Selective Fuel Injector Cutoff

[lowbridescape]

New thought experiment of the day.

I have two cars that are seriously overpowered for efficient operation. Both are V-6s. 3.0L Escape and 3.8L Grand Caravan. I have thought for some time about the possibility of selectively shutting off some of the injectors (they're both multiport fuel injection) and letting the cylinders freewheel. Ages ago GM was working on something like that by controlling valve actuation and I think there are some production cars out there now that change number of operating cylinders. I don't know if they change valve actuation.

Here's my take on the subject:

1. Put disconnect switch on wire going to the switched injectors.

2. Would need to be able to switch between three and six cylinder operation on the fly. Three cylinders would idle badly and sometimes, you need that power (1% of the time?)

3. The ECU knows when an injector goes bad. So the injector wire needs to switch to dummy load to keep the ECU happy.

4. The Caravan at least only has an O2 sensor on the back bank. If the front cylinders were shut off (assuming the firing sequence is correct), the O2 sensor and mixture control would still be operational.

5. The freewheeling cylinders are going to be pumping cold air into the cat converter. How bad can that be?

6. Maybe put a diverter valve on exhaust pipe from front bank forward of the Y and when freewheeling, pump air overboard to avoid chilling the cat.

So how crazy is this? Ya' know what they say, "A mind is a terrible thing to loose."

[theholycow]

Common idea, commonly discussed here. Consensus is that it won't work like you describe. You have to close the valves for the dead cylinders. I think there are some people here experimenting with it anyway.

GM failed with it 30 years ago but they're having success with it now. They call it AFM (Active Fuel Management, I think). Other manufacturers are doing it too.

[bobc455]

5. The freewheeling cylinders are going to be pumping cold air into the cat converter. How bad can that be?
It will prevent the cat from ever coming up to temp.

Many people throw "universal" cats on their cars, not realizing that the are application-specific. Too small, and they can become a restriction- too big and they never heat up (and therefore never act as a catalyst) (and therefore become clogged quickly).

-BC

[05rolla]

One that is doing this now is Honda, with our New Honda Odysee. Whenever you are cruising along at a steady MPH it cuts down from 6 to 3 cylinders. It still doesn't get all that great MPG, but I suppose it is better than if it didn't do this.

jeff.

[hybriDatsun350]

GM, Dodge, and Honda are all doing it now. It's a cool technology, but it involves more than just shutting off injectors. It seems like if you drive them right you could get some impressive mileage, but as soon as you get into the throttle your back to full cylinder operation again. Well, that's true for the Chevy's and Dodge's I've driven. I don't have any experience with the new Honda's.

[lowbridescape]

Never one to give up a bad idea quickly...I reviewed some of the earlier discussions.

Why do the valves have to be closed? The pumping losses would be the same or less (dead cylinders pushing cold, low pressure air instead of hot, high pressure exhaust, throttle opened farther). But the thermodynamics should be improved (higher pressure/temperature in hot cylinders for same output power).

It did occur to me that it may require a kickdown solenoid or throttle by wire to keep the engine from surging when the extra cylinders kicked in.

I did look up the firing sequence today. The three forward cylinders can be turned off. That means the exhaust pipe coming off the front of the engine can be diverted to keep the cold air from the cat. Also, the O2 sensor is just downstream of the aft cylinders, so it stays hot.

So exhaust mods would include a diverter to send cold exhaust to either a parallel muffler or to bypass the cat to the original muffler and maybe a variable restricter to maintain pressure on the O2 sensor and cat. That shouldn't be hard, right?

2,000 rpm should sound like 1,000. That's not rough.

Idle may be a problem. But really, the engine shouldn't idle at all. That leads me to think about how to externally drive the oil pump in the auto tranny, but that should be a different discussion.

So.
1. shut off three injectors

2. kickup throttle setting to more or less equalize power

3. divert front exhaust away from cat.

4. close restrictor to maintain pressure in cat.

Heh?

[usedgeo]

I have done it both ways. Pumping air and closing valves. Of course these were fixed things that could not be switched. Pumping air is more drag than not pumping air but what you describe would work. I never thought about doing it this way on a V6. There are a lot of variables your computer will adjust that you can't take into account so I don't think the savings will be as great as you want but it would be fun to try.

[RoadWarrior]

I've been thinking of getting an "Exhaust Cutout" and sticking it in the pipe from the front bank... then just running it out somewhere... don't think a muffler would be needed on it, it's the combustion that makes the pressure waves and the noise.

I'd also like to get a good look at the dual throttle body setup in the LH cars (Vision, Concorde, Intrepid) since the motors in those are more similar to the 3.8, than the 3.0, you might find you can do a direct transplant of the linkages and intakes. Then arrange a solenoid cutout of the front bank throttle body, and loop the exhaust cutout into the intake that side and that should be a nice pumping loss reduction.

IMO pumping and frictional losses aren't all that bad below 3000 rpmish, 3000 rpm is to frictional losses as 60mph is to aero losses, due to a square law relationship, beyond those points they go up BIG, but aren't too terrible up to that point.

It's all about driving speed and load though for whether cylinder cut will save gas. If you can cruise at the point of minimum BSFC, then increase load, by cutting cylinders, BSFC will decrease more (to a point) and you save gas. But if you increase load too much on the 3 clyinders, or cruise at the wrong speed, you can pop out the other side of the minimum BSFC island and be using just as much as on 6.

[theholycow]

If you close the valves, you won't have to worry about diverting exhaust...

[RoadWarrior]

Just thinking actually, if you had one TB/intake per bank, connected the exhaust cutout to the active bank, closed that TB with a solenoid and had the inactive bank TB controlled by the throttle pedal... that might be an interesting thing to try... your inactive cylinders would be a supercharger for the active ones.

[lowbridescape]

I don't think going over the best FE point is an issue. The point of this exercise is to cruise efficiently at low power where the engine is so far behind the power curve, there is nowhere for the efficiency to go but up. I'm an engineer. Hypermiling violates the engineer's code of being lazy. So after improving the aerodynamics of the body, big changes have to made by improving the thermodynamics of the engine.

The goal is to modify what I've got. Closing the valves would be no small feat. If the engine was non-interference, the cam shaft could be clutched(?)...but that would still be seriously difficult since it would have to engage like a dog clutch to re-time the valves. I'm going to go with the free flow open valve approach for now and accept the pumping losses.

Remember that the pumping losses are always there. Whether it's six banging or three banging. The question is what the thermal efficiency is that's making up for those losses. On one of the previous threads on this subject, someone mentioned that they had a four banger that became a two banger and it ran like cr@p. The primary issue is not how well it runs, but if it runs like cr@p efficiently. A six banger should be more tolerant of running half dead than a four.

Diverted exhaust would still need a muffler or it would sound like an air compressor, which generally are not that quiet. It wouldn't need to be big since the air volume is much smaller than hot exhaust. It could even be routed back into the main pipe ahead of the muffler.

The dead cylinders would not act as a supercharger. The volume of the dead cylinder is the same as the volume the active cylinders. It would just move air through a more complex path to accomplish the same thing.

If the transmission shifts based on vacuum or throttle position, it would tend to undo some of the advantages of this system. How to lie to the tranny...?

[05rolla]

GM, Dodge, and Honda are all doing it now. It's a cool technology, but it involves more than just shutting off injectors. It seems like if you drive them right you could get some impressive mileage, but as soon as you get into the throttle your back to full cylinder operation again. Well, that's true for the Chevy's and Dodge's I've driven. I don't have any experience with the new Honda's.

I will be going on a road trip on 27JUN and will be travelling interstate much of the way... entire tanks of gas at a time I am sure. I can then let yall know if there is a significant diffrerence at least with the honda's when it comes to shutting down from 6 to 3 cyl. Not that this is the point of this thread but I thought I would throw it out there. Knowledge can't hurt...