18 replies to this topic

[manolis]

In the typical reciprocating piston pump a crankshaft rotates inside a crankcase; a piston is slidably fitted and seals one side of a compression chamber, or chamber, formed into a cylinder; a cylinder head seals the opposite side of the chamber; a connecting rod is rotatably mounted, at one end, on a crankpin of the crankshaft and is pivotally mounted, by a wrist pin at its other end, on the piston; the piston with the piston rings separates / seals the chamber from the crankcase.
Valves on the cylinder head (like poppet valves, reed valves etc) control the working medium flow.

In the Pulling Rod Engine below (combustion chamber at bottom-left, scavenging pump at right), a blade valve (or blade) is secured to the connecting rod:



The blade moves inside the chamber while the connecting rod moves inside the crankcase.
The chamber and the crankcase are sealed from each other.
Following the connecting rod motion, the blade covers and uncovers (in synchronization to the piston motion) inlet and exhaust ports properly formed onto the cylinder head.

The geometry / shape of the blade and of the ports define the timing of the opening / closing of the inlet and exhaust ports. Larger ports can increase the "overlap" and the flow capacity in high revving applications.

Lightweight materials, like Carbon-fiber, Kevlar etc, seem more than adequate for the blade valve that sees only low pressures and temperatures.

The application of the blade valve on the PatOP engine could be like:



The controllable windows exe animation is at http://www.pattakon....e/PatBlade1.exe



The controllable windows exe animation is at http://www.pattakon....e/PatBlade2.exe

The scavenging-pump cylinder-head adds "height" to the engine, the long-stoke of the exhaust piston subtracts "height" from the engine.
The combustion pistons are both crosshead-pistons, i.e. they can operate without touching the combustion-cylinder-liner (lube-oil consumption / scuffing resistance / emissions).
The scavenging-pump cylinder-head operates at low temperatures and pressures, allowing cheap and easy to mold materials.

Among the advantages of the blade valve is the high flow capacity, the small dead volume, the reliability, the progressive opening and closing of the ports, the lower noise and the simplicity (there are only three moving parts: the crankshaft of the engine, the piston of the engine and the connecting-rod/blade of the engine).

Objections?
Comments?

Thanks
Manolis Pattakos

[gruntguru]

What kind of seal is used where the con rods penetrate the scavenge pump piston?

[manolis]

What kind of seal is used where the con rods penetrate the scavenge pump piston?

The seal is the curved, holed, dark-brown rectangle (one per con-rod):



It is “locked” into a niche at the backside of the piston. The "top" of the connecting rod is formed "cylindrical" and remains permanently in contact with the seal.

But this is the easy seal. The challenge is elsewhere: Can the combustion pistons seal the combustion chamber without using piston rings? Think how many things would change.

The crosshead architecture (the combustion pistons can operate without touching the combustion cylinder) fits much better than the conventional architecture with the ring-less pistons.

In smaller sizes and higher revs (until 20mm bore, above 20.000 rpm) the ring-less piston is quite common, efficient and reliable even for the compression-ignition model engines.

Thanks
Manolis Pattakos

[MatsNorway]

Wild uneducated guess: 500rpm tops before the red thingy starts to flapp around and breaks loose. You could give it a guide, but the mount is still too flimsy.

What program is that drawn in?

If you use Inventor you can use dynamic simulation to find peak G loads etc. and do a FEA analysis on those readouts.

Edited by MatsNorway, 22 February 2013 - 13:07.

[manolis]

Wild uneducated guess: 500rpm tops before the red thingy starts to flapp around and breaks loose. You could give it a guide, but the mount is still too flimsy.

Thanks MatsNorway.

Here is a brief calculation / approximation of the forces:

The long stroke PatOP of the animation has: 80mm bore, 64+128=192mm stroke.

At 4,000 rpm the mean piston speed of the exhaust piston is 17m/sec, at 5,000 rpm it becomes 21.3m/sec, which is set as the rev limit.

The surface of the blade valve is 100cm2 (10000mm2).

With 1mm thickness, the blade valve volume is 10cm3.

With 1.5gr/cm3 density (Kevlar, Carbon fiber) the blade valve mass is 15gr or 0.015Kg.

With 2.8gr/cm3 density (aluminum) the blade mass is 28gr or 0.028Kg.

At 5,000 rpm the maximum acceleration of the scavenging pump piston is 1150g (11300m/sec2) upwards and 650g inwards (i.e. towards the crankshaft).

The maximum force on the blade valve at the 5000rpm rev limit is (approximately):
case of Kevlar : 1150g*0.015Kgr=17.3Kp (170Nt or 39lb).
case of aluminum: 1150g*0.028Kg=32.2Kp (316Nt or 71lb).

So, the extension of the connecting rod has to apply to the Kevlar blade valve less than 20 Kp at the rev limit, while the combustion pressure loads the connecting rod main body by 6000Kp (6ton): 50cm2 piston surface times 120Kp/cm2 peak combustion pressure.

It seems there is no problem with the inertia forces and the support of the blade valve. Even if it is made out of steel, even it is made thicker.

At the 500rpm (i.e. at 10 times lower revs) the inertia forces drop 100 times (10*10=100), i.e. at 500 rpm the blade valve needs to receive a force of 0.17Kp (or 170p or 1.7Nt or 0.4lb) from the extension of the connecting rod.

More important than the size of the forces (which is not big as shown above) is their direction. The blade moves only parallel to itself , i.e. wherein it has its maximum strength, and rotates about an axis normal to the blade (the axis of the wrist pin). i.e. wherein it has its maximum strength.


By the way, this is a Diesel engine, and the pump is a scavenging pump (rather than a supercharging pump) aiming to minimize the friction and the pumping losses of a Diesel engine and make it work at VERY lean Fuel-Air-ratios where the cycle-efficiency is over 60% (the goal is to get this 60% out on the crankshaft by minimizing the heat loss through leaner burn, eliminating friction losses through ringless pistons, reducing pumping loss through better aerodynamics.

What program is that drawn in?

The slides are made with AutoCAD (and AutoLisp programs when necessary).
The slides are combined into a "controllable windows exe animation" by a program written in Visual Basic.
The slides are combined into a GIF animation by the Microsoft GIF animator.

Manolis Pattakos

[MatsNorway]

Mighty impressive that your able to do that in Autocad. here we went over to Inventor somewhere in 2000-2004 i think. Its the future. Jump on it. You can still do lots with dwg files etc.

Its tempting to ask for the DWG and see if im able to do some dynamic simulation on it in inventor. Very likely im not.

[manolis]

Its tempting to ask for the DWG and see if im able to do some dynamic simulation on it in inventor. Very likely im not.

MatsNorway,
email me ( my email is at pattakon contact ) and I will reply with the dwg.
Thanks
Manolis Pattakos

[Powersteer]

How about using reed valves instead...

[manolis]

How about using reed valves instead...

Thanks Powersteer.

The first PatOP prototype engine is using reed valves:



There are two inlet reed valves (the dark steel-sheet leaf of the one is shown at right), and another two transfer - or exhaust - reed valves (the basis of the one is shown at left, the steel-sheet leaf is shown behind the holes of the basis).

In the two stroke engines the disk valve is/was, depending on the application, the alternative of the reed valves (in the model engines, for instance, wherein the rev limit is over 20,000 rpm, the reed valves were soon phased out; in their position they are used disk - or drum - valves).

The disk valve is secured to, and rotates with, the crankshaft, while the crankcase is used as the "compression chamber" of the scavenging pump.
The blade valve is secured to the connecting rod, while the "compression chamber" of the scavenging pump is sealed from the crankcase allowing like-four-stroke lubrication, like-four-stroke lube consumption, way smaller dead volume etc.

Each type (reed valve, disk valve, piston valve, blade valve) has its own advantages and characteristics.

Manolis Pattakos

[Powersteer]

So you split the breathing element of a 2 stroke engine from the crank area which needs lubrication, clever. Have you designed the pump piston on a 1:1 ratio with the combustion piston? If so have you thought of changing the ratio?

[manolis]

So you split the breathing element of a 2 stroke engine from the crank area which needs lubrication, clever. Have you designed the pump piston on a 1:1 ratio with the combustion piston? If so have you thought of changing the ratio?

Thanks Powersteer.

The simplest case is the 1:1 ratio between the diameter of the scavenging pump piston and the diameter of the combustion piston.
The OPRE I prototype was made with this ratio, i.e. the bore of the cylinder/casing is the same from the one end to the other.
But this gives a small (1:1) scavenging ratio:




The next OPRE and PatOP prototypes ware made with a more than 1:1 scavenging ratio.
With combustion bore 80 and scavenging bore 86, the OPRE II prototype has a - theoretical - scavenging ratio (86/80)^2 = 1.15




With combustion bore 80 and scavenging bore 90, the OPRE III prototype has a scavenging ratio (90/80)^2 = 1.26




In the PatOP prototype there is only one scavenging piston.
With combustion bore 79.5 and scavenging stroke 64+64=128,
and with scavenging bore 130 and scavenging stroke 64,
the "geometrical" scavenging ratio of the PatOP prototype is ((130/79.5)^2) / 2 = 1.33



The crosshead architecture of these engines allows different combustion and scavenging bores, i.e. scavenging ratios different than 1:1.
More important characteristic of the crosshead architecture id that it offers "like four stroke" lubrication": the combustion piston is rid of thrust loads (the thrust loads are taken at the "scavenging pump piston, at the wrist pin height, wherein there are neither ports, nor hot cylinder walls); the combustion piston skirt can even avoid touching with the combustion cylinder wall. Think about it.

Manolis Pattakos

[Powersteer]

If you were to install a large supercharger to minimize reciprocating parts...can that be make it even more efficient?

[gruntguru]

Reciprocation is not inherently "inefficient". Mannolis' scavenge piston design, with its crosshead guided piston and no contact with the cylinder is low friction. Being direct drive, eliminates losses in external drive mechanisms.

In addition, the reciprocating pump would have a higher efficiency than a Roots.

Not sure about the packaging. A Roots running at high speed may be slightly more compact - Manolis?

[manolis]

If you were to install a large supercharger to minimize reciprocating parts...can that be make it even more efficient?

Thanks Powersteer.

The number of reciprocating parts with, or without, the scavenging pump is the same.
The combustion piston together with the scavenging piston is a body.
The piston skirt of the scavenging piston takes the thrust loads (it is wherein the torque is created), leaving the combustion piston to deal only with the sealing of the combustion cylinder and the cylinder ports.
The scavenging pump "directly" absorbs from the "combustion" the required energy. The reciprocation of the combustion piston needs not first to change to rotary motion of the crankshaft and then back to reciprocation of the scavenging pump piston. This advantage is more apparent at the:



(the windows exe animation is at http://www.pattakon.com/pre/PRE19.exe )

Besides being an efficient pump (it, actually, adds no friction to the engine), the piston-type volumetric scavenging pump (especially when it has small dead volume) maintains its volumetric efficiency along a wide rev range, from very low to very high revs (think how the four-stroke engine operates: during the suction cycle and the exhaust cycle the 4-stroke engine is actually a volumetric piston type pump; it is exactly this "pump" that enables "flat" torque curves in the 4-strokes).

The built-in scavenging pump of the OPRE and PatOP add neither extra cost, nor extra weight, nor extra friction to the engine. Compare this to the case with external scavenging pump that increases substantially the cost of the engine, increases the complication, increases the weight, increases the external dimensions and the friction.

For instance, take the Commer-TS that uses an external Roots compressor for the scavenging (in practice the Roots compressor is comparable in size to the engine):



The Roots compressor has sealing problems (especially at low revs).
Looking how the two rotors and the casing of the Roots cooperate, you can see the cause of its poor efficiency: air from the exhaust of the compressor initially "expands" (i.e. returns) into the space between the two successive teeth that “just arrived”, and later it is pushed back to the exhaust.

What I am saying is that the built-in scavenging pump of the OPRE and PatOP is a significant advantage and not a problem.
In case of turbocharging, the TwinCharger version:



at http://www.pattakon....ttakonPatOP.htm is interesting.
The compressed air from the turbocharger goes to an air-cooler.
The throttle valve (green) allows or stops the communication of the air-cooler with the space behind the intake piston.
When the turbocharger pressure is low (cranking, low revs, light loads etc) the throttle valve is kept closed, air enters through the one way (reed) valve into the scavenge cylinder and is trapped there for the scavenging.
When the turbocharger pressure increases, the throttle valve opens, the one way valve remains constantly closed (less noise, improved reliability) and the scavenging is made by exploiting the energy of the exhaust gas.

Manolis Pattakos

Edited by manolis, 10 March 2013 - 06:45.

[manolis]

Reciprocation is not inherently "inefficient". Mannolis' scavenge piston design, with its crosshead guided piston and no contact with the cylinder is low friction. Being direct drive, eliminates losses in external drive mechanisms.

In addition, the reciprocating pump would have a higher efficiency than a Roots.

Not sure about the packaging. A Roots running at high speed may be slightly more compact - Manolis?

Thanks Gruntguru.

An external compressor adds weight and increases the external dimensions.

The full-balanced single-cylinder direct-injection-Diesel PatOP prototype has a combustion chamber stroke of 128mm (64+64=128).
The bore is 79.5mm.
The capacity is: 635cc.
The scavenging pump capacity is: 850cc (i.e. the geometrical scavenging ratio: 1.33).

The height of the engine is only 500mm.
The weight of the engine, without the flywheel, is 20Kp / 44lb.

Isn't it one of the most compact reciprocating piston engines?

Manolis Pattakos

[Powersteer]

Besides being an efficient pump (it, actually, adds no friction to the engine), the piston-type volumetric scavenging pump (especially when it has small dead volume) maintains its volumetric efficiency along a wide rev range, from very low to very high revs (think how the four-stroke engine operates: during the suction cycle and the exhaust cycle the 4-stroke engine is actually a volumetric piston type pump; it is exactly this "pump" that enables "flat" torque curves in the 4-strokes).

The built-in scavenging pump of the OPRE and PatOP add neither extra cost, nor extra weight, nor extra friction to the engine. Compare this to the case with external scavenging pump that increases substantially the cost of the engine, increases the complication, increases the weight, increases the external dimensions and the friction.

Think I need to correct myself here,....use a supercharger to minimize reciprocating mass that load up bearings, friction seal parts with piston pumps but you have already adressed it, sorry.

Edited by Powersteer, 03 April 2013 - 20:09.

[manolis]

Think I need to correct myself here,....use a supercharger to minimize reciprocating mass that load up bearings, friction seal parts with piston pumps but you have already adressed it, sorry.

Thanks Powersteer.
By the way, the tilting valve at http://www.pattakon....akonTilting.htm is (just like the blade valve) a "connecting rod valve" for 2-stroke engines:



Here is the version for "Pulling Rod Engines" (PRE):



wherein the dead volume of the scavenging pump is even smaller.

Compare the tilting valve with the conventional drum or disk or reed valves.

Manolis Pattakos

[malbear]

Thanks Powersteer.

The simplest case is the 1:1 ratio between the diameter of the scavenging pump piston and the diameter of the combustion piston.
The OPRE I prototype was made with this ratio, i.e. the bore of the cylinder/casing is the same from the one end to the other.
But this gives a small (1:1) scavenging ratio:




The next OPRE and PatOP prototypes ware made with a more than 1:1 scavenging ratio.
With combustion bore 80 and scavenging bore 86, the OPRE II prototype has a - theoretical - scavenging ratio (86/80)^2 = 1.15




With combustion bore 80 and scavenging bore 90, the OPRE III prototype has a scavenging ratio (90/80)^2 = 1.26




In the PatOP prototype there is only one scavenging piston.
With combustion bore 79.5 and scavenging stroke 64+64=128,
and with scavenging bore 130 and scavenging stroke 64,
the "geometrical" scavenging ratio of the PatOP prototype is ((130/79.5)^2) / 2 = 1.33



The crosshead architecture of these engines allows different combustion and scavenging bores, i.e. scavenging ratios different than 1:1.
More important characteristic of the crosshead architecture id that it offers "like four stroke" lubrication": the combustion piston is rid of thrust loads (the thrust loads are taken at the "scavenging pump piston, at the wrist pin height, wherein there are neither ports, nor hot cylinder walls); the combustion piston skirt can even avoid touching with the combustion cylinder wall. Think about it.

Manolis Pattakos

what type of injector are you using ? it looks like a standard one from some other small engine. does the spray pattern suit the opposed piston combustion chamber?
Mal

[manolis]

what type of injector are you using ? it looks like a standard one from some other small engine. does the spray pattern suit the opposed piston combustion chamber?
Mal

Malbear,

The fuel injector and the high pressure fuel pump are from a cheap Chinese electric generator set (single cylinder, 400cc, 4-stroke, 3000rpm).

The injector nozzle has four holes, all at the wrong direction for an opposed piston engine, i.e. the spray pattern does not suit the combustion chamber of the OPRE and PatOP engines. On the other hand, the cheap standard injector is OK for a low-budget proof-of-concept prototype.

One of the injectors was modified by closing the existing four holes (0.2mm diameter, each) and by opening a central hole (0.4mm diameter). In some videos the OPRE and PatOP prototypes operate with the standard injector, in some others with the modified single hole injector.

In the last version of the PatOP at http://www.pattakon....P.htm#CrossHead



the injector nozzle gets deep into the combustion chamber, softening the problem.

Thanks
Manolis Pattakos