Valveless Pulsejet Engines

 

Unlike the rest of the website, this section was prepared by Bruno Ogorelec.

Comments and other input can be addressed to me via the link on the top of

the page, to the forum on this website, or to Bruno directly at terna@zg.tel.hr

  

WHY PULSEJET?

The pulsejet was one of the first operational jet engines.  While the opposed sides in World War II were busily trying to put
together the first jet-powered fighter aircraft in 1944, the Vergeltungswaffe 1 (or V-1 for short) was regularly buzzing its
way to England with its 1,870 lb load of explosives.  It was powered by the Argus As 109-014 pulsejet engine designed by
Paul Schmidt.

For some time after the war, the simplicity and low cost of the pulsejet looked hugely promising.  Scores of researchers and
engineers threw themselves at it to see how far it could be developed.  After a while, some of them seemed to be on the trail
of significant improvements.  However, real success proved elusive and during the 1950s and 1960s, most researchers
gradually abandoned their efforts and turned to other things.

On almost all the important counts at the time, the turbojet eclipsed the pulsejet.  Even the improved pulsejets were
comparatively inefficient, horribly noisy and vibrated madly.  If they depended on reed valves, they were short-lived and
unreliable.  Also, there was little they were really good for.  For a while, it looked like they would power small helicopters
(but never made the grade), there was some French use on powered flying drones, and a modified version of the German
Argus tube was briefly commissioned to power a few missiles for the US military.

By the mid-1960s, even those limited uses petered out and only a few isolated enthusiasts still considered the pulsejet as a
potential aircraft powerplant.  The noisy tube was perceived as a blind alley and relegated to the role of model aircraft engine
and such humdrum applications as an efficient combustor for central heating systems, a power unit for agricultural spray
dusters and a blower and shaker for industrial slurry drying machinery.

The reason I am looking at the pulsejet as a propulsion unit now is the change of circumstances.  Sometime in the early
1980s, ultralight fun flying started getting increasingly popular due to the availability of good, simple and affordable flying
platforms – hang gliders and paragliders.  When provided with motor power, these machines offered unprecedented freedom
of flight to anyone interested.

However, all the piston engines currently used in ultralight flying are heavy and cumbersome, even in their simplest form. 
They also require much ancillary equipment, like reductors, prop shafts, propellers etc. etc.  Having all that gear mounted on
a lightweight hang glider virtually defeats the original purpose.  Turbojets are just as bulky and also horribly expensive.

In this light, the simple, cheap and lightweight pulsejet engine starts making a lot of sense.  Its relatively high fuel
consumption and noise levels need not be of major importance for such applications.  In addition, together with vibration,
they can possibly be designed out of the concept.  It is perhaps time to blow the dust off the old tube.

WHY VALVELESS?

The ordinary pulsejet is already a very simple engine.  It is just a piece of tube cut to the required dimensions, with a few
small flaps and a fuel jet at one end.  So, one might ask, why go that one small step further and eliminate the valves?

The prime reason is that in most designs the use of flap valves limits the reliability and longevity of the engine.  Valves of
the As 109-014 lasted for only about 30 minutes of continuous use.  Given that its role was to destroy itself in the end, this
was not considered a big fault, but today you might have a flying model that is your pride and joy up in the air, or you may
even want to fly yourself.  You really need your engine to last a bit longer.

Admittedly, development has improved the design in many ways and stretched its working life from minutes into hours, but
the fundamental problem remains.  It is difficult to get around the fact that the valves in a pulsejet are supposed to satisfy
conflicting demands.

To open and close quickly (in the interest of efficiency), they have to be as light as possible.  Because they have to endure
great mechanical stress (bending open and slamming shut at high-speed) and do it in a high-temperature environment, they
have to be very tough.  If something has to be light, yet exposed to great abuse, it either spells short life or exotic
technology.  The former is impractical and the latter is expensive.

There are other problems.  Theoretically, flap valves can be arranged so that they are not too much of a nuisance in terms of
fluid dynamics, but in most practical designs they do represent a big obstacle to the incoming air.  The reed valve array
produces aerodynamic drag.  A lot of valve area provides relatively little effective intake area.  Instead of smooth passages,
the valve gear offers various edges and projections to the air stream, producing terrible turbulence.  OK, you want some
turbulence to help the atomization of fuel and mixing with air, but not so much that it consumes half your power.

Finally, there is a question of elegance.  I find the idea of a jet engine that is actually just a cheap empty metal tube without
moving parts very appealing.  Making the various gases jump through hoops and produce useful tricks without resorting to
any mechanical complexity is a nifty thing that will be appreciated by all lovers of simplicity and elegance.  (I am talking of
elegance in the mathematical sense -- desired result achieved with minimal complication.)

     HOW DOES A VALVELESS PULSEJET WORK?

The picture shows one of the possible layouts of a valveless pulsejet engine.  It has a chamber with two tubular ports of
unequal length and diameter.  One port, curved backwards, is the inlet pipe and the other (flared at the end) is the tail pipe. 
In some other valveless engines, it is the exhaust pipe that is bent into the U-shape, but the important thing is that both ports
point in the same direction.

 
 

When the fuel-air mixture combusts in the chamber, the pressure inside rises very suddenly.  To an observer it looks and
sounds like an explosion.  The rising pressure forces the hot gas to expand out of the chamber and pass through the two ports
at high speed.  As it leaves the engine, the hot gas exerts thrust.

As the gas expands, the pressure inside the chamber drops.  Due to inertia, the expansion continues even after the pressure
falls back to atmospheric.  At the lowest point, there is partial vacuum in the chamber.  At that point, the momentum of the
expanding gas is spent and the expansion stops.  The process reverses itself and fresh air starts rushing into the ends of the
two ports to fill the vacuum.

At the intake side, it quickly passes through the short tube, enters the chamber and mixes with fuel.  The tailpipe, however, is
rather longer than the intake, so that it takes incoming air longer to reach the chamber that way.  One of the prime reasons
for the extra length is to have some hot exhaust gas remain inside the tailpipe at the moment the suction starts.  This
remaining hot gas will now be pushed back towards the chamber by the incoming fresh air.  When it enters the chamber and
mixes with the fuel/air mixture, the heat and the free radicals in the gas will cause ignition and the process will repeat itself.

It took me almost 250 words to describe it, but this cycle is actually very brief.  In a small (flying model-sized) pulsejet, it
happens 100 to 250 times a second.  The cycle is not much different, really, from that in the conventional flap-valve pulsejet,
like the Dynajet.  There, the rising pressure makes the reed valves at the front of the chamber snap shut and there is only one
way for the hot gas to go -- into the exhaust tube.  In the J-shaped and U-shaped valveless engines, the hot gas spews out of
two ports.  It does not matter, because they both face in the same direction.

Some valveless engine designers have developed designs that are not bent backwards, but employ various tricks that work in
a similar fashion to valves -- i.e. they allow fresh air to come in but prevent the hot gas from getting out through the intake. 
We shall describe some of those tricks at a later point.

     KADENACY OSCILLATION vs. ACOUSTIC RESONANCE

Before getting into details of actual engine designs, let’s get another theoretical point out of the way.  In the explanation of
the working cycle, I described how inertia drives the expanding gas out until the pressure in the chamber falls some way
below atmospheric.  The opposite thing happens in the next part of the cycle, when the outside air pushes its way in to fill
the vacuum.  The combined momentum of the gases rushing in through the two opposed ports causes the chamber briefly to
be pressurized slightly above atmospheric.

 
 

There is thus an oscillation of pressure in the engine caused by inertia.  The pressure swings from way above atmospheric to
partial vacuum and back again, in damped oscillation (see the graph on the right).  This is called the Kadenacy Effect.  It is
what makes the aspiration (removal of burned gas and replacement with fresh fuel-air mixture) possible.  Without it,
pulsejets would not work.

However, some people prefer an acoustic explanation of the same process.  They say that the pressure swings are caused by
acoustic resonance.  The explosion gives the impulse that makes the engine tube and the gas within it resonate, just like a
hammer blow makes a bell ring.

A standing wave forms due to the reflection of waves from both ends of the tube and creates a peculiar distribution of gas
pressure and gas speed.  At any point of the tube, the speed of the gas and its pressure are out of phase by 90 degrees (a
quarter wavelength on the sine curve).  The consequence of this fact -- in very simple terms -- is that the combustion
chamber is the place of the greatest swings in pressure but the smallest changes in gas speed.  The intake and exhaust ports
are the places where the pressure swings are minimal but the speed changes are the greatest.  So, the chamber provides the
pressure peaks and troughs and the ports allow the gases to rush in and out at high speed.

Undoubtedly, both descriptions are true.  Both effects are obviously at work.  In a very roundabout way, they are just
different manifestations of the same thing.  I will risk being controversial here by saying that – theory notwithstanding --
they do not appear the same in practical use.  Pulsejets do not always behave exactly like acoustic resonators.  The classical
acoustical phenomena take place at low pressure changes, low gas velocities and very little gas displacement (mass
transport).  In pulsejets, we see great pressure variations, high gas velocities and great mass displacement -- at high
temperatures.

I think that pulsejets follow their own, distinctive, Kadenacy-like cycle of compression and rarefaction powered by the self-
excited explosive combustion process.  There is no doubt, however, that the acoustic laws can and do modify the inflow and
outflow of gas, often significantly so.  Because of that, one should watch out for resonance, knowing that the regular
pressure impulses will inevitably set up standing waves and influence the flows.  The negative influence of resonance must
be avoided and – if possible – the positive influence harnessed to help the engine along.  Some designs do this better than
others.

 
 

     ENGINE DESIGNS

     Marconnet

The first valveless pulsejet engine ever -- the grandfather of all subsequent valveless designs -- was the one designed and
developed by Georges Marconnet, a Frenchman, in 1909, as he was trying to develop a simple combustor for a gas turbine.

He figured that a blast inside a chamber would prefer to go through a bigger exhaust opening, rather than squeezing through
a relatively narrow inlet.  In addition, a longish diffuser between the inlet and the combustion chamber proper would direct
the charge strongly towards the exhaust.  He tolerated what hot gas did escape from the intake.

In their descriptions of the Marconnet engine, F. H. Reynst and J. G. Foa (both noted experts on pulsating combustion)
agreed that it could not have worked very well, really requiring forced air at the inlet (by a fan or a similar device) if the
blowback was to be avoided.  Foa actually called the Marconnet “a bad ramjet” on account of the need for some ram
pressure at the inlet.  It does resemble ramjets of a few decades later rather closely.

While it may not have been a very practical jet engine, the basic idea behind Marconnet was good and it spawned almost all
the other pulsejets in the 20th century.

 
 

Schubert

One of the simplest successful valveless pulsejets of all was the 1944 design of Lt. William Schubert of the US Navy, called
the “resojet”.  (Later the same term was applied to other valveless pulsejets, most notably the Lockwood-Hiller design.)  At
first sight, it is just a more angular Marconnet.  However, Schubert carefully calculated the geometry of the intake tube so
that the exhaust gas could not exit by the time the pressure inside fell below atmospheric.

The resistance of a tube to the passage of gas depends steeply on the gas temperature.  Thus, the same tube will offer a much
greater resistance to outgoing hot gas than to the incoming cold air.  (I guess that those more at home with the gas
thermodynamics will prefer to say that the acoustic impedance is inversely proportional to the square root of the gas
temperature.  This is not my forte -- I am only quoting this from a textbook.)

With such tuning, the cool air necessary for combustion managed to get in during the “vacuum” part of the cycle, but the hot
gas theoretically did not manage to get out during the high-pressure part.  In practice, while it worked better than the
Marconnet, the Schubert engine also needed to move forward at some speed (or to have air blown in by a fan when
stationary), so that ram pressure helped push the fresh air in and kept the hot gas from getting out of the intake tube. 
Without it, the machine would still spit some of the hot gas forward.  The intake tract long enough to prevent the blowback
completely choked the air supply too much for good performance.

The intake of a Schubert-type engine can be kept shorter and made more effective by employing a surge chamber.  A surge
chamber is a part of the engine that provides an enclosed volume of air, which acts as a damper on pressure oscillations in
the intake tube.  It communicates with the tube through holes or slits.  The device also lowers the noise to some extent.  (One
notable engine designer that experimented with surge chambers with a measure of success was Paul Schmidt, the inventor of
the modern reed-valve pulsejet.)
 

Logan

This is a very interesting American design whose principle of operation is identical to the Schubert engine.  Yet, it is said to
work substantially better than Schubert in practice.  Having the fresh mixture enter from the side appears to offer very good
mixing.  Reportedly, the Logan is sensitive to the exact placement of the intake tube.  I have not found more information in
the available literature.

In the 1950s and 1960s, a man named Joseph G. Logan was a prominent wave engine researcher, leaving behind a legacy of
notable designs (e.g. the Hertzberg-Logan pulsating combustor with rotary intake and exhaust valves).  I have not managed
to establish any tie to the Logan pulsejet, but there is a high probability that he also designed this simple device.

Trick Intakes, Baffles, Serrated Tubes, Convoluted Passages…

Like Schubert, many developers hoped to have gases moving through the pulsejet in one direction only.  Without a
mechanical non-return valve up front, one must employ some kind of a trick, like Schubert's choking intake tube.

Paul Schmidt also toyed with the idea of replacing the short-lived reed valves with a kind of aerodynamic flow rectifier -- a
device that would use the behavior of the gas itself to make it pass in one direction only.  He tested a number of designs
featuring umbrella-shaped baffles in the intake tract, which offered great resistance to back flow but let fresh air in easily. 
However, all the configurations he had tried produced lower thrust and consumed more fuel than the equivalent engines with
valves.

Many other people tried baffles of various shapes and configurations.  One of them is pictured to the right, with the arrow
showing the direction of fresh air during intake.  Air coming in encounters a series of concave baffles with increasingly
broader openings.  It flows easily down the concave slopes that form a kind of a diffuser.

In the opposite direction, however, hot exhaust gas will be trying to expand as it travels forward and increasing amounts will
be trapped in the pockets between the baffles.  Only a relatively small amount will ever be likely to escape.  At least, that
was what the designers hoped would happen.

Alas, the sad story of almost every valveless pulsejet that employs some kind of a trick flow rectifier is that such devices
never work as well as their designers hope for.  They pitch in only at high gas speeds, meaning that the engine will suffer
from at least some blowback at the beginning of each cycle.

Numerous versions of tubes with similarly serrated walls have been tried, sometimes with baffles/serrations awaiting
exhaust gas on more than one side.  The next picture shows a typical design of that family.

There are some obvious problems with the design.  The first is surely the bulk of the intake tract, especially of the central
serrated spike.  In addition, the problem with most serrated designs is that the return flow is not impeded as much as their
inventors would like because the exhaust gas quickly fills the concave ‘pockets’ in the tube sides and forms cushions of dead
air for the passing stream.  Under some gas speeds, the flow of gas in one direction will actually be very similar to the flow
in the opposite direction, as the gas cushions will allow the formation of tidy flow in both cases.

Few people will be surprised to hear that the amazingly prolific and spectacularly inventive researcher of things electrical,
Nikola Tesla, also turned his mind to the problem of pulsating combustion.  He wanted to have a good gas generator for his
neat smooth-disk rotor turbine that used the viscosity of the working fluid to transfer energy to a rotating shaft.  He
immediately saw that mechanical valves would not offer the simplicity and reliability he had sought.  So, he studied the
ways to rectify the gas flow aerodynamically.  Eventually he came up with arguably the best aerodynamic ‘valve’ ever.  Its
cross section is shown below.

At first glance, it looks like another serrated passage, but if you take a closer look, you can see that it does not really employ
either baffles or dead air pockets.  Instead, it just changes the direction of the gas and turns it upon itself.  At each turn, a side
blast of gas will push the main flow towards the side passage that eventually turns backwards.  The harder you blow into that
tube, the harder it will resist.

While undoubtedly ingenious, the device never found practical application to the best of my knowledge.  Tesla himself
probably did not have time or inclination to pursue its development after applying for patent, being busy with his
experiments in electromagnetism, and the patent was mostly forgotten.  As the inventor has recently become the center of a
cult following, his modern disciples have revived the idea, but I have never heard of anyone actually building the ‘valvular
conduit’ (as Tesla called it) and testing it on a pulsejet.

Escopette

Dissatisfied with the baffled intake designs, engineers at the French SNECMA (Societe Nationale d'Etude et de
Construction de Moteurs d'Aviation) corporation simply turned the intake tract backwards.  That way, blowback could easily
be tolerated, as it contributed to thrust.  To its designers, the machine looked like one of the old-fashioned musket guns and
they called it Escopette (which is French for musket).  It looks very similar to the picture we used in the introduction, with
one significant difference.

 
 

Namely, the intake does not curve backwards from the combustion chamber.  It points straight ahead.  What turns the hot
exhaust gases backwards is a separate curved tube mounted at some distance from the mouth of the intake proper.  So, the
engine breathes through the gap between the intake and the 'recuperator', as the designers called the curved tubular deflector.

If I am right, this extremely neat design deftly exploits the resonance by behaving as a closed-end tube in the expansion part
of the cycle and an open-ended tube in the intake part, using the 'recuperator' during the former and ignoring it in the latter. 
The gap between the 'recuperator' and the tube act like an aerodynamic switch between the two modes.  One of the neatest
designs I have seen.

It seems to have been effective, as Escopette is the only pulsejet ever to be commercially available to power people-carrying
aircraft.  It was delivered as an auxiliary engine on the French Emouchet SA 104 sailplane sometime after World War II,
enabling the pilot to take off and achieve soaring altitude without a tow plane or a winch (see the above picture).  It was a
small engine, producing only 22 lb thrust, so three were used in a cluster under each wing, looking as if somebody had stuck
three furled umbrellas at each side.

Kentfield

The idea of a 'recuperator' or deflector found several adherents who produced variations on the theme, some simple and
others complex.  J.A.C. Kentfield, one of the most recent researchers in the pulsejet field, tried to make up for the energy
lost in turning the gas flow around by introducing thrust augmentation to the 'recuperator'.  Instead of a simple tube bent
backwards, he employed a gently flaring cone, which let fresh air be sucked in by the hot gas stream as extra reaction mass. 
(I will talk about thrust augmentation in more detail in a later chapter.)  According to Kentfield, who patented the idea, the
gain more than offset the drag and turbulence losses incurred by the 180-degree turn.

Others experimented with variations on the theme.  Most deflectors were symmetrical and used internal vanes to help control
the flow and lower the turbulence.  Two such designs are shown on the next two pictures.  On the left is an obvious
derivative of the above augmenter, simply branching to left and right.  Reportedly, it is a very effective design.  There are
indications that it was also developed by Kentfield, but I am not certain.

The one on the right – whose provenance is unknown to me -- is rather more ambitious.  It attempts to harness ram pressure
(or ‘velocity head’) of the incoming air.  Ram pressure gives a welcome boost to the power output at no cost.  The J-shaped
and U-shaped engines as well as most engines with deflectors up front must forgo that advantage, as their intake ports are
either turned in the wrong direction or masked by the deflector structure.

This deflector has an almost straight path for fresh air from the front intake to the combustion chamber (see the lower half of
the picture).  The trick that prevents the exhaust gas from escaping through the same route may have been borrowed from
Tesla, but similar methods are also used in various other pneumatic flow control devices.

 
Note the two small airfoil-section vanes in the central passage, right behind the intake wedge.  When the hot exhaust gas is
pushed forward by the blast, the part blowing into the gap between the vanes is divided into two flows, one going upwards
and the other going downwards.  Each flow forms a kind of a gas curtain that cuts across the path of the main flow (see the
upper half of the picture).  The curtain deflects exhaust gas flow towards the curved passages that turn the flow around and
eventually eject it backwards.  As a result, almost all the exhaust gas that would normally be blown out of the intake port
gets deflected and contributes to the thrust.

Comparing the two is difficult.  What one gains from ram pressure, the other makes up for with thrust augmentation. 
According to Kentfield, the simple one on the left outperformed the more complex one on the right in laboratory testing.

Foa

Joseph G. Foa, another well-known pulsejet researcher, devised a very simple and compact deflector allowing direct
entrance of air and thus giving the benefit of ram pressure.  He took advantage of the inertia of gases and simply gave the hot
gas a direct entry into the deflector, while fresh air entered the front passage at an angle.  Despite the utter simplicity, the
device apparently functioned very well.  Given the simple layout, I am surprised that the design did not gain popularity
among amateur enthusiasts.

Chinese / Thermojet

As we said right at the beginning, some engines turn their intake backwards, rather than rely on trick aerodynamic valves or
deflectors.  Thanks to Don Laird who made a drawing according to a factory-built example in 1993 and Kenneth Moller,
who published the plans on this website at the end of the 1990s, a Chinese-manufactured valveless pulsejet engine has
become a fairly popular design among amateur engine builders.  Its origins are murky, but there are some indications that it
was actually designed in Europe.  In the 1960s, the engine was apparently imported to the USA and Europe from China and
offered to aero modelers.

The novelty of this engine is that its intake port (which also serves as the auxiliary exhaust) is not bent backwards from the
front of the chamber, but branches out near the exit.  That makes the combustion chamber into a dead-end vessel.  Instead of
mixture entering the chamber on one side and hot gas exiting on the other, the ‘Chinese’ has the chamber “breathe” in
reciprocating fashion.

This may help increase what little compression these machines are able to achieve, as the incoming mixture and the reflected
shocks from the tube ends both slam into the bottom of the chamber and generate some kind of a hammer effect.  If this
hammer is timed well, it can usefully increase the pressure of the fresh mixture at the moment of ignition.

However, this is theory.  In practice, by most accounts, it is not a very effective design.  One builder I know described the
output of his engine as “a hamster blowing through a straw”.  I have found references to the fact that pulsejets having intakes
and exhaust ports on the same wall of the chamber usually have problems with mixing of fuel and air.

This is really the only reason I can see for this engine to work any less effectively than, say, the Schubert or the Logan.  At
first glance, there is nothing particularly wrong with the layout.  In fact, to naked eye its auxiliary exhaust looks more
effective than that of the Logan.  At least it points in the right direction.  However, valveless engines are very sensitive to
small details.  It is quite possible that the Chinese manufacturer simply did not develop its geometry carefully enough and
the amateur builders, following the crude plans faithfully, simply replicated all its faults.

Here, a vignette from the history of the Dynajet, the most famous flying model pulsejet of all, might come in handy to
explain what I mean.  Mark ‘Thixis’, a pulsejet enthusiast and builder, found it in a book somewhere but neglected to tell me
which one, so I cannot credit it properly.

Early on, Tenney and Marks, the developers of the Dynajet, were shown the big Argus engine that powered the V-1 flying
bomb, and decided to build something similar themselves.  The first effort was rather big and “didn't make enough thrust to
make you blink”, according to the builders.  The second, built “after two weeks of burning the midnight oil”, didn't work at
all.  When they finally did get it to run, thrust was measured in ounces.  They came up with what we know as the Dynajet
only after building hundreds of different configurations and combinations, and spending long hours in the workshop.

Somehow, I just don’t see the developers of the Chinese engine burning the same amount of midnight oil.  My feeling is that
the potential of this design is rather greater than the practical results have shown so far.

 
 

Obviously, I am not the only one, as other people have tried to explore the same route.  At about the same time as the
‘Chinese’, model builders could also purchase the so-called Thermojet from selected model shops.  Manufactured by the
Thermo-Jet company, it had between two and four short parallel intakes flanking the long exhaust.  The most common
model, crudely drawn on the picture below, had two.  Effectively, it is a ‘Chinese’ with a stubbier combustion chamber and a
surfeit of intake tubes.  Splitting the intake area into several smaller tubes allowed the intakes to be shorter.

Like the ‘Chinese’, the Thermojet endeared with its simplicity and disappointed with its effectiveness.  Consuming vast
amounts of fuel, it generated rather poor thrust.  Again, I can offer no obvious explanation for this and must assume the lack
of proper development.  I could find neither the ‘Chinese’ nor the Thermojet described in any textbook or research paper and
have had to rely on modelers’ lore and descriptions from people I know who have built and tested those designs.

Some researchers thought the above layout offered great promise with a bit of tinkering.  One of the typical exercises of the
kind is the mid-1960s effort of a Frenchman, Rene Malroux, on the above picture.  I have no data on its performance.

Melenric
The most ambitious "thermojet" by far seems to have been the engine of John A. Melenric of the early 1970s.  The author
claimed his engine was good not only in the speed range normally achieved by pulsejets (i.e. under Mach 0.5) but almost up
to sonic speeds.  This was achieved by careful streamlining and ducting of fresh air.  (Melenric is not the only one to claim
successful high-speed application – a specially streamlined Escopette is said to have achieved Mach 0.85 powering a drone.)

One can see the basic layout of the Thermojet in the cross section, but it is also clear that Melenric projected the intake tracts
(three of them, radially disposed around the central exhaust tube) well into the combustion chamber.  The reason must surely
be the problem of bad mixing that apparently arises if the intake and exhaust ports are too close together on the same side of
the engine.  The front wall of the combustion chamber is profiled to resist pressure shocks, rather than to offer any aid to the
mixing of the incoming charge.  The front dome is empty and is cooled by air passing through a small inlet and outlet.

The second notable feature of the Melenric engine are the fresh air deflectors around the three-intake/exhaust tracts.  Their
scoops force fresh air to curve 90-180 degrees and either enter the intake tract (in the intake part of the cycle) or get blown
rearwards again through the back end of the scoops (in the expansion part).  This saves the engine the effort to suck fresh air
against the direction it wants to go and even provides a measure of ram pressure.  To some extent, the benefit is negated by
the increased drag, but at very high speeds, this is probably the only way such an engine will function at all.

 
 

In a similar way, air is ducted at the end of the main exhaust tube, so that at high speed, there is enough fresh air pressure at
the end for the engine to suck back during the intake phase.  During expansion, the duct serves as a thrust augmenter.

The third notable feature of the Melenric engine is the effort to utilize the waste heat of the engine to gasify the liquid fuel
and inject it into the engine as gas.  Fuel is led to the engine through a metal pipe coiled around the exhaust (see the central
part of the engine on the picture).  Exhaust heat boils and vaporizes the fuel, and the vapor is led to injectors that poke into
each intake tube.

This simple device saves one the trouble of providing some kind of a powered pressure injection system or from using
propane gas as fuel.  Liquid fuels are usually cheaper, and easier and safer to store and handle.  For flying applications, a
liquid fuel tank is much easier to engineer than a propane tank.  Finally, liquid hydrocarbon fuels are better power producers,
giving more heat per unit volume than propane.

Lockwood-Hiller

Despite the existence of neat deflectors and aerodynamic valves, the U-shaped Lockwood-Hiller engine is arguably the most
familiar valveless design by far.  This is possibly due to the effort of Ed Lockwood, inventor Ray Lockwood’s son, to
preserve his father’s legacy and keep it in the public eye, but there is no denying that the Lockwood is an effective valveless
pulsejet engine, quite possibly the best developed ever, despite its deceptively simple appearance.

Ray developed it in the 1960s, partly at the Fairchild and Hiller companies and partly on his own.  (The 1964 patent is in his
name.)  Despite its tubular looks, it works almost purely by the Kadenacy Effect, showing less sensitivity to exhaust tube
length than most other pulsejets.  The mixture is generated by the mixing of propane gas, which is injected through a jet
either built into the side of the combustion chamber or on a strut projecting into the chamber, or on two crossed struts
spanning the front part of the chamber.  The chamber is the drum-like broad part of the engine.  The short tube to the left is
the intake.  The flaring cone end is the exhaust.

Lockwood was prepared to admit that many of the features of his engine had been inherited from the Marconnet.  Some of
his papers even show a lineage of design from the Marconnet to his final U-shaped layout.  I am not convinced.  Some of the
French designs of 1950s vintage, themselves based on the Marconnet engine, must have been a more direct influence.  To
me, the Lockwood design looks like a brother to SNECMA’s Ecrevisse (‘Crab Claw’).  I have found this confirmed by other
authors.

The pictures above and below show two versions of the Ecrevisse.  One is a basic early model and the other a streamlined
later version with thrust augmenters.

Reportedly, the “standard” Lockwood/Hiller engine (Model HH 5.25-7) offered by Hiller pushed out 280 lbs of thrust.  It
weighed only 30 lbs.  I believe the latter figure, for it is just an empty tube.  However, judging by the efforts of enthusiasts
building Lockwood-type engines, the former figure is somewhat suspect.  It may have been available under laboratory
conditions, but no one else seems to be able to get anywhere near such high power outputs even when the layout was copied
slavishly.

Some observers have opined that Lockwood/Hiller may have been intentionally optimistic with their figures at the time, as
their efforts were obviously pitched at defense contracts.  Higher figures were more likely to produce R&D grants.

Anyone interested in building a Lockwood engine is advised to visit the Lockwood website (www.blastwavejet.com) and
Paul Sherman’s website (www.brainvirus.org).  The Lockwood site is rich in historical information and offers the possibility
to purchase engine plans.  Apparently, this is not a snake-oil site but a genuinely enthusiastic effort -- fully on the level.

Paul Sherman’s site is a good illustration of how far an enthusiast can get.  Paul built and developed two or three different
Lockwoods (by the time of writing) and has provided interesting photographs and some spectacular video footage.  Highly
recommended!

The way the Lockwood engine functions is outwardly simple and fits the explanation I have given at the beginning very
closely.  However, we have to remember that it is by far the most extensively developed valveless pulsejet and thus features
some fine details that one can develop only after a lot of tinkering.  Some of those evade simple explanations and have
generated some controversy in the circles of pulsejet enthusiasts.

One of the more controversial features is the fact that the exhaust tube is very narrow where it exits the combustion
chamber.  Its area is only about a half of the intake tube area.  In most other pulsejets (including valveless ones), the intake is
smaller than the exhaust.

One reason for this unusual ratio that I can see is that the rather large diameter of the intake tube is necessary to facilitate
breathing.  However, the total area of both apertures (front and rear) must be in certain relationship to the volume of the
chamber in order for the chamber to work like a Kadenacy oscillator.  If that area is too large, the Kadenacy effect will be
lost.  So, if the Lockwood is to have a big intake for good breathing, it must have a small exhaust aperture.

Bruce Simpson says that the trumpet shape of the exhaust -- a small hole at the chamber flaring out to a fairly large diameter
rear end -- is also a good “Kadenacy pump”, which enhances the pressure swings in the chamber.  I have no opinion on this
but will trust Bruce until learning otherwise.  Graham Williams and Larry Cottrill have pointed out that it makes for good
ignition, shooting a slug of retained hot gas deeply into the center of the chamber, from where combustion can spread
through the chamber easily and evenly.

Another curious detail is that the Lockwood engine, if built properly, does not have a single 'straight' (cylindrical) part.  All
its parts are tapered and its cross-section changes constantly from front to back.  The reason here is probably the wish to
escape the constraints of acoustic resonance.  Such a shape will be willing to resonate in a fairly broad spread of
frequencies.  Among other things, this will allow the engine to be throttled up and down easily.

What Lockwood may well come to be remembered for in history will be the fact that he was a prominent champion of thrust
augmentation.  A thrust augmenter is a device that uses hot exhaust gas to suck additional fresh air into the exhaust jet gas
stream.  Though the device was known long before his application, it gained general attention in the field by increasing the
thrust of the Lockwood engine very noticeably.

Hot gas streaming out of the end of the exhaust tube (on the left of the picture) enters a venturi-shaped duct on the right,
sucks fresh air behind it, mixes with it and heats it up.  Heating makes air expand.  Expansion increases static pressure on the
walls of the augmenter, and a component of that pressure is aimed forward, pushing the augmenter.

Aerodynamics of thrust augmenters is a bit more complex than this.  Apparently some forward “lift” is generated by the
curved front lip of the duct.  However this simplified explanation will suffice for our purpose -- more so as I am convinced
that heat plays a much greater role here than aerodynamics.

The principle of thrust augmentation is very old and many a designer has tried to exploit it.  The majority of those
experiments were performed either with compressed air (notably by NACA) or with turbojet engines.  The practical results
never justified the added complexity and bulk.  Yet, when Lockwood tacked augmenters on his engine, the gain was quite
appreciable.

That has proved to be the case only with pulsejets.  On other jet engines, thrust augmentation is of marginal use at best.  The
reason is that other engines use more air much earlier in their operating cycles.  A little hot gas pushes along a lot of air.  In
pulsejets, a small amount of air is sucked into the combustion chamber and used for combustion and a slightly larger amount
is sucked back into the exhaust tube between explosions.  That is all.  There is very little propulsion mass for the
considerable thermal energy to act on.

Because of that, only a small part of the heat liberated by the process of combustion can be converted to kinetic energy.  The
energy-mass transfer ratio is very low and the thrust is lower than it could be for the available energy.  The temperature of
the exhausted gas is much higher in pulsejets than in the other jet engines.  So, the super-hot exhaust of the Lockwood
machine simply cried out for additional propulsion mass to heat up and accelerate.

Tharratt

I have little or no information on the extremely interesting valveless pulsejet designs of C. E. Tharratt.  What I have is oral
tradition.  Among other things, Tharratt experimented with pulsejets that consisted of a main tube and a kind of a cap partly
closing one of its ends.  Spacers held the cap at a set distance from the walls and the end of the tube, so that the passage
between the cap and the main tube acted as the intake/exhaust tube.

The above picture shows the most basic configuration of this kind.  It is just a tube capped with a tin can.  The thin tube in
the front is the propane supply tube.  Tharratt apparently favored the main tube that tapered gently inward almost all the way
to the end, so that the narrowest part was at the rear, with the very end flaring out.  I have no data on this and cannot really
describe it in any detail.

However, logic would suggest that the most basic layout can be improved in a number of ways – introducing shaped 'cap'
that would help the internal gas flow, a waisted main tube that would provide additional confinement to combusting gases
etc.  The drawing below shows a possible developed configuration following the same basic principle.

To me, the capped tube designs look like just about the easiest pulsejet of them all to tinker with, because the cap and the
main body are connected by brackets that are easily undone, so that most modifications are much easier to do than in most
other pulsejets.

However, the simplicity is there mostly if you use gaseous fuel, like propane.  For a number of sound reasons, many
enthusiasts prefer to use liquid fuel.  I do not doubt that a way to improvise carburetion of liquid fuel can be found upon
some thought, but I will have to leave the details of such setups to the readers' imagination, as I do not have any concrete
data on actual working engines.  I can only offer speculation.

Myers

Another engine that has provoked considerable discussion from enthusiasts is the unusual propulsive tube patented by Elman
B. Myers in 1946.  I can provide a drawing of the cross-section of the engine and attempt an explanation, but I have never
seen Myers's claims independently verified.  I do not know of anyone who has seen it work.

 
At first glance, it looks conventional enough, with the intake aperture perhaps too broad.  However, in this engine,
appearance is misleading.  The large chamber up front (to the right on the picture) is not the combustion chamber.  It is a
mixing chamber in which fresh air arriving through the broad front opening mixes with fuel, which is injected continuously
through the jet shown on the lower right.  The real combustion chamber is the narrower and longer tube on the left!

The way the curious layout is supposed to function is the following.

Let's imagine the tubular combustion chamber filled with the fuel/air mixture.  One can fill it by turning the gas jet on and
blowing lightly into the mixing chamber from the front.  Next, one turns on the ignition and the several spark plugs situated
near the rear end of the tube in radial fashion produce sparks that ignite the mixture.  Combustion progresses through the
mixture from back to front in the form of a flame wall.

Pressure builds up progressively with the flame propagation until the pressure front reaches the venturi-like constriction
between the combustion tube and the mixing chamber.  The pressure wave reflects off the constriction and reverses its
direction, moving now towards the exhaust end, creating a low-pressure area behind it and sucking fresh mixture from the
mixing chamber.  When the tube is refilled with the mixture, the combustion cycle is completed.

While this wave rebound is taking place, the expanding hot gas produced by combustion escapes through the exhaust and
generates thrust.

Spark is not necessary for the next cycle.  In his patent application, Myers said that ignition "follows from the flash back of
each preceding explosion".  It is difficult to say what that really means, as his engine does not have the long exhaust tube
that would retain a portion of hot exhaust gas for ignition of the next slug of mixture.  It does not have 'dead air' pockets that
would retain hot combustion products either.  One must assume that the ignition was initiated by the thin boundary layer of
burnt gas that clings to the smooth chamber walls

Another unusual feature is the sharp front edge of the venturi between the mixing chamber and the combustion tube.  The
reason for this feature must be the generation of vortices.  When the pressure wave travels from back to front of the engine,
some unburned mixture will be pushed from the combustion tube back into the mixing chamber.  With a smooth, sloping
transition, it might shotgun through the mixing chamber and out of the engine.  The sharp transition, however, generates a
strong toroidal vortex that expands towards the chamber walls, rebounds from the front constriction and stays within the
mixing chamber.

Again, I must stress that the above explanations are speculation based partly on Myers's patent application text and partly on
deduction.  I can only hope that the debate on the possible merits of this unusual design results in someone actually building
one and testing it.

Reynst

With the Myers, I have come to the end of the list of notable valveless pulsejets of the past that I had something to say about
– with one notable exception.

Diverse as the 15-odd engines described above may have been, they had one feature in common: two openings.  One is the
intake and the other the exhaust, with the intake most often serving also as an auxiliary exhaust.  The intake is always
controlled in some way -- devised so that hot gas is prevented from flowing forward during explosion, or bent around to face
in the opposite direction, or at least choked so that the opposed flow is reduced to a minimum.

Enter Francois H. Reynst.  As a youngster, he played with alcohol in jars and accidentally discovered an interesting
phenomenon.  If a closed jar filled with a little alcohol on the bottom had a hole in the lid, and one ignited the alcohol vapor
emerging from the jar, a curious pulsation ensued.  Flame would shoot out of the hole, only to be sucked inwards and then
ejected out again.  It looked as if the jar breathed fire.  What Reynst discovered was a peculiar pulsating combustor he would
later develop into a serious industrial product.

What was happening in the jar was a relaxational oscillation driven by the combustion of the mixture of air and alcohol
vapor.  Combustion would generate a great amount of hot gas (mostly carbon dioxide and water vapor).  Pressure inside the
jar would rise and the gases would expand out through the hole in the lid.  Because of inertia, the gases would overexpand
and the resulting partial vacuum would suck fresh air in.  This fresh air would whirl inside the jar, mixing with alcohol
vapor.  The mixture was ignited by the remaining free radicals present in the hot gas that would remain clinging to the walls
inside the jar.

This simple experiment is very easy for everyone to reproduce at home.  Just make sure that the lid is sealed tight and
preferably put the jar into some water, so that it gets cooled, or the heat will crack it within several seconds.  Of course,
various metal vessels can also be used.  Reynst was fascinated by this and eventually published a research paper on the
behavior of such a combustor.  Later still, he developed an industrial combustor working on the same principle.  A
simplified sketch is given on the picture.

 
 

The chamber has a conical intake/exit part that serves as the accelerating jet in the exhaust phase and a decelerating diffuser
in the intake phase.  A mixing chamber, into which air flows from below and gas is added through a feed on the upper left,
encloses the mouth of the chamber.  Fresh mixture is sucked into the pot through the narrow slit between the chamber mouth
and the mixing chamber.

The chamber sucks in the mixture from the sides rather than air from above because of the distribution of pressures.  Above
the mouth, due to inertia, air is still moving away from the engine, trailing the flow from the previous exhaust pulse, when
the suction starts.  The pressure above the mouth is below atmospheric, while the pressure in the mixing chamber is at least
atmospheric, possibly slightly higher.  So, the flow that replenishes the chamber commences through the slit, rather than
through the big intake aperture on the top.

The unusual item in the middle of the chamber is an additional intake diffuser that helps form the kind of flow that works the
best.  This is another peculiarity of the Reynst 'pot'.  As the air is sucked from the sides into the chamber, it forms into
doughnut-shaped (toroidal) vortex that travels towards the bottom of the chamber.  The diffuser entrains it on its way,
preventing it from expanding too early.  When the vortex hits the bottom it rebounds and – expanding to the walls of the
chamber – climbs back upwards.

The diffuser thus separates the downward part of the internal flow from the upward-moving part.  Unlike most other pulsejet
combustion chambers, this one has a tidy two-way flow, with the mixture traversing the entire chamber twice – from the top
down and from the bottom up.  As it whirls all the time in the toroidal vortex, all of the mixture makes contact with the
chamber wall at some point.  No part remains isolated in the gas 'core' as happens in some other engines.

This makes for very reliable ignition by free radicals that remain in the thin boundary layer of hot gas clinging to the
chamber wall.  They are intimately and vigorously mixed into the flow.  So, the ignition starts at about the moment the
vortex touches the bottom and continues on the way towards the top.  At some point of travel, internal pressure created by
combustion makes the vortex explode.  The internal pressure jumps steeply and the whirling mass of hot gas is ejected from
the engine at the top.

The filling of the chamber is helped by the behavior of the vortex.  As ignition starts inside the vortex, the heat adds energy
to the spin and the vortex accelerates its whirling and contracts.  This lowers the volume of the ingested mixture and still
more mixture is sucked in.  The inventor also described a thermal side to the cycle, with contractions strengthened by the
transfer of heat to the walls of the chamber.  As a result of this complex process, the Reynst engine has the most efficient
Kadenacy pumping of all pulsejets.

Young F. H. Reynst thus gave birth to a pulsejet engine that – unlike any other – only has one aperture.  This is actually a
logical layout – after all, it is only with continuous combustion that intake and exhaust must be separate.  With pulsating
combustion the separation is not essential, as the intake and exhaust parts of the cycle are not simultaneous.  The combustion
chamber thus 'breathes' in and out through the 'mouth'.

The Reynst pot is not an acoustical resonator and its working cycle is much slower (by an order of magnitude) than the
natural resonance frequency of the chamber.  However, by pure luck during testing, Reynst discovered a way to make it into
an acoustical resonator and improve performance very notably.  He used a flue to take away the hot gas, as his tests were
performed inside a building. One day, the flue split across at some distance from the pot. The combustor suddenly switched
into high gear, quadrupling the frequency and increasing the pulse amplitude notably. After that, Reynst instigated a program
of development that perfected the resonator exhaust for practical applications.

So, the complete, fully developed Reynst engine looks similar to the simplified picture below.  The combustor and the
exhaust combine their resonance, with the 'pot' resonating in quarter-wave mode and the resonator in half-wave, so that the
pressure antinode is situated at the gap between the two.  (Again, this is a simplified explanation, but close enough to the real
events to be true.)  The whole thing was about a meter long.  "The noise produced,” wrote Reynst, “can be heard, depending
on the wind, at about six miles."

What must be remembered is that Reynst never intended his combustor to be a jet engine.  It was an industrial heater,
designed to work as a blower furnace.  So, he never tried optimizing it for thrust, but for a great volume of high-temperature
gas and the ability to consume cheap fuels.  (It works well on heavy oils and even on coal dust!)  However, there is no
reason that I can see that would make it any worse than the more common valveless pulsejets.  Indeed, it might easily be the
best of them all, given its excellent pumping and mixing ability.

Just imagine the blunt bottom of the pot rounded into a more aerodynamic shape (closer to the teardrop form).  Imagine also
the exhaust resonator extended forward to enclose most of the chamber and serve as the fresh air duct.  The resulting form
gives a very nice jet engine shape, while remaining about as simple as any other valveless pulsejet.  Other parameters would
need to be played with as well, like the aperture size.  A broad aperture is excellent for burner purposes, but for propulsion
the open cross-section of the chamber mouth should probably be made smaller, so that the pressure amplitudes and the flow
speeds increase.

 
 

Blast Compression Valveless Pulsejet (BCVP)

Publicly the best known engine concept by Bruno Ogorelec, who put this section of the website
together, is the Blast Compression Valveless Pulsejet (BCVP).  Having a concept of one's own share
the space with the designs described in the previous chapters may be presumptuous.  So far, we have
been talking mostly of proven engines.  What we have here is a concept that, at the time of writing, still
has some way to go towards a practical engine.  However, given that few other places on the Internet
offer information on valveless pulsejets, so be it.

A speculative paper describing the idea was first published on this website in October 2000.  Since
then, several BCVP prototypes have been built, tested and developed by a cooperative of enthusiasts –
Nick Ibbitson, Vivian Collins and Graham Williams in the UK, Gary Robinson in Australia and Bruno
Ogorelec in Croatia.

The results of testing have proved the soundness of the concept, but also the difficulty of putting the
idea into practice, especially with the limited resources at the disposal of amateur enthusiasts. 
Nevertheless, so encouraging have the results been that there is reasonable hope of having a fully
operational engine of notable performance ready for showing to potential investors, sometime in 2003. 
I am keeping my fingers crossed, writes Bruno.

The major detail that makes this engine different from others is the attempt to harness the energy of pulsation to boost the
efficiency.  In most pulsejets, most of this energy is wasted.

Conventional jet engines, like the turbojet and fanjet, use the constant pressure generated by combustion to drive a large
amount of gas rearwards and thus create thrust.  In a pulsejet, the pressure pushing the exhaust gas is not constant, but comes
in explosion-like peaks alternating with partial-vacuum troughs.  Plotted as time vs. internal pressure, the process is
represented by a sine curve.  The peaks provide thrust and the troughs provide aspiration.  The troughs also detract a bit from
performance, as the sucking of retained exhaust gas back creates a small amount of negative thrust.

The pulsating nature of the flow is partly beneficial, as the explosive combustion provides some compression during the
process.  Expanding products of combustion compress the unburned remainder of the fuel/air mixture.  The need for the pre-
compression provided in turbojets by compressors is thus reduced.  However, the gain is not great enough.  The mean
effective pressure inside the engine – the best measure of efficiency of an internal combustion engine – is still too low
compared to the competing engine types.

One of the reasons is that the pressure wave created by the explosion has no practical use in the engine once the combustion
is over.  It is simply dispersed into the outside.  Its dispersal manifests itself as combustion noise.  It is both unpleasant and
wasteful.

The pressure troughs, or negative pressure fronts, or rarefaction waves, are a similar case.  No matter what we call them,
they are mostly wasted once the combustion chamber is replenished with fresh charge.  The door is slammed shut on the
incoming fresh air by combustion pressure, just like it is slammed shut in the valve-equipped engines when the reeds close. 
Only a part of this energy is utilized in the hammer effect that helps boost pre-combustion pressure a little.

The BCVP engine – one of whose possible layouts is shown above – attempts to close the system so that the pressure waves
either do not escape from the engine at all, or do so only partly.  Instead, they are harnessed to do useful work on increasing
the pressure in the working cycle.

Low pressure is the Achilles' heel of the conventional pulsejet.  To start with, there is no pre-compression to speak of.  In
fact, some researchers have taken the absence of compression before ignition to be one of the basic, innate features of
pulsejets.  What little pre-compression is achieved is passive -- due only to the inertia of fresh air rushing into the
combustion chamber.  The compression ratio normally does not exceed the 1.2:1 ratio, meaning that the pressure before
ignition is barely 20 percent above atmospheric.  In turbojets, it is up to 30 times atmospheric and in piston engines, it is 10-
20 times atmospheric.

My concept introduces a rather complex "active" mechanism, using the force of one exploding charge of the fuel/air mixture
to compress the next charge.  It is a strong force and the process should raise the compression ratio closer to the range
associated with piston engines.

Another area in which conventional pulsejets score poorly is the energy-mass transfer rate.  The thermal energy generated by
combustion should be made to accelerate as great a mass of ambient air as possible to the highest possible speed.  In
pulsejets, this is difficult to achieve.  As ambient air does not really pass through the engine but is ingested in small discrete
'gulps', there is little air on which the generated heat can act.  Some pulsejets manage to recover a part of the otherwise
wasted thermal energy by the use of thrust augmenters and I fully intend to exploit that route.  All my engine concepts
involve a great increase in the through-flow of fresh air through the engine compared to conventional designs.  Their energy-
mass transfer rate should be considerable.

The envisaged engine has three main components – the fresh air intake tract, the combustor tube and the thrust tube.

INTAKE TRACT -- Given that the intake tract will be an elongated duct in intimate contact with the combustor, it is to be
expected that the power pulsations of the combustor will try to establish a sympathetic resonance in this ducting.  The
resonance will affect the intake flow in a notable way.  Because of that, the tract should ideally be tuned to resonate in a way
that will promote airflow in a pattern suiting the combustor.  Resonance is paramount here.  The exact shape is less
important.  One only has to take care to proportion the intake so that a sufficient amount of fresh air reaches the combustor
mouth and that there are no obstacles and discontinuities that would promote harmful turbulence.

 
 

COMBUSTOR – This is essentially just a bent tube.  A curved section connects two straight sections to make a teardrop
shape.  Each straight side of the teardrop serves as a combustion chamber.  Each 'chamber' ends in an abruptly tapered
constriction and has a narrower tubular intake/exhaust nozzle The nozzles project into the main body of the engine at the
point between the intake end exhaust tracts.  The curved part connecting the two 'chambers' serves only to allow the pressure
pulses to travel freely from one chamber to the other.

This is not the first time a pulsejet has had two combustion zones in a single tube.  In 1906, a Frenchman, Robert Esnault-
Pelterie, patented a pulsejet in which one 'chamber' fired against the other (see the picture below).  His machine had valves at
each end of a straight tube.  Hot gas was ejected from the center point between the two 'chambers' into an exhaust branch set
at a right angle to the main tube.  The machine was not an engine in itself, but a gas generator for a gas turbine.

As in the BCVP, explosions alternate between the ends.  In practice, a column of hot gas constantly shuttles back and forth,
playing the role of the piston in the reciprocating engines and of the turbine and compressor in the turbojet.  Experiments
have shown that for a brief period the interface between the hot and cold gas fronts (i.e. between exhaust gas and fresh
mixture) indeed behaves like a piston.  Gas pressure rises in front of the "piston" and falls behind it.  In the Esnault-Pelterie's
combustor, the pressure at one end of the short tube would drop far enough for vacuum to open the valve at that end and
suck air in, even though it was rising quite steeply at the other end at the same time.

 
 

My design is different in being valveless, having a curved tube instead of straight, and in ejecting hot gas from the ends of
the common tube, rather than from the middle, but the basic principle of interaction of two combustion chambers in one tube
is the same.

When the fuel/air mixture within one end of the combustor ignites and explodes, hot gas is driven in two directions.  One
part of the hot gas blows through the nozzle (and into the thrust tube).  The other part goes through the curved section
towards the other combustion 'chamber'.  The process at the other end is 180 degrees out of phase with the first.  At the
moment when the expansion starts in one 'chamber', it is already over in the other.

 
 

Just as in a conventional pulsejet, the pressure in the chamber drops below atmospheric by the end of expansion.  Low
pressure sucks fresh air in.  The air enters the chamber through the nozzle and mixes with propane provided by the fuel
feed.  The mixture fills the combustion chamber.

So, at this moment, there are two pressure fronts traveling fast towards each other.  A cold pressure front of fuel/air mixture
is moving inwards from the nozzle, while a hot pressure front of burning gas is moving towards it from the other chamber. 
The two slam into each other.

In the collision, the mixture is compressed.  As the hot front is stronger, it drives the mixture back towards the nozzle.  Here
is where the sharp conical constriction comes in.  The relatively tight funnel does not allow a massive flow rate.  In addition,
the 90-degree taper of the constriction acts as a wave reflector.  Thus, most of the mixture driven into the end of the chamber
will not be able to escape and will be compressed further.

Meanwhile, the heat transfer and migration of free radicals from the hot gas to cold will degrade the interface and ignite the
mixture.  It will explode, and the cycle will be repeated in the opposite direction.  The new explosion also drives hot gas in
two directions.  A part blows into the thrust tube, while the rest travels as a hot pressure front back to the first chamber to
compress and ignite the mixture that has formed there in the meantime. Explosions alternate between the ends of the
combustor tube.

THRUST TUBE – On the diagram to the right, one can see the place at which the combustor ends poke into the main body
of the engine.  This is the point at which the end of the intake tract meets the beginning of the thrust tube.  Both are split
vertically into two branches by the central double-wedge deflector -- one branch for each end of the combustor.

Those branching passages to either side of the deflector are D-shaped (or crescent-shaped) in cross-section.  Their cross-
sectional area gradually increases.  Tubular nozzles of the combustor take up the center of each passage, but fresh air can
pass around them.

Hot exhaust gas produced in the combustor blasts from those nozzles into the passages.  The gradually broadening passages
function as gas ejectors or thrust augmenters.  Here, pulses of hot gas hit the passing fresh air and push it towards the
exhaust.  The air is mixed with gas, heated and accelerated further.

Finally, the two ejector/augmenter passages merge into a single thrust tube (or exhaust tube).  Alternating blasts of gas from
the two converging passages are entrained here to blow in the same direction.  If the length and shape of the
ejector/augmenter passages is right, they will also resonate in a sympathetic manner to the combustor and the intake tract.

There will be a pressure anti-node at their junction that will mark the stage from which the exhaust gas should propagate
without major pulsation.  The two sine waves converging at that spot should cancel each other out, with the peak of one
coinciding with the trough of the other, producing a constant-pressure flow.

At the time of writing, I have no idea how high a compression ratio can be achieved by the pressure increments that follow
each other in this system.  I have never seen any data on the compression in the Esnault-Pelterie's engine.  The only other
twin-combustor engine I know of that attempted mutual compression was an experimental device built at the University of
Calgary in 1983 under the guidance of J.A.C. Kentfield.  According to the paper published at the end of the program, it
failed to achieve notable compression -- but its concept was completely different from that of the BCVP.  (Encouragingly,
though, it did run much quieter than conventional pulsejet engines.)

What I am certain of is that pre-compression will be far greater than the paltry 1.2:1 ratio achieved by the standing wave in a
conventional pulsejet.

Anything over 5:1 will push the engine into the range achieved by conventional piston engines and – together with the usual
benefits of a pulsejet – make it highly competitive with any other small jet engine – including those costing an order of
magnitude more.

Offhand, looking at the geometry of the combustor and the forces involved, I have a feeling that this (and more) should be
readily achievable.  Some of the evidence from prototype testing would tend to confirm this.  However, I cannot back the
conviction and the subjective impressions with numbers yet.  Not enough tests and measurements have been performed.

The above description ignores the fact that some of the mixture will start escaping from the combustion chamber into the
thrust tube as the hot front hits the fresh charge and pushes it backwards.  The funnel-type nozzle will act as a barrier only
above certain gas flow rate, however.  Before that rate is achieved, the mixture will escape.

It should not matter much in practice.  The pressure and speed build-up will be very quick and inertia will continue to push
the fresh mixture forwards for some time even after the collision of the two fronts.  Only a small part of the mixture is thus
likely to escape.  It will not be completely lost, either, but will burn up in the thrust tube, contributing to overall thrust, if not
as effectively as the part that remains in the combustor and gets compressed before ignition.

This raises an interesting possibility of having the thrust tube function as an afterburner.  It is supposed to have more or less
constant pressure inside.  The alternating hot blasts will be able to serve as an igniter-type of flame holder.  Plenty of fresh
air will be passing through the tube at any point.  Additional provision of fuel into this environment, coupled with a
convergent nozzle at the tail end, may well produce effective afterburner reheat.

Going a step further, one can observe that there is a free, unobstructed flow of fresh air through the engine between the
intake and exhaust apertures, especially if the flow around the combustor nozzles is streamlined.  This raises an interesting
possibility of the afterburner becoming a ramjet at high airspeeds.

This section was last updated in October 2002