COVER STORY ADAM HADHAZ Y | adamhadhazy@gmail.

com

FUEL-F
SPACE
TRAVEL
16 | FEB RUARY 2017 | aerospaceamerica .org

16-23_February_EmDrive_v1.indd 16 1/23/17 2:42 PM

 FREE
E
EL NASA researchers rattled the propulsion community last
year when they reported generating thrust inside a vacuum
chamber with electromagnetism instead of propellants. Was
the effect real or a product of an unknown source of error?
Adam Hadhazy spoke to scientists who aim to find out.

aerospaceamerica .org | FEB RUARY 2017 | 17

16-23_February_EmDrive_v1.indd 17 1/23/17 2:42 PM

 I
s it a breakthrough, or baloney? Not long
after the turn of the millennium, the EmDrive
catchy shorthand for electromagnetic
drive was a mere glimmer at the fringes
of space-propulsion research.
The idea was to generate thrust by harnessing a force was generated on the test article, roughly the
strange effect that seems to happen when electro- equivalent of a honeybee landing on a flower.
magnetic energy is circulated in an enclosure. Thats Unlike EmDrive tests abroad that appeared to
where things stood until November, when the con- detect force, the NASA EmDrive experiments took
cept began commanding mainstream attention place in a vacuum chamber, this one at the Johnson
after a team of NASA researchers reported in AIAAs Space Center, home to Eagleworks. Conducting the
Journal of Propulsion and Power that they may have test in a vacuum chamber slashed a major possible
generated an anomalous force with their version of source of bogus thrust: air warmed up by the ex-
an EmDrive, a force that would fly in the face of periments equipment could jar the test article.
conventional physics. How can it be?
Even the papers lead author isnt sure whether The paper, Measurement of Impulsive Thrust
the mysterious phenomenon is real. Its possible from a Closed Radio-Frequency Cavity in Vacuum,
that an error source remains unaccounted for, and went through about a year of peer review involving
that the test article moved a distance imperceptible five expert referees instead of the more typical
to the human eye on the order of 4 to 10 microm- two referees. Even so, some physicists remain un-
eters for completely mundane reasons. The body convinced about the EmDrive. They have a huge
of work we did in the paper helped us eliminate a hill to climb here, because regardless of how it works,
number of those [error sources] fairly definitively, the EmDrive would violate deeply fundamental and
says aerospace engineer and physicist Harold Sonny well-verified physical phenomena, says Brian Ko-
White, who led the experiment. For other possible berlein, a professor of physics at Rochester Institute
sources, although maybe we put a little bit of a of Technology in Rochester, New York.
pencil mark through them, they are certainly not Another skeptic put it even more bluntly. My
black-Sharpie-crossed-out. perspective is, theres no there there, says Eric
Nevertheless, White and his colleagues at NASAs Davis, a senior research physicist at the Institute
Eagleworks advanced propulsion physics labora- for Advanced Studies at Austin, Texas, as well as a
tory in Houston are confident enough that they are former consultant on NASAs defunct Breakthrough
busy planning more tests, despite sometimes skep- Propulsion Physics Project and who visited White
tical reviews by other technologists who have gone at the Eagleworks lab. That conical cavity of theirs
over the paper. is a microwave oven, Davis continues, and we
Why has the paper generated so much intense know from physics that microwave ovens cant fly.
scrutiny? If the effect turns out to be genuine and Part of the problem might be the breathless head-
the technology can be expanded to a meaningful lines touting NASAs verification of an impossible
scale, it could portend a revolution in the space in- or Star Trek drive. Seeing this coverage, White is not
dustry. Spacecraft would no longer need hundreds at all ready to pop a champagne cork. Were not
of kilograms or even tons of propellant to stay in there yet, he says. Although confident about the
orbit or explore deep space. The International Space EmDrive on one hand, he calls for careful incremen-
Station, for instance, burns through approximately talism on the other. There are no shortcuts in sci-
4 tons of propellant each year, and more fuel must entific investigation and in doing empirical obser-
be delivered to it regularly at a cost of about $20,000 vation, White says. You just have to have patience.
a kilogram.
Trendy research
The test Patience seems to be in short supply, though. Re-
Specifically, the researchers found that per every searchers at the China Academy of Space Technol-
kilowatt of microwave energy they pumped into ogy say they have been testing their own version
their test article, which has the humble appearance of an electromagnetic drive in space, according to
of a tapered, copper bucket, 1.2 millinewtons of a December article in Science and Technology

18 | FEB RUARY 2017 | aerospaceamerica .org

16-23_February_EmDrive_v1.indd 18 1/23/17 2:42 PM

 NASAs EmDrive
experimental setup:
Microwaves moving
around inside this copper,
bucket-shaped test
article in Houston appear
to have generated thrust
toward the articles
narrow end, hinting at a
potentially propellant-free
mode of propulsion.

NASA

Daily, the newspaper of the Chinese Ministry of thruster. The concept had occurred to him while
Science and Technology. Meanwhile, the Pennsyl- evaluating closed-system gyroscopes as a means of
vania-based company Cannae Inc. has announced thrust for missile control. Despite critics who say
plans to launch an EmDrive-like, also-propellant- his experiments are imprecise and his physical the-
less thruster into space on a cubesat as a technol- ories questionable, Shawyer has claimed success
ogy demonstration. in experiments over the years. Along the way, his
All in all, it is a meteoric rise for an outr pro- efforts have inspired the aforementioned research-
pulsion concept that is not even two decades old. ers in China, Germany, and also White and his col-
Roger Shawyer, a British aerospace engineer, left leagues, to delve further into the potentially
Matra Marconi Space, forming his own company game-changing prospects of the EmDrive.
in 2001 to develop an electromagnetic drive, tech- White and his colleagues recognized that the Em-
nically known as a radio frequency resonant cavity Drive was a long shot, but long shots are the Eagleworks

aerospaceamerica .org | FEB RUARY 2017 | 19

16-23_February_EmDrive_v1.indd 19 1/23/17 2:42 PM

 THE EXPERIMENT
NASA researchers think they may have moved the copper cone below just by circulating microwaves inside it. The results
raise the possibility of generating spacecraft thrust without propellants. In a series of test runs, an optical displacement
sensor detected subtle movements of a test article attached to a torsion pendulum arm.

Test article Linear exure bearings

Pendulum arm pivots on these

Direction of
thrust
Ballast
Rotation
Mirror Assures neutral
position of
pendulum arm

Electrostatic ns
Provide force calibration

Torsion
pendulum arm
Supports test article
Optical displacement sensor
Measures movement of test article
and aids in calibration
Magnetic damper
Dampens harmonic
motion
Source: NASA Eagleworks

Illustration by John Bretschneider

labs raison detre. Indeed, the EmDrive might be tame the torsion pendulum from its rest position. White
compared to other work at the lab, including theoret- and his team needed to describe this displacement
ical and experimental forays into the possibilty of in terms of force, which meant they had to know
developing a faster-than-light warp drive, courtesy how far the pendulum would move with a given
of isolated bubbles of spacetime. amount of force. They placed two electrostatic
Not long after the founding of the Eagleworks combs opposite each other, one on the pendulum
lab in 2011, Whites team decided it was time to take arm and one on some gear near it, so that the fins
the EmDrive for a spin. What ensued were years of of the combs interleaved. Before each test run, they
work to set up the experimental apparatus, report charged the combs with particular voltages to in-
preliminary findings in 2014 and conduct a year of duce specific levels of force, and then with the
peer review for the paper published in November. optical sensor measured the separation of the fins.
The result was a touchstone they could consult to
The tests at NASA know how much force was generated on the test
For his EmDrive, White hewed to Shawyers design article on each run.
to construct the 23-centimeter-long, copper test Direct current power and data signals traveled
article in the geometric shape of a frustum a cone from the fixed portion of the torsion pendulum into
with its pointy end lopped off. A small antenna the movable portion through brass screws into liquid
inside the test article floods its interior with reso- metal contacts made of the trademarked alloy Galin-
nating microwaves. stan, a mixture of gallium, indium and tin that stays
This test article was mounted to the aluminum liquid at room temperature. These connections elim-
arm of a torsion pendulum. This arm was free to inated cable interfaces and their confounding forces.
rotate horizontally around a linear, vertically ori- To offset the weight of the test article on the torsion
ented flexure bearing, should it experience a force. pendulums one end, a ballast weight was placed on
An optical sensor transmitted a laser to measure its other, while a magnetic dampener converted some
the displacement of the movable arm portion of oscillatory motion into heat. Thats a way to keep

20 | FEB RUARY 2017 | aerospaceamerica .org

16-23_February_EmDrive_v1.indd 20 1/23/17 2:42 PM

 ... MAYBE WE
the system fairly well-damped so its a nice, clean
signal that we can look at, says White. The experiment

PUT A LITTLE
was done on a shoestring budget in line with other
low-technology-readiness-level, university or NASA

BIT OF A PENCIL
center lab efforts. The cost was on the order of tens
or hundreds of thousands of dollars, not millions.

MARK THROUGH
You just try and be as frugal as you possibly can to
make forward progress, White says.

POTENTIAL SOURCES
The researchers ran the experiment at varying
power levels of 40, 60 and 80 watts for periods of

OF ERROR, BUT THEY

30, 60 and 90 seconds. The optical sensor recorded
any displacement of the copper frustrum in both

ARE CERTAINLY NOT

forward and backward orientations on the torsion
arm. When all was said and done, on average, a

BLACK-SHARPIE-
small, but persistent force unattributable to other
aspects of the experiment emerged.

CROSSED-OUT.
Whites paper addressed nine possible sources
of error for this force, ranging from air currents to
vibration to electrostatic interactions of experi-
mental components. Although White is confident Harold Sonny White, NASA Eagleworks
that most were ruled out, perhaps the most difficult
to counter was the thermal expansion and contrac-
tion of the various items mounted on the torsion
pendulum arm. That expansion or contraction
could tip the center of gravity and yield spurious Marc Millis, the former head of NASAs defunct
thrust signals. That was the thing we worked the Breakthrough Propulsion Physics Project, and
hardest to understand and put in a box, White co-authors wrote about the paper on Cenaturi
says. As we move forward, we want to do some Dreams, the website for his interstellar travel ad-
more due diligence. vocacy group, the Tau Zero Foundation. The new
study is an improvement in fidelity on the prior
Peer reviews tests and may be indicative of a new propulsive
Rochesters Koberlein applauds White and his team effect, according to the authors, but they went on
for applying such rigorous peer review. The whole to observe what they called some shortcomings.
point of peer review is it puts [research] out there Philosophically speaking, the papers authors
publicly so people can rip it apart, said Koberlein. lacked impartiality, given their assumptions that
Thats how we do science and thats the only way the EmDrive somehow or another worked. On the
were going to get to the truth. methodological front, Millis and colleagues write
That said, Koberlein is quick to argue that that Whites team failed to fully characterize the
pre-publication review is, of course, not any guar- performance of their torsion pendulum; quanti-
antee of a studys results, just that its general meth- tatively assess possible interactions with the vac-
odology passed muster with at least one, if not a uum chambers wall; and exhaustively account for
few, independent scientists. Its important not to thermal effects. The analysis suffered, Millis and
overvalue peer review, because lots of things that colleagues say, because it could not credibly dis-
have been subjected to peer review have been criminate which part of the data was due to actu-
wrong, says Koberlein. al impulse, versus thermal effects. Questionable
A recent swing-and-a-miss was BICEP2, short for subjective techniques are used to infer the thrust
Background Imaging of Cosmic Extragalactic Polar- from the data, they write.
ization, an experiment that gathered relic light from Davis, who was one of Millis co-authors, says
the Big Bang with a telescope near the South Pole. In that before the experiment, Whites team cleaned
March 2014, researchers announced collections re- up a lot of the potential sources of error , but they
ported finding evidence for a period of rapid expan- still engaged in a lot of wishful thinking.
sion of the universe following the Big Bang. Less than
a year later, the team withdrew the claim, and the Reaching for an explanation
data was chalked up to mundane, cosmic dust. As for the physics behind the EmDrive, White and
At this early stage of digestion of Eagleworks his team speculate they may be seeing a manifes-
EmDrive paper, the review from the wider commu- tation of whats known among physicists as a quan-
nity has ranged from lukewarm to skeptical. tum drive or Q thruster. Such a thruster works on

aerospaceamerica .org | FEB RUARY 2017 | 21

16-23_February_EmDrive_v1.indd 21 1/23/17 2:42 PM

 THE WHOLE POINT OF PEER REVIEW
IS IT PUTS [RESEARCH] OUT THERE
PUBLICLY SO PEOPLE CAN RIP IT APART.
Brian Koberlein, Rochester Institute of Technology

the well-accepted principle that empty space is not vacuum and induce a slightly preferential flow in
a vacuum but is frothing with fluctuations in quan- it. A terrestrial analog to think of is a submarine
tum fields, with pairs of virtual particles perpetu- that sits in the water, White explains. It takes ad-
ally forming and mutually annihilating. vantage of the fact its embedded in its propellant.
Based on an interpretation of quantum mechan- A submarine uses a propeller to couple with this
ics known as pilot wave theory, these vacuum fluc- liquid medium and generate a hydrodynamic pres-
tuations can be thought of as a dynamic medium sure gradient. The water moves in one direction
that guides, or pilots, individually observed, real and, to conserve momentum, the submarine moves
Cannae announced that particles. And as a medium, the quantum vacuums in the other, just like the EmDrive goes one way
it will launch its propri- constituents can interact and exchange momentum, while a wake of virtual particles streams through
etary thruster which it
not unlike a fluid, or a plasma. the quantum vacuum.
says requires no onboard
propellant into space
As White sees it, the EmDrive can leverage elec- According to this explanation, one of the biggest
on a cubesat. tromagnetic fields to couple with the quantum slams of the EmDrive that it violates conservation

Cannae

22 | FEB RUARY 2017 | aerospaceamerica .org

16-23_February_EmDrive_v1.indd 22 1/23/17 2:42 PM

 of momentum, a bedrock principle of physics is
averted. Yet other physicists are not buying that the
Q-thruster argument conserves momentum, famous-
ly conveyed as Newtons third law: For every action,
there is an equal and opposite reaction. Rochesters
Koberlein says extending the third law to the quantum
vacuum in effect, the entire cosmos, anywhere and
everywhere cannot work, because then what does
the cosmos move against? he asks.
Furthermore, some question that the quantum
vacuum operates at all like a plasma, as White con-
tends. The paper also does not present any equations
or mathematical predictions for how to relate mea-
sured thrust back to a theoretical paradigm, even a
wildly conjectured one. They have nothing to go
on theoretically, Koberlein says. That means they
have no convincing way to pull any signal they might
have out of the noise [of the experiment].
But backers posit entirely different theories for
how they might work. Shawyer in the U.K. points to
electromagnetic waves momenta being linked to
velocity, influenced by the geometry of the tapered
EmDrive, thus creating asymmetric forces at its wide
and narrow ends. Some speculate on theoretical

Ames Research Center

Unruh radiation, a hypothesized inertia phenomenon
caused by virtual particles and in this case felt by
microwaves to differing degrees at opposite ends of
the EmDrive. An even more outlandish conjecture
is that an EmDrives microwaves might interact with
elusive dark matter particles, which although unde-
tected to date, theoretically outnumber particles of
regular matter by five to one. Optimistically, a few years ago, White and a col- Next, Harold Sonny
league ran some figures on what a Q-ship outfitted White and NASA
Eagleworks want to test
What the future may hold with a fully-fledged EmDrive, could do throughout the
their EmDrive concept
White seems unfazed by the hubbub sur- Solar System. The researchers assumed a conservative on a Cavendish balance.
rounding his experiment and is planning his next thrust of 0.4 newtons per kilowatt about 300 times Additional findings
move. To further tackle the possible bugaboo of that hinted at in the recent, small-scale experiment, might help to define the
EmDrive thermal expansion and contraction, he but a tenth of what they ballpark as achievable sup- underlying physics.

and his team want to run similar tests on a type plied by a modest nuclear reactor in space, producing
of apparatus called a Cavendish balance. In such a megawatt or two of electrical power. In this scenario,
a setup, the EmDrive could rotate out to much a one-way trip to Mars could take 75 days a far cry
larger angular displacements, such that the thrust from the approximately eight months with current
force would dominate over any thermal effects. rocket technology. More eyebrow-raisingly, trips to the
Additional findings also might help to define the neighborhoods of Jupiter or Saturn would take around
underlying physics. Those are two major brush- 200 and 260 days, respectively, opening up the outer
strokes that well be applying to the canvas, solar system to human exploration.
White says. Beyond these next steps, White says White realizes this is pie in the sky for now, con-
it is premature to consider, say, altering the shape sidering where the EmDrives development is. Weve
of the test article or the frequency of the micro- got some work to do, he says. Were still potential-
waves to try and squeeze out more oomph. We ly even in the mode of definitively eliminat[ing] all
really dont have a good sense yet of what partic- false positives.
ular dials there are for us to be able to grab onto Whether or not the EmDrive will ultimately prove
and turn and be able to say you can do this, that itself in the years ahead, many will be rooting for the
or the other, he says. Were very much in the Eagleworks team to upend physics as we know it.
early phases of trying to understand the engi- I want them to succeed. I want warp drive, or
neering and physics and how they interact with EmDrive, or whatever. I want to go to the stars, said
one another. Koberlein. But that doesnt make it so.

aerospaceamerica .org | FEB RUARY 2017 | 23

16-23_February_EmDrive_v1.indd 23 1/23/17 2:42 PM