather than a clear hull. His report also records a later observation from Dupo at a much greater apparent altitude, and his return to the station to sketch the object and document the encounter.

The record is compelling partly because it preserves those differences. This wasn’t a set of isolated accounts assembled years later: officers were exchanging information during the event. But their communication also means their expectations weren’t independent, and the reported dimensions, altitude, and shape didn’t become instrument measurements simply because several departments participated. The Illinois case supplies something more useful than a perfect match to a rendering: a specific set of observations that any proposed aircraft must explain. How could something appear so large, move so slowly, and produce so little sound?

How the TR-3B Acquired an Engine

The names attached to the alleged aircraft have their own history. Aviation Week reporting in June 1991 discussed a purported TR-3A Black Manta reconnaissance aircraft. Edgar Fouché’s later account described a different machine, the TR-3B, which he called Astra. Modern references to a “TR-3B Black Manta” blend distinguishable stories. Fouché’s widely circulated 1998 presentation is an identifiable source for the exotic version and its proposed internal machinery.

Fouché described a nuclear-powered triangular vehicle with a mercury-based medium circulating inside a magnetic-field disruptor. The numbers repeated in Yates’s presentations are striking: approximately 250,000 atmospheres, 150 kelvin, and 50,000–60,000 revolutions per minute. Fouché claimed an approximately 89 percent reduction in gravitational effects or effective mass, with separate propulsion units moving the lightened vehicle. His account also invoked Mach 9 performance, a silver-blue corona, and vehicles described at different scales, including hundreds of feet across.

The distinction between weight and inertia is crucial to that story. Reducing weight would lower the force needed to support an aircraft. Reducing inertia would change its resistance to acceleration. Protecting occupants during a violent maneuver requires an explanation of the forces acting on their bodies. Fouché’s language moves among these effects, but they aren’t interchangeable accomplishments. An engine that solves hovering hasn’t necessarily solved the high-speed departure Barton described.

Reverse engineering adds another proposition. A classified aircraft could exist without exotic propulsion; exotic propulsion wouldn’t by itself establish extraterrestrial ancestry. Yates’s work instead asks how far recognizable human research might take the explanation. The question becomes whether a power source, field generator, propulsion mechanism, and control system can be assembled into a coherent machine—not whether several impressive-sounding technologies can be placed inside the same triangular outline.

Following the Trail into Winterberg’s Archive

Yates approaches the mystery as a technician trying to understand a machine. In his Alt Propulsion Engineering Conference (APEC) presentation, he described 11 years in the Air Force and Air National Guard working on F-15 integrated electronics, an associate’s degree in integrated electronic systems, and further studies in electrical engineering and physics. He subsequently worked in aerospace as a mechanical-electrical engineering technician. He also acknowledged the limits of his expertise in advanced theory. His motivation was less qualified: “I’m going to see a warp drive before I die.”

The investigation eventually brought Yates and Ventura into Winterberg’s family archive. Their interview describes approximately 50–100 chronological binders, another 30–40 topical binders, and more than 30 boxes of papers and related material. They found scientific correspondence, requests for reprints, teaching material, and handwritten calculations. For Ventura, the annotations mattered as much as the publications. A photocopied article might exist elsewhere, but Winterberg’s equations in the margin could preserve an unanswered question or the beginning of another proposal.

The Paperclip connection has independent support, although its implications require care. A 2009 University of Nevada, Reno profile states that the government invited Winterberg to America in 1959 under the same program that brought Wernher von Braun from Europe. Yates identifies that program as Project Paperclip. The university account supports the recruitment history, not a TR-3B assignment. Nor does recruitment justify the “Nazi scientist” label used in the interview’s title: Winterberg, born in 1929, was a teenager when the war ended.

On the aircraft connection, Yates’s most important statement was an admission: “We didn’t find a single literal smoking gun.” He believed the correspondence pointed toward further leads, but didn’t report finding a document naming Winterberg as the triangle’s designer. The archive nevertheless offered a substantial subject in its own right: a physicist’s sustained attempt to solve the energy and propulsion problems of spaceflight. Following that work provides a firmer starting point than guessing what an absent classified file might contain.

The Real Physics Behind the Search

Winterberg began his fusion research in 1954, earned his doctorate under Werner Heisenberg, and received the Hermann Oberth Gold Medal in 1979 for nuclear-propulsion work, according to the university profile. He belongs in the wider history of fusion-powered spaceflight, but not as the sole creator of Project Daedalus. The British Interplanetary Society identifies that interstellar probe as a 13-member design study conducted from 1973 to 1978. The distinction preserves both Winterberg’s contribution to the field and the separate history of the spacecraft study.

A particularly relevant comparison is the 1992 paper “Space Propulsion by Fusion in a Magnetic Dipole,” coauthored by Edward Teller, Alexander Glass, Kenneth Fowler, Akira Hasegawa, and John Santarius. Its conceptual system had a specific power around 1 kilowatt per kilogram, with calculated possibilities including Mars missions in approximately 90 days and Jupiter missions in a year. Those weren’t completed flights, but they show the ambition of identifiable fusion-propulsion research. A magnetic dipole, however, isn’t a gravitational dipole, and a fusion rocket isn’t automatically an atmospheric hovercraft.

Yates also follows the Strategic Defense Initiative, announced by President Ronald Reagan on March 23, 1983. He sees its work on lasers, pulsed energy, and demanding space systems as a possible technological ancestor of the triangle, particularly the nuclear-pumped X-ray laser concepts associated with Excalibur. The important engineering distinction is between producing an extreme event and operating a reusable system. An aircraft needs repeated pulses, durable components, controllable energy delivery, and a way to dispose of heat throughout a mission.

Winterberg’s Planck-aether hypothesis and negative-mass proposals introduce a further theoretical layer. Yates sees them as possible routes beyond conventional propulsion, but their predictions must be assessed separately from his fusion research. A fusion rocket releases energy and transfers momentum through its propulsion system. Altering gravitational interaction or inertia requires another mechanism. A reactor could supply that mechanism with power; the reactor’s existence wouldn’t explain the effect itself.

Inside the Alleged Mercury Ring

Fouché’s numbers become more revealing when treated as operating requirements. The claimed 250,000 atmospheres correspond to approximately 25.3 gigapascals, while 150 kelvin is approximately −123°C, or −190°F. Composition and ionization state remain essential missing details. A nonequilibrium plasma can contain particle populations at different temperatures, and the temperature of a cold component doesn’t describe the entire medium. “Mercury-based” isn’t enough information to determine what material could operate inside the proposed ring.

Consider a hypothetical circulation radius of 30 meters. At 50,000 rpm, a material element traveling around that ring would move at approximately 157 kilometers per second, with centripetal acceleration near 84 million g. These are consequences of the stated assumptions, not measurements of a TR-3B. If “rpm” instead describes a rotating electromagnetic pattern or plasma mode, it means something different. The distinction determines whether the problem is moving enormous amounts of material or controlling the propagation of a field disturbance.

Magnetic pressure supplies another scale estimate. Using p = B²/(2μ₀), a magnetic pressure comparable to 25.3 gigapascals requires approximately 252 tesla. That corresponds to a magnetic energy density near 25 gigajoules per cubic meter. Pressure balance is only the beginning: the system would still need a stable configuration, suitable currents, and structures able to withstand the forces. These figures describe the magnitude of the confinement problem rather than a complete solution to it.

The circulation claim also converts to approximately 833–1,000 revolutions per second. Yates notices a possible relationship to reports of humming, but a revolution rate doesn’t automatically become an audible tone. A connection would need a mechanism that couples the system’s motion to the air, followed by an actual spectrum to compare. Stevens’s faint buzzing is an observation worth preserving. Assigning it to a particular machine frequency requires another measurement.

Plasma Mirrors and the Power of a Short Pulse

Yates’s later hypothesis goes well beyond the mercury ring. He proposes structuring a plasma, driving it with fields and laser pulses, and using moving plasma surfaces to compress and shift radiation toward much greater field strengths. From there, he connects particle-pair production to gravitational effects. The sequence makes his proposal more testable: each stage has an input, an expected output, and a question about whether the next stage can actually use it.

One genuine connection is the 2003 paper “Light Intensification towards the Schwinger Limit,” by Sergei Bulanov, Timur Esirkepov, and Toshiki Tajima. It proposes using a relativistically moving plasma structure as a reflecting and focusing surface, compressing an incident pulse and shifting its frequency upward. The driven plasma transfers energy into the reflected radiation. That’s a specific route toward extreme-field experiments, not simply a claim that shining a laser into charged gas makes it more powerful.

Peak power and continuous output are very different quantities. An illustrative 1-petawatt pulse lasting 30 femtoseconds contains 30 joules. At an assumed 1,000 pulses per second, that would be 30 kilowatts of average optical power, with greater electrical demand because of inefficiencies. Lawrence Livermore achieved its first petawatt laser in 1996, but that early system could fire only once every few hours. Repetition rate, not just the headline pulse power, matters enormously to an aircraft expected to hover continuously.

Twisting light offers another possible control mechanism. Optical orbital angular momentum concerns a beam’s spatial phase structure, and it can be transferred to matter. That makes it relevant to manipulating particles and plasma, but it doesn’t create extra energy. Likewise, changing an optical medium’s refractive behavior isn’t automatically changing gravitational spacetime curvature. The value of these effects lies in what they can predictably do to a beam or a plasma—not in treating every unusual optical behavior as a miniature warp drive.

Crossing the Schwinger Threshold

The characteristic Schwinger electric field is approximately 1.3 × 10¹⁸ volts per meter, associated with nonperturbative electron–positron pair production. Its corresponding magnetic-field scale is roughly 4.4 billion tesla, although the two field configurations aren’t interchangeable pair-production mechanisms. The units matter: volts describe a potential difference, while volts per meter describe field strength. Geometry, duration, and interaction volume also affect an experiment. The threshold isn’t a universal percentage-conversion dial for producing antimatter.

SLAC’s E144 experiment supplies a concrete measurement. In results published in 1997, researchers reported 106 ± 14 positrons above background from interactions involving 46.6-gigaelectronvolt electrons and 527-nanometer laser pulses. Energetic electrons first generated high-energy photons through scattering; those photons then interacted with multiple laser photons to produce electron–positron pairs. This demonstrated remarkable particle physics, but it wasn’t a measurement of reduced inertia or the same process as pair creation in a strong static field in empty space.

Making the particles costs energy. An electron and positron require at least 1.022 megaelectronvolts of combined rest energy, before their motion is included. The input comes from the interacting fields and particles. Both members of the pair have positive rest energy; a positron isn’t a packet of negative gravitational mass. Creating antimatter could supply an energy-storage medium, but it doesn’t bypass the requirement for an energy source.

Yates’s largest proposed leap comes next: oscillating the particle populations to generate gravitational effects capable of supporting or accelerating a craft. The missing link is a quantitative relationship between those controlled conditions and a measurable force. How much lift should a given field, frequency, and geometry produce? Does changing those parameters change the force as predicted? An experiment answering those questions would connect extreme-field physics to the aircraft problem far more directly than another resemblance between a plasma image and a UFO’s glow.

Can Superconductors Close the Gap?

Superconductivity is central to Yates’s attempt to make the field requirements attainable. His mercury comparison draws on real solid compounds: published work on Hg-1223 reported a superconducting transition temperature increasing from approximately 133 kelvin at ambient pressure to 164 kelvin at 40 gigapascals. Those conditions understandably attracted his attention. But a mercury-containing crystalline ceramic isn’t the same medium as circulating mercury plasma. Its particular structure and composition are part of the result, not details that can be discarded when applying it to an engine.

The proposed shortcut also requires distinguishing Cooper pairs, electron–hole pairs, and electron–positron pairs. A 2023 Nature Physics paper, “Mesoscopic Klein-Schwinger Effect in Graphene,” reported a condensed-matter realization involving electron–hole creation. Such experiments reproduce aspects of relativistic physics at accessible scales. An electronic hole isn’t a free positron, however, and the experiment doesn’t establish that superconductivity lowers the actual vacuum pair-production requirement by the enormous factor Yates suggests.

Comparisons with Eugene Podkletnov involve separate experiments as well. His 1997 rotating-superconductor report described weight changes reaching approximately 2 percent during deceleration, with the disk below 70 kelvin. His 2001 paper with Giovanni Modanese instead described high-voltage discharges, surface potentials above 1 megavolt, and a claimed directed impulse. These are different apparatuses and reported effects. A common theory would need to predict both, rather than combine the voltage from one experiment with the percentage from another.

A useful next step would therefore be smaller than an aircraft: an experiment predicting and measuring a repeatable force under specified conditions. Controls could separate any proposed inertial or gravitational effect from current interacting with external fields, ion thrust, vibration, and thermal drift. Superconductors, plasma sources, and lasers would then have defined roles in a test—not simply places in a speculative parts list. That would give the broader TR-3B hypothesis something it currently lacks: a propulsion result that another laboratory could try to reproduce.

What the Navy Patents Actually Describe

Salvatore Cezar Pais’s Navy-assigned “Craft Using an Inertial Mass Reduction Device” provides an actual document to examine. Filed in April 2016 and granted as US10144532B2 in December 2018, it explicitly permits a “lenticular triangle/delta wing configuration.” The triangle comparison isn’t purely an internet invention. The document does not, however, identify its proposed craft as Fouché’s TR-3B or connect it to the Belgian sightings.

Pais describes cavity walls, electrically charged structures, and electromagnetic excitation intended to produce a locally polarized vacuum state. His proposed emitters span 300 megahertz to 300 gigahertz. The goal is to reduce inertial resistance through controlled electromagnetic conditions around the vehicle. Cavities and emitters are recognizable engineering elements; the unusual claim concerns the effect their arrangement is supposed to produce. Establishing that effect requires performance measurements beyond the patent’s description.

John St. Clair’s “Triangular Spacecraft” is a separate proposal. Filed in December 2004 and published as US20060145019A1 in July 2006, it describes an equilateral triangular hull and electrostatic and electromagnetic features intended to produce propulsion through spacetime effects. St. Clair is the inventor, not Pais, and the application isn’t the Navy’s filing. Its operating concept also differs from Fouché’s circulating-plasma disruptor.

The useful comparison is therefore architectural: what creates the fields, how they interact, and where a net force is predicted to arise. Similar outlines don’t answer those questions. Nor can a later filing independently establish what flew decades earlier. A documented connection would require test records, attributable correspondence, or identifiable hardware linking a particular proposal to a particular program. Until then, the patents are technical hypotheses to examine alongside the sightings.

The Whole-Aircraft Engineering Test

Return to the slow-moving object Stevens described. Suppose, for illustration, that it were a 100-metric-ton aircraft using ordinary aerodynamic lift, with 1,000 square meters of effective lifting-flow area in sea-level air. It would need approximately 0.98 meganewtons of upward force. Ideal actuator-disk theory gives an induced hover-power requirement near 20 megawatts and airflow through the disk around 20 meters per second, before real losses. These assumed dimensions don’t identify the Illinois object. They show why conventional hovering would have consequences beyond the aircraft: a substantial amount of air must move.

Now consider an extreme departure, again as a calculation rather than a reconstruction. Accelerating that unreduced mass from rest to 1 kilometer per second at 40 g takes approximately 2.55 seconds and adds 50 gigajoules of kinetic energy. The average mechanical power is nearly 20 gigawatts before aerodynamic losses. An actual reduction of inertia could change the calculation; reduced apparent weight alone wouldn’t. Neither effect automatically eliminates the shocks and heating associated with high-speed atmospheric flight.

Heat is another unavoidable accounting problem. A hypothetical plant supplying 500 megawatts of useful output at 40 percent efficiency would reject 750 megawatts of waste heat. Radiation alone would require approximately 15,000 square meters of emitting surface at 1,000 kelvin and emissivity 0.9, facing a cold background. Atmospheric cooling, hot exhaust, or temporary heat storage could change the design, but each brings consequences. An engine explanation must account for energy that doesn’t become useful output as carefully as the energy that does.

Unfamiliar propulsion can still operate within established physics. In 2018, MIT researchers flew a five-meter-wingspan aircraft using approximately 40,000 volts to generate ionic wind, without moving propulsion parts. The system pushed air rather than eliminating momentum transfer. Buoyancy, distributed fans, and aerodynamic lift likewise address different parts of the problem. The decisive question is whether one particular design can explain the entire encounter: apparent size, low speed, maneuvering, lighting, little sound, and whatever heat or airflow witnesses should have noticed.

What Would Turn the Hypothesis into Evidence?

Yates’s most striking visual argument concerns footage in which a triangle’s central luminous region appears to change shape before departure. He interprets the changes as possible plasma resonance modes, while acknowledging that shutter behavior and image saturation could affect the appearance. Original recordings, camera characteristics, accurate timing, and known range would make that interpretation testable. Without them, the footage provides an idea for an experiment rather than a measurement of an engine.

The Illinois encounter points toward what better evidence would look like. The officers’ reports preserved locations, an observation sequence, different descriptions, and an attempt to photograph the object. Synchronized cameras at separated sites could add distance and dimensions. Calibrated sound recordings could turn “quiet” into a quantitative limit, while spectroscopy and thermal imaging could test a plasma or power-source explanation. Those measurements would distinguish mechanisms more effectively than assigning every light to a component in advance.

The archive offers another route. A dated drawing with secure provenance, a test report, or correspondence explicitly connecting a named subsystem to a named program would strengthen Yates’s historical hypothesis. Even without that discovery, Winterberg’s working papers have value. They preserve attempts to solve difficult problems, including approaches that were revised, abandoned, or never developed. A scientist’s unfinished question can be worth recovering without assuming that someone secretly built the answer.

The black-triangle story ultimately connects two kinds of uncertainty: what witnesses encountered and what engineers might be able to construct. Yates has proposed a path between them, drawing on research far more substantial than a mysterious silhouette alone. Its crucial links remain to be demonstrated. Stevens returned to his station with a sketch and a report; Yates and Ventura left the archive with leads rather than a blueprint. The next breakthrough may be a measurement—or a document—that finally connects the shape in the sky to an engine that can be tested.