Cylinder deactivation discussion
 

i missed the title, this more like cylinder decommissioning

ive always wondered if i take half the pistons,rods and valvetrain out of an engine will it get better mileage? or will the remaining cylinders use more gas working twice as hard?

yes & no or no & yes ?

^YES and no!

It has been done before. I've seen where some college kids back in the day pulled 2 cyls. out of a VW Bug and coaxed 58 mpg out of it, IIRC.

2 Points
 

diesel_john: 2nd point -- I kinda have some experience with cylinder deactivation. A couple years back, I deactivated 2-cylinders in the 'Teg and ran it 50 miles. Too bad it shook too much and the O2 sensor dumped more fuel in the operating cylinders.

If you make a 2-cylinder from a 4 (or 3-cylinder from a 6), you'd have to pull the whole connecting rod and cylinder with the top-end components from the opposing cylinders. Now you have a huge hole in the cylinder top where oil can splatter out. If you can seal it up and balance it out, then you may have a winner -- providing the cylinder action is balanced. Next you'd have to tell the ECU to expect half the burned fuel for stoich. I've heard that it's possible, and has been done.

On a side note: I've driven the new Chevy Impala 3.9L with the 3-cylinder mode (of 6) and the SS with 4-banger of 8. LOD, RPM, and speed determined activation and deactivation with the ECU -- a 5% FE savings was a REAL stretch. I'd like to force the whole thing to work in half-mode all the time. I'm sure it could be hacked. Chrysler's HEMI has a similar system. Extra weight, I s'pose, but half power should result in a big increase if driving properly.

RH77

"Now you have a huge hole in the cylinder top where oil can splatter out."

Valves closed and a dummy spark plug, where's it gonna go?

The problem area is on the crank throw- should a stationary or rotating cover be affixed and how? And doesn't the ECU already know how much fuel to send via the remaining injectors and O2 sensor?

Welp, the cover is tricky bit. If you leave it untreated, you have to disable the valve action (grind the cam lobe down) or else the valve will open and leak oil. If the valves are seated firmly then there should be no problem.

I reported in my experiment that the ECU throws a CEL and a "Cylinder Misfire #1, or #3, code etc.) and dumped the fuel into the operational cylinders to acheive the proper air/fuel ratio for the oxygen sensor. In my application, the ECU didn't figure it out. YMMV, but OBD-II is fairly similar in operation in most vehicles. Pre-OBD vehicles, with carbs, have noted to defeat the system and run well.

BUT, the problem also resides in hot/cool spots with operations vs. dummy cylinder areas. Coolant has a tough time regulating bores with no combustion and hot spots with where the action is conducted.

I'm not trying to downplay the deactivation concept -- just reporting my experience and research. It can likely be defeated with proper preparation and ECU mods.

Sorry to hijack the thread -- we should get back to Atkinson action. Otherwise, maybe a cylinder deactivation thread is required? Feel free to start one, anyone. I failed at that experience, so I'm moving on...

RH77

I wouldn't ruin the cam when all's I gotsta do is pull a few lifters (or followers as the case may be) out. Besides, the valves need to be deactivated so that the flow to the working cylinders doesn't get all screwed up, not to mention crankcase pressures. Then, later on, if I want to re-activate the cyls for some reason (experiment failed? want the power back? want to switch cyls for extended life?) nothing has been wrecked; it's a simple matter of slapping the parts back in.

You misunderstood the "cover" I mentioned, for as mentioned, it goes on the CRANKSHAFT to cover the oil hole and to provide better balance. What I have in mind is to get a few junkyard rods and cut 'em off at the big end, and bolt that onto the crank. What I'm wrestling with right now is, should they be free-spinning (depends on if they'll hit the block) or indexed and solidly bolted on? I guess I can't answer that until I have it disassembled, with parts in hand.

"BUT, the problem also resides in hot/cool spots with operations vs. dummy cylinder areas. Coolant has a tough time regulating bores with no combustion and hot spots with where the action is conducted."

I don't believe you. Neither do any of the manufacturers that offer cylinder deac from the factory.

YOUR ECU screwed up because your "deactivated" cyls were still pumping air into the exhaust stream. John and I are talking about REAL deactivation -maybe ours should be called "decommissioning"- no air being pumped, no fuel being squirted, no pistons going up and down. O2 and ECU function should be "normal" then... unless the TPS being open wider throws something askew.

[1] Why would you want the extra weight spinning on the shaft?
wouldn't that create a parasitic load or even an off balance situation?
I would opt for a small core plug or maybe drive a solid pin all the way through.

[2] You don't believe the coolant would use the empty holes as heat sinks?
I read an old manual when I was in the Army, that gave instructions and part #'s for the conversion of 4cyl jeep into 2cyl run and 2cyl pump to build an air compressor.
Very neat idea. I am sure someone in the Pentagon thought about the heating and cooling factors.

[3] No extra air means the o2 reads right.

Have fun.S.

1. REMOVING the con rod CREATES an unbalanced condition! Now the counterweights on the crank are too heavy if there's nothing on the throw. This is an attempt to re-balance. I've read that the bottom 1/2 of the rod's weight is used for balancing cranks.

2. Absolutely I do. I don't believe it causes problems.

A quick quote from the Chrysler info pack when the vehicle was released:

"Chrysler C300 has a four cylinder / eight cylinder engine.
Four cylinders at normal running and eight on demand.
Utilising the four cylinder demand option results in a seven percent fuel saving."

I would have thought it would be more than that.

Pete.

I wouldn't go through all the work for a measly 7%. GM boys claim up to 20%, depending on conditions. I would go through all the work for 20%. And since this way, there is no access to all cylinders at any time, I'm going to hope that there's a chance for greater than 20% improvement when it's all said and done.

Yes and no... Depending. You can get the same effect more or less from using your left foot and right hand, with maybe some gearing changes.

"Oh duh grampa", My bad I hadn't thought of the weight versus counterbalances. :o I'm not really a troglodyte, I just sound like one sometimes.
S.

Still have all the internal friction and it hasn't changed the actual loading... I'd rather have a better transmission. Especially when you consider at leats in the case of the Odyssey which was the only cyl-deact vehicle I've driven just how much other stuff is added including noise cancellation through the speakers... way too complex! Make the motor smaller, or some kind of on demand boost, but deactivation is too much complication and risk for too little gains IMHO.

To me it would seem to be easier to create more power in a 4 cylinder when its needed , then it would to deactivate half the cylinders and drag them along, tryin to get a 10% improvement/.This to me has been shown particularly true with fuel injection.The early carbed v8 to v4 conversions got around 26 mpg with dummy pistons in the dead holes. If you don't use dummy pistons then you are dragging the rings along. especially if the valves are disabled, and creating both compression and vacumn situations the other 4 cylinders must fight against. In addition you are carrying a lot of unneccessary weight, that would net you free mileage. In my mind it was tryed and was a waste.

Frank -

This makes me think that a Subaru flat-four boxer would also be a candidate, since it is "naturally balanced", yes?

CarloSW2

good discussion modders. i just found out i have to grind the cam lobes off a bit because they hit even with the buckets out. the valves have worn down into the seat. i am thinking cut the rods off and take out the clearance, so they clamp on and can't turn. not much room between the rod and block. next i need decommisson two ports on the injector pump. that should be interesting. all rod throws are individually counter weighted on this crank.
How many votes for even fire, uneven balance?
How many votes for even balance, uneven fire?


1.6 D short block
1.6 D head
third pic is for comic relief.

I vote for even fire. You can re-balance later as needed.
S.

If all it is good for is 10% I wouldn't futz with it either. I have seen evidence it can be worth more.

What kinda damn fool would drag 4 dead pistons around in a V4 conversion??? No way would I do that, rings gone or not. Imagine the massive rushes of air past the pistons in both directions without rings! That takes ENERGY.

Yes, if that's what they tried it was a waste.

I'd guess it would be equal to the VW. As far as naturally balanced, the flat fours still have counterweights on the crank; but flat sixes don't, making them truly naturally balanced.

Totally agree.
A smaller four with super / turbocharger would have been a better choice in my opinion.

Pete.

Here's the deal...
 

So, Chevy makes a V-8, front-drive Impala (which I will admit is fast).

People want a HEMI Charger, but they don't want to pay for more fuel.

The answer?

Cylinder Deactivation. LOTS of cake, and eating it until you're sick. Good idea. Poor implementation.

I've driven each of these applications, hooked up the SG, and found minimal advantages: 5% at best. The base engine in both examples have plenty of power and achieve much better FE. The Impala 3.5L pushes 30 mpg with mostly highway driving.

There has to be a paradigm shift in engine size perception. A 4-banger with a turbo is a good solution, but much of America still sees V-8 as "I've made it". That's where the manufacturers are. I wish I could have manually cut the cylinders -- it would have helped in-town driving for sure, but how much? Dunno.

I wish the best to those testing this experiment -- just make sure to perform ample research.

RH77

would two smaller engines, in essentially one box, connected by a clutch, so one or both could run work better? almost half the engine fictional and pumping losses. both could be kept warm by one, ready if needed at the next mountain.

Good, good, good, good vibrations! (P.S. I really mean bad vibes)

-If you've ever had to synchronize multi-engined equipment, you know what I mean.

Somewhat off-topic
 

Does anyone remember the "Push-Pull" Chevy Citation X11 with 2 engines: front and rear ('85 I think)?

This is an interesting idea -- as AWD hybrids are generally driven at the rear wheels by electric motors along with the CVT/Gas-Engine up front.

I wonder how 2 small engines would fair in a real-world application... A small diesel on both ends: both work for acceleration from a standing start and one for cruise. Brainstorming over here...

RH77

I have considered such a beast before and decided that the engine(s) either need to be split from within a common block, or be close enough together to share coolant and oil so as to keep the "part-time" stuff hot for when it's called to action. Letting the part-time stuff go cold all the time would be bad.

Then there is the vibes problem, they really need to be indexed to each other somehow or very disagreeable vibes will result... I think.

do not grind off the cam. set the valves. i once fad a shop grind down the fact shims. waste of money, the valves were shot

just about everyone has Cylinder Deactivation. it works. long ago ford said they were working on one that only shut off fuel to 3 of 6 Cylinders i've been tring to find info on it. i think (maybe) you could ground out some of the injecters and Deactivate that way.

i have extra cams. #2 chamber has a coolant leak any way. so i'll just weld or epoxy that. i going to use 1 & 4 so even fire, and a nasty rocking couple.

John: I've disabled 1 & 4, and 2 & 3, and there was no noticeable difference in "rocking couple" vibes.

A related story. Not an FE project, I'm sure - but two engines, front and rear.

Pop Mechanics or Car and Driver or one of those... built a two engine car based on a fwd, 4-cyl compact. Probably n the '80's. I thought it was a Honda or maybe an Escort. Anyway, they added a duplicate engine in the rear, complete with stock fwd drivetrain. I think it had to be an automatic, since coordinating two linkages + clutches would be a project in itself. As I recall they had to move the gas tank and also did some suspension work so it would handle decently. Apparently it went like a bat out of heck.

No link between the two engines/drivetrains other than a throttle cable. And the tire/road interface.

There has been more than a couple of those projects. Moser made Cadillacs for sale IIRC.

cool

Here's an excerpt from a really interesting message I had from an older XFi owner. I invited him to participate, of course, but in case he doesn't show up, I don't think I'm breaching any privacy issues by omitting his name from this quote:

Frank Lee,
I just realized that the two injection pulses that I was going throw away are EXACTLY equal to amount of fuel that i would be burning. So if i put these two together and measure the volume, i will finally get around using two flow meters to measure total and subtract the spill volume. HOT DOG!

Cylinder "decommissioning"
 

John,I think your question requires context.If you are willing to settle for slower acceleration,reduced load-carrying capacity,greater engine and powertrain wear,etc.,then the smaller displacement,lower-power engine may be acceptable.The Eaton valve dis-enablers used on prior Diesels(don't know what Dodge is using today) did allow for full power on demand,and then for economy,during light load conditions.They were plagued with reliability issues and subsequently dropped from GM.Today,I think Chrysler claims only modest fuel gains for their technology.Only a life-cycle-cost analysis and crystal ball will determine the viability of the technology.

Quick question
 

how many hours are you willing to spend to avoid buying a cheap(?) used smaller engine and replacing it with your larger engine (it doesn't sound like anyone has proposed a 1.0l 2-cylinder ecotec). Last I heard, swapping GM-to-GM engines was pretty easy (maybe not not-really-GMs like the metro, but bear with me), and would be a better place to start (many examples on this thread mentioned GM products).

From the sound of it, chevys use of extremely tall gears with big V8s seem to work a lot better than cylinder deactivation. Maybe trying to use a "nerd gear" (term from a request for special green models on this site) combined with a low rpm cam.

aerohead,
let me put this project into context.
i am bored with 5% here and 5% there.
iam more interested in the outside limits. i have extra engines some of which have dead cylinders. So i am starting with a 1.6L diesel, 52 hp, 71Lb-ft. I am taking out #2 & #3 pistons and cam followers. Cutting off the big ends of the rods, taking out the clearance and clamping them soild on the crank. Then measuring the unused injection pulses from 2 and 3 for mpg info. see post#34
My limit of acceptable performance presently is 60 Lbs per Lb-ft of torque.
I arrived at that number like this 80,000 Lbs./1300 Lb-ft = 61 Lbs per Lb-ft
My car is 2100Lbs/60Lbs/Lb-ft= 35 Lb-ft of torque minimum.
The 0.8L diesel will have approx. 35 Lb-ft I am taking almost half the friction out of the engine. And I now have an aircleaner, exhaust, water pump, oil pump and radiator designed for 52Hp but i am only putting 26Hp through it. So i can run full rack for hours if i want to, without overheating.

Just don't climb hills, unless you don't mind going up at 15mph in first. ;)

on second thought i'll redo that calculation with ft-lbs of torque instead of hp.

whoops, looks like 0.8 Liters is as low as i can go, without losing weight.

i'll be able to run with a loaded truck. hopefully without turning up the fuel on the two remaining cylinders.

I don't think so, those trucks have way more gear ratios than you do. If by trucks you mean semis that is...