Cylinder deactivation attempt
 

Ok, well, not exactly.

I bought this '91 Geo Storm with a bad rod knock (assuming it's a rod) that you can hear while cranking the engine, it's so bad.

I've got a few options that I'm considering playing with, since the car is just a toy to me, really. I had thought originally that I was going to put a 3.0 V6 turbo into it, and make it a "fun" car, but I decided otherwise, for now at least.

So the current thought is that in a couple weeks, when I have time to actually go play with it, I'm going to go pull the head off and pull the pan off, and see how bad the damage is.

I don't want to rebuild the engine, and I'm not going to replace it with another stock engine, either, but I was thinking:

Maybe I remove two cylinders, and run it like a parallel twin? Every 360*, I get a power stroke, while the pistons shake the hell out of the car because they're always either up or down, and the counterweights on the missing pistons add to that shaking, because they're in the same locations as the remaining pistons at TDC and BDC... this poses a concern for the engine literally shaking itself apart.

I have two options, though. I can remove #1 and #4, and leave #2 and #3, giving me a power stroke every 360*, but leaving the pistons in sync with each other at all times through the rotation of the crank.

I could also remove #1 and #2, leaving #3 and #4, which are 180* out of phase with each other, canceling out the huge power pulses and associated vibration, but creating it's own funky vibration by having offset power strokes, where the second power stroke occurs at 180* from the first, leaving a full 640* of slack rotation between them, where no power is produced.

There is a less likely option to run #2, #3, and #4 in their original firing order (3,4,2), which provides a power stroke at 180*, 360*, and 640*, but then leaves the remaining 360* of rotation as a slack.

EDIT - This is not really an option, because of the strange balance that would come of it. Maybe if I get bored one day, I'll try it, but I find it highly unlikely that I'll ever bother.

Whatever I do, whichever piston(s) I remove, I will also remove the valves and associated hardware, and leave the cylinder an open bore with no plug.

EDIT - Removing the valves will leave the intake/exhaust open to oil spray, so the valves should/will be left in, and the rockers removed to prevent them from being in contact with the cam. (Thank you, Frank.)


Anyone have any questions, comments, etc?

Looking for some feedback on this, guys.

That is EXACTLY what I was considering to do to a Tempo!

I think it would work. Disable 1 & 4, or 2 & 3. Did you know pistons in parallel twins go up and down in unison? Yeah, so when you block off the crank journals on the dead cylinders, do it so that your block-off is heavy enough to offset some of the counterweight. I was considering cutting the rods off at the big ends but hadn't come to a solid conclusion about if they should be free to rotate or if they should be affixed to the crank somehow. I think clearances to the block would answer that question.

Ignition: I figured on disabling two plugs by hammering the ground onto the center and welding it. Or making a place to ground the two wires externally.

Lifters: There needs to be something in the lifter bores for the oil gallery to function on my OHV. I thought I'd just pull the pushrods and let the lifters float. If that proved to be unacceptable for some unforseen reason I'd affix them in a neutral (no cam contact) position.

EFI: I figured on just pulling the wires off the injectors for the two dead cyls.

Exhaust/O2: Do nothing. Should be fine.

Here's something that will help some of you understand what it is I'm asking, exactly.

http://static.howstuffworks.com/gif/engine-inline-4.gif

This is how a traditional Otto-cycle 4 cylinder works, the firing order of this engine would be 1,3,4,2.

If you want to properly envision what would be going on with each scenario I posted, just stick your finger over the missing piston from my description and watch the rest of the engine. Notice where the #1 and #4 crank weights are in relation to the #2 and #3 cylinders.

I do think I might be able to alleviate some of the problem with the crank weights by either weighting the journals with something similar to the weight of the piston/rod assembly, or by removing weight from the counter weights on those throws.

Frank -

Ignition - remove two unused wires (what about ignition advance, though? It won't be optimal with only two cylinders operating.)

EFI - I want to find a way to "spoof" the injector return signal so that the ECM thinks that they're still there and functioning, so that the CEL doesn't turn on.

This engine is SOHC, so I planned on just removing the valves/rockers from the engine for those cylinders, leaving the cam spin freely (which could also create some unbalance).

You've actually made me think when talking about oiling, I will probably just weight the crank, because there needs to be something preventing oil from just flowing out of the rod journals! Maybe I cut the rods off and just add some weight to them?

I think you will want to leave the valves in; you don't want any flow through there.

Re: Ign: Dunno about your engine but I don't think mine would notice two are offline. If there is an ignition indexed in some way to #1, then I'd have to disable 2 and 3, for the easy way out.

Re: Injectors: Perhaps short them out, or measure an injector and stick the appropriate resistor across there.

Ah, yes, I hadn't thought about that... removing the valves also leaves intake/exhaust open to oil spray! Good catch, Frank.

I will remove the rockers for the dead cylinders, though, so the cam doesn't move the valves at all.

That, and the exhaust and intake charges are free to migrate across log and even tube manifolds.

I'd go for shorting out the two unused high tension ign wires because coils don't like conflicting signals. If you don't short out the high tension side, then remove the triggers for the induction side.

On mine I was planning to make it 100% reversible, as I would be starting with a good engine and on the chance this is a failure, would want to restore 4 cyl operation. You don't have that problem though.

Ah, yes, charging the block with exhaust gasses might create some odd noises, eh?

Might play games with the PCV too.

Meh. :P

What do you think about the 180* Parallel twin?

Supposedly, compared to the 360* PT of the same specs, the 180 has a flatter torque curve, which might be something to think about for a 2,000lb car that will now have less than 50HP. (The OE config only has ~90, and it's got 100,000+ miles, so probably closer to 70HP... divided by 2 and then some.)

Thoughts?

Maybe when you tear into it, the damage will tell you which cyls should stay and which should go.

As far as I can tell, it's only #1 that's damaged. The three cylinder thing is a long shot, I'll have to be really bored to want to do that.

When I really get into it, though, you're right, it'll probably lead me into a decision, rather than having to think about it.

I do sorta wish I wasn't going to spend any money doing this, but I guess a head gasket is cheap enough...

Also -

Once this thing is running (if it gets running at some point, satisfactorily), I will probably have to completely block the grille and remove the radiator, replacing it with a smaller core inserted into a forced air duct.

I say this because I'll have half an aluminum engine making heat, and the other half acting like a heat sink... it'll be hard enough to get it to warm up, so I don't imagine that I'll have any cooling issues.

Do NOT attempt a 3 cyl!!! :eek:

The pulses are not even like a Metro triple; the exhaust note alone will make you want to blow it up within a few miles. I know.

Keep a rad, it'll need it. The "heat sink" will cause a longer warm-up but from then on it needs to be able to reject heat faster than a block will.

I lol'd a little.

If you just look at 2 and 3 in your animation, that is what the Honda CB450 I have does. I could only wish for the auto 2 banger to run that nice!

This was before your time here...http://ecomodder.com/forum/showthrea...ssion-968.html

I was trying to get a CB450 Honda-Matic awhile ago, but the guy wouldn't budge on the price, which was about $1000 more than I wanted to pay (he wanted $1750... not even close to reasonable), so I passed on it.

You ever flattened out a wheel weight? You can get them pretty flat, and that might be enough to insert it between the rod journal and bearing on the big end of the rod, to keep it from rotating on the bearing surface at all... squeeze fit, once you tighten/torque the nuts.

You gonna finally do this if I start working on it, Frank?

I wouldn't trust that flattened weight. The forces are huge.

I'm not even going to be near my home for a few months yet. Even then, doing it to one of my current cars kind of interferes with my long-time, hard-earned mantra: If It Ain't Broke, Don't Fix It. All the Tempos are running great at this time.

If I had an engine like yours, with what's likely one problem cylinder, I'd be all over it!

I think you should try to do it without taking the heads off or it will never get done :)

so:
1. pop off the oilpan.
2. remove the big end caps on the two cylinders you want to deactivate.
3. pull the sparkplugs on those cylinders
4. take a washer and a long drywall screw and drill it into the piston through the sparkplug hole to "secure" it at the top of its travel.
5. sawzall off the bottom end of those con rods so the crank doesn't hit it (perhaps do this earlier so you don't stress out your drywall screw).
6. remove lifters on those cylinders
7. enjoy 2 cylinder goodness :)

I was thinking about something like this last night, except cutting the conrod right below the wrist pin, then topping out the piston in the bore and fastening it in there. Alternatively, if I wanted to break the piston, I could take it out the bottom, but the wrist pin would fight me every step of the way.

I don't want to be putting screws and such into things, though, especially when that might introduce metal shavings to the oil.

Have you considered re-using the head gasket :) Or is my neck to red for you! A propane torch and some gental heat will help anneal the metal ring surounding the cylinder and it "may" seal up again, at least only two of the four would have to seal, so your odds just doubled! I've never tried it, but have contemplated it when the funds were low. I have done it with copper head gaskets and it worked every time.
On the issue of counterbalance, consult with an engine shop that does balancing of rotating assemblies to determine how much wieght should be attached to the the crank in the position of the missing cylinders. I'd just use the old con rod and weld an attachent on that would be ok to float, ie. cylindrical.

I've never had good luck trying to reuse head gaskets, especially if they had any kind of miles on them. First of all, you'd have to get it off cleanly to reuse it, which probably ain't gonna happen with over 100,000 on the engine.

I had considered using a sheet of copper to make a gasket from the old template, though I'm not sure about the combustion rings (would copper hold up, or require plating?)

Remember, it still has to seal oil returns and water jackets that are around the dead cylinders.

The SOHC Geo Storm fuel injectors run 12.7 ohms resistance, so I'd imagine that if you put some 12.7 ohm resistors in the wiring harness on the dead cylinders you could trick the ECM into thinking that all 4 injectors are working.

Copper gaskets should hold up fine a while back there was a guy who made a few for these cars and he's been running one for a couple of years without any trouble.

I'll try those resistors.

I use copper gaskets on race engines, but I have no need to pull the head in this case. I don't intend for this engine to ever be reused, so I'm going to draw the pistons down, cut the rods, then shove 'em back up in there and see about fastening them in place.

Not having to pull the head will be nice, It's never a fun job. I think and hope this plan will work because the Storm gets decent mileage as is, and with two less cylinders I would be interested to hear the results.
The one hurdle that may be hard to overcome is the MAP sensor because, and my physics are a little rusty here; the manifold pressure may be out of the sensors operating range bringing you back to limp mode. I'm not sure if this would actually occur, but it was just my first thought.

If that occurs, and it really shouldn't, as far as I can tell, I'll have to pull the intake and plug the two empty runners with expanding foam or something to make up the volume difference. Since the MAP is only detecting vacuum, it should still be in range, though.

Claimed fe benefit for current deac is up to 20%. Since this would cut down internal friction and whatnot my WAG places potential improvements at 10 to 25%??? 30? It'll be something but it won't be like WOW my fe doubled.

This, and the "no 100%" throttle mod that will accompany it, I'm hoping will yield about 20%. 25% is probably pushing it.

The throttle mod is going to have to be there, though, because we both know I'll be putting the go-pedal through the floor the way this thing is going to run.

If it runs, but doesn't quite make enough power to hold 45 through my paths, I'll see about putting a small turbo on it or finding another way to increase the power output slightly. Or, I'll eventually scrap the idea and put a diff engine in it. It needs body work, so I don't mind playing with it for now.

I don't really think it will happen either, just a thought , but if it does than plugging the runners would certainly do the trick.

Even a 10% increase would be pretty nice for a pretty cheap although labor intensive process; especially on a car that already gets pretty good mileage and you get to keep one more car out of a junkyard.

Unfortunately, if it's only 10%, I'm going to have to call it a fail. 10% is a gain that you'd see from doing the basic stuff that we all do when we first get our cars. It's a little more difficult on a car that's already somewhat efficient, but still within reach.

I'm going to set an upper limit of 40%. If I see more than that, I'm really going to have to analyze it to find out why. Unfortunately, I have no baseline to go from, either, other than EPA estimates and whatever real world data I can glean from other owners. There are no other Storms on this site, that I know of. (Of the same body/config)

Only 10% would be a definite fail. It's going to be soooo slooooooowwwwww, soon you'll be dreaming of more power.

I first read about a project like this when I started collecting Mother Earth News back in the 90's. The magazine I was reading it in was considerably older, and I can't remember exactly how old. They did pretty much everything you guys have mentioned, with the exception of EFI considerations. I think they switched down to a smaller carb to cope with the lower airflow. I also can't remember what kind of results they got, more's the pity, but I think what they got was stronger than what you guys are talking about, closer to 40%. But Mother was likely to shout exaggerations quite a bit back then, and the stock mileage was atrocious.

I don't think they cut conrods though. The article mentioned being able to undo the mods and have the vehicle in its original configuration with one quick afternoon's work.

I've seen a couple of projects that used a VW engine shortened by two cylinders to make an opposed twin. Performance goes out the window. Thinking of that though, does anybody know if anyone's tried the same trick with a Subaru EJ? If you could have even just 30% of the power, that's still a solid 50hp in a compact powerplant - not bad for a flyweight commuter.

El High -

There are 2cyl boxer conversions on Youtube, but I normally don't save them when I stumble on them.

I've seen a few different iterations of the 2cylinder idea, too, one of which involved heated water injection into the 2 non-functional (still moving) cylinders where exhaust heat would turn it into steam, causing expansion.

The only reason I'm cutting rods is so I don't have to cut/balance the crank. I'm going to cut the big-ends of the rods and just put some sealant in the bearings and over tighten them a bit so they don't swing around. I'm gonna make an attempt to keep both cut rods the same weight, so that there isn't as much variation in the fly weight balance.

I wouldn't expect the water injection to work. I don't think there'd be enough heat leaking into the two idling cylinders for a decent steam expansion. I say this with no experience, just my gut. My guts aren't that smart, so I may be full of baloney at this point.

I would like to see more from Crower on his six-cycle process. A couple of intriguing news articles, followed by a long and disheartening silence. He suggested his steam cycle could take the place of a radiator in the engine's cooling regime. Still need one for a condenser, but anyway.

The "idle" cylinders had exhaust gasses pumped from the exhaust of the running cylinders into the intakes of those cylinders, along with water injection on the "combustion" stroke of those cylinders. The water was also pre-heated with coils on the exhaust piping to take advantage of any left-over heat in the exhaust stream.

Basically, imagine it as two 2cylinder engines.

Run your first engine normally, except pipe the exhaust into the second engine's intake manifold, then run that engine, except instead of spark plugs use pipe connections with differential valves that spray water into the cylinder. Preheated water sprayed in, exhaust heat warms it up enough to turn to steam, it expands, then exhaust stroke happens on the second motor. Make them both run at the same speed. It's just another way to extract heat from exhaust gasses and re-use it as mechanical work.

well, I'm thinking ust for simplicities sake, drop the pan and cut the crank with a cutoff wheel just outside the middle bearing so you have three bearings for the two cylinders near the flywheel.

Then pull the rockers on the disconnected cylinders
hook the alternator up to a wheel/inboard axle shaft via belt/something
put an electric motor on the water pump
aeromod the heck out of it
figure out a lockup switch if appropriate
and see if you can't make something that can cruise at hiway speed at a real decent mpg.

replace alternator with altermotor and install a trolling battery to help a tiny bit with getting up to speed, after initial hiway tests.

There's nothing simple about cutting the crank, my friend.

How would I then drive the cam?

The other issue is that I still don't know what a damage inspection is going to reveal, and that using just the #3 and #4 cylinders would give me a two combustion events 180* apart with 640* of nothing.

I already intend to do some aeromods to see if I can get it down to .22-.20 or so. I'll settle for .25, if I have to without going too extreme.

In theory, the engine should have at least 20 HP. We already know that a heavier Escort wagon can go 40 and get 40MPG with a 6HP Robin single cylinder (Thanks Frank.), so I should be pretty good to go. Slowly, but still able to go.

ah, yah, cam :) There is probably some oiling issues there too.

Yup. It's a good thing I don't just run & cut, eh? ;)