The compound-expansion internal-combustion engine was patented in the late 1800's by Gottlieb Daimler. The most basic form is an inline 2-cylinder with a passive third, larger, cylinder wedged between. The exhaust strokes, rather than leading to the outside, are pumped into the larger cylinder for further expansion. The two (4-stroke) cylinders timeshare the one (2-stroke) expansion cylinder.

A number of efforts were made <link> but it seems efficiency suffered due to loss of thermal energy during the transfer.

This idea proposes the reasonably obvious¹: use a sleeve valve (or any other design that can put a port right into a shared cylinder wall) to effect a lossless transfer between cylinders, instead of poppets on top with external transfer pipes.

The result should be an engine that provides the same power and efficiency as a modern four cylinder "Atkinson" style engine, but in a smaller and lighter package².

The engine can be made smoother running than its 4-banger competition as well, by using crankcase chambers as air-springs: compression for the two regular cylinders, tension for the expansion cylinder; strength set with simple one-way flap valves.

Unique to this design, there's even more...

• Vehicular pneumatic regen/launch that actually works:

Driven by the driveshaft during braking, the I-2+1 can fill (and later empty) a small air storage tank using the expansion cylinder as pump. Since the piston in the expansion cylinder goes right to the top there's no pressure limitation³.

• Coolant thermal recycling:

Put the cooling system under enough pressure to keep the water below its boiling point then inject it, when it's hot enough to flash into steam from the pressure drop, into the extended expansion stroke.

Recover the water from a magic box on the exhaust (hopefully one that can harvest energy from the phase change), filter to remove large particulates, possibly treat for dissolved NOx's, and return to the water sump.

---

¹ Obvious in hindsight, that is : sleeve valves didn't come into common usage until the 1920's.

² The most annoying thing about modern Atkinsons is they spend quite a bit of time pumping air in and out for no purpose.

³ Normal cylinders can only pump as much as their geometric compression ratio - eg: a 10:1 can pump to 10bar - because the piston stops short of the top of the cylinder.
-- FlyingToaster, Dec 17 2015

Compound internal combustion engines http://www.douglas-...ndIC/compoundIC.htm
from the late 1800's and early 1900's. Aesthetic. [FlyingToaster, Dec 17 2015, last modified Dec 18 2015]

Animation, called the "5-stroke", as repatented by some guy in the 1990's, who talked Mercedes Benz into getting their pet racecar engine company to build a one-off. https://www.youtube...watch?v=rpw0SfBcSRM
It's the same as Daimler's patent, of course. They seem to have gotten decent results but, since they don't mention the previous century's work anywhere in their blurbs, I'm a bit skeptical as to their reporting. Also, of course, they're still using poppet valves. [FlyingToaster, Dec 19 2015, last modified Dec 20 2015]

Too much frothing at the mouth ?
-- FlyingToaster, Dec 19 2015


Can you make a drawing to clarify? Just a little sketch showing the sequence? Don't be shy, I put up stuff that looks like it was drawn with crayons all the time.
-- doctorremulac3, Dec 19 2015


Well, hmmmm...

For the basic flow I can give you somebody else's take (who's also trying to flog the basic design). Hang on a bit... okay, <link>.
-- FlyingToaster, Dec 19 2015


That's cool, a picture is worth a thousand words but an animation offers complete clarity.

So this improves on the efficiency by using a different valve that has less friction? Sorry to be slow. Coffee just kicking in.
-- doctorremulac3, Dec 19 2015


The poppet (or most others) valve is on top of the cylinder head. In order to move the hot gases from the initial cylinder into the expansion cylinder, the gas goes up, around and down again through a transfer pipe. Lots of heat gets radiated or convected to the outside through the pipe for no reason, the volume of the pipe takes away from the expansion volume, and what you mentioned: friction along the walls of the pipe.

It's like going to the office next door by walking into the corridor, taking an elevator up a floor, then getting on the elevator next to it to come back down again, then walking into the new office... by which time you're tired and have mostly forgotten why you wanted to go there in the first place.

A sleeve valve basically operates through a hole in the wall of the cylinder. Now you're walking to the office next door by opening a door between rooms.

No apology necessary; I may have pared the original novella-sized draft down a bit too much.
-- FlyingToaster, Dec 19 2015


The crankcase air-springs bit puts a high-pressure spring under the outside cylinders and a vacuum spring under the middle cylinder. This is done easily enough by one-way valves opening inwards on the outside cylinders, and a one-way valve opening outwards on the middle cylinder.

What they do is put some more energy into the half revolution where the middle cylinder is going down and the other two are going up.

You could also do this on a regular I-2 engine, or on a 1cyl 2-stroke engine, to smooth them out.
-- FlyingToaster, Dec 19 2015


Ok, got ya. If you can take some parts out of a mechanism and have it do the same thing that's always a good deal, usually for multiple reasons. [+]
-- doctorremulac3, Dec 19 2015


Well, the basic principle's the same, of course. Assuming a compression to (over)expansion ratio of 1:2, it's going to be as powerful as a 4cyl Atkinson (with combusting cylinders the same size), while being smaller and lighter.

Adding the air-springs should bring it up to par vibration-wise.

Adding a little more plumbing to an engine, to get pneumatic regeneration using the engine's cylinders to pump, has certainly be talked about, but never been seen; probably because it can't pump more than the cylinder's compression ratio. However this engine's expansion cylinder isn't limited by that.

Water spray (steam spray, actually) might work as well. Of course it might also bring the air's temperature down enough to be not worth it, but hey that's what experimentation is for (or simple physics math). It would also serve to filter out some particulates and NOx out of the exhaust.
-- FlyingToaster, Dec 19 2015



random, halfbakery