10-15% reduced consumption after "engine split" (physically deactivating 2 of 4 cyl.)

[ElectricZX2]

I did a similar thing on my 90 Honda civic 4-cylinder. I modified the lifters on one of the cylinders that had a bad valve guide such that the valves wouldn't open. Was surprisingly smooth on 3 cylinders and fuel economy went from the low-mid 40s to over 50 once I got the fuel/air ratio right. It was fuel injected but with a single injector so I put a pressure regulator (from grainger) in-line with the fuel line and could adjust the pressure with the twist of a screw. Since the computer was pretty dumb I achieved a fairly stoich mixture and actually saw close to 55mpg experimenting with "lean-burn" by adjusting the pressure a bit lower than stoich.

[Christ]

Originally Posted by metromizer

Thanks for posting this, I have a few comments.

only 5-10% reduction in fuel consumption just isn't enough motivation to justify the work, in my mind. By disabling a couple cylinders, but not removing the 'dead cylinder's' friction doesn't seem worth the trouble. If you leave the pistons in, you need to maintain some valve action. If you don't, the piston is compressing air needlessly. Worse, it has no place to go. Ring friction is also a big deal.

Most of the disabling a cylinder or two FE experiments I've read about, have started with a worn out and/or broken engine. I'd like to see someone test a good running engine on a dyno, then remove two pistons and run it.

Pushrod engine: there is a couple ways of dealing with loose lifters after removing pushrods, pistons, permanently disabling a couple cylinders. If you completely remove the lifter, on many engines, you must close off the lifter oiling port or replace the lifter with a plug of somesort. If you remove the camshaft, and precision grind the lobes off round, you can just leave the lifter in place. Another way to remove pistons without having to m9odify the crankshaft, is to modify the connecting rods. Saw the beam and upper end completely off, grind it smooth, and re-install what is left with bearings. That will control the oil leak from that journal. If you don't want it to turn, just add a little .003 shim stock under the bearing shells of the modified rods.

The compression loss is actually very little, the compressed air doesn't go anywhere on the compression stroke and acts as a spring pushing back down on the piston on the expansion (combustion regularly) stroke.

The pumping losses come from actually moving air. the best bet in that case is just to remove the valve actuators (rockers, usually) and let the cylinder continue to compress and expand the air already existent.

To minimize the air in the cylinder (which will naturally dissipate over time anyway), remove the rockers while the cylinder is at TDC with either the intake or exhaust valve open.

Someone mentioned modifying the pistons by putting holes in the crowns:

While this /will/ work to reduce pumping losses, it isn't necessary as I pointed out above, because in order to do so, you would also need to prevent the valves from opening to keep oil from flowing into the manifolds from the oil mist that occurs naturally in the crankcase.

The best bet to remove friction is either to completely remove the piston from the cylinder, affecting balance in some engines, or just deal with it and move on to larger/lower fruit.

To negate (most of) the balance issue, leave the cylinders that are 180* opposed to each other. This will result in a smoother engine with a better torque curve, similar to a 180* parallel twin common in Japanese motorcycles.

It is best, I've been told, to leave the cylinders closest to each other as well in the case of 4 cylinder engines, to prevent resonant vibration as much as possible, but I'm not necessarily sure it really makes a difference.

If you really want to utilize the two cylinders you've removed, you can run pod filters on their intakes and modify their exhaust ports to feed back into the remaining two cylinders' intakes to act as a "supercharger" of sorts to maximize power for the remaining output of the engine. You would need to modify the cams by adding an additional lobe so that every down stroke were intake and every up stroke were exhaust. The result would be a compressed charge in the intake due to 2 "charge" strokes per "power" stroke of a utilized piston, and a required retuning of the fuel/ignition systems to handle that.

[JohnAh]

This thread may beginning to get too long for newcommers to read it all before replying so I´ll try to summarize what I actually did to my Fiat engine:

I removed the two pistons on one end next to each other (piston 3 and 4). I chosed the 180+540 configuration to reduce vibrations from mechanical imbalance. I know from my earlier experiences that single cylinder engines as well as paralell twins with 360 degree crankhaft have quite unpleasant vibrations so I was curious how a 180+540 would run.

My resulting engine was very smooth at any speed but as expected it gave stong vibrations when giving torque at low speed. The long and uneven separation of powerstrokes makes such an engine quite annoying and demands a lot of work with the gearbox. I guess however the 2x360 config would have been more cruel to the engine in the long run and also would make unplesant vibration at any speed.

The fuel saving was 10-15% with no major changes in average speed. The original engine was a worn-out carburetted 1050cc OHC of 50hp. When I sold the car I had run it for over 40.000 km in this crazy way. There was a lot of blow-by caused by worn cylinders and rings. I guess a modern injection engine in good condition could give higher fuel savings.

--------------

Yesterday I came up with a possible method to estimate the fuel savings for a 4-2 conversion with valves disengaged but all pistons still in place: Hammer two of the sparkplugs together and meassure the fuel consumption. (a cathalytic converter may get ruined by the unburnedt fuel passing by) The consumption should be ROUGHLY doubled and any diffrerence from this theory must be quite equal to the savings when the engine is modified later. If the consumpton is more than doubled I would be surprised and the only reason for such a result that I can see is if the piston and bearing losses are MUCH bigger than the pumping losses together with the less efficient cylinder filling.

If I hade the time and a scrap-engine to do this to again I would start with the comparison above, then proceed with disabling intake and outlet valves on two cylinders. After this I would try the modified pistons with LARGE holes in the crown and without rings to see if piston friction is a separate meassurable factor. Finally I would remove the pistons completely to se if I the bearing and last bit of piston friction is of any importance.

The suggestion of keeping two pistons after modifying them may be a tricky step to learn from. With pistons in original shape and rings still in place the compression/expansion will act as a spring and not waste much energy in theory but in practise there may be losses from blow-by and thermal losses from the compression. Or is all that heat returned during the expansion to follow?

With only the rings removed to reduce friction I suppose the blowby will increase significantly. This is why I suggested that the piston crowns should be modified with LARGE holes. If the holes are too small the pumping losses will increase. From what I know about "impedance matching" in electronics I suppose medium sized holes will cost more energy than large holes or no holes at all.

The only reason to keep the pistons at all is to act as counterweights. An alternative way suggested by somebody earlier in this thread is to keep only the big-ends of the conrods or even better to make pairs of heavier C-shapeds counterweights to replace the conrods. Unfortunately this approach is exactly the same as when balancing any single cylinder engine: A rotating counterweight can never cancel out the linear movement from the piston. (I guess the conrod can be theoreticaly divided in one part belonging to the piston and one part to the crankshaft)

I wrote earlier that I have experiences from p-twins with 360 deg. crankshaft. They fire twice as often as a single 4-stroke but they have the same dynamic balancing problem. -They vibrate a lot and balancing them is always a compromise. The only way to make them mechanically smooth is to install a balance shaft or symmetricaly placed linear counterweighs to make it run like a symmetrical boxer engine with fork-shaped conrod on one of the cylinders (and a cranshaft with three throws). Such an engine is btw the only way to have a 2x360 twin with good balance.

A 180-crank is offcourse not perfect. The pistons will counter-act in a single direction but a rocking vibration from the crank and conrods will remain.

A straight six or a single disc wankel is probably the smallest configurations that run perfectly smooth unloaded but none of them are known to be the most fuel efficient among ICE:s... The most efficient config I can imagine is a single cylinder 4-stroke, preferably with some kind of balancing device, and a transmission with many gears to keep the rev up at any speed. I guess a balancing device cost less fuel than the friction of another set of piston, valves and bearings.