Note: Descriptions are shown in the official language in which they were submitted.
3136~
This invention is a new internal-combustion engine whose purpose it is to
reduce greatly all the major losses inherent in present internal combustion
engines, and that with less pollution, less weight7 and a cheaper fuel mixture
To clarify, " reduce greatly " means here an overall increase in thermal effi-
ciency by at least 3 ~, and in the specific case of car engines in slow city~
type driving which, according to official sources, amounts to about 55% of all
car mileage, this means giving 300 to 50 ~0 more kilometers per liter than
present equally powerful engines can deliver.
As this new engine combines many well known parts and functions with sev-
10 eral new or modified features, so that it cannot be qatisfactorily distinguishedby any one new feature, I name it "Combination Engine", or "Coengine" for short.
The words "internal-combustion" in the title may be inappropriate, for heavy duty
ensines of this type will include a steam turbine. This becomes practical precise-
ly because of the higher thermal efficiency and an already coupled exhaust turbine.
~ efore I show how the Coengine is built, functions and reduces the losses,
I have to explain briefly the nature Or these losses and why present engines can
not reduca them more effectively. Search of literature and the fact that new very
large engine plants are now being constructed to produce multicylinder engines
for cars and trucks have shown to me that the fundamental principles of how to
20 overcome the great losgeg in car engines are not yet well understood by experts,
and that the Coengine is therefore a true invention regardless of the fact that
it employs many well known parts and functions. me use of more than one piston
in the Coengine for use in cars and trucks would hardly represent an improvement,
and it would certainly increase los~es.
The main losses in present reciprocating engines consist of two classes.
GAS CYCLE LOSSES WORK LOSSES
1. Low compression ratio 1. Friction of pi~tons and rings
2. Incomplete oylinder filling 2. Friction of crank bearings
3. Delayed combustion 3. Friction of main bearings
4. Incomplete combu~tion 4. Cooling fan
5. Heat losses in high pressure range 5. Electric generator
-- 2 --
&'~
Losses deriving from valve gear, oil ~np, water purnp, fuel injec-tion
systems etc. are of a smaller magnitude than the enumerated types, and I shall
therefore exclude them from further explanations.
The standard air cycle with constant volume combustion based on the
compression ratio has, as published, the following efficiency:
Compression ratio 4s1 6:1 8:1 1Oil 12:1 14s1 16:1 18:1 2051
Efficiency in % 42.6 51.2 56.5 6002 63.0 65~2 67.o 68.5 69.8
The standard air cyole efficiency is reduced in linear fashion in the work cyclebased on fuel addition by 0 to 33.4% proportionately with 0 to 10C~ combustion
10 of available oxygen. Constant pressure combustion reduces the work cycle effi-
ciency further. ~hese work losses can be largely recovered with the aid ot the
right type of exhaust-gas turbine.
According to the work cycle efficiency engines should have a higher ther-
mal efficiency at lower loads or fuel addition than at higher. Present engines
fail in this respect glaringly. ~he overall efficiency of an engine must be eval
uated factually in connection ~rith its use, that is the car engine must be evalu-
ated when delivering power to the wheels. And also its fuel consumption when
idling must be considered. The tractive force (rolling resistance plus aero-dynamic
drag) on a flat road of a medium streamlined compact car weighing one ton with
20 driver has been measured with a decelerometer to be 51.1 kg at 48.3 km/h (30m/h)
and 81.7 kg at 96.6 km/h (60m/h), which is a power demand rekpectively of 6.7
and 21.5 kW. Work required for accelerations in town traffic is generally over-
estimated and I therefore give a few figures for a one ton car.
Acceleration toEqual to lift height of car Work
km/h, m/s m kWs
18 5 1.25 12.25
54 15 11.25 110.3
go 25 31.25 306-4
The accelerating work required for a city trip is therefore approximately
for 100 accelerations up to 18 km/h = 1225 kWs = 0.34 kWh
for 10 accelerations up to 54 km/h = 1103 kWs = 0.31 kl~
- 3 -
8~;6
'~he gre~ter part of this work is recovered as rollin~ resistance during the ordi-
nary deceleration runs, including ~he braking distance7 These figures also show
that sy~tems for recovery of decelerating work for cars will not be cost effeotive.
The work demand for one kilometer at 48.3 km/h is 6.7/48.3 = 0.139 kWh/km,
13.9 kWh per 100 km. The 10 000 kcal per liter fuel ~ivided by 860 tkWh=860kcal)
amo~lts to 11.62 kWh per liter. At 4 ~0 overall efficiency, which the best present
diesel engines can deliver, a liter would give 6.5 kWh. 100 km would require
accordingly at 407~ efficiency 1309/6.5 = 2.14 liter. Present comparable cars re-
quire about five times that much fuel in slow city traffic9 2nd the thermal
10 ef~iciency of their engines is therefore only about 8~. At lower average speeds
the wor~ demand per kilometer is even lower as stated and thus can cancel trans-
mission losses.
Con~idering that much of city driving is done well under the speed limit,
when decelerating for adJusting to traffic flow, &nd in down-hill run3, I
estimate that over half the dis-tance traveled in city traffic requires a power
of no more than 2 - 4 kW for the one ton car. And for about half the total running
time the engine delivers power at a rate of under five kW. ~herefore, any car
engine which has a low overall efficiency at this output level is extremely waste-
ful and unsuitable as a car engine at a time of scarce &nd expensive fuel. The
20 overall thermal efficiency, or brake e~fficiency, of most present car engines drops
to under 5~ when delivering less than 4 kW.
The enumerated losses individually are generally well understood and have
been measured. Less understood, and proven even more difficult to overcome, is
the vioious circle-like nature of these losses. ~or example, the heat losses in
the high-pressure range through well conducting metal walls do not merely take
away otherwise recoverable work from the highly co~pressed gas, but they also delay
combustion, thereby reducing further the reooverable work. And furthermore9 delayed
co~bustion involves so~e condensation of (diesel) fuel on the relative cool lubri-
cated cylinder walls~ thereby diluting the lubricant and reducing the amount of
30 fuel available for combustion. And in the oase of the ordinary low-pressure gaso-
line engine, where power output is regulated by throttling of the air-fuel mix-
- 4 -
ture, the well cooled wall surfaces may, due to the insufficient evaporation ofthe fuel-mist droplets, reduce combustion speed so greatly that much of the f'uel
burns only in the second half of the power stroke, and partly not at all, being
instead expelled with the exh~ust a~ hydro-carbon pollution. The ~uch lower
efficiency then demands higher speeds, thereby causing greater f'riction losses
which increase approximately with the square of the piston speed, and most impor-
tant, are almost independent of loading. The late and inefficient combustion
requires then more fuel which raises the tempera-ture of the low-pres~ure gas
more than it produces more work. ~he exhaust temperature climbs to over 1000C.
lO This leads to the almost unbelievable point where some modern car engines, in
spite of their large cooling surfaces, overheat (water boils) during long idling
periods and therefore require a complicated ignition-advance system. This raises
the idling speed, normally already 800 - 900 rpm, upon overheating so that the
faster and more power consuming fan now can reduce the temperature to the set P~int
when normal idling speed returns. mermostatically controlled electric fans are
another solution to avoid this type of overheating.
The widely accepted rough statement that the gasoline engine converts about'
30~o of the heat of oombustion into work~ loses 25~o through the cooling system
and 35~o in the exhaust, plus 1 ~o as friction, is for the present car engines in
20 prnctical use utterly wrong. I did show this already for city traffic.
The lO~o friction losses seem at first glance small, till it i9 realized
that these losses are taken from produced work, when they become one third of
the work delivered at the most favorable output level. The friction of the
close fitting pistons with their large friction areas is due to the viscous
friction of the rather tightly enclosed oil layer and increases approximately
with the square of the piston speed~ changing little ~ith load. The same applies
to the even tighter fitting beaFings~ with the difference that higher engine
speeds increase greatly the mass forces which have to be carried by the large
bearings, for in the four-stroke engine only one of four piston accelerations is
30 compensated by gas pressure. ~ere again, engine speed rather than power output
determines the magnitude of friction. With a bearing circumference of about
-- 5 --
1.7 times 'he stroke length for the gasoline engine, 2.5 times for the heavy
duty Diesel, it can be shown that for high-speed engines the produot of bearing
travel times bearing load is nearly equal to the friction a car would have if
simply pulled along, like a sleigh, on a well lubricated smooth surface.
The friction losses are most visible du~ing idling. one to four liter ~uel
with a work equivalent of 12 - 45 kWh are required just for one hour of idling.
If friction increases 90 strongly with speed and is high also at light loading,
~hy then is engine speed not simply more lowered for smaller outputs, why must
engines idle at 300 - 900 rpm~ The vicious circle consequences of losing heat
10 to the cool cylinder walls I mentioned already, and lower speed includes more
heatloss per stroke. Increased escape of the air-fuel mixture passed the piston
rings into the crankcase at high pressure is also unavoidable at low speed. And
a type of fuel burning fast enough for high-speed and low-pressue operation will
detonate when used at low speed and high pressure. ~or the high-compression
Diesel, bearing failure at low speed is likely. At high speed the mass forces
compensate much of the gas pressure, and the oil film cannot quickly enough be
squeezed out from between the gliding surfaces. At low speed the compensating
mass forces are tiny and the oil film has time to be squeezed out~ causing thin-
film lubrication and consequently quick wear and failure of the bearings.
Insulating piston faces and cylinder heads has been tried, and precombustion
ohambers made of ceramics are in use~ ~he complicated shape of the piston crown
with cutouts for valves, whioh would have to be even larger if the valves also had
insulation added, and the not compensated mass forces in four-stroke engines,
whioh easily could shake lose brittle~ heat resistant insulation and thereby
destroy the cylinders~ plus problems with heat stresses of the tightly fitting
pistons have made such insulation impractical. The large number of cylinders and
the small production runs of such engines would render such insulation also too
costly. And, obviously, spark-ignition engines cannot use them.
Direct admission into the cylinders by metering valves of gaseous fuel at
30 hi~h pressure, cold or hot, has been tried. But in the four-stoke versions a~d
with multiple cylinders having large manifolds and interconnected open valves,
-- 6 --
6~
and with direct access to the large crankcase passed sometimes failing pi~ton
rings, even small irregularities of the fuel admission system, insufficiently
supplied with safety features, led to frequent fires and even explosions, and
direct ad~ssion of high pressure gas had to be abandoned.
~ ighly loaded, and especially supercharged diesel engines should, according
to the work cycle efficiency, contain much recoverable work in the exhaust gas.
Why then is so little of this work recovered? Xeat lost during the compression
stroXe is part of the heat to be rejected in accordance with the gas cycle. The
exhaust temperature will be lowered accordingly. ~ut heat lost faster than added
~0 by compression, that is near the dead point of piston travel~ and all heat lost
thereafter is real work lost according to the gas cycle efficiency at the point
of loss. And as heat transfer of a gas to a wall increases proportionately with
its temperature (ener~y) and density (number of molecular hits per cm2 per sec),
most heat is lost in the high-pressure period, when the piston moves but slowly
near the dead point, that is precisely when the capacity to produce work is
greatest. The enlarged surface areas due to cutouts in the piston orown and for
precombustion chambers~ncrease these losses. ~esides the direct loss of wox~ taken
from the piston, the exhaust pressure i8 also lowered by this heatloss, and thereby
some capacity to produce work in the exhaust turbine is lost also. The large mani-
20 fold and the lon~ exhaust stroke combine in most present supercharged engines toincrease the entropy of the gas, that is they take away from the gas much or most
of its remaining capacity to produce work, without, however, hardly reducing its
dnmaging high temperatureO And the exhaust volume enlarged by temperature has to
be forced out against the atmospheric pressure. I had to explain this, for the
Coengine can avoid most of these losses~
~ s mass or g-forces play such an important roll in reciprocating engines~
rather than estimating~ I have figured some out when they are greatest at the dead
point. About 23~o has to be added or subtracted respectively to oompensate for the
angular action of the connecting rod. A piston stroke of 7 cm (7 cm bore) can be
30 representative of the one-piston Coengine for small and medium cars, delivering
up to 80 kW, whereas the 9 cm stroke is ample for the heavy car or light truck.
- 7 -
g-forces for given strokes a~d given speeds
rev/s 7 cm stroke 9 cm stroke 30 cm stroke
14~08 g18009 g 60.~ g
56~3 72~36 241.2
352. o 452. 26 1507 . 5
100 140~.2 1809.0 6030.2
150 3168-4 4070.3 13567.9
The area of the 7 cm bore cross section is 38.48 cm2. With a gas pressure
of 100 at near the dead point that amounts to a force of 3848 kg as against
lO a force o~ 3168.4 kg for a one kg reciprocatin~ mass at 9000 rev/min. Heavier
reciprocating parts would therefore be better for an engine which uses all
g-forces for countering of the gas pres~ure, or higher speeds respectively for
large engines with constant running speeds.
These fi~res also clearly show that at low engine speeds most of the gas
pressure has to be carried by the bearings. A one-piston engine has the great
advantage that the two main bearings are end bearings, therefore suitable for
roller bearings. And the one crankpin bearing can be made thinner and longer,
thereby reducing bearing travel. For high-speed heavy-duty engines a one-crank
crankshaft can be heat-shrunk and welded together from two parts after instal-
20 lation of a roller bearing on the crankpin, with the end bearings fitted afterthe welding.
~ igh-speed onecylinder motorcycle engines of 250 and 500 cm3 oylinder
volume demonstrate that even light engines can function easily with only one
piston. And a heavier engine with an attached gearbox sufficient to transmit
well the much greater power will therefore likewise have no excessive vibration
problems in this respect.
The specific ways in which the Coengine avoids the~arious losses and how
the new features support each other in achieving so great savings in fuel con~
sumption will be explained after the parts and the fun¢tioning of the Coengine
30 have been explained with the help of the drawings. Then also will be given the
reasons for less pollution.
- 8 -
~8~36~
Fig. 1 shows a cross section of one embodiment of -the Coen~ine with some
parts shown schematically.
~g 2 shows a longitudenal section through a piston and a foreshortened
piston rod and cylinder with overcharger~ showing mode of cooling,
lubricating, insulation and the way the overcharger works.
~ig. 3 shows schematically a cylinder head with kidney-shaped valves and
rectangular valve stems.
Fig. 4 shows in sectional view a cylinder-head wall with insulation, insulated
valves with valve stems, and a bearing protection.
0 Fig. 5 shows in sectional view the con~ollable admission valve with lever and
membrane, and schematically the joining and corresponding metering system.
Fig. 6 shows schematically in cross-sectional view the starting burner with
heating ducts surrounding the admission valve, fuel pipe, and gasifier
with conduction fins, inside insolating walls.
~ig. 7 is a cross section through a cylinder wall to which is bra~ed a lubri-
cating pipe, and it shows the openings into the cylinder.
Fig. 8 is a longitudenal section through a one-piece counterweight with ball
bearing, and a cross section through the round-faced crosshead thereof.
In ~igA 1 the number 1 shows the cylinder wall9 2 the double-acting piston,
20 3 the valves, 4 the piston rod, 5 the crosshead with bearing, 6 the connecting
rod~ 7 the crankpin bearing~ 8 the crankshaft9 9 the main bearings, 10 the small
fly~heel with balance weight~ serving as cOuF1ing disc~ 11 the coupling or
clutch. 12 is the small-diameter gear with ball bearings 13 transferring power
through the only schematically shown gearbox to output shaft 14. 15 is a larger
flywheel~ also with balancing weight. 16 are cams on the crankshaft which serves
also as camshaft, for each revolution requires a full valve movement. 17 shows,
representative for several, a valve-train. 18 i8 a multistage exhaust turbine
which drives the singlestage compressor 19 that delivers the air to the inlet
valves after having it drawn in through the airfilter 20. 21 are turbine-shaft
30 bearings. ~he alternator 22 and the cooling fan 23 cooling the radiator 24
are driven through reduction geaxing by the turbine shaft. ~he fan can drive
g
inversely the turbo-compressor at high car and engine speeds when the gas is
suddenly taken away and the ~inimum gas would then not be sufficient to keep the
turbine spinning at high engine speeds~ for this type of engine piston has no
pump action. Fluid flow directions are indicated by arrows. In the gasifier 25
liquid fuel is heated above its critical temperature and becomes thereby a gas,
not merely a vapor. An insulation wall 26, here only shown partially ~nd schemati-
cally, surrounds the gasifier 25 and the following hot pipes leading to the cyl-
inder, leaving a free space in between through which the gasifier and pipes can
be heated up with a hot flame for starting. High-pressure air pump 27, high-pres-
10 sure fuel pump 28, oil pump 29 and low-pressure fuel pump 30 are driven through
reduction gearing 31 by the crankshaft. The gasifier 25 is filled before first
starting with liquid fuel through and approximately up to the height of fill and
check screw 32. Insulation is shown in all drawings exed - xxxxxO 33 is insulation
in the lower part of the gasifier through which fuel and air is delivered by the
respective pumps through the corresponding pipes. 34 is a li~htly spring-loaded
safety valve whose for only moderate speeds streamlined valve stopper will let the
gas pass at up to the intended maximum working speed3 but it will close if for
any reason that maximum speed is exceeded. 35 is a sprin~-loaded check valve which
can provide a somewhat lower pressure in the following gas pipe 36 and admission-
20 valve casing 37. Its use allows a reduction of possibly other~ise excessive fuel
flow from the gasifier into the pipes after engine shutdown with following induced
cooling of the hot pipes and gasifier when the gas quickly condenses. ~1el pipes
36 and controlled admission valves 37 (Fig. 5) serving as metering valves, are
required for both cylinder heads. mey are shown only single in order not to
overcrowd the drawings. Pressurized air is delivered from the upper, the greater
cylinder end through oooled and check-valve protected pipe 38 to the small-volume
high-pressure air pump 27~ which can be a membrane pump~ which delivers it into
the ga~ifier for partial combustion and for gasification of the fuel brought into
the gasifier by the controlled fuel pump 28.
39 is a heat exchanger whioh reduces the temperature and the volume of the
e~haust gas before expulsion through exit 40. Dhe thereby produced hot ~rater, or
_ 10 -
8~66
a vapor, can be used for car heating as heating with air from the radiator is not
enough, or for driving of a vapor turbine. Turbine, flow through narrow cooling
passa~es, and exit port combine to act also as silencer, just as the air filter
also has the function of silencing the compressor.
Due to the limitations of the fuel admission valve used, the absolutely
required speed regulator (Fig. 5) must set the minimum speed below which no
fuel can be admitted into the cylinder. This demand requires a special higher
speed starting system. As the heating-up of the gasifier 25, pipe 36 and admission
valve 37 before starting requires a heating period of about 7 - 9 seconds, this
10 time is also available to have simul~niously the small electric starter motor 41
wind up the small flywheel 42 whose energy is~ after achieving the set speed9 trans-
mitted through heavy strong spring 43 to coupling wheel 44 which turns the crank~
shaft by means of switched enforced contact with coupling disc 10. The required
~inimum speed is thereby reliably enforced within 1 - 2 revolutions of the crank-
shaft. ~lywheel 42 and spring 43 are somewhat heavy but inexpensive to manu-
facture~ whereas the starter motor 41 can be considerably lighter, smaller and
cheaper than ordinary starter motors, for its efficiency need not be very high,
for the required overall starting work, due to much less friction, is much smaller
than that for multicylinder engines. And the starter motor has 7 - 9 seconds to
20 deliver the work to the flywheel. ~he first few cylinder fillings are achieved
with natural aspiration -- there can be no vacuum with open valves --, after that
the highly efficient~ light-wheeled and quickly starting turbo-oompressor takes
over. 45 and 46 are tightly sealing bearings (Fig. 2) through which the piston
rod 4 moves back and forth. A gap between them is connected to the air intake
at 47,so that any gas having forced its way passed the rings of bearing 45 is
directly returned to the cylinder, leaving the orankoase free of combustion gas.
~ he flap valve 48 prevents the spreading of the exhaust gas into the
low-pressure section of the exhaust manifold. It should be noted that the two
exit ports widen as they enter into the exhaust pipe, thereby increasing the
30 speed of the exhaust before it enters the multistage impulse tur~i~e, for highly
loaded and supercharged engines achieve supersonic exhaust-gas flow. mese pipes
are also very short, so that a speeding axial-flow turbine immediately after
the high-speed pulse produceg suction before the next pulse from the other
cylinder end begins. And it is in or~er to avoid that this advantage of lower
~ressure gets logt, as lower pressure means faster flow to the turbine, especially
also under very small load conditions, that flap valv~ 4a i~ sO valuable for
the Coengine.
49 represents a thermostat and 50 a pressure regulator, both working to-
gether to keep the temperature and pressure in the gasifier at a set levelO ~he
thermostat is thereby the overriding element. As long as the temperature in the
10 gasifier i9 below the set level, a relay sygtem leaves the admisgion valve to
the air pump 27 open till combustion achieves the set temperature. During this
time no l~uel is pumped in. When~ after a short running period, with the tempera-
ture still above the set minimum point, the pressure drops below the set point,
a relay system leaves the admission valve to the high-pressure fuel pump 28 open
and more fuel is added. As only a definite small amount of o~ygen is required to
achieve the required gasification and temperature for a given amount of fuel
passed through the well insulated gasifier, the relative mixture ratio of fuel
and air remains nearly constant, differing only slightly with power output, for
a slower fuel flow has a relatively increased heatloss, requiring therefore a
20 little more oxygen.
Valves, thermostatic and other control and relay systems are in daily
u~e in thousands of variations and I therefore need not explain them any further.
me inventive idea lies in where to employ them with best results, not how they work.
~ ig. 2 shows piston rod 4 and cylinder wall 1 shortened in order to sa~e
space. 51 shows the single and therefore larger inflow conduite through piston
rod 4 for the passage of oooling and lubricating oll into and through the piston
and returning through several smaller conduits 52 into the crankcase. The circu-
lation of this oil is achieved by the pumping action enforced by the changing
mass forces and a check valve in the crosshead which allow~ the oil t~flow in
30 from a small reservoir in the crosshead, but not out a~ain. ~he smaller retu~n
outlets are always open. ~he oil for this reservoir and for the lubrication of
the cro3shead with connecting rod bearing and the piston rod i8 supplied by the
oil pump. Some oil flows inside the piston into the small pump housing 53 wherein
a free piston with a small p~ump plunger5~, driven by the mass forces, meters and
dalivers with each second stroke lubricating oil through corresponding passages
and outlets 55 into the oil groove 56 of the piston si~uated between the rings.
In a horizontal cylinder at least two oil outlets 55 are required on top of the
piston from where the oil flows do~m both sides along the oil groove 56 which is
provided with several crosscuts 57 made with a thin small circular saw. Due to the
shape of this oil groove, only the droplets just in front of these crosscuts can
10 ~et thrown out and against the cylinder wall by the mass forces~ leaving other
oil to flow further down. The rings then ensure further and even spreading of
the oil all over the cylinder wall. Vertical cylinders require more oil outlets 55
and a zig-zag-shaped oil groove. Only the compression rings 58 touch and glide firmly
on the porous chromium plated cylinder wall. The piston itself does not seal
tightly and fits only close enough for vibration-frae guidance, and it therefore
does not have the great friction losses of ordinary pistons. The explanations of
why this piston does not need to glide firmly in the cylinder and other particulars
~ill be given following the descriptive part illustrated with drawings.
The compression rings 59 within bearing 45 are inserted as two parts in the
20 upper and lower bearing halves. As, due to the small diameter of the piston rod,
the play for heat expansion can be small, the jointed gaps between the ring halves
G~n be very small. The simple construction of the piston rod can be recogni~ed by
the threading 60 on both ends~ where they are screwed into the crosshead and the
piston~ Qnly the center passage must have a tight, separating fit on both ends.
m e well-spaced passages, produced simply by casting or pressing, allow a stress-
free through and through hardening of a low-cost but high~strength alloy steel,
and therefore a small rod diameter. The outside of the rod is best also plated
with porous chromium for better oil retention and less corrosion.
The construction of the short and light piston presents no unusual diffi-
30 culties, as is shown by the center line 61 around which the piston can be manu-
factured in two main parts and then fastened together before the insulation is
- 13 -
added. This cylinder wall does not have to take a side loading by the piston andit can therefore be made thinner and have ribs around the high-pressure, high-
temperature ends which can easily reinforce and better cool it. By cladding the
outside, including the ribs, wi-th copper, any required degree of cooling can be
achieved, for it is the heat transfer between coolant and wall rather than the heat
flow through the thin metal wall that limits most the cooling of the inside.
62 shows the two combustion-space enlarger valves, or simply the overchargsr,
in which the tightly fitting valve pistons 63 are located which have insulated
faces 64 where they come in contact with the hot working gas which can enter
10 through open slots 65. The position of the valve pistons is controlled by oil,
delivered from the oil pump. It can enter and leave through oil pipes 66 through
the controlled oil valve with double opening 67, or instead (not shown) through two
separately controlled oil valves~ In the inflow position the inflow branch 69 with
check valve is in front of oil line 70, allowing oil to be forced under the valve
pistons 63 and closing them during the periods when the gas pressure is lower than
the oil preqsure. The backflow of oil is prevented by the check valve. As oil is
very little compressible, only tiny amounts of air can force itself bet~een
the tightly fitting valve-piston face and the also insulated head wall of the
valve cylinder. ~his small amount is excluded from the combustion prooess, so
20 that heatlosses from this source are negligibly small. AB the pressure above and
below the valve pistons is equal, bypassing of gas around the tightly fitting
pistons is practically impossible, for capillary pressure enforces an oilfilm
between the tightly fitting walls. The control of the oil valve 67 i8 achieved
automatically through xelay action when the air pressu-re delivered by the compres-
sor surpasses the high-pressure point and the oil valve is changed into the open
or outflow position. Gas pressure then forces the valve pistons in the open position.
~lore air with higher pressure can now be entered into the working cylinder, and
also more fuel, without, however, producing a much higher and possibly destructive
pressure increase during a longer constant-pressure combustion period. ~he relative
30 lower thermal efficiency within the cylinder with such overcharging -- as again3t
- 14 -
-
8~i6
merely supercharging -- is in larger engines (truck engines), with the turbine
capable of getting coupled through gearing to the engine, largely recoverable
by the e~haust turbine and the also coupled gteam turbine driven by the vapor or
steam produced in the heat exchanger 39 (Fig. 1), whose output i9 in this case
increased by higher temperature and greater flowO 71 shows the fuel-ga~ inlet
nozzle situated near the open 910t 65, so that with longer lasting fuel injection
the air contained in the enlarged combustion space of the overcharger can more
quickly participate in the combustion. The quicl~y increasing combustion pressure
forces some air and with it much of the simultaneously injected fuel gas through
10 the slot into the overcharger.
72 shows insulation on piston faces and cylinder heads. As the valves are
open only when the piston is at the other end, the piston faces can be flat. And as
the piston itself does not closely seal and touch the cylinder wall, small differ~
ences in heat expansion do no damage. The insulation can therefore consist of two
parts, the insulation layer proper, and the covering plate or lid. Important is in
this connection that the two layers are constantly pressed together and against the
piston, by gas pressure in the upper part of the stroke and by the reversing
acceleration in the lower part. ~he insulation layer proper can therefore be made
of a set of ceramic seotions fitted on cylinder faces and cylinder heads and on
~0 the valves. ~he individual pieces can be screwed on or center-welded to the under-
lyin~ cooled metal. The small gaps between the pieces relieve heat expansions. The
ceramic can be made solid, porous or contain hollow spaces for better insulation.
Ground ceramics heat-bonded with a small amount of metal powder, giving a porous
insulator, has high stability against frequent temperature changes and can be
easily shaped. ~ll these types of insulation layers must be protected against
penetrating gas. Simple metal coating, with a plasma torch for instance, ca~ be
done effectively and cheaply in large production runs. ~or the piston faces at least,
a prefitted coverplate or lid, the rim of which being bra~ed or welded -to the piston
crown, is preferable. A few concentric rings and curves embossed in the lid 72
30 compensate for heat expansions. Super alloys are required for these relative
thin covers, for high-t0mperature corrosion can be severe~ Eut unlike in turbine
15 -
blaAing, high mechanical stresses and erosion by fast gasses with avalanchingeffects are not a problem in this application. I shall explain after the des-
criptive part why the peak temperatures that these cover plates will have to
withstand are actually lower than might be assumed at first glance.
Fig. 3 shows the cylinder head containing piston rod 4, kidney-shaped
valves 3, open slot 65 leading to the overchargera and the fuel-gas noz~le 71.
73 is a rectangular valve stem which is better suited to transmit the required
forces to the large area of the kidney-shaped valves than a round one, and even
more important, these flat, thin valve stems with their bearings, made of two
10 L-~haped halves~ fit closer and parallel to the piston rod and its sealing bear
ings. The valve springs can be placed easily in another position in the valve-
train, having a shorter stroke, than on the valve stem.
Fig. 4 shows in sectional view the insulation layers 72 on valves 3 and on
the inside of the cylinder head. ~he valve stem 73 and its bearing 74 are shown
protected from the hot exhaust by the chambered gas seal 75. ~his seal does not
glide on the valve stem, but rather the penetrating gas from the short pressure
pulse is lowered in pressure in the first chamber before it can, now slower, pass
under the next wall into the second chamber, and after further pressure reduction
passes on at still lower speed into the third and following spaces then esoaping
20 into the low-pressure flow of the exhaust stream. mis leaves the lubricated part
of the stem surrounded by nearly stagnant and cool exhaust gas which does not
destroy the oil film. me insulated, cooler valve body is made hollow to make it
slightly elastic for better valve seating.
~ he rather thick layer of insulation on valves and cylinder heads has be-
sides insulation another important function. Fbr best results,the exhaust valve
should be opened late, but then ve~y quicXly and fully~ and the inl~t valve should
remain open late but then close very quickly. mis demand rl~s counter to the
demsnd for reduced acceleration forces for the larger valves. In the arrangement
as show~ in Fig. 4, the exhaust valve is closed whereas the inlet valve i9 ~till
30 fully open. For about the first and the last 2 ~ of the valve strokes only the gaps
between the cylinder-head insulation and the valve body with its insulation are
- 16 -
B~6
open to the gas. The passage around the 90 degree angle oYer the flat valve seatfurthermore about ha]ves this outflow. In this way, starting and seating of the
valves can be done slowly and with moderate forces~ ~Thereas the full opening of
the va]ves occurs just at -the rig~ht moment very quickly. Even the gas escaping
through the gaps of the valves has its forethought f~ction and is therefore
part of the invention. During normal medium to high speed running the time element
for returning of an appreciable amount of already lightly compressed air during
the compression stroke through the still open gaps is too short. However, at very
low idling speeds a considerable part of the air in the cylinder already lightly
10 compressed can return throu~h the gaps into the intake manifold~ which holds but
a very low overpressure when idling. This results in less filling~ and therefore
a lower maeimum pressure when idling~ and it reduces thereby the load on the
bearin~s Nhich at low speeds~ when counteracting mass forces are sm~ll, are in
danger of destructive thin-film lubrication. Fuel in~ection with quick combustion
well after the dead point of the piston reduces further the time of maximum load_
in~ of the bearings when idling. Lower idling speeds are better achievable with
the Coengine because of the precise combustion~ because of the insulation, and
because of the lower and shorter lasting maYimum gas pressure when idling.
~ig. 5 shows in sectional view the admission-valve body 76 with the valve
20 needle 77 within its housing 37 (Fig. 1) surrounded by insulation 26~ with valve
lifter 78~ valve lever 79 and valve membrarle 80. Lever and membrane can be cast
or pressed as one piece, or the lever can be welded into the membrane. The
membrane is fastened wnth screws to the housing 37 after insertion of a gasket.
The quite thick and therefore quite rigid membrane resists the gas pressure in-
side the valve housing without hardly any bending, so that the valve body with the
valve lifter in form of a wire or thin strip remain in the centered position at
full gas pressure. This " indireot ~ membrane functions instead as a pivot for the
lever. When the cam 16 on shaft 8 by means of the valve-train lowers the lever at
the right-side end~ the valve body is raised from its seat and the valve needle
30 opens more or less~ depending on the height of the lifting~ the exit of the valve
from where the gas leaves through vPlve no~zle 71 into the cylinaer. In order
to understand better the small pivoting movement and thus the small bending re-
quired by the membrane, it may be noted that for the already mentioned Coengine
with 7 cm bore and stroke cylinder, the required length of the valve needle, and
therefore its maximum lift height, is only about 2 mm, and its maximum diameter
at the top is only about 1 mm with a minimum free space between needle and valve
opening of only ~.01 mm around the needle. Approximately 0.3 mm of the upper
needle ha~ a constant diameter, 90 that the minimum outflow speed remains nearly
constant whether the valve is lifted 0.1 or 0.3 mm. mis tolerance allows for
10 sufficient valve-gear play. The space between the valve body and the encircling
housing is over the lower 2 mm not greater than the space between valve needle
and valve opening at the same lift height, 80 that the valve needle never can
touch the exit wall. ~nis rather close fit of the valve body hardly decreases the
outflow speed, for the length of the gap around the valve body is at least three
times longer than the gap around the needle. And it has the great advantage that
it serves as a gas bearing which automatically centers the valve body in the middle
of the ou~flow stream. Given the high pressure and temperature with the high gas
speeds, this centering force is strong enough to overc~ome the small bending force
of the thin valve lifter 78 and to seat the valve body in the center position
20 without touching of the surrounding wall.
The metering of the outflowing gas must be done by changing the lift height
of the valve. mis changes with the running engine and the form or shape of
cam 16 the cross section of the valve opening and the duration of the opening.
The lift height and the duration of opening in degree angles i9 changed by vary~
in~ the length of the valve-train rod where it pa~ses thrcugh regulating units
81 and 82. ~nit 81 is moved by a speed regulator, which can be a mechanical or
electronic type, in such a way that with increasing speed the length of the rod
is continuously shortened so that the cam follower is lifted by the cam higher
and over more degrees with inGreasing speed. This unit limits the maximum flow
30 of fuel with respect to speed. ~elow the minimum setting, with a speed below
about 150 to 200 rpm, the rod is 80 much lengthened that the cam follower does
- 18 -
6~
not a-t all touch the cam, ~herefore no fuel can f]ow below thig speed. During
the minimum flow setting, which corresponds to low idling, the top of the cam
lifts the cam follower for about 3.6 degrees, or for 1/100 part of one revolution,
and thereby lifts the valve body by 0.1 to 0.3 mm for the 7 cm cylinder over the
same period. Unlike ordinary diesel injections or ignition with spark~ that occur
well in advance of the dead point, this valve opens several degrees after the
dead point of the piston, for combustion is very fa~t. With the minimum flow
setting sufficient for an idling speed (warm engine) of 450 rpm, this minimum valve
setting under speed regulator control will deliver three times as much fuel into
10 the cylinder at 150 rpm than at 450 rpm, but less per power stroke if the speed
is hisher. With unit 81 limiting the maximum flow for a given speed and setting
the minimum flow for idling, unit 82 shortens the rod by command from the gas pedal
or the governor. As 81 and 82 are coupled or cooperating units, 82 can shorten
the rod only so much as is permitted by the maximum setting of unit 81 which iB
governed by the speed regulator. As the movements of valve body and cam follower
are quite small, ordinary screw elements, or spindels and nuts~ are sufficient to
effect the changing of the rod~s length. The controlled fuel injeotors of ordinary
Diesels are much more complicated and their pump stroke is much longer than the
lift height of the valve body.
Fig. 6 shows schematically the relative positions of gasifier 25, admission-
valve housing 37 and fuel pipe 36 within insulation walls 26, topped by controlled
insulation door 83, which is closed during the running of the engine, open when
not running. Heater 84 receives liquid fuel directly from the fuel tank or from
the low-pressure fuel pump through pipe 85 and backfire protector 86, which is just
an insert of metal fibers~ and through one-way valve 87. ~he outflow of fuel is
restricted by spring-loaded valve 88. m e pressing of the starter button begins
a series of events that heat instantly the heater spiral 89 and the ignition
spiral 90, and it starts the starter motor. ~he small amount of starter fuel
contained in the heater, whose inside is coated with insulation, is very quickly
30 heated to the point where its lowest boiling fraction evaporate~. ~his vapor pres_
sure forces the lower still liquid fuel up the insert pipe 91 and after overcoming
- 19 ~
66
the spring-loaded valve ~9 sprays the liquid fuel passed ignition ~piral 90
which ignites it, upwal~s into the combustion and heating space æurrounding
gasifier, fuel pipe and admisgion valve, and by means of heat transfer fins 92
quickly heats them to a temperature close to 500C~ that is above the critical
temperature of the fuel used. The exhaust leaves through open door 83. After all
fuel in the heater is evaporated and has left as vapor, after the spirals 89
and 90 have been shut off and the starter motor has started the engine? door 83
closes and the heater fill9 again with fuel. When the engine stops, door 83 opens
and the upflowing air quickly cools and condenses the fuel contained in the
10 gasifier, fuel pipes and admission-valve housings.
Fig. 7 shows a cross section through a cylinder wall 1 surrounded by a
thickwalled oil pipe 93 whose central pipe space 94 is connected by drill holes 95
with the inside of the cylinder wall. Lubricating oil can be pumped thereby into
the cylinder by means of a metering pump. This oil pipe is simply bent and brazed
or welded around the cylinder. This oiling principle is not new, but that it can
be effectively employed for oar engines with only one cylinder, that idea is new.
Fig. 8 shows in longitudenal section a one-piece co~mterweight with ball
bearing 96,crosshead 97~ and a cross section of said crosshead in the indicated
cut.line 9S~ its curved gliding surfaces placed within the lubricated guide way
20 99. As the side pressure exerted by the crosshead is reduced proportionately with
length of the connecting rod 100, and as the required curves of the two gliding
sides of the crosshead become shallower the longer the rod, it i9 clear, that with
higher engine speed, when the side pressure becomes greater, the effective oil
film area under the shallow curves also become greater and more effective with
a longer connecting rod. At lower speeds with small side pressure, the oil film
resistance also becomes smaller. Such a one~piece counterweight iæ much less
costly than a second piston9 and its friction losses are only a small fraction
that of a second piston. The ball bearing is ~imply in~erted in a Gorresponding
hole of the fly~heel, or the wheel can have a crankpin. If such a counterweight
30 i9 used, the flywheel will be brought close to the main bearing~ Cams and other
~ear can be arranged on the other end of the crankshaft. Instead of placing the
- 20 -
crosshead in a guide wayl it ~imply can be car~ied by a swing arm of severaltimes the length of the piston stroke. This adds but a minor rocking force
to the crosshead and the engine.
S m a 1 1 C ~ e n ~ i n e s of about 25 kW power for use in motorcycles,
minicars, tractors, outboards etc.~ where lower wei~ht and co~t is more i~portant
than a few percent higher overall thermal efficiency~ the turbocompressor can be
replaced with a piston compressor, and the overcharger is unnecessary. The com-
pressor piston, driven directly from a second crank, or by means of a crankpin
and a ball bearing from the flywheel, serves also as the counterweight. As the
10 pressure provided by such a piston is quite small, only ordinary one-way valves
are required, and only one spring-loaded sealing and oil-control ring suffice~.
No long piston skirt i9 necessaryS for the side presæure of the connecting rod is
transmitted to ~he wall by small piston fields. Lubricating oil pushed by the
piston to the cylinder head is collected in a bottom groove and is pushed out of
the cylinder by a small protruding oil piston on the piston crown. Cost and fric-
tion of such a compressor piston are therefore much smaller than that for an ordi-
nary piston. Important is that also for this type of Goengine the compression ratio
at low idling speeds must be reduced~ and at high speeds there should be super-
charging. For this purpose the compressor piston delivers about 50 - 8Cr/o more air
20 into the large intalce manifold than is re~uired for full cylinder fillings. ~he
intake manifold se w es also as air rese woir, for only every second compressor
stroke delivers air, whereas each stroke is a power stroke. During low idling
speed, the time period when exhaust and intake valves are both open is quite long,
and a small overpressure i~ sufficient to drive all extra air out as scavanging
air. As engine speed increases, the period for scavanging gets shorter and the
amount of air retained in the cylinder gets higher, till near maximum speed hardly
any fresh air can escape as scavanging air, the engine runs supercharged.
The m u 1 t i s t a g e exhaust turbine seems at first glance difficult
to produce, and therefore expensive. High co~t is, however, not necessary, for this
30 turbine can be easily assembled from simple pressed or stamped parts. In a multi-
stage singleshaft turbine, losses from the first stages are largely recovered in
_ 21 -
~8~6
the following stages. Simpler rotating ~heels ~nth blades can therefore be producedas single stamped discs with the rim being formed as blades, and the thickness of
the blade~ made thinner near the outer rim. 'rhe stator blades are similarily stamped
as half-discs and fastened to the respective halves of the tu-rbine housing. The
multy stages and the l`act that only low-pressure exhaust is passed through the
turbine result in moderate turning speeds and considerable suction after each
high-speed pulse. The work required to produce this suction, e~pecially also during
idling, is mostly recovered by the higher speed of the exhaust due to such lower
pressure. Most important, especially for æmall turbines, is the limiting of the
10 average maximum temperature of the exhaust, for the strength o~ the nickel-
chromium alloys used for such discs with integral blades diminishes quickly in the
upper temperature range of the exhaust. In the free-wheeling (not coupled) version,
the speeding turbocoDpressor produces more air of a higher pressure than is re-
quired or used in the cylinder with the overcharger in action. As the impulse
turbine blading takes the exhaust as fast as it is thrown at it~ lower pressure is
produced immediately after the high-speed pulse when exhaust and inlet valves are
still both open. With the flow speed through the valves at this pressure difference
being near the speed of sound, and with the great valve openings, a considerable
amount of scavanging air can pass into the turbine, thereby reducing decisively the
20 average temperature of the exhaust. It is the cool soavanging air, reduced further
in temperature by lowering of its pressure, that contacts most of the time the
blading. With the coupled turbine, when power is transmitted to the engine, the
delivered air has somewhat lower pressure and less air can bypass in a shorter
period of simultaneously open valves. But the suction pulse can hereby be used to
suck cool air through a thermostatically controlled one-way valve into the space
directly in front of the blading. This valve is normally locXed~ but it opens
when the thermostat, sensing the set high temperature, unlocks it. Ordinary by-
passing OL air from the compressor directly into the turbine is also a means to
reduce the exhaust temperature peaks. But as this turbine is a pulsed turbine,
30 the bypassing must also be pulsed. A rotating or a slide valve must time the by-
passing to coincide with the suction pulse. The problem of turbine lag applies to
~ 22 -
all turbocharged engines. In the Coen~ine the turbine speed quickly increaseswhen fuel addition is increased. To prevent excessive fueling b~fore the turbine
has reached sperchar~ing speed, an inertial~ hydraulic (similar to a shock ab-
sorber), or magnetic delaying device must be added to the gas pedal. As the power-
ful Coengine when used as car engine normally runs with much excessive air, there
is a great b~lilt-in loading reserve~ and the delay time can therefore be very
small~ hardly noticeable in ordina~y driving.
P o 1 1 u t i o n and combustion are closely related. C0 and hydro~carbon
pollution is practically impossible in the Coengine, for hot gaseous fuel will
10 always burn completely, provided only that the stochiometric ratio is not exceeded,
which is most unlikely, for the Coengine always works with an air surplus. N0 for-
mation is an endothermic and therefore reversible process which is negligibly
slow below 1600C and increases geometrically in speed and formation ratio with
temperature. Qnly nearly perfect stochiometric gas mixtures, which usually burn
with detonating speed~ achieve temperatures high enough for quick N0 formation. The
spark-ignition engine usually has a low-speed range when the highly charged hot
filling has had time to evaporate fully the fuel mist and when the spark causes
superfast combustion or detonation. With the combustion oompleted well in advance
of the dead point of the piston~ there remains much time for N0 to form. In the
20 diesel cylinder detonation is restricted to small pockets. The heat of evaporation
of the fuel droplets first cools lightly a small volume of air, keeping it from
instant ignition, and when after full evaporation a near stochiometric mixture i~
achieved, this volume explodes~ and the pressure wave given off ignites neighbor-
ing similar mixtures. The combustion in the Coengine cylinder is completely dif~
ferent~ and the high-temperature part is much shorter. The hot fuel gas injected
at high speed in a thick beam burns the instant it comes in contact ~ith the hot
compressed oxygen of the air. ~ut just as fast the now oxygen-free gas surrounds
the continuing fuel beam and separates it for a little while from the surrounding
air. ~he momentum of the injected fuel, increased by exp~nsion due to combustion,
30 starts a rotating movement of the air mass in the disc-shaped combustion space,
and the air now hits the straight tangentially flowing beam sideways and thereby
- 23 _
very quickly -- but never with detonating speed -- dispexses and intermixes
air, fuel and combustion productg, causing quick and complete combustlon. Unlike
in oxdinary engin~s, combustion beginq only after the dead point? 90 th~t even
befoxe achieving the peak temperature some work is already delivered to the piston.
And during the peak temperature the piston moves already at a good speed, so that
the peak temperature lasts only a very short time. There simply is not enough
time for formation of ~0 at the moderately high temperature ~he Period from 9 de-
gree before the dead point to 9 degree after the dead point amounts to 1 ~ of the
time for one stroke, but the piston moves only 1% of its working stroke during
10 this time. ~he extra friction work requixed for so long a bearing travel with high-
pressure loading is about as gxeat as the work produced in this 1~o stroke length.
But the heatloss during this 1 ~ time of one stroke to the close and well cooled
walls in ordinaxy engines is much greater, for heat transfer increases proportion-
stely with pressure, and even worse, this heat is lost before and at the highest
point of thexmal efficiency in the gas cycle. ~he long time of bearing overloading
at lower speeds is also saved when combustion starts only after the dead point in-
stead of 20 degree before~ Without sparks, no N0 is formed this way.
~ he late combustion start also greatly reduces the tempexature of the in-
sulation covers. Instead of heating over the long period of high pressure, thexe
20 is cooling~ for the -temperature of the compressed gas lies below the temperature
of the insulation covers. ~his is especially the case with overcharging, when the
maYimum endpressure is lower, but the air charge, due to the enlarged combu3tion
space~ is aotually greater. And the period of very high temperature is in this
oase very short because it occurs later. For heavy-duty engines, like truck engines~
charge-sir coolers will be used, which9 at least at higher loading, will greatly
reduce the compression temperature before fuel addition9 so that the cover temper_
ature is thereby also reduced. And as heat radiation increases with the fourth
power of absolute temperature~ heat lost from the insulation covers to the cooler
lubricated cylinder wall, being very small at moderate temperature differences~
30 will become considerably greater as the temperature of the covers increase~
strongly. Likewise~ there is a somewhat increased heat conductivity of insulation
- 24 _
layers at hi~her temperatures. And the scavanging air also helps cooling. This all
adds up to the point, that the temperature of the insulation covers will never
surpass the limits of the super alloy employed.
The insulation layers not merely reduce heatloss9 they also pro-tect the
cylinder heads. The little heat penetrating can easily be transferred to copper
plating on the cylinder-head metal and transferred by the copper to the cooling
medium. This reduces the complexity of the cylinder heads. And as one cylinder
with cooling ribs can easily be cooled with oil~ cooling water cc~n be avoided
altogether, only a small oil cooler being required.
W h y i s a i r u s e d to gasify the liquid fuel) why not simply heat
with electricity or even with exhaust heat? ~irst, a fuel mixture with heavy
fractions will crack partly at the required temperature near 500C. Cracking
itself is not undesirable, for it is merely a preliminary step of combustion,
and heat added to -the fuel in the gasifier is reco~ered in the cylinder. But
cracking produces unsaturated chemical bonds, and carbon could bond to the walls,
causing blockages and insulation. As oxygen bonds are much stronger, bonding with
o~ygen will prevent heavy deposits of carbon. Second, heavy fuel fractions are
extrenely dense at the required pressure of over 120 at, and the spaed of sound,
which is the limiting entrance speed into an open valve, is Iower for heavy mole-
20 cules. ~he presence of light balast gag~ mostly nitrogen~ makes the mixturelishter and increases its speed of sound and therePore its flow speed. ~his im-
proveq the metering precision, and also the temperature can be slightly reduced,
for the balast gas will compensate for any condensate droplets whioh can occur
at a lower temperature. Third, the presence Or balastgas reduces the maximum
combustion temperature by which N0 formation is determined. N0 formation requires
the presence of free oxygen, which is not available in the fully combusted ~one
surrounding the injected fuel beam as long as there remains gaseous fuel to burn.
When the mixing of this high-temperature combusted gas with the remaining free
oxygen occurs~ it is the temperature of this new gas mixture and the remaining
30 time of high temperature which determines how much N0 can be fo~med. As with balast
gas the oritical maximum temperature is reduced, and as this mixing occurs only
_ 25 _
66
after combustion, when the piston already has developed speed, and work is al-
ready transferred to it, the maximum temperature and the time of peak te~perature
in the mixture are both drastically reduced in the Coen~ine. As NO production is
an endothermic and therefore reversible process, some produced NO will break up
again before the gas is exhausted~ The reversibility is bloc~ed Yhen the hot NO is
suddenly cooled below 1000C~ like ~rhen coming in contact with cooled walls. The
hot insulation covers, by far the largest wall areas of the combustion space at
high temperature, should reduce this blockage of reversibility strongly. Gverall
then, pollution~ including NOx, should be only a small fraction that of ordinary
10 internal-combustion engines. Lead or other fuel additives are not required. Anti-
pollution devices are therefore superfluous for the Coengine.
N a t u r a l g a s and other gaseous fuels can be burnt in the Coengine
with hi~h thermal efficiency by simply pumping it with a high-pressure membrane
pump into the gasifier~ where its temperature is slightly increased. The gas for
this pump comes either from a high-pressure natural gas main~ or from a low-pres~
sure pump after passing through a cooler. As the temperature increase due to com-
pression does not grow proportionately with the numbers of increased at, but rather
is nearly equal for each doubling of pressure, the temperature increase due to the
doubling or quElupling of the gas pressure in the membrane pump does not over-
20 stress the membrane. And as this pump runs with reduced speed and does not re-
quire any metering except for a controlled admission valve~ which is either open
or closed~ its production difficulties and durability should lie within acceptible
limits. Ordinary spark-ignition engines burning natural gas can do 90 only with
a low compression ratio~ that is with low efficiency~ for otherwise damage and
losses caused by detonation are unacceptible. Gas turbines can therefore have a
higher thermal efficiency and have been preferred when burning natural gas.
L a r g e s h i p e n g i n e s burningcrude oil can also be built as
Combination Engines. ~hey will have a gasifier, and a fuel admission valve for
each cylinder head, plus ordinary liquid fuel injectors. The thick crude is heated
30 to a point where about 1 ~ of the fuel~ that is the light fractions, is evaporated.
m is condensed fuel is pumped to the gasi~ier and admitted as gas during~ the first
_ 26 -
B~6
part of the combustion stroke, whereby the combustion also starts much later thanwith ordinary liq~ùd injection. Into the already very ho-t and quickly rotating
air is then injected the heated liquid fuel which now al~o bu~ns very quickly
and in a prediotable ~ashion in the swirling gas~ producing a good portion of
constant_pressure combustion. The longer combustion period for the very large
engineY ensures that all liquid fuel is also combusted in the high-pressure range
of the cycle. For idling and low power, for instance in the harbour, only the sul-
fur free light fractions and the gasi~ier are used. l~ith all mass forces compen-
sated by gas pressure, large Coengines can run much faster than ordinary large
10 eneines with 1 o n g heavy pistons of the same power, and they are therefore
much lighter and smaller, besides having hi~her overall thermal efficiency. The
reduced high-temperature period per stroke reduces the lubricating problems.
Most two-cycle diesel engines open their exhaust valves very early and
keep their entrance openings ~uncovered by the piston very late, so that only about
60~o of the stroke is usable as working stroke. Worse~ a large amount of scavanging
air, from a few times to several times the cylinder volume, is forced through the
cylinder, requiring much work. ~o wonder9 then, that the two-cycle engine is not
widely employed. How does the Coengine overcome these problems? First it should be
understood that even at relatively small pressure differences ~0.2 - 0.4 at over-
20 pressure) the flow speed of air is very much greater than the piston speed, forkinetic energy increases with the square of speed, which means that speed of flow
increases very quickly in the beginning of a pressure difference and its growth
rate slows as the difference increases. As the piston moves fastest in the middle
o~ its stroke~ it is clear that the time gained for exhausting already during
the high-speed movement of the piston is relatively quite small, and very good
reasons should exist to waste 30 - 4 ~ of the piston displacement during its high-
speed movement. The strangest thing is~ that it i~ precisely the high speed of the
scavanging air which causes the demand for more scavanging air in Diesels. Due to
the entrance through several separated 310ts in the cylinder wall in partly oppos
30 ing directions, tremendous swirling movement of the gas develops~ which is then
transmitted to the gas to be pushed out of the cylinder. Exhaust gas and sca~anging
- 27 _
air ar~ n~ixed by this wild swirling and keep on mixing~ so that actually several
times the cylinder volume of scavanging air is required to achieve a high per-
centage of removal of the combusted gas. If not only the filling capacity is in-
fluenced by remaining exhaust gas, but if such remaining gas limits speed and
completeness of combustion of a fuel mist~ then an argument for a purer filling
can be made and has been accepted, as is shown by diesel engines with high con-
sump~on of scavanging air~ The Coengine has a completely different gas flow. In-
stead of several opposing openings, there is one large inflow opening with the
inflow directed straight for the other end of the cylinder, this flow spreading
10 only over a larger cross section by the partly streamlined kidney-shaped valve.
At the same time the likewise large exhaugt valve openlng sucks gas into the
turbine, so that there is a simple direct in- and outflow of the gas with little
mi.Ying. The turbine is driven only during the short pulse when the exhaust valve
opens very quickly and the fast hot gas, with little heat having been lost to
the walls, rushes out into the light vacuum previously produced by the turbine
before the inlet valve is opened. And if some exhaust gas remains during idling
and slow running wi-th low power output, when the turbine turns relatively 910wly~
so does this not cause losses, for only a small percentage of the oxygen in the
cylinder filling is then required. And gaseous fuel finds the free oxygen ~uickly.
20 The resulting higher combustion temperature and the faster outflow speed of a
hotter exhaust increase thermalefficiency. ~pproximately 95~0 of the power stroke,
as much or even more than in ordinary four-cycle engines, can therefore be used
for work production if the quiokness of valve opening, as already explained, is
¢onsidered, and if little work is lost for scavanging, for in higher pressure scav-
anging the turbine recovers most of that work. The ~mall amount of exhaust escap-
ing first through the gaps of the slow starting exhaust valves starts the opening
or changing of the flap valve (Fig. 1, 48), so that it later does not interfere
with the fast moving exhaust pulse.
S e a 1 i n g and 1 u b r i c a t i n g of the piston inside the cylinder
30 is probably the least understood, and misconceptions in this respect have more than
anything else delayed the decisive improvement of present reciprocating engines.
- 28 -
I have fo~nd it stated i~ literature that the long, closely fitting piston skirtor sleeve is necessary for sealing, in support of the piston rings. The increasing
use of crosshead pigtons, lately evsn in smaller diesel engines, clearly shows that
these long, much fliction producing sleeves are quite unnecessary. The only reason
for which they were required is ~or oil control in splash-lubricated cylinders.
And that can be achieved with much less friction by a couple of oil-control rings.
It is generally understood that compression rings must have on their underside
high pressure which can force the rings against the cylinder wall. It is leqs real-
ized that there must also be low pressure behind the ring into which the gas from
10 the sealing side can escape. If therefore the long, tightly fitting piston sleeve
with its oil layer does the sealing, then the gas behind the piston ring cannot
escape, the gas pressure is equal on all its sides, and it is then that not the
piston rings but rather the piston itself does the sealing~ ~wo or three rings are
required in order to reduce the pressure with which a single ring would be pressed
against the cylinder wall. Theintsrmediate pressure between the rings reduces the
pressure of the first ring. The backside pressure of the double-acting piston
functions in this respect like another piston ring, so that two rings are probably
enough if porous chromium coatings are applied which oan accept high pressure with
out losing all oil from between the gliding surface~. Ordinary two-cycle en~ines,
20 including those with double~acting pistons, require pistons longer than the stroke
length to cover the inlet slots. ~he great friction caused by so large gliding sur-
~`aoes makes them therefore utterly unsuitable for fast running when lightly loaded.
Oil runoff during long running pauses, and even complete evaporation of the
oil coating, is a problem that must be faced. Lubricating with excessive oil
splashes solves this problem for ordinary four-cycle engines~ the oil coating be-
ing reestablished within a few revolutions. In the Coengine quiok oil runoff is
prevented by the porous chromium coating, and evaporation is strongly retarded
by the fact that the one-cylinder Coengine -- unlike ordinary engines -~ always
stops with the valves closed. First high pressure at both ends enforces the stops
30 in the middle of the stroke, when th0 valves on both ends are still closed, and
later the strong valve return springs, which have to be overcome for valve opening,
- 29
keep the valves closed when the gas pressure is gone. me normal running pau~es
of a few days to a few weeks are therefore not enough for the cornplete 108s of
the necessary oil coating. Longer pauses require berore starting an oil injection
by means of an oil can with pump and a long pouring spout with spray nozzle which
is inserted through an open valve into the cylinder. As there is only one oylinder,
and as the most easily accessible valve can be used, this requirement is not much
of a bother. And, of course, such a start oiler can be added as an integral part
of the Coengine, reducing the start oiling to a mere pu~hing or pulling of a
button.
Lubricating oils with less additives can be used, for it is but little re-
cyoled -- piston rod -- from the hot cylinder back into the oil pan~ and it is
therefore less diminished in quality by oxydation. me surplus oil is consumed
as fuel and is therefore part of the total f`uel consumption. As oil is constantly
consumed, only oil addition is required, oil changes are unnecessary. Only the
oil filter needs cleaning or changing at long intervals. me higher heat loading
of the lubricated cylinder wall can easily be accepted by thinner walls. Very
large diesel engines capable of running as slow as 60 rpm demon~trate that the
heat reserve and heat conductivity of the cylinder wall metal is great enough to
keep an oil film intact even over heating periods that are a hundred times longer
20 than in the high-speed Coengine. Arguments according to which the cylinder walls
need cooling by long piston sleeves, by oil splashes -- which never reach the
highest loaded parts anyhow -- and with scavanging air are not statements of real
needs, but they say merely that some cooling occurs this way. The hot insulation
covers of the Coengine are also cooled by scavanging air~ but only little, becau~e
its pressure is so low.
B r a k i n g with the Coengine is ineffective for the same reasons that
much less friction makes thi3 engine so much more efficient. Gear brakes are al-
ready required for heavy trucks and buses, and a completely independent second
braking system with practically no wear and noise is ~ery much desirable fo-r all
30 cars. In Fig. 1 I show the gear box as integral part of the engine. One good reason
for this is that the greater weight and dimensions of such an engine combined with
- 3o
8~~6
the gear box reduces strongly the maximum g-forces to which attached parts can
be subjected, and it reduces the amplitude of the vibration enforced by unbal-
anced mass forces. Another good reason is that a gear box has to contain the said
gear brake, whereby for cars the mass and normal cooling surfaces of the gear box
with the total oil in the cooling system, and the single o:il cooler, i8 suf-
ficient to absorb the heat produced by braking. A small gear brake with several
small lubricated discs with a shorc activating stroke, which also functions as
brake coupling, is gufficient for car engines. Hea~y trucks require an oil-swirler
brake and a separate oil cooler with a thermostatically controlled high-speed
10 coolin~ fan. ~he permissible high oil temperature has high heat exchange capability
for hi~h-spaed air. This oil cooler can therefore be quite small.
The g o v e r n o r enforces the speed limit of the engine in two ways.
First, when approaching the set top speed, the fuel-gas flow is reduced in the
same way that the governor reduces the maximum flow of the fuel admission valve
at low speeds. Second, when the set maximum engine speed is reached by the en-
gaged gear driven by the speeding car, the governor activates a relay which en-
~a~es the gear brake. The driver can brake either with the normal brake or with
the gear brake. In actual usage only the silent and wear-fr~e gear brake will be `
employed, the ordinary braking being reserved for maximum emergency braking and
20 for the final part of full stops~
~ e a v y c i t y b u s e s with frequent stops can use the already
mentioned steam system for recovery of brake energy. For this purpose the steam
production is done in two stages~ which anyhow is better able to recover more
exchange heat. For braking~ a one-oylinder double- or single-acting steam pump
pumps steam from the low-pressure container into the high--pressure container. ~his
pump is driven by a continuously variable gear, for instance by an automatic
hydraulic transmission recovered from a scrapped heavy car, whereby the variably
applied speed of the pump determines the strength of the braking. ~he advantage of
this arrangement is that no further fittings or gearing are required, for the
30 starting energy itself is derived in the usual way directly from the Coengine.
~he steam contained in the high-pressure storage container i~ sent through the
- 31 _
~8~6
coupled small steam turbine which delivers power in a continuous way during the
normal medium speed runs. And the heat of condensation i8 therefore likewise
available in a continuous way for heating of the bus. As this heating is a vacuum
system, no damage can be caused to passengers when heating pipes should break.
That -this system should be considered as part of tne improvement made possible
with the Coengine is ~ounded in the fact that the Coengine produces more exhaust
heat in a higher temperature range, but minus the damaging super temperatures
caused by low compression or delayed combustion in ordinary engines. And heat de-
livered at low temperature to cooling water and oil is economically not recoverable
10 except for heating~ which can be done just as well by condensing steam. The pre-
heatin~ of the relative cool condensed water can be done in the cooling jacket
surrounding the one cylinder of this Coengine before pumping it into the low-
pressure container, so that practioally almost no heat is lost in winter.
P o l i c e c a r s requiring instant starts can use the truck engine
with steam turbine, ~hich is not heavier than an ordinary powerful engine. In
this case the coupled steam turbine (coupled with engine), deriving power from
the steam con-tainer~ starts moving the relatively light car instantly. And during
the first acceleration of the car, the start heater has time to heat the gasifier
and the fuel admission valves to the requured temperatuxe, ~hen the Coengine takes
20 over. The more powerful engine allows strong acceleration in the high-speed range.
The heat reserve of the hot water in the steam container also can provide heating
during frequent long stops, so that the engine does not have to run continuously.
~ r a n s m i s s i o n s can be simpler with the Coengine~ especially for
cars, because the demand for a limited favorable engine speed is less important
with the more powerful Coengine, and because the speed of the smaller Coengine is
generally higher. The use of wasteful and expensive automatic hydraulic trans-
missions with the Coengine would be a folly. A simple three~speed transmission
will do. First gear is used for starts and crawling speeds~ with a maximum speed
of about 30 km/h. Second gear is used in city-type driving and when towing a
30 trailer, it has a speed limit of 96 km/h (60 m~h). ~hird gear is used on the high-
way~ its speed limit is 135 km/h. These limits are required, becallse the over-
- 32 -
81~6
charger becomes effective only in the upper third of the engine speed. '~o reduceunnecessary foot-pedal work, the clutch pedal has a coupled gear-switching function
added. The firs-t half stroke of the pedal disengages and engages in the usual way
the clutch. The second half of the stroke, ending wath a click-oatch holding the
pedal in the fully depressed position, gwitches the gear into neutral. The click-
catch is disengaged when the pedal is depressed a second time, firmly but only over
a very short distance. During its first half of backtravel the spring-loaded pedal
then causes switching into first gear, and during the final half stroke the clutch
is engaged. For changing into second or third gear, the clutch is disengaged in the
~0 usual way by depressing the pedal over half its stroke, this limit being indicated
by an intermediary resistance, and after changing of the gear in the usual way~ the
pedal is allowed to return, thereby engaging the clutch. In this way the driver
does not need to talce his hands from the steering wheel for stops, and clutch and
gear are both disengaged dul~ing stops. The mechanics for such coupled switching
gear are straight forward, and work is already being done in this direction. It
has been the demand for 5-speed and higher stages (tru~ks~ for ordinary engines
that has made the general application of such coupled switching systems look less
attractive. Nany engines require more resistance for better idling.
A i r c r a f t e n g i n e s do not require an overcharger, for they run
20 already continuously supercharged. ~he cranksh2ft with two roller bearings and the
pulsed exhaust turbine can accommodate easily two cylinders. With the cranks opposed
by 180 degree, and with each stroke of each cylinder being a power stroke, the
pulses of the two cylinders occur together, so that there remains sufficient time
after each scavanging pulse to develop a light vacuum before the next exhaust pulse.
And very important, during this longer period it is the cooler scavanging air that
comes in contact with the turbine blading. Reducing the pressure of a gas alæo
reduces its temperature. The steady high speed of these engines cancels all mass
forces with gas pressure, so that the bearings have to take only the work load.
~nd as the exhaust energy is recovered~ the inJection of the fuel gas into the
30 supercharged cylinders can be even more delayed than in car en~ines. The peak
combustion pressure and temperature arrive in this way later and last therefore
- 33
a shorter time, stressing less bearings and cylinder walls. The larger, more
efficient e~laust turbine -- only one stage -- drives only the compre~sor which
produces cooled charg?e air with a hi~her pressure than slower engines. As such
aix moves near the speed of so~md, that is with about 400 m/s at th~ higher
temperature, through the large valve openings, a good percentage of the charge
air is passed as scavanging air through the turbinet thereby lowering the other-
wise too high temperature of the turbine blading.
For propeller aircraft with a speed of less than r~ach 0.5, the speed of the
hi~h-temperature exhaust leaving the high-speed turbine is too high for efficient
10 recovery in a reaction jet noz~le. Jet thermal efficiency increases with greater
mass and lower exit speeds. ~herefore, a fan driven directly by the fast Goengine
delivers one part of the air to the turbocompressor, thereby compensating for
differences in air pressure due to altitude -- and for idling --, and the greater
~art is delivered into a mix or combustion chamber - afterburner -- in which the
remaining heat of the exhaust is exchanged with the fan air. ~his results in a
large air mass leaving at low speed, and therefore with high efficiency, the jet
no~zle. For additional push, mainly in emergencies, or for starts from shorter
I~nways, hot gaseous fuel delivered from the gasifier can be injected into the
combustion chamber of the afterburner. In consideration of this demand, the gasi-
20 fier is made larger, having a greater throughput. The high fuel-gas temperature
ensul~s reliable combustion under all working conditions.
~ e a t p u m p s driven by the Coengine with -total heat recovery is prob-
ably the most important application in which, wGrldwide, the most fuel can be saved.
And although heat pumps~ just like cars, ships or aircraft, are obviously not the
subject of this invention~ I consider it most important that the application and
intertwining of the Coen~ine with heat pumps should be understood as widely and as
quickly as possible, for such understanding can help to avoid the wasteful, mis-
directed investment of many billions (109) of dollar in expensive but inefficient
heatin~ systems. Ordinary inbernal-cornbustion engines have so low overall effi-
30 ciencies when driving heat pumps for the same reasons that they are so inefficientwhen driving cars. Their use for the driving of heat pumps could therefore not even
~ 34 -
B8~
be considered seriously~ For most of the time, hi~hly efficient heat pumps requir~ little power, but double~ triple~ 10~fold and even 20~fold such power de-
r..and also occurs during winter. ~or a well desi~ned hea-t pump that always can
operate at the greatest efficiency between continuously changing temperature dif-
ferences, the temperature difference between the heat exchanger surfaces and the
surrounding air and the working f`luid must always be as small as possible. This
requires that the engines runs continuously in winter. ~he recovery of heat from
waste water, including part of i-ts heat of freezing, and the recovery of heat from
w~ter partly heated already by the St~ should not be omitted from highly effi-
lO cient heat pump operations. The higher temperature of the engine cooling waterd from the heat exchanger in the exhaust stream is recovered as hot water, after
this water has already been preheated by the heat pump and by the lower exhaust
temperature in countercurrent flow. ~he demand for low temperature differences
requires heating with warmed air. Hot water radiators would be wasteful. The
filtered warmed air entered into apartments or houses by metering pumps with
electronic heat counters is preheated by the outflowing exchange air in counter-
current flow before entering the heat exchangers of the heat pump. Preliminary
calculations allow me to state that such a heat pump system combined wi-th the
Coen~ine would reduce ordinary heating oil or gas consumption by about 7 ~o, and
20 ~0~ is not impossible where the winters are mild. With such savings, large-scale
production should be assured, and this would bring down the cost Or this new
system to below the cost of ordinary~ highly wasteful heating systems, thereby
enlargin~ the demand and therefore the production scale7 thus further reducing costs.
The light wei~ht and high overall efficiency which changes little with
speed or load, just as friction and combustion quality change little over the
speed and output range, make it unnecessary to produce Coengines of many sizes.
Just a few sizes can cover the whole range of maximum power demand for each task.
This permits engine building in numbers sufficient for automatic production. With
the cost of these engines accordingly low, intermediate sizes can then be sup-
30 planted more efficiently ~nd less costly from two or three engines. Coengines withsingle-acting pistons would merely reduce this economy of scale, saving nothing.
~ 35 -