The goal for this idea is to have an engine which has high efficiency, high power density, and low vibration.

The design is as follows: two or more cylinders are arranged radially around a central crankshaft.

In each cylinder, on the side toward the crankshaft, is a piston; all of these pistons are connected to a single crank on the shaft (just like in a regular rotary engine).

In each cylinder, on the side away from the crankshaft, is a second piston, whose face is aimed inwards, towards the first piston. The rear of each of these outer pistons is connected, via a short connecting rod, to a pair of long connecting rods (one along each side of the cylinder), which in turn connect to the crankshaft.

The two cranks to which the long connecting rods attach to are in phase with each other, and about 160 degrees away from the crank which drives the inner pistons.

The inner pistons, the ones closer to the crankshaft, cover and uncover the intake ports. The outer pistons, the ones further from the crankshaft, cover and uncover the exhaust ports. In the part of the cylinders between the outer and inner pistons, there would be fuel injectors (gasoline or diesel), and either glow plugs or spark plugs.

Forced induction would be provided by a hybrid turbocharger, with a variable geometry turbine.

Advantages:

Because there are two sets of moving parts (inner pistons and outer pistons), their motions largely cancel out, minimizing load on the crankshaft bearings, and minimizing engine vibration. This results in lower weight (smaller bearings, smaller engine mounts) and a longer service interval.

Because each cylinder has two pistons, the pistons need to move half the distance a single piston would in a regular radial. Peak piston speed is cut in half, which means that the maximum engine RPM is higher, which increases the engine's maximum power, which increases the engine's power density.

Due to the uniflow scavenging, the amount of fuel that can be burnt per cycle is higher than if loop scavenging were used; however, unlike most uniflow scavenged engines, the exhaust is through a port which a piston covers and uncovers, instead of through a valve. This reduced parasitic loss, and increases available engine power.

If the engine is to be built to use diesel fuel, then direct injection is natural; if the engine is to use gasoline, then the advantages of using direct injection should be obvious.

As with all two stroke engines, forced induction is necessary; by using a hybrid turbocharger with a variable geometry turbine, the engine can operate efficiently over a wide range of speeds, and turbo lag is largely avoided.

The reason why the crank for the inner pistons would be between 160 degrees out of phase from the cranks for the outer pistons is as follows:

If the cranks were 180 degrees out of phase, then dynamic equilibrium can be trivially achieved, the load on the crankshaft bearings could be zero, and vibration minimal... but the time the pistons spend at TDC would be short.

If the cranks are 160 degrees out of phase, the relative speed of the two pistons slows down close to TDC, allowing more time for fuel to burn, which results in more complete combustion, thus more power and less pollution.

Also, I think it allows the port timing to be such that expansion stroke is longer than the compression stroke, producing something like the Miller cycle.

The advantage of this design over OPOC is of course that it's not limited to just two cylinders. And I'm pretty certain that the OPOC design uses a crank angle of 180 degrees (but I'm not sure since it's not mentioned in the text of their website, nor is it clear in the pictures & animations).
-- goldbb, Sep 14 2011

OPOC engine home http://www.ecomotors.com/engine-design
Opposed Piston Opposed Cylinder [goldbb, Sep 14 2011]

Napier 'Deltic' Diesel http://en.m.wikiped.../wiki/Napier_deltic
Similar? [8th of 7, Sep 14 2011]

Six stroke engines http://en.wikipedia...i/Six-stroke_engine
Less complex means to couple the second piston. [Twizz, Sep 16 2011]

Torsional Vibration http://www.enginehi.../NoShortDays/TV.pdf
Not trivial [Twizz, Sep 19 2011]

What is the idea?

The OPOC engine (not a new design) is exactly as described in the post, phase angle included.

The principle of opposed pistons phased to create asymmetric 2 stroke porting goes back to 1900. Even then, engines were built with single crankshaft, long rods and superchargers, as described.

Arranging more than 2 cylinders in a radial pattern would require a complicated crank and rod arrangement.

Radial engines suffer from issues with master rod loads and torsional vibration. Adding more throws to the crank (for the long rods) and trying to manage the forces in the master long rods will make the torsional vibration issues considerably worse.
-- Twizz, Sep 14 2011


//160 degress... vibration// an 8 cylinder model would be balanced in that respect. [edit:oops]

hmm, what about... instead of two pistons, in one cylinder, each connected on opposite sides of the crankshaft... have just one piston in one cylinder, both connected to opposite sides of the crankshaft.

Basically the same advantages but without lengthy pushrods.
-- FlyingToaster, Sep 16 2011


FT - I'm not sure you've grasped the concept of the opposed piston engine. Essentially, it's a 2 stroke with improved timing. This is achieved by the use of 2 pistons out of phase with one another.

Torsional vibration is the result of power pulses exciting the torsional frequency of the crankshaft assembly. The usual way to avoid this problem is to make the crankshaft stiff enough that it's torsional frequency is higher than the frequency of power pulses at max RPM. The 3 throw crank required for this type of opposed piston engine will be substantially less stiff than a conventional single throw crank. As a 2-stroke, the frequency of power pulses will also be twice as high as in a 4-stroke. Adding more cylinders will further increase that number.

If I wanted to design a radial opposed piston engine, I would probably use a much shorter stroke for the 'upper' piston, using it primarily as a valve. This would allow for actuation with a small crank or camshaft operated in a similar manner to overhead cams in 4 strokes. The layout ends up very similar to Beare's 6 stroke design, but with the primary and secondary cranks operating at the same RPM.
-- Twizz, Sep 16 2011


The Jumo design (IIRC) was an inline which used a small(er) crank for the (smaller)upper pistons which also ran all the auxiliaries: fuel pumps etc.

Many-power-pulses is not necessarily a bad thing: 5 evenly spaced .2L cylinders makes for a smoother, if lossier, ride than 1 1L cylinder, Vernonian maths aside.

I admittedly don't really get the 'balance' thing.
-- FlyingToaster, Sep 16 2011


/What is the idea?/

The idea is basically the same as OPOC, but with the possibility of using more than two cylinders; also combining it with a hybrid electric turbocharger, with a variable geometry turbine.
-- goldbb, Sep 18 2011


/Arranging more than 2 cylinders in a radial pattern would require a complicated crank and rod arrangement./

Perhaps I'm missing something; I would expect to use basically the same crank and rod arrangement, just more of them.

/Radial engines suffer from issues with master rod loads and torsional vibration./

Why does there need to be a master rod? Why can't all of the connecting rods attach to their crank directly?

As for torsional vibration... I'm guessing that you're talking about the long connecting rods flexing, yes?

/Adding more throws to the crank (for the long rods) and trying to manage the forces in the master long rods will make the torsional vibration issues considerably worse./

What about getting rid of the long connecting rods, and using some alternative?

For example, if the outer piston were connected to the crankshaft an internal rod, which passed *through* the inner piston (like the rod which drives the displacer piston in a beta configuration Sterling engine), there would be no cause for this rod to flex.
-- goldbb, Sep 18 2011


Radial engines (mostly) use a master rod (with a 'big end' running on the crankshaft) and slave rods (with smaller 'big ends' attached to the master rod. This allows for a short, stiff ctrank design. If each rod is of conventional design, with a big end running on the crank, the crank gets longer (and floppier) as you add more cylinders.

Torsional vibration is twisting of the crankshaft. An inline engine has regular power pulses exciting the crankshaft, but they are all at different locations along the crankshaft. Vibration can be prevented by using the correct firing order. A radial engine has regular power pulses exciting the crankshaft, all at the same location. Changing the firing order makes no difference in this configuration.
-- Twizz, Sep 19 2011


Radials only have one 'throw' and all the conrods are connected to one master conrod/hub which is on the throw. However if there's only a couple cylinders they can use forked conrods (I'd guess 2, possibly 3 cylinders would be the maximum) and retain the single point of force transfer, which means the radial remains vibrationless in all but the axial plane.

I have no clue what you'd call a radial with a segmented crankshaft (though I assume they exist).
-- FlyingToaster, Sep 19 2011


"radial remains vibrationless in all but the axial plane"

Any device with any kind of motion will have vibration. A simple electric motor experiences vibration within the windings, the shaft, the commutator, bearings and frame even if it is perfectly balanced.

See link re torsional vibration.
-- Twizz, Sep 19 2011



random, halfbakery