Cylinder Deactivation......

[GasSavers_RoadWarrior]

I understand that for best efficiency, valve closing is desireable, but in looking into this for a while, I am wondering if it is actually 100% necessary for some savings, or more of a somewhat historical convenience dating back to manufacturers DOD efforts from the late 70s and early 80s... just the way they have done it...

In the beginning... manufacturers were experimenting with this on carburated engines... it's real hard to stop fuel getting into a cylinder from a carb without closing the valve. TBI motors have the same problems. Then when MPFI came along, most of the experimentation with valve closing had been done, it was some piece of technology they had sitting on the shelf as it were, too expensive to implement, and too complex to mechanically control satisfactorily within customers NVH expectations. As ECUs got more sophisticated, it became more of a practical proposition to put it under electronic control.

Bear in mind that while even fairly crude DOD schemes were quite effective at a steady state cruise, none of them have performed very well on the EPA highway cycle test cycle, due to very little of it being steady cruising. Ergo the cost/benefit numbers didn't appeal to the bean counters.

Another thing to remember is that Detroit knows of a far better Displacement on Demand scheme, it's been trying to get customers to accept it fully since the 80s, it's where you start out with a small motor, and give it the ability to burn as much charge and hence produce as much power as a large motor, it's called turbocharging. So if you've got a V8 windsor in a ford truck, you can do it this way, get the 4cyl mustang turbo motor in there instead.

Mainly with an MPFI car, the real obstacle to getting DOD working acceptably enough for an MPGgeek is the vacuum and fuel getting out of shape due to airflow metering, whether by MAF or MAP being wrong, or the O2 reading lean and making the ECU dump fuel. These problems might be worked around in a way that a manufacturer would consider a poor production solution when they have the ability to make mechanical and other changes for a valve closing scheme, which would avoid these issues.

Manufacturers also have to plan for "worst case scenario" type situations, where for instance the evap canister dumps fumes into the manifold at the same time the PCV is blowing in a lot of fumes and oil mist... if allowed into a cylinder not receiving fuel this could be enough to detonate lean, possibly causing damage, also minor injector leakage could have this effect. Now if only 10% of their customers mistreat their engines enough for this event to happen during the warranty period, this would be a heck of a lot of engines they would have to replace.

Anyway, I'm just trying to point out that there is a lots of room to play underneath the holy plateau of what is a perfectly engineered, marketable and profitable consumer product.

However... DOD isn't for all engines... yes some 4 cylinders may be a little too weak for it to be worth the effort and have inherent balance problems before you start deactivating anything. If the vehicle isn't a little overpowered, then the motor will have way too much work to do on 2 cyl , and you'll have the gas to the floor to keep it moving. I would regard DOD as a final stage modification on many vehicles rather than something to do to a stock 4cyl economy vehicle. First the vehicle needs to be easily driven at highway speeds, work on the aero, next the engine needs to be efficient and have some power in hand, properly sized and wrapped headers, crank scrapers, porting, are some of the modifications that might be desireable, 80% of performance modifications are efficiency modifications. So now if the motor is making about 130 or 140 HP where the base engine made 90, and is doing so efficiently, we might be able to think about screwing with DOD on it. The reason most people think performance mods kill MPG is because they can't keep their foot off the pedal.

Bear in mind that economy oriented 5th gear ratios will likely not be friendly to DOD, if the 4cyl motor is turning 1500 on the highway, it's close to lugging as it is, autos, or manual transmissions without available ratio swaps, might fare better when they're turning 2500-3500 RPM on the highway. In other words, if you have a car that's only acceptably powered with a 4cyl, if there's a way to get your gear ratios more favourable for highway cruising, do it instead, it's far simpler.

I have DOD plans for my minivan, I am inspired in this by knowing of a guy with the same van who has got 40mpg doing it. More on that another time.

Road Warrior

[GasSavers_RoadWarrior]

In looking into how O2 sensors work, I've just realised that the situation with crude DoD by injector deactivation and the O2 sensor reading lean may not be as bad as it seems like. The pressure in the exhaust system where the O2 sensor is, will be lower when there are cylinders not firing, even though they are pumping air through. Therefore the partial pressure of the O2 in the exhaust is lower than if it had all cylinders firing lean and putting that much O2 in. This makes the O2 in the exhaust stream influence the sensor cell less, so it doesn't read as lean as it is... from reports of folks who have tried this, it appears to be just enough such that the O2 isn't faulted as dead by the ECU, but it's likely that it does cause some enrichment, though one might say this is necessary for making enough extra power for the live cylinders to haul the vehicle along...

[spacepilot]

Originally Posted by RoadWarrior

The pressure in the exhaust system where the O2 sensor is, will be lower when there are cylinders not firing, even though they are pumping air through.

That's probably why proper variable displacement is not easy to DIY because it requires the intake and exhaust valves to be kept closed.