25.

September 2008

Alfred Evert

07.09. Schauberger - Repulsine - Redesign

Viktor Schauberger
At 25. September 2008, fifty years ago, Viktor Schauberger died. This Schauberger-Year
inspired to reconsider work of that great nature scientist and public interests increased.
Especially his Ufo (see known photos, e.g. here at picture 07.09.01 upside) is most striking,
as its told it did tear away from founding and flow off through roof of workshop.
Central constructional elements were corrugated iron disks (see picture 07.09.01 at the
middle), some of still exist in original state. However up to now, no working reconstruction is
known. Between two such disks, mander-shaped canal is build, within which air is
centrifuged (see picture 07.09.01 at bottom). Schauberger used that construction- and
motion-principle at diverse machines,
e.g. also at a Klimator for heating or
cooling rooms, for activate living water
and a home-power-station.
Schauberger all times pointed out
function of suction and nature-conform
motion. Many of his applications for
water-handling and agriculture were and
are successful. However his intension to
build an engine for decentralized power
supply, running without consumption of
common fuels, up to now was not
realized. At this chapter, movementprocesses of that Repulsine are
analysed and proposals for reconstruction respective new
construction are presented.
There are lots of assumptions
concerning function of these doubledisks with its concentric valleys, flanks
and mountains, e.g. also concerning
kind of material and working medium.
My workouts however are concentrated
only at pure mechanical aspects and at
movements of just normal air. Effectiveness of that construction, by view of my
considerations and experiences of last months, are based at only two well known effects.
Water-Jet-Pump
At picture 07.09.02 left side, general principle of well known water-jet pump is sketched,
which analogue works by air streams. Flow A (at the following named main-stream) exists
within a pipe. Aside of pipe an influx B (at the following named side-stream, in German
called wrong-air) exists, where additional air flows into pipe. At first, molecules of influx B
just stand locally and represent no stronger resistance than main-stream is affected by
friction on inner faces of pipe.
Particles of that resting air at B are trembling, based at normal molecular movements, thus
are flying into all directions from one collision to next. Just by pure chance, any particle will
be pushed into direction of main-stream. That particle moves ahead within main-stream, i.e.
will not come back to influx-area. At its original location, that particle no longer is available as

collision-partner, so is leaving some empty area there, into which following particles will fall.
As many particles behave likely, flow C comes up within side-influx.
That pipe downside at D shows wider cross-sectional surface than upside at A. Normally
(see Bernoulli formula) such an extension results deceleration of flow and thus destroys
kinetic energy. Here however that wider cross-section if filled up by air mass of side-stream.
All air-particles any time are moving by speed of about 500 m/s, also within a flow. When air
is flowing, vectors of particle movements only are showing some more forward in flows
direction, i.e. instead of local trembling these movements are wandering some forward. The
faster a flow is, the more all movements show into likely directions. Finally at well structured
flows, particles fly nearby parallel and come forward one nearby next. The stronger a mainstream is, the more side-stream will enter with less resistance. Up to sound-speed of about
330 m/s, additional influx aside can follow original main-stream and is integrated without
problems.
The original stream practically works like
suction, into which particles of additional
stream disappear. These pumps work really
effective because mass throughput of sidestream practically is done by no additional
energy-input. There is no acceleration of
particles of side-stream (via power-input) nor
deceleration of main-stream. Only some of these
influx-particles (by given speed) fall into mainstream and move forward within - and other
particles of influx area again fall behind into that
direction (with their given speeds).
So if an original main-stream exists, mass
throughput can be strengthened by skilled merge
of flows, where static pressure of side influx
partly is transferred into dynamic pressure of structured flow. So kinetic energy of that sideflow is usable e.g. for generating mechanical turning momentum.
Laval-Nozzle
Second effect used here is based on function of Laval-nozzles, where flows of (ultra-) soundspeed are generated. At picture 09.09.02 right side, schematic is sketched a longitudinal
cross-sectional view of Laval-nozzle. Motion-processes of such nozzles in details are
described at chapter 06.03. Ultra-Sound-Engine and once again at chapter 07.06. WindTower-Electric-Generator. At the following thus important function of Laval-nozzles is
mentioned only in brief.
Within a pipe E (grey) a flow must exist e.g. by 10 to 30 m/s, so corresponding to strong-wind
or storm, nevertheless far below sound-speed. Despite of, air particles move by about 500
m/s and tremble ahead on much longer ways, like here marked by zigzag-lines. At area of
narrowing (at F) particles are rejected from inner face of pipe each shorter distances and
thus particles collide after shorter ways and corresponding shorter time. Air there is more
dense and particles collide much more frequent.
So probability is strongly increased, several particles collide same time, e.g. two particles
transferring their kinetic energies onto a third particle. That third now flies off by over-size
speed, while both previous particles remain rather weak. As these multiple-collisions occur
within general forward-motion of flow, that third racing particle flies through bottleneck G
e.g. by sound-speed, while these both stand-around particles remain within bottleneck.

With their slow motions they hang around and thus show few resistance for following
collisions, so are good candidates to become next racers.
At outlet H these Laval-nozzles show wider cross-sectional surfaces, so particles fly into
relative empty space and thus come forward long distances until next collision. These
particles fly into likely directions and if there again multiple-collisions occur (by acute angle)
particles can move by ultra-sound-speed.
Function of Laval-nozzles is approved and used in many technical applications. That effect
might be calculated by formula of flow-sciences by some efforts (or mathematical tricks) however main-stream sciences wont agree thats a case of energy-surplus, not allowed by
energy-constant view. Finally by previous explication (respective detailed descriptions of
mentioned chapters) real understanding of that process is achieved.
Laval-nozzles manipulate spreading of speed of all particles. Normally, air particles move by
about 500 m/s only in average, however each single particle actually has an other speed
(where Gau-spreading is assumed). Within Laval-nozzle that normal spreading is changed
as some particles actually move much faster and other particle actually move much slower.
Mass-throughput ahead of and behind of Laval-nozzle is constant, however kinetic energy at
outlet is increased. There at output, particles fly much faster, within one time-unit however
less particles cross that section.
That flow is ideal for merging-in side-stream respective for combination of both effects. Mainstream (off Laval-nozzle) shows relative less density, so particles of side-influx can dip into
main-stream with minimum resistance. Particles of side-stream occasionally are pushed into
fitting direction by their normal speed of about 500 m/s (respective come forward by their
zigzag-ways at least with sound-speed). Behind Laval-nozzles thus comes up not only flow
by sound-speed but also multiple mass-throughput is achieved.
Side-Stream and Laval-Nozzle
Based on these considerations,
combination of both effects should
result an engine with sufficient
performance. Now should be checked,
whether and how these effects were
realized within Schauberger Repulsine.
Picture 07.09.03 upside left shows a
view on these waved disks where
downward showing flanks are marked
red and upward showing flanks are
marked blue. Black zigzag-line
represents upward-downward track of
air at its way from centre to border of
that rotor.
That picture upside right shows part of
canal between both disks (grey), which
are mounted at plan supporting disk. When disk is started turning, also air will become
turning around system axis, based on friction of air at inner faces. Canal between both disks
practically is a disk-shaped space with wave-shaped contours. From centre (right), air flows
downward-outward (A, red), can fly outward unhindered within valley (B), afterward flowing
upward again (C, blue) towards border (left) of disk.
Here, cross-section of canal is drawn constant. The further outward air wanders, the wider
space is available, i.e. density of air decreases. Air particles all times wander towards area of

less density, because there they can fly longer distances between collisions. Schauberger
had installed some openings D within upper sheet, through which wrong air was allowed to
enter canal. So mass-throughput is increased by that side-stream, corresponding to previous
considerations.
Picture 07.09.03 at bottom shows canal, where contours of both disks (nearby) are sinusshaped. Resulting are sections with increasing cross-sectional surface (light blue) and
sections with decreasing cross-sectional surface (dark blue). At middle of each up- resp.
downward showing flank, bottlenecks are build. So that canal from inside towards outside
represents a sequence of Laval-nozzles.
Behind a bottleneck, air can flow into relative empty space (like e.g. at E) and especially
through valleys or over the mountains, air can flow outward unhindered. Afterward, air is
dammed-up at area of more narrow cross-sections (like at F) and finally can move through
bottleneck (like at G) by sound-speed, again flying into wider space. If both wave-disks were
build by such contours, effect of accelerating flows by Laval-nozzles is realized at that
Repulsine.
Laval-Nozzle plus Side-Stream
Schauberger mostly mounted these waved-sheets at plane supporting disk (grey beam at
previous picture), so openings for wrong air could be installed only at one side. Naturally
canal could be build also between disks more thick with corresponding contours, like
sketched upside of following picture 07.09.04.
Here contours are mirrored (each phase-shifted), however no longer sinus-shaped. Again
there are sections with increasing cross-section-surface (like at A, light blue), where upside
and downside air can fall outward relative unhindered. At the other hand, now bottleneck is
arranged more narrow (like at C), so in front of comes up area of high density (like at B, dark
blue). Additional influx now could come through openings from both sides (like at D).
That construction thus would
allow effect of Laval-nozzle
within canal and also increased
mass throughput is achieved
by additional influx at both
sides. However, at least at this
drawing, wrong air enters little
bit too late, because near most
wide cross-section. Now must
be checked whether that
waved tracks are necessary at
all or if no clearer design and
easier construction is possible.
Flat symmetrical Disk
A suggestion for solution is shown at picture 07.09.04 at bottom. Canal here is build between
two symmetric constructional elements. Air wanders from centre to border (here form right to
left) at rather straight way. Cross-section of canal is a sequence of narrowing and extending
sections (here probably over-drawn).
The air is dammed-up at area of narrowing (like at E, marked dark blue), air passes
bottleneck (like at F) and flies outward into following area of wider space (like at G, marked
light blue). Thus Laval-nozzles were build, all around, where air is accelerated up to soundspeed. Short behind bottlenecks, side-stream from both sides is sucked in (like at H), so
stronger mass throughput is achieved.

Suction works up to sound-speed, so one could use relative small number of Laval-nozzles
(like here these three instead of fife at upside zigzag-track). At the other hand here is shown,
also influx aside could be nozzle-shaped, so wrong air enters by fitting speed and direction
into main-stream.
General Track of Flow
By that version more simple, now one can understand
why Schaubergers Repulsine was self-accelerating.
Picture 07.09.05 shows a view from aside towards rotor,
which is assumed left-turning (like always here). Ringshaped areas of narrow canal-sections are marked dark
blue, areas of extended cross-sections are marked light
blue. General track of an air particle, representing all air
movements, here is marked by black, yellow and red
arrows.
When starting system, air within canal also becomes
rotating around system axis, based on friction at canal
inner faces. Air moves forward within space (in turning sense), at first however air moves
backward relative to rotor. Afterward in running mode, air however will turn faster than rotor,
as discussed later.
Air enters canal by central opening A and is pressed into area more narrow (dark blue).
There exists relative high density and particles are accelerated strongly in turning sense of
system, direct by friction at inner canal surfaces or indirectly via mutual collisions. These
particles thus in principle move at circle-track, from which they can move outward through
bottleneck rather slow. As a whole, thus particles move at an opening spiral track, like
marked by yellow arrow B.
Within area of bottleneck (transit from dark to light blue) movement of particles is directed
outward-forward, so tangential or some more outward. Flow strongly becomes accelerated
within bottleneck and particles can fly outward into following area of wider cross-sectional
surface (light blue) relative unhindered, like marked by red arrow C, in nearby tangential
direction. At this area now additional influx from aside enters main-stream (here not drawn),
so wider space further outside is filled up by additional air masses.
Generating Turning Momentum
That air-mass with its high speed now hits onto dammed-up air in front of next narrowing
(dark blue). That stream can not move further tangential-outward, but is redirected some
more into circled track (with corresponding increased angle-speed). Opposite, that air-mass
by itself presses air in front of narrowing forward in turning sense. Air can escape through
bottleneck outward only relative slow, like here marked by yellow arrow D.
That pressure within bottleneck is directed further forward and consequently now air-particles
fly off bottleneck straight forward into next wide space (light blue), as marked by red arrow E.
That movement process is repeated. However finally at third narrowing (dark blue), air no
longer will be accelerated by friction of rotor surfaces. Air-masses here move faster forward
than rotor is turning there. Via friction now rotor is accelerated by that fast moving air with its
high density in front of next bottleneck, like marked by long yellow arrow F.
After relative free flight through last extended section (light blue, with red arrow G) surplus
speed via friction is transferred into turning momentum within a final narrow part of canal
(grey, like upside at pictures of wave-shaped Repulsine).

That movements process describes reason, why Repulsine was able to turn up until soundspeed. If rotating air of outlet of that rotor is guided back to central inlet, main-stream there
already is turning and no longer must be accelerated in turning sense. If that backflow is not
controlled, system can turn up self-accelerating until self-destruction. If that outlet-air is
guided downward, that Ufo-Repulsine can fly off (where additional effects of ether-motions
come up, based on fast rotation of solid bodies - however this can be discussed only by
many new chapters of Ether-Physics).
Repulsine with flat Rotor
Picture 07.09.06 shows previous flat rotor with its symmetric constructional elements, now
with some additional elements. At picture left, side-view at rotor is sketched, at picture right
side longitudinal cross-sectional view through machine is sketched (only upper part by some
larger scale). That rotor is drawn with horizontal shaft, which is beard within housing GE
(grey). This rotor functions as turbine (red), some parts however function as pump (green).
These drawings only show construction in general and are not true to scale. For example,
canal here has only three bottlenecks, while in real machines more (and smaller) nozzles
could be used. At this version, canal is build between four ring-shaped elements T1 to T4
(light red). These constructional elements are fix connected by spokes SP (dark red, her e.g.
six).
Inlet of main-stream
enters through
openings between
turbine-shaft TW (dark
red) and first ringelement T1 (see arrow
A). Between shaft and
that ring, pump-blades
PS (light green) are
installed, transporting
and pressing air from
both sides into first
bottleneck (dark blue).
Afterward, air flows into
wider space, where mass throughput is increased by influx of wrong air through openings
between first and second ring-element (T1 and T2, see arrow B).
These air masses are moving forward-outward into next narrowing, again are accelerated
within next bottleneck, afterward flying into next extension area, where additional air enters
from aside (see arrow C). That side-influx thus occurs through ring-shaped openings (light
blue) between these ring-elements (T1 and T2 plus T2 and T3), only interrupted by spokes
SP.
As mentioned upside, air is accelerated within nozzles in direction forward-outward, sooner
or later beyond turning-speed of rotor. These fast flows affect thrust at canal-surfaces (by
friction, direct or indirect), especially at area of narrowings and relative high density there.
Before air is leaving turbine, once again that thrust-situation should be organized. In addition,
here that last narrowing is curved, so flows of radial level are redirected into axial direction
towards both sides of rotor.
Air flows off turbine (see arrow D) through that outlet (dark blue) between ring-elements T3
and T4. Both ring-elements again are connected by spokes SP. However, within that outlet
well could be installed additional turbine-blades TS (dark red). Via redirection of flow at these
blades, additional turning momentum is generated at long lever arm.

Air of main-stream disappears through central inlet-openings (transported by pump-blades

PS) and also side-stream is sucked through openings of wrong-air. Air following from
outside towards centre (see arrow F) will not flow radial inward but at spiral tracks inward
curved. Air still is turning at outlet of turbine. That rotation however should be slower, so a
whirlwind comes up. Static pressure at circumference thereby is transmitted into kinetic
pressure of faster turning flows near centre. System should not be closed hermetical, but
housing should be partly open, so atmospheric pressure (see arrow E) weights on inward
directed vortex, pressing and accelerating flows towards centre.
Controlling-Problems
Based on previous considerations, Schauberger Repulsine achieved acceleration of flows by
Laval-nozzles and achieved increased mass throughput by influx of wrong air, both
demanding no additional energy-input. By previous relative small and symmetric rotor, these
effects are used straight and consequently.
Bibliography of Viktor Schauberger tells, diverse versions of Repulsine (or Repulsator) were
build in 1940 and following years, in the fifties also several Heimkraftwerke (home-powerstations), however all times regulation of revolutions made problems. Obviously its hard to
build well controllable engine by only one rotating constructional element. It might make
sense to use one (controllable) component for drive and
an other component for taking off turning momentum. So
functions of pump and turbine should be realized by
different assemblies.
Picture 07.09.07 once more shows previous longitudinal
cross-sectional view, however now that disk-like rotor
functions only as turbine T (red). Ring-shaped elements
of turbine still are connected by spokes SP, which now
reach inward to turbine-shaft. At both side now pumps
are installed at separated hollow-shafts (dark green) with
corresponding pump-blades PS (light green). These
pumps transport air into first narrowing of turbine and air
already is rotating in turning sense of system.
Mass-throughput of main-stream is controlled by revolutions of pumps. If pumps stand still,
air through pump-blades moves contrary to general turning sense, so system will slow down
and stop.
Re-Design und Re-Production
Viktor Schauberger nowhere had clearly defined that terminus Repulsine and also had
never explained understandable decisive movement processes of that conception. Instead of
he mentioned, quality of material would be important and transformation of working medium
would occur - still not to grasp by common knowledge. So its not astonishing, many most
different explanation of Repulsine exist and some of still sound rather mysterious.
I suggest at first to concentrate at simple fluid-mechanic. I think these two well known effects
of water-jet-pump and Laval-Nozzle respective combination of both processes is decisive
secret. That will do for accelerating flows and increasing mass throughput - without
corresponding strong input of energy. It would be fine if based at that new understanding,
now finally a running Repulsine could be re-constructed.
Evert / 25. September 2008 / www.evert.de