/^k I ^f^ I j IC ^ i f C ^ 0-F THE (j O/^L^^^Dj/^L

When a liquid or gas flows along a solid surface, it tends to stick

to the surface. This effect, named for a Romanian aircraft pioneer,
has potential uses in such devices as burners and hovering vehicles

by I m a n t s R e b a

veryone knows that it is not easy causes the flow to bend and cling to born engineer named Henri Coanda

E to pour a liquid smoothly out ol:

a teapot spout or the mouth of
a bottle. The liquid lias an annoying
the vessel's surface? Is it simply surface
tension, as has usually been supposed?
The answer is not quite so simple, and
it has significant technological conse-
tested a new flying machine he had
built himself. He had been a student of
Alexandre Gustave Eiffel, designer of
the Eiffel Tower and one of the pio-
iendeucv to flow around the lip and
dribble down the curved surface of the quences. neers of modern aerodynamics, at the
spout or the side of the bottle. What In the year 1910 a young Romanian- Ecole Superieure Aeronautique in Paris,

" T E A P O T E F F E C T " is a low-speed form of the Coanda effect: glass in the same way that tea sticks to the spout of a teapot. High'
water poured slowly from a glass tends to stick to the side of the speed fluids similarly adhere to a surface of suitable shape.
 and the young engineer's approach to
flying was both bold and imaginative. SHOULDER
Coanda's machine was a kind of jet air-
craft. It was equipped with an air com-
pressor powered by a reciprocating
engine; thrust was obtained by injecting
fuel into the compressed air to provide
a kind of afterburner. At the outset
Coanda was confronted with the possi-
bility that the flames from his two jet ex-
hausts might set fire to the plane's fuse-
lage, which was made of plywood. He
therefore installed metal plates to de-
flect the exhaust flames away from the
fuselage.
RESERVOIR
Coanda took his vehicle out for a
COANDA E F F E C T was named for the tendency of a fluid, either gaseous or liquid, to
test run on a field at Issy-les-Mouli-
cling to a surface that is near an orifice from which the fluid emerges. In this representa-
neaux, a Paris suburb. As he taxied
tion air from the reservoir emerges through the slit and, as shown by the dark color,
across the field he saw with dismay closely follows the sides of the vessel. An important part of the effect is the tendency of
that the deflecting plates, instead of the primary flow of air to entrain, or draw in, more air (light color) from the environment.
oanda diverting the exhaust away from the
3 had body of the plane, were actually suck-
ent of ing the flames toward it! Preoccupied
ier of with this surprising development and
s pio- with handling the engine throttle, Co-
it the anda did not notice until the last mo-
Paris, ment that lie was approaching the Paris
city wall at high speed. He jerked back
on the control stick, became airborne
long enough to clear the wall, then
crashed.
As far as is known, this was the first
flight of a jet airplane. Reflecting on
the incident, however, Coanda was less
interested in the takeoff than in the
strange phenomenon of the wrong-way
bending of the jets by his deflector
plates. At the first opportunity he dis-
cussed the event with Theodor von Kar-
man, the foremost theorist of aerody-
namics, who was then working at the
University of Gottingen. Von Karman
realized that the phenomenon was a
new discovery and named it the Coanda
effect.

'Pwenty-hve years or more passed be- E X T E R N A L NOZZLE was designed by the author in early tests of the Coanda effect as a
means of generating lift. Pressure over the upper portion proved to be less than atmospher-
fore any other investigator gave seri-
i c ; hence a vehicle with the shape of such a nozzle could be lifted by low pressure at top
ous attention to the Coanda effect.
and higher pressure at bottom. Step near slit proved to be crucial in maintaining good lift.
Coanda himself went to work as chief
engineer of the Bristol Aeroplane Com-
pany in Britain and became involved in
various technical researches having
nothing to do with this subject. In his
spare time, however, he carried on
experiments to examine the basic fea-
tures of the effect and its possible uses.
Consider a sheet of fluid (gaseous or
liquid) that is discharged through a
slit onto an extended and rounded lip.
As in the case of the teapot spout, the
sheetlike stream will attach itself to
the curved surface and follow its con-
tour. C o a n d a f o u n d t h a t b y c o n s t r u c t - I N T E R N A L NOZZLE takes its name from the fact that the flow of air is through the in-
High- I ing a shoulder c o m p o s e d of a series of side of the nozzle. Otherwise the Coanda effect is the same as that in the external nozzle.
ihape.

85
short, flat surfaces, each at a certain conventional propulsion systems. In- pressure at the nose and higher pr e s .
angle to the one before and each with stead of applying force at the rear of sure at the rear. This suggestion is S o
a certain length, he could bend a jet the vehicle (such as a rocket or a boat) completely contrary to ordinary con-
stream around an arc of 180 degrees to propel it against the resistance of the cepts about the (low of a fluid around
[see top illustration on preceding page]. air or water in front, one could use the moving bodies that it has understand-
Moreover, by experimenting with air Coanda effect to generate a vacuum at ably met with great skepticism among
blown through the slit he discovered the front, so that the higher pressure at aerodynamicists.
that the deflected airstream sucked in the rear would impel rather than pro- Undismayed by skeptics or by the
air from the surroundings. As the jet pel it. Similarly, a Coanda-generated general lack of interest in his ideas,
flowed around the shoulder it entrained vacuum above the vehicle would pro- Coanda proceeded to construct various
up to 20 times the amount of air in the vide lift. In conventional aircraft-pro- nozzles embodying his effect and to de-
original jet. Coanda measured the air pulsion systems the necessary vacuum sign vehicles to be propelled by them,
pressure at various points along the above the wings is generated by driving The fact that some of his designs took
surface over which the stream was the vehicle forward at high speed; the the shape of flying saucers has not
flowing, and he established the signifi- Coanda system would accomplish the helped matters. The Coanda effect
cant fact that the pressure on the sur- same purpose without any forward mo- might have been dismissed as an eccen-
face was less than atmospheric pressure: tion or moving parts simply by blowing trie notion had it not attracted the at-
in other words, it represented a relative air over a rounded shoulder. tention of the team of Allied scientists
vacuum. Near the slit the existence of In aerodynamics the problem of pro- that was assigned to investigate the
this vacuum accelerated the flow of the pulsion is reduced to the problem of German wartime research after the lib-
jet after it emerged from the slit. overcoming drag, a resistance whose eration of Paris in World War II. It was
From his experiments and calcula- components are high pressure at the alleged that the Germans had conscript-
ions Coanda derived a remarkable con- nose of the vehicle and low pressure at ed Coanda into their jet-propulsion in-
clusion. The Coanda effect, he argued, the tail. Coanda's proposal, to put it search. In any case, the Allied scientists
could lift and propel a vehicle by a briefly, consists of propelling the ve- examined Coanda's devices and suggest-
process that was the reverse of that in hicle by means of "negative drag": low ed that the Coanda effect merited seri-

]IMPORTANCE OK GEOMETRY in optimizing the Coanda effect slit ( m i d d l e ) the air tends to follow the surface, but with distor-
is demonstrated by these shadow photographs. At left is the tions. A step at the exit of the slit (right) produced a better flow
flow of an undeflected jet of air. With a curved surface outside the but not an ideal one because the air impinged on the surface.

F U R T H E R MODIFICATIONS showed (left) that too high a step surface ( m i d d l e ) the lion was undistorted. A well-balanced geoinf
can cause the flow to separate from the surface. With a grooved try (right) provided an ideal flow that was tangent to the surlace.

86
(,us sludv. A theoretical analysis by von
Karman in 1949 and experiments at
the Cornell Aeronautical Laboratory in
1952 supported Coanda. Several other
investigators, however, reported that
his claims could not be confirmed, and
this discouraged a thoroughgoing eval-
uation of his ideas.

T Tappily unaware of the controversy

among the experts, 1 was attracted
to Coanda's concept by an article about
him in a popular magazine in 1956. As
a student at the Polytechnic Institute of
Brooklyn, I was particularly interested
in propulsion systems, and the unortho-
dox Coanda approach seemed exciting.
When I went on to graduate studies at
the institute, working as a research as-
sistant in its newly established Rocket
Propulsion Laboratory, I had the good
fortune to obtain more technical infor-
mation about the Coanda effect from
T. Paul Torda, director of the labora-
tory, who had worked with von Karman
on the theoretical analysis.
R E O R I E N T A T I O N OF SURFACES was necessary to progress from nozzle-like applica-
I set out to test the effect with a tions of the Coanda effect (top) to the wide-based shape ( b o t t o m ) desirable for a ground-
nozzle of my own design. Coanda had effect machine. The changes involved much experimentation with different configurations.
experimented with both "external" and
"internal" nozzles. In the former the de-
flecting shoulders were outside a tube; for propulsion. Coanda had said that a My first model had had a small step
in the latter, inside [see middle and bot- Coanda nozzle, by increasing the mo- down from the mouth of the slit. This
tom illustrations on page 85], My nozzle mentum of the gas flow, could produce small step at the aperture, only about
was of the external type: a roughly cone- twice as much thrust as an ideal straight a thirty-second of an inch deep, made
shaped arrangement in which air blown nozzle with the same input of energy. the difference between the airstream's
outward through a slit circling the body Experimenters testing his claim had clinging to the shoulder and separating
would cascade over a shoulder angled failed to achieve any augmentation of from it. With such a step the new
60 degrees from the slit. This form of thrust by means of the Coanda effect. model also worked.
the device was inspired by the thought Their test devices, however, were quite
that the Coanda effect might be applied simple in design, whereas Coanda had VV7 e then proceeded to extensive test-
to a "ground-effect machine," the type developed rather sophisticated nozzles ing of some 30 different shoulder
of vehicle that moves over the ground of complicated geometry. It therefore configurations. With the help of shadow-
on a thin cushion of air. For such a ma- remained possible that with proper de- graph photography that showed details
chine one of the main problems is steer- sign a gain in thrust could be demon- of the flow, \ve obtained several in-
ing and related control, since it has no strated. teresting insights into the Coanda effect
traction with the ground. It seemed that [see illustrations on opposite page]. The
As a start in our testing program I
Coanda nozzles might provide such small step at the very lip of the slit is im-
designed a one-piece nozzle six inches
controls. portant because it turns the flow from
in diameter with a shoulder in the form
The first tests of the nozzle were suc- of a series of successively angled flat the slit into an eddy, or vortex, which
cessful. With a nozzle positioned up- surfaces, according to Coanda's original in this device rings the entire nozzle
right in a model ground-effect machine, design. The results of its first test were and therefore forms what is called a
the Coanda phenomenon caused the most disappointing. The airstream from ring vortex. It is the vortex—or rather
model to take off and rise to a height of the slit did not flow around the shoulder the low pressure generated within it—
several inches when a fast stream of air contour; it promptly detached itself that causes the stream from the slit
was blown through the slit. Emerging unless the slit opening was made very to bend and thus follow the contour of
from the slit at the speed of sound, the small or the flow was very gentle and the shoulder. The curving of the stream
air accelerated to supersonic speed as undisturbed. I went home from the around the shoulder produces a force
soon as it began to flow over the laboratory frustrated and discouraged. directed radially outward; this force,
shoulder. But my assistant, Edward Wohlthausen, tending to pull air away fiom the
In 1961, having joined the Armour stayed on and carefully went over every shoulder surface, apparently accounts
Research Foundation, now the IIT Re- feature of the nozzle's construction. He for the relative vacuum, or suction, next
search Institute associated with the discovered a seemingly trivial differ- to the surface. To maintain the suction,
Illinois Institute of Technology, I was ence, which turned out to be crucial, which prevents separation of the stream
provided with funds for an investigation between this model and the one I had from the surface and causes the flow
of the possible use of the Coanda effect built originally at Brooklyn Polytech. to accelerate, the stream must hug the

87
shoulder surface closely without actually Obviously the effect depends pri- zle. The fluid in a manometer attached
impinging on it or adhering to it. marily on the maintenance of a balance to the model promptly rose, showing an
Our experiments showed clearly that between the centrifugal force and the increase in the suction pressure. By
the creation of a stable Coanda effect suction force as the stream flows around simply forming a shroud around the
depends on the appropriate adjustment the shoulder and afterbody of the nozzle with his hands Coanda had
of many factors: the diameter of the nozzle. We found by experiments that increased the device's production of
slit, the strength of the jet, the depth of the efficiency of the effect could be en- thrust!
the step and various other details of hanced by certain adjustments of the The shroud, Coanda explained
the shoulder contour. Particularly cru- orientation of the slit and the slope of guides the streamlines of the entrained
cial is the ratio of the slit diameter to the shoulder; these controlled the dis- flow in a direction parallel to the sur-
the diameter of the nozzle as a whole. tribution of pressure within the stream face of the model and thus enhances
If the slit is too wide in relation to the so that it did not separate from the sur- the suction pressure. Coanda went on to
shoulder breadth, the stream will tend face. emphasize also the importance of proper
to detach itself from the surface; if it Although we obtained high thrust design of the afterbody, or tail, beyond
is too narrow, the stream will tend to efficiencies with such devices, we were the shoulder. A suitably tapered after-
stick to the surface. Also important is not able to produce the gain in thrust body, he said, would cause the flowing
the texture of the surface: a properly (over the theoretical capabilities of con- stream to exert a positive pressure on
roughened surface helps to prevent ventional jet devices) that Coanda had the rear of the body and thereby add to
distortion of the stream flow. In short, said should be achievable. At this junc- the propulsive thrust.
it takes a complex set of arrangements ture, in the spring of 1963, Coanda him-
to produce a useful Coanda effect. It self visited my laboratory. (He was then TVTot entirely convinced of this last
is no wonder that most investigators of 77 years old.) He examined the model argument, because it was so con-
the effect had found it to be an elusive with which I was working, bent over it trary to accepted ideas in aerodynam-
phenomenon. and cupped his hands around the noz- ics, I nevertheless proceeded to build

C O N F I G U R A T I O N AND SHROUDING affect the performance of over the same airfoil is shown. An airfoil of different geometry ap-
an airfoil. At a the Coanda effect over the airfoil at left produces pears unshrouded at c and shrouded at d. In each instance the data
the forces represented at right. At b the result of placing a shroud were obtained from an airfoil that was fixed rather than moving-
several models based on Coanda's rec-
ommendations. The shroud concept im-
mediately proved its value in enhancing
thrust. The afterbody theory, on the
other hand, turned out to be harder to
confirm. At length, after more than
a year of experimenting with various
afterbody configurations, we finally pro-
duced one that bore out Coanda's pre-
diction [see illustration on opposite
page]. This nozzle, with a shroud and
a steeply tapered tail, produced a thrust
19 percent greater than that of an ideal
jet lacking the augmentation of the
Coanda effect.
Independent studies at the University
of Toronto's Institute of Aerospace Sci-
ence, at Pennsylvania State University
and at the Research Center of the
Huyck Corporation in Stamford, Conn.,
have also verified that Coanda nozzles
are capable of augmenting thrust. Gains
of up to 38 percent have been reported.
Indeed, it can be said that Coanda's
hypothesis of propulsion by negative
drag has been confirmed.
This is not to say that the Coanda
effect is fully understood. There is no G R O U N D - E F F E C T M A C H I N E , which is known by the acronym GEM, rests on a table at
doubt that the gain in thrust with the top and rises from the table at bottom. It achieves lift by means of air, but in an unusual
Coanda type of nozzle derives from way. Because of the Coanda effect the pressure at GEM's top is less than atmospheric. With
the fact that a jet sheet flowing around this negative drag, or reduced resistance, over the upper portion of the surface, the higher
a curved wall entrains more outside pressure at the bottom lifts or drives the vehicle. The hole on the left side is the air intake.
fluid than a straight jet stream does.
But the mystery remains: Why does it
do so? The answer probably lies in the
turbulences set up within a curving jet
sheet; presumably they promote vigor-
ous mixing of portions of the fluid and
so increase the rate of entrainment at
the outer edge of the sheet. In any case,
we now know from experiments that the
Coanda effect does work to augment
thrust, and we therefore have a sound
basis for exploring possible useful ap-
plications of the phenomenon. Indeed,
it has already shown its usefulness in
another field, namely the device known
as the fluid amplifier, whose functioning
depends on the Coanda effect [see
"Fluid Control Devices," by Stanley W.
Angrist; SCIENTIFIC AMEBICAN, Decem-
ber, 1964],

rP h e first device we built to test the

practical application of Coanda
thrust was a small model called GEM
(for ground-effect machine). This dome-
shaped model, two feet in diameter, has
a ring slit at the top, a shoulder con-
figuration in the form of an ellipse with
its major axis parallel to the ground and
a short, collar-shaped shroud around
the lower part [see illustrations at right].
The slit is inclined at an angle of 45 T O P V I E W of the G E M shows the slit from which the air emerges. The slit is the thin light
degrees from the horizontal, so that the circle near the outer ed«e of the dark central area. T h e dark circle just outside the slit
is the step. This model, designed by the author, is approximately two feet in diameter.
jet is directed upward—contrary to all
conventional uses of jet thrust to gen- that of the environment. The primary of the model can be tilted so that ij l e
erate lift. This vehicle is raised from jet, generated by a small ducted fan, .slit is widened on one side and nar -
the ground simply by the jet of air that represents a comparatively modest ap- rowed on the other; this maneuver, by
produces a vacuum at the upper sur- plication of power, the strength of the skewing the distribution of force <,Ve
face, flows around the shoulder and jet amounting to a pressure of four the top surface, serves as a means of
emerges at the bottom, thus producing inches of water. banking the vehicle in the air. By a
the ground effect: pressure higher than The cover hooding the slit at the top simple modification of the vehicle'

H Y D R O F O I L CRAFT using the Coanda effect has achieved speeds long. The wings on the hull are for stabilizing the craft and for
of more than 19 knots in this pool. The model is about four feet generating a ground effect; propulsion is from an underwater foil.

K E Y P A R T of the hydrofoil is this foil with several slits for

e
tends to curve around the foil because of the Coanda e f f e c t , thei'
achieving the Coanda effect. Water emerging from the slits (right) by creating a negative drag similar to that in the airborne t-b

90
 jopn it has been endowed with forward
propulsion. In this model the configura-
| (jon of the body is asymmetrical, and
I fjje resulting asymmetrical distribution
Several thousand scientists and engineers
| 0f forces in the jet sheet provides for-
| ^ai'd thrust to the vehicle. will see this message, but...
The Coanda system offers several im-
i provements over the present versions
1 gf ground-effect machines. It should
I j-educe the amount of power needed for
| propulsion, reduce the weight of the
ONLY
I (jjaehine, permit steering of the vehicle
I by banking and make it easier to mount A FEW
CAN
| slopes.
The estimated lift efficiency, of the
I present rather crude model is about 25

QUALIFY ,
j pounds of lift per horsepower for a lift-
! off height of about a twentieth of the
| diameter of the vehicle. This perform-
ance is comparable to that of ground-
• Because of the nature of our work, not everyone can meet our
effect vehicles operated by conventional
exacting requirements. We are engaged primarily in analytical
means. It remains to be seen, however, and theoretical studies dealing with questions of strategic and
if a vehicle of the Coanda type will tactical warfare, economic development, and space systems —
retain comparable lift efficiency when it essentially problem solving—requiring disciplines in the physi-
is scaled up to full size. cal sciences, engineering, mathematics, behavioral sciences
and economics.
I Another vehicle currently under de- For those who meet our high standards and join our growing
I -• velopment in our laboratory is a company, the rewards are extraordinary. Our professional staff
is small. Each member has great creative freedom in the quiet
| boat of the hydrofoil type, that is, a
of his own office. Each finds personal pride—and endless chal-
I craft in which the foil lifts the hull out lenge—in his work. And there is tremendous professional satis-
! of the water. In this case, employing
faction in contributing to the solution of problems of national
| the Coanda effect, the hydrofoil is self- and international import —problems that must be solved.
: propelling: it provides not only the lift Moreover, you will be encouraged to continuayour education
: and controls but also the propulsion of or teach (our Research Center is less than a mile from the
j the craft. Nozzles that emit jets of water campus of the University of California at Santa Barbara). The
j are incorporated in the leading edges working climate and personal benefits are among the most
j of the foil, the strut and water-intake satisfying and progressive to be found anywhere.
j pods. The craft has a hull with down- During the next few months, we will add to our staff profes-
; swept wings that I have designed to sionals capable of solving problems in these areas of activity:
i minimize drag, improve the craft's sta- SYSTEMS ANALYSIS BEHAVIORAL SCIENCES
RADAR THEORY A N D ANALYSIS ECONOMICS
bility on the water and provide some ad- ELECTRONIC COUNTER MEASURES PHYSICS
; ditional lift by the : ground effect [see OPTICAL SYSTEMS THEORETICAL AERODYNAMICS
illustrations on opposite page].
If you possess intellectual curiosity and an outstanding recfrd
Our present model, four feet long of academic and professional achievement, we'd like to hear
and weighing 22 pounds, cruises at bet- from you — whether or not you are presently considering a
i ter than 20 miles per hour at the end change of association. We will send you a brief brochure which
of a tether ill a circular pool 10 feet describes DRC in more detail and introduces present members
in diameter. It can lift its hull out of of our professional staff. We'll also send you Evidence of the
delightful life available to those fortunate enough to live and
the water and ride on the hydrofoil at
work in world-famed Santa Barbara.
very low speed and without using any
Please write in confidence to Mr. John S. Adams, Assistant
more power than is required for cruis- to the President.
ing. Control of the craft is accomplished
by means of surfaces that do not pro-
trude into the flow. Thanks to the large
entrainment of fluid from the vicinity of
the nozzles, the foil generates essential-
ly no wake, very little spray and little
noise. We believe the use of a "ventilat-
DEFENSE RESEARCH CORPORATION
ing flow"—achieved by injection of small ... contributing to the solution of problems of national importance
amounts of air—will minimize undesir-
POST OFFICE BOX 3 5 8 7 • SANTA BARBARA, CALIFORNIA S3105
able effects caused by cavitation (the
(Opportunities also exist in the Washington, D.C. area)
formation and collapse of bubbles
around the foil). If we are right, a full- AN EQUAL OPPORTUNITY EMPLOYER

scale version of the craft should be ca-

pable of speeds of more than 80 knots,

9I
COANDA N O Z Z L E is shown disassembled at left and assembled at fect. The conical part is the afterbody; pressure over it is slightly
right. The plastic part is a shroud that enhances the Coanda ef- above atmospheric when the nozzle is operating to achieve thrust.

with a considerable saving of power In some applications most of the air for It is in the field of propulsion, of
over conventional high-speed hydrofoil combustion is supplied by the nozzle's course, that the prospects for employing
boats. entrainment of air. In addition the noz- the Coanda effect look most attrac-
Aside from the uses of the Coanda zle can be used as a pump to recirculate tive. The promise of attaining a propul-
effect for propulsion systems, which are unburned gases and thus produce blue- sion system that would produce its own
still in a highly experimental stage of flame, smokeless combustion. To these dividend of power from the very nature
development, there is another type advantages are added mechanical sim- of its design, that would have no mov-
of application in which the principle plicity, compactness, comparatively low ing parts such as propellers or fans, that
has already proved its value. Nozzles cost and the absence of moving parts. could be blended into the contour of a
based on this principle are extraordinar- Attempts are now being made to adapt vehicle without wasting space or sig-
ily efficient as components of burners the Coanda nozzle to the burning of nificantly disturbing aerodynamic char-
[see illustration below]. When properly heavy fuel oils. Nozzles of this kind acteristics, that would be smaller, lighter
shrouded, they produce an almost in- might also be employed in vehicle- and quieter than present engines—is in-
visible flame with no smoke and essen- propelling engines, and this possibility deed an incentive to further exploration
tially complete combustion of the fuel. is now being considered. of the Coanda effect.

T Y P I C A L F L A M E in a burner using a Coanda nozzle is blue, shrouded. Almost total combustion of fuel is achieved. Much of a11'
smokeless, quiet and well concentrated when the nozzle is properly for combustion comes from entrainment of air by the Coanda effect.

92