Picture a pair of cylinders in a boxer-configuration engine. For simplicity's sake, ignore linkages and offsets: it's just one long enclosed tube with 2 pistons inside. For clarification's sake we'll add nomenclature specific to the relativity of the pistons: in-stroke and out-stroke.

Put an intake port on each cylinder head, an exhaust port on each piston, and one supplementary exhaust port right in the middle of the long cylinder.

The intake and piston-exhaust ports act exactly as they usually do in 4-stroke boxer: for each in-stroke one side is combusting/expanding while the other side is intaking fresh air; for each out-stroke one side is compressing, the other exhausting.

Of course the difference here is that the piston exhaust-ports are "exhausting", still full of energy, into the middle section ghost cylinder. During the out-stroke, the combustion products continue to expand, continuing to power the engine.

The ghost-cylinder's exhaust, into the engine's exhaust manifold, is open during each in-stroke and closed during each out-stroke.

The resulting engine has a native uniflow 2:1 Atkinson ratio, at half the geometric displacement(ie: size/weight) such an AR ratio requires in a standard configuration.

The ghost cylinder's a neat place to put a water spray, as well.

The "ghost cylinder" is also, of course, the crankcase, or at least the crankcase is enclosed by the ghost cylinder, whatever. Maximum effectiveness is achieved when the cylinders aren't offset at all (ie: a complex linkage). The "one long cylinder" would be a bit bulgy in the middle to accommodate the linkages' movement. The oil reservoir would probably have to be external.
-- FlyingToaster, Apr 09 2014

Boxer Engine http://en.wikipedia.org/wiki/Boxer_engine
is a flat engine (a line of pairs of cylinders that face each other) in which each pair of pistons come together in the middle at the same time. [FlyingToaster, Apr 09 2014]

4-Cylinder Pogo Stick 4-cylinder_20Pogo_20Stick
As mentioned in an annotation. [Vernon, Apr 09 2014]

Coaxial Boxer Engine http://hildstrom.co...ts/boxer/index.html
[FlyingToaster, Apr 15 2014]

Sorta like this, except not a http://www.youtube....watch?v=ISpKyOmISs4
V engine. And I think some of the tiny arrows are going the wrong way. [FlyingToaster, May 12 2014]

Sounds interesting. I think I understood what you intended, but it took a while. Perhaps a link to a concise description of a normal boxer and a sketch of yours would make it easier.

I thought of two concerns:
1) If you are using your crankcase as your ghost cylinder you'll need to think more about the seal around your drive shaft.
2)It also seem like there might be heat issues in the crankcase. I think I heard that the cylinders and pistons normally receive some amount of cooling from intake air. The ghost cyliner will have a new batch of hot air every cycle, and I worry about the bearings on your linkages in that case. But I'm no engine expert...

[+] Since it sounds like the concept harnesses quite a bit more energy with only marginal increase to the size, assuming you can work out all the details.
-- scad mientist, Apr 09 2014


I'm not an engine-expert either, but the challenges don't seem insurmountable.

The design is for a new-build, rather than a retrofit, so the crankcase bearings and seals are from-scratch. The crankcase is as small as possible of course, but the water-jacket or air-cooling-fins extend around it as well.

Thanks for the reminder: I'd meant to link to a boxer-engine explanation. The pistons moving in and out at the same time (thus the "boxer" moniker) is the key to the whole thing.

When the pistons are as near to each other as they're going to get, the exhaust valve in whichever piston is at the end of its combustion stroke opens (the other piston's port is closed as it's just starting its compression stroke), allowing the combustion product to flow freely into the confined space in between pistons. When both pistons then retreat from each other, the combustion product continues to expand insde the ghost cylinder which gets bigger as the other piston retreats. the hot-piston's exhaust valve remains open during this stroke.

When the pistons are at their furthest extension from the center, the main exhaust valve opens, releasing the spent exhaust into the manifold. This exhaust valve stays open until the pistons get back together again.

So, while the "real" cylinders go through their 4-stroke operations - combustion, exhaust, intake, compression - the "ghost cylinder" goes through 2 cycles of a 2-stroke operation: expansion, exhaust, expansion, exhaust.
-- FlyingToaster, Apr 09 2014


//a 2-stroke operation consisting of expansion and exhaustion.//

You've just summarised my life.
-- MaxwellBuchanan, Apr 09 2014


Sorry, I am not agreeing with the visualization presented. The linked article specifically states the two cylinders of a Boxer engine cannot be directly across from each other, but:
"just one long enclosed tube with 2 pistons inside."
OK, here is the stuff on the inside of that tube, without including the tube itself (the = represents the shaft, and the underscore represents empty space):
(cylinder head) |_1_|==2==|===C===|=3=|__4__| (cylinder head)
The 4 combustion chambers are numbered. The Crankcase area is represented by "C". The vertical bars between 1 and 2, and between 3 and 4, are the pistons. The vertical bars between 2 and C and between C and 3, are interior "cylinder heads". The overall shaft passes through these "heads" and must be sealed, just like the outside of each piston must be sealed.

In the ASCII sketch the pistons are located such that this could be happening inside the engine (pistons and shaft moving toward the left side of the sketch):
1: exhaust stroke
2: power stroke
3: compression stroke
4: intake stroke

The Volume Of The Crankcase Space Does Not Change, unlike what the Idea in the main text describes. What I would do, in that Crankcase area, is use a magnet and a coil to generate electricity (fairly high-frequency A.C.) from the reciprocating shaft, and then use the electricity to move the vehicle containing this engine.

Major mechanical things would not exist to wear out in that Crankcase space. Valves and spark plugs for regions 2 and 3 could be located on the body of the overall engine cylinder. Note the Alternating Current produced could have certain points in its cycle perfectly associated with opening the valves electrically, or firing the spark.

I should mention that this description comes from an idea I had many years ago, but never pursued because I found out that others were working on it (including generating AC power). So far as I know, though, it is not widely known, and perhaps I can post this as a separate Idea here (hadn't really thought about that before). I did post a variant of it as "4-Cylinder Pogo Stick", because it always bothered me that the shaft, passing through those interior "cylinder heads" needed to be sealed. (Still, they DO succeed at sealing piston edges, so sealing a shaft should be quite workable....)
-- Vernon, Apr 09 2014


III ·combust· I>I ···exhOaust··· I<I ··intake·· O

III ·exhaust· <O ····expand···· I>I ·compress· III

O ··intake·· I>I ···exhOaust··· I<I ·combust· III

III ·compress· I<I ····expand···· O> ·exhaust· III

Legend:
III - cylinder heads
I>I - piston travelling --> way
I<I - piston travelling <-- way
O - open valve: intake (on the cyl heads) or exhaust (on the pistons or ghost cylinder)

1. L cyl combustion stroke; R cyl sucking in air from outside; pistons on their way to BDC, pushing spent exhaust from the ghost cylinder.

2. L cyl exhausting into the ghost cylinder through a piston port, R cylinder compressing its charge; pistons on their way to TDC with the ghost cylinder expanding, pushing against the underside of the R piston.

3. Same as 1, except from the other side. Now the R cyl is expanding, the left intaking, while the ghost cylinder exhausts.

4. Reverse of 2: the R cyl is exhausting while the L is compressing, the ghost cylinder is helping the L to compress.
-- FlyingToaster, Apr 15 2014


OK, so you are assuming appropriate mechanisms to allow actually-aligned cylinders in a Boxer-style engine, just so you can use the crankspace as an alternate route for the exhaust of the engine's combustion gases.

The only "fly" I see is that you will need an extra set of valves controlling the exhaust of gas from the crankcase to the outside world. Otherwise, when the volume of that space expands (as one Boxer cylinder is compressing fuel-air-mix and the the other has its piston-valve open for exhausting the cylinder), air will be sucked backward from the main exhaust pipe into the crankspace, when maybe you ONLY want air from the exhausting cylinder to enter the crankspace.
-- Vernon, Apr 15 2014


I find it helps to read the ASCII art from top to bottom for each cylinder (left, ghost, right), to see when the strokes are taking place.

The "O" in the middle of "exhOaust" represents when the main exhaust valve is opened to the exhaust manifold (the two real cylinders only exhaust into the ghost). That valve is closed during each of the ghost cylinder's expansion strokes, which is what gives the contraption the extra power: the expansion is pushing against the opposing piston's underside. When it's open, the pistons are simply pushing the ghost-cylinder's exhaust into the exhaust manifold.

There's 3 possible "appropriate mechanisms": an exo-crank, which concept annoys me too much to consider even though it would allow the pistons to butt up against each other at BDC(which is optimum) or, on a normal internal crank, a complex linkage<link> which may be too fragile, or keeping the cylinders off-set and simply have a bend in the middle.

The disadvantage of an internal crankshaft (for this engine) is that the journals and crankpins and stuff need clearance to spin around, resulting in dead space which compromises the extra expansion somewhat. This can be minimized of course by filling all available unused space.
-- FlyingToaster, Apr 15 2014


With a bit of input from Max's "Backside Supercharger" post and annos, this'd work without having to mess with the crankcase, or designing an exo-crankshaft arrangement, by having each cylinder as a double-acting cylinder (closed at both ends, a piston rod (crosshead arrangement) protruding), with an open pipe (with an exhaust valve in the middle) connecting the in-side cylinder halves.

It's lossy, at the start of the extended expansion stage, in proportion to the volume of the connecting pipe, but it works. (Perhaps too lossy considering that the required crosshead arrangement means the pipe will be quite long.)
-- FlyingToaster, May 07 2014


Link added to a YT vid that shows a V2 engine with the same general idea: using the crankcase as a shared compound cylinder. The obvious differences are that this'un's a boxer layout. with exhaust valves in the pistons - much more flowey.

It seems there were quite a few designs for compound IC engines back around the turn of the previous century: 2 firing cylinders sharing a single larger expansion cylinder, all in an I-configuration. I imagine the vibrations were pretty severe, moreso than a regular I2 because of the expansion piston. (The vibrations on this are the same as a regular boxer, ie: not much)

Run it backwards and it's a supercharger @ 3/2 atmospheric.
-- FlyingToaster, May 12 2014



random, halfbakery