At Cal State Fullerton, a revolutionary propulsion program aims at the unthinkable: thrust without propellant, a steady shove drawn from the physics of inertia. For over thirty years, James F. Woodward led that effort, assembling the tools and techniques to test the claim. Since his death on August 9, 2025, the lab has pressed on under Curtis Horn, Michelle Broyles, and Hal Fearn, who are refining the devices to prove it one careful run at a time.

The Mach effect premise: inertia as a handle

Woodward’s starting point was a century-old itch in gravitation: Mach’s principle, the notion that an object’s resistance to acceleration (its inertia) might emerge from the gravitational influence of the universe’s distant mass. If that’s true in a concrete, calculable way, then—just maybe—tiny, rapid changes to a device’s internal energy while it’s being pushed and pulled could produce transient mass fluctuations. Time those “breaths” in mass with the push–pull, and you can ratchet momentum from the rest of the cosmos. In Woodward’s shorthand: a Mach Effect Gravity Assist (MEGA) drive that lets you exchange momentum with the gravitational field of the universe.

This wasn’t a rejection of Einstein so much as a change in emphasis. General relativity already ties matter and motion to spacetime geometry. Woodward’s program leaned into the idea that inertial forces—so often called “fictitious”—may reflect real, relational physics. The technical claim: when energy sloshes in a device on microsecond timescales while the device accelerates, additional time-dependent terms appear that look like effective mass changes. If those terms are real, they can be worked with in the lab.

“The way I think of it is not as propellantless propulsion, but as recycled propellant propulsion, because you’re pushing on something when it’s heavier and then pulling it back when it’s lighter. So basically what it does is it gives you a way to exchange momentum with the gravitational field of the universe. And it works.”

An engineer’s translation of philosophical physics

Physics history is littered with beautiful ideas that never touch a workbench. Woodward was unusual in that his Mach-inspired hypothesis came bundled with testable engineering claims. If transient mass fluctuations scale with the second time derivative of internal energy, you want a component that can store and release energy extremely fast and in a controllable phase with a mechanical oscillation. Enter piezoelectric ceramics: humble, off-the-shelf disks that convert voltage to strain and back again at tens of kilohertz, with high quality factors and well-behaved resonances.

The prediction became a blueprint. Stack piezo disks, clamp them between asymmetric “reaction masses,” and drive the stack electrically so it vibrates strongly at a chosen resonance. As the disks store electrical energy and convert it to mechanical strain, their internal energy density changes rapidly; if Mach-effect mass fluctuation is real, the stack’s effective mass should modulate in sync. Couple that breathing mass to a push–pull oscillation and bias the phase just right. Over many cycles, any tiny asymmetry integrates, yielding a net time-averaged force.

Building a thruster out of piezos and patience

Strip away the heady physics, and Woodward’s rigs are deceptively simple: a stack of piezoelectric ceramic disks sandwiched between end masses—often with one end heavier or engineered with “ears” and bearings to tune its dynamics. A high-voltage amplifier drives the stack; a function generator sweeps the frequency to find resonances, typically somewhere between the mid-30s and mid-40s kilohertz. The device rides on a torsion balance or pendulum so that minute forces deflect a long arm by measurable amounts. The assembly is instrumented to the hilt: oscilloscope channels for drive voltage and current, a strain gauge to track piezo deformation, and a position sensor or high-resolution video to monitor deflection.

“The devices have a couple of minor design improvements over the earlier devices, but all of them are designed as a PCD stack clamped between two masses, where the masses are asymmetric and produce force.” – James Woodward

Small details matter. A stack might weigh on the order of 150 grams, while the balance’s effective support mass is many times larger—a deliberate asymmetry. If the device merely shakes internally, momentum conservation demands the support move the opposite way, an equal-and-opposite jiggle. If, however, the device experiences an external force, both device and support should drift in the same direction—the telltale signature the team now looks for in high-speed video analysis.

How do you weigh a whisper of thrust?

Measuring micro- to millinewton forces is tricky business. Early on, critics pointed out that a high-frequency shaker can create illusions: air currents, magnetic coupling, thermal buoyancy, stick–slip friction. The response from Woodward’s crew—now led day-to-day by Horn, Broyles, and Fearn—has been methodical: turn each “usual suspect” into a controlled knob and show that the candidate artifact misbehaves differently than the observed signal.

One practice that now feels routine is video-based motion tracking. The team places a tracking dot on the device flange and another on the fixed support. A free, open-source program follows both dots across thousands of frames. If the motion is momentum-neutral internal vibration, the device dot and support dot will “breathe” oppositely. If a genuine external push is at play, both dots drift together in the direction of the force. Picoscope traces and FFTs of the mechanical response run in parallel, letting the team correlate when a resonance is hit with how the deflection evolves.

Sweeping, holding, and the reality of heating

Near a high-Q resonance, even minor heating shifts the center frequency. That’s why frequency sweeps—rather than parking on a single tone—became standard: sweep through the neighborhood and you virtually guarantee you’ll graze the peak. In several runs, that sweep produces a crisp, repeatable deflection right as the trace crosses resonance, a sharp “snap” the team has learned to expect when tuning is right.

Once a hot resonance is located, the group sometimes shifts to constant-frequency holds: drive at the peak and watch the balance settle into a deflection plateau that can last seconds. Then, as the stack warms and the resonance drifts away, the force fades without any changes in wiring or mounting—a behavior that feels mechanical rather than electromagnetic, matching the intuition that this is a vibration-phase phenomenon married to the putative mass fluctuation.

Ruling out the usual suspects

Artifacts have a way of impersonating breakthroughs. Perhaps the most humbling lesson came from the air-pump effect. A vibrating device in open air can act like a tiny bellows, pushing and pulling the surrounding fluid, producing a real force on a balance with all the right timing but none of the right physics. After several dramatic “discoveries” collapsed under the weight of simple baffles and enclosures, the team moved to vacuum runs and redesigned housings that reduce or eliminate any airflow-driven thrust mimics.

Then there’s stick–slip friction, capable of masquerading as one-directional action. If a device sits on supports that alternately catch and release, the whole rig can “ratchet” in a preferred direction. The antidote is to suspend the device with minimal contact and monitor the support’s motion directly—again using dual-dot tracking. When stick–slip is happening, the support moves opposite the device during each micro-jerk. When a thrust-like force is in play, both move together.

A third defense is the null device: make the device symmetrical so that if Mach-effect thrust depends on asymmetry in reaction masses, the effect should cancel. Currents, fields, and vibrations remain. In the team’s best-controlled trials, the signal collapses when symmetry is enforced and reappears when asymmetry returns, a strong (if not yet definitive) argument that the geometry matters in precisely the way the hypothesis predicts.

From micro-newtons to newtons?

A decade ago, the reported signals lived in the micro-newton realm: whisper-faint and hard to separate from confounders. Over many cycles of redesign—freeing the device to vibrate without shaking the world, refining isolation, dialing in the harmonic content, improving the balance—the lab reported climbing orders of magnitude. In their most optimistic reading, peak forces crept toward newton-class deflections with watts-class electrical input during the most favorable runs.

Even inside the group, caution remains. Peak deflections during resonance sweeps aren’t the same as steady, DC-like thrust, and translating a swing on a torsion balance into a calibrated force demands care. Still, the trajectory matters: when multiple instruments point the same way at the same time, when flipping the device reverses the deflection, when symmetry kills the signal and asymmetry revives it, the case for “something there” strengthens. The team’s aim now is to turn eye-catching runs into routine, repeatable plateaus that survive new chambers, new bearings, and new labs.

“It should be possible to get something that will self levitate, which means eventually that you’ll be able to build these things and simply turn them on and have them elevate themselves without blowing toxic chemicals into the atmosphere and so on.” – James Woodward

The analytics culture: everything on one screen, saved forever

One thing that stands out is the all-digital workflow. The lab treats each measurement not as a single number but as a bundle: synchronized oscilloscope traces, drive parameters, temperature logs, strain gauges, and time-stamped video. That bundle can be replayed and reanalyzed days later as a kind of experimental “flight data recorder,” with new filters, cross-checks, and independent reviewers who weren’t present during the run. It’s a culture that acknowledges past missteps and seeks to de-risk human enthusiasm by inviting outside eyes—while also recognizing that fragile setups don’t travel well without trained hands.

Bearings, “Big Ears,” and the razor’s edge of high-Q hardware

Much of the program’s drama lives in mechanical details. Bearings that look benign on paper can bleed energy, detune a resonance, or introduce microslips. The group’s evolution—from rubber pads to washers to sliders to miniature linear ball bushings—reads like a textbook in learning by breaking. A larger-bearing geometry nicknamed “Big Ears”, championed by Curtis Horn, promises smoother motion and less parasitic drag, at the risk of changing the device’s vibrational modes. Every such change is now treated as its own experiment: swap one element, rerun the sweep, document whether plateaus survive or vanish.

Likewise, drive strategy is a science of its own. A common protocol is to sweep from the upper 40s down into the mid-30s kilohertz at a known rate (say, 1 kHz per second), repeat, and then, after mapping the response, sit on the hot spot with on–hold–off pulses. The goal is to observe three things in the same run: a resonance-aligned kick, a sustained hold at that frequency, and a graceful fade that tracks the resonance’s temperature drift.

Replication: why “plug-and-play” isn’t

Everyone agrees: a claim this counterintuitive needs replication—preferably in a lab that tries to break it. Straightforward loaners, however, have backfired. Groups received a setup, plugged it into different balances or mounts, hard-clamped what should be softly supported, bypassed enclosures, and—predictably—either generated spectacular artifacts or suppressed the very oscillation that makes the effect visible. Woodward had been burned by this more than once, and the team shares that caution.

The path forward is clearer now: provide documented geometries, null devices, dummy stacks, and detailed mounting and vacuum protocols, plus a menu of artifact demonstrations—“here’s what the air-pump effect looks like,” “here’s stick–slip,” “here’s magnetic coupling with a coil in proximity.” The most convincing replications will be those where independent teams can produce the artifacts on demand and then turn them off, revealing the thrust-like signature only in the narrow window where the setup is correct.

What would “real” look like?

“Real,” in this context, isn’t a headline; it’s a checklist. The thrust vector reverses cleanly when the device is flipped relative to the balance arm; the signal remains under hard vacuum and in magnetically quiet environments; the amplitude scales with phase and harmonic content of the drive in ways the model predicts; a symmetric device kills the effect. Independent labs see the same patterns using different balances and different piezo suppliers. The team publishes open bundles of run data so that critics can re-plot and re-fit without fear of cherry-picking. And in the best case, someone takes a device into an inertial sensor testbed with no moving air, no buoyancy gradients, and shows a force trace that integrates to a measurable impulse over a long hold.

Short of that gold standard, there are intermediate milestones: a thrust-to-power curve with reproducible plateaus; a thermal model that predicts the resonance drift and the associated decline in force; and a parametric map showing where in frequency–voltage–load space the effect appears and where it vanishes.

The bigger picture: an impulse engine before a warp drive

Pop culture sees “propellantless propulsion” and jumps to warp drives and wormholes. Woodward’s vision was almost the opposite: start small. If a device can deliver even millinewtons per watt reliably, it changes how we station-keep satellites, deorbit hardware, and nudge small spacecraft. Continuous, miserly push across months beats a short, high-thrust burn for many missions. Imagine a deep-space tug that sips power from a solar array and inches cargo onto interplanetary trajectories with no propellant tanks to refill. For small operators, there’s a business case even if the physics never graduates beyond “impulse engine.” Horn, Broyles, and Fearn have embraced that near-term framing.

Only later, if the effect’s roots in gravitation are established, does the conversation wander back to metric engineering, to deeper links between mass, inertia, and spacetime. That’s a different mountain. The first hill is getting an unequivocal, reproducible force from a bench-top device that everyone can see and no one can dismiss as air, friction, or wishful thinking.

What about energy and momentum conservation?

The natural objection to any “reactionless” drive is conservation law anxiety: Where does the momentum go? In the Mach-effect narrative, the device exchanges momentum with the distant mass-energy of the universe via the gravitational field. That’s not hand-waving; it’s a claim that the boundary conditions of the universe matter. In practice, the lab doesn’t need to touch cosmology to do honest metrology: track forces on isolated balances, examine phase relationships, measure energy flow into the piezos, and ask whether the observations square with purely local couplings. If they don’t, and if every local path for momentum leakage is closed, the case for a nonlocal exchange—however unsettling—gains weight.

Energy questions are different. The device is powered; there’s no free lunch on the table. What’s at stake isn’t violating conservation, but rather exploiting a previously unrecognized coupling to trade electrical input and mechanical phasing for a slow but steady push without expending propellant.

Lessons from false positives

The last decade has seen its share of propellantless propulsion mirages. Microwave cavities produced signals that later evaporated under better controls. Ionic winds hid in plain sight. Temperature gradients tricked balances. To inoculate against déjà vu, the team now stages these artifacts for visitors: deliberately creating a stick–slip ratchet, building an air-pump into a dummy device, showing how a magnetic field near a current loop can tug a pendulum. The aim isn’t to embarrass anyone; it’s to equip would-be replicators with a catalogue of failure modes so they can tear through them quickly and see what—if anything—survives.

A second lesson is cultural. Experiments at the edge of detectability reward slowness and boredom: change one variable, log it, revert, repeat. The hardest part isn’t often the math; it’s resisting the urge to improve three things at once. The progression from early, buzzing contraptions to today’s quieter, higher-Q assemblies reflects that discipline. Most of the “news” is hidden in a dozen small refinements that remove one source of confusion at a time.

If it works, who needs it first?

The earliest adopters aren’t starships; they’re operators of small spacecraft who care about grams and watts. Picture a CubeSat constellation that needs continuous drag-makeup at ~400 kilometers without carrying propellant. Or an on-orbit inspector that can rise and fall between shells over months. Earth-observing platforms could use propellantless trims to hold view geometries for years, then deorbit cleanly using the same device as a brake against trace atmosphere while slowly “pushing” retrograde.

In deep space, an impulse engine changes the calculus for missions that today rely on Hall thrusters and limited xenon. A modest-thrust, propellantless tug could shepherd small payloads from Earth-escape to Mars transfer or the outer planets, and it could do so with only solar-array power until sunlight thins, then hand off to modest nuclear power. None of this requires science fiction. It requires a reliable millinewton-per-watt machine.

Skepticism is a feature, not a bug

Scientific American and Popular Science readers know the drill: extraordinary claims, extraordinary evidence. The best friends of a project like this are people who don’t want it to work and try to break it in public. Woodward’s group increasingly internalized that stance, moving from “we saw a thing” to “here is the exact artifact you think we missed—and here is how we generate it on demand.” That shift—from defensive to demonstrative—may be the most important cultural evolution of the entire effort, and it’s the culture Horn, Broyles, and Fearn are now stewarding.

There’s also a humility that comes from fighting the same gremlins for years. It’s hard to keep enthusiasm alive when a triumphant plot turns into a lesson about airflow. But in the same way superconducting qubits were a curiosity before they became a platform, subtle mechanical effects can stall for a decade and then lurch forward when the right materials, bearings, electronics, and data culture finally line up.

A note on theory’s moving target

Mach’s principle has many faces, from philosophical slogans to formal implementations in field theory. You will not find universal agreement about which “Mach effect” equations are canonical. That’s okay. In engineering disciplines, the best theories are the ones that tell you where to look and how to tune. In this program, theory points to: (1) rapid energy storage/release in a material with high strain energy density, (2) mechanical oscillations at resonant frequencies, (3) asymmetry in how that oscillation couples to the wider structure, and (4) precise phase control. Devices built to these constraints show the most interesting behavior. Whether the final explanation uses the original Mach-effect framing, a more orthodox GR calculation, or something entirely different, the hardware lessons will survive.

What failure would look like—and why that’s valuable

Another virtue of the current program is that it’s falsifiable in clear ways. If, under vacuum, with symmetric devices, rigid isolation, magnetic cleanliness, rock-solid metrology, and third-party hands, the phase-dependent thrust signature refuses to appear, the hypothesis will have met the kind of wall that serious ideas must survive. That would be disappointing—but not wasted effort. The discipline and tooling developed here would still advance precision force metrology at high frequency, feed into other micro-propulsion efforts, and sharpen the community’s instincts about how subtle artifacts sneak into beautiful plots.

The human factor

It’s easy to forget how much of experimental physics is craft. The first time you watch a balance settle after a careful sweep, you notice the body language around the workbench: hands hovering over knobs, someone reading the scope aloud, someone else marking timestamps on a log, a quiet chorus of “hold it” and “there.” The best labs feel like workshops and the best practitioners like instrument makers—people who can hear a chilly bearing and sense a bad ground. Whatever the final verdict on Mach-effect propulsion, the craft on display here is the kind that builds ecosystems. Students learn to love careful work. Visitors leave with respect for slow, meticulous science.

Beyond the bench: community and openness

For years, advanced propulsion research lived in a fragmented cottage industry. That’s changing—and so has the makeup of the Woodward effort itself. What began as a small, fragile setup has steadily grown into a collaborative program anchored by a team that blends craft, metrology, and systems thinking. Curtis Horn has driven mechanical refinements—bearing geometries, isolation schemes, and “Big Ears” experiments—to tame parasitic friction without detuning the device’s high-Q behavior. Michelle Broyles has championed measurement rigor and data discipline, from synchronized “flight-data-recorder” bundles to video-based motion tracking that makes subtle signatures visible and repeatable. Together with long-time collaborator Hal Fearn and a rotating cast of students and volunteers, the lab now operates more like an instrumentation group than a lone skunkworks, with checklists, null devices, and vacuum-first protocols replacing ad-hoc heroics.

“If Curtis and Michelle hadn’t come along and insisted on cleaning things up, getting everything working right and so on — the project would be dead in the water.” – James Woodward

That team ethos extends outward. Talks now linger on how to run the experiments—artifact demonstrations, mounting jigs, calibration interleaves—so that would-be replicators can deliberately create the known confounders (air-pump effects, stick–slip ratchets, magnetic tugging) and then turn them off. The culture aims for transparent rivalry: labs vying to be first while converging on shared procedures and publishing negative results. If the field avoids the trap of hero experiments and doubles down on procedural literacy, the quiet accumulation of trustworthy runs—across people, chambers, and balances—will do what charisma never could: make the effect, or its absence, unmistakable.

A cautious coda—and a legacy

What should a rational reader take from all this? Not a guarantee. Not a revolution next summer. But certainly not indifference. The problem is well-posed, the artifacts are getting boxed in, and the measurement culture is maturing. If the data keep aligning—and if independent groups can make the good runs routine—the first practical impulse engine may arrive quietly: a shoebox that sits in a vacuum chamber and nudges a balance predictably every time you dial in the right phase. From there to orbit is an engineering marathon, not a leap.

James F. Woodward spent decades pushing this once-esoteric idea from a philosophical curiosity toward an experimental craft. With his passing on August 9, 2025, the responsibility—and opportunity—to finish the demonstration falls to his colleagues. Curtis Horn shepherds the mechanical refinements, Michelle Broyles drives measurement rigor and data discipline, and Hal Fearn continues to sharpen the theoretical throughline. If an impulse engine truly exists, it won’t first appear as science fiction. It will look like a well-kept lab notebook, a pile of boring nulls punctuated by stubborn outliers, and a small team willing to be wrong loudly until it’s right. That is the legacy they are carrying forward.

This article draws on interviews with Dr. James F. Woodward, (“MEGA Drive Propulsion Experiments”), and (“Mach’s Principle and the MEGA Drive”), as well as the conference presentation, (“Mach Effect Propulsion”).