Shade tree cylinder deactivation?

[roflwaffle]

I've seen this idea tossed around the web, supposedly w/ some success w/ V engines. I'd love to see it done w/ my V6 sedan, since it would allow me to cruise at a relatively slow speed where possible (~45-55mph) w/o being in an undesirable portion of my engine's BSFC map. The downside to doing it this to a vehicle not designed to close the cylinders, aside from the possibility of undesirable vibration, would be too much cooling of the deactivated bank from the continuous air at whatever the outside temperature was.

My question is, would jerrying a system similar to Toyota's TVIS in order to close the intake runners to the deactivated cylinders result in the bank staying hot enough from the exhaust to avoid significant wear, or is trapping the hot exhaust from the last power stroke charge in the cylinder needed for that?

 

[BrianPetersen]

Pumping losses with the engine completely unthrottled are less than with the intake throttled, but still not as low as leaving the valves shut, but there's another thing to think about in that situation - the catalytic converter. If you collect the exhaust from the cylinders involved in this de-activation together with the exhaust from the cylinders that are running, you will send the catalyst (and lambda sensor) a wacky lean and probably too-cold exhaust stream. If you keep the exhaust from the cylinders involved in the de-activation separate somehow, any catalyst would go wayyyy too cold after any length of time with the cylinders running de-activated, so you'll get a big shot of un-catalyzed emissions nasties when the cylinders start up again.

 

[roflwaffle]

So long as I cut-out the heated oxygen sensor that checks on how well the cat is working, and the sensor associated w/ the bank I cut, emissions shouldn't be too bad. The oxygen sensor in the active manifold would still be working fine and I'd be at ~14.7:1. In terms of EGTs, why would running just one bank drop 'em to the point where that cat would go cold? I mean, I'm still running roughly the same amount of fuel, less any efficiency increase I would see compared to at lower loads, so while one bank may see temps that are only slightly higher (compression) than ambient, the other would see a proportional increase in EGTs due to twice as much load, and when the two streams mix pre-cat the overall temp should be roughly the same, shouldn't it?

 

[BrianPetersen]

A three-way catalyst is not a three-way catalyst if it is being fed a "lean" exhaust stream, and if you allow the exhaust from the "running" cylinders to mix with that from the "de-activated" cylinders, that is exactly what will happen. For the three-way catalyst to function, it needs to see the exhaust stream from stoichiometric combustion - not a lean stream.

You can DO that, if you segregate the exhaust systems from each bank completely with their own lambda sensor and catalyst. But then, the trouble will be that if the engine runs for any length of time, the catalyst will cool off below its light-up temperature, and when the bank switches on, it'll let out a spike of nasties until the catalyst re-lights.

Keep in mind that on a V8 engine that has a normal 90 degree crankshaft, if you want an even firing order, you have to shut off the two inner cylinders on one bank and the two outer cylinders on the other bank (with Chevrolet V8 firing pattern, may differ with others). But on vehicles with dual exhaust, that ain't the way the exhaust systems are divided up.

I think we are exploring the reasons why the OEM's do it the way they do it ... Shut off the valves completely, with no flow at all through the dead cylinders, and you simply don't have any of these issues.

 

[patprimmer]

Um DING DING

Did anyone listen and think

All comments from anyone who has real knowledge on this subject says.

Just cutting fuel wont work as the engine will drop power and you need to add extra fuel and air to the other cylinders. Also you really mess up the emission controls and don't save ANY fuel at all and probably use more.

To save fuel you need to improve engine efficiency by converting a greater portion of the heat generated from the fuel burnt into power at the crank or by reducing friction or mechanical losses or by reducing pumping losses.

That requires deactivating the valves and fuel flow on the deactivated cylinders.

The extra throttle opening to maintain power and the reduce surface area (but at a greater temperature difference) will give a slight improvement, I would say on a magnitude of about a sparrows fart in a gale.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[roflwaffle]

I see what you're referring to Brain, but then the problem isn't catalyst light off, since most of the heat via the exhaust stream will be the there, the problem is that in a lean mixture HC/CO reactions would be heavily favored, in other words NOx emissions would increase running on just three cylinders w/ the exhaust stream from the other deactivated cylinders, until of course the catalyst saw stoich again, at which point it would operate in a balanced manner again.

Pat, as I have explained before, and illustrated via the BSFC map for my engine, only using three cylinders will double load and decrease the pumping losses associated with throttled operation below ~half load. Ideally I would like to just drop in off the shelf gearing to do the same, but unfortunately that isn't available. If as you stated, this is roughly comparable to a sparrow's fart, then IMO, going from ~35mpg@50mph to ~50mpg@50mph, not counting the reduction in idle fuel consumption at the expense of a rougher idle, is more than enough gas for me.

 

[patprimmer]

Dream on

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Tmoose]

"pumping loss"
Owners of vans and big rwd cars report something like 10-12 inches of mercury vacuum cruising on the highway.
To me this means that lovely 9-1 compression ratio is more like 9*12/29 or something like 4-1 in the cylinder. That represents efficiency of about 40% instead of the 55% or so at 9-1. A solid 25% decrease on my thermodynamic ROI.

The extra 15 or 20 degrees or so of ignition advance that get tossed in under light load is not confirmation of the monster's mantra that "mmmmm.... advance good." Rather it is necessary because the lower density mixture leaves the burnable bits so far apart we have to heave our torches at noon time to ensure the castle is burned down by suppertime.

The "negative torque" apparent in pv diagrams during the intake stroke indeed looks to be made worse by closed throttle. But when I look at the compression stroke, the lower cylinder pressure would seem to cancel some of that loss.

 

[roflwaffle]

That's good to know Tmoose, and makes a lot of sense compared to results I've seen people report. Pat, I'm not adverse to your statement that it won't work, however Brian has been explicit about throttling losses, for instance how just blocking the intake wouldn't work, and according to the BSFC map there are significant gains to be had from minimizing those losses. Granted, there may be greater gains from shutting a bank off completely, but going from the ~600-400g/kWh regions to the ~300-260g/kWh regions represents significant improvements as well.

 

[BrianPetersen]

More random thoughts here.

The best way to get closer to the favorable BSFC region during normal driving is to NOT OVERSIZE THE ENGINE. Don't undersize it either - right-size it. Most auto companies in North America have been selling people oversized engines, and the media hype this up by complaining about anything taking longer than 8 sec to get to 60 mph as being "slow" (meanwhile, I can keep up with traffic on my 12 hp 125cc motorcycle everywhere except steep uphills and in strong headwinds!), and people have been buying it all.

If you right-size it, you don't have to de-activate cylinders. There won't be significant operation in load/speed regimes where it would be an advantage. In any case, in most right-sized cases, this is going to be an engine with 4 cylinders or less, and you really don't want to be de-activating cylinders on that.

Idle conditions bear thought. Best thing is to just stop the engine completely. I hope engine stop/start systems become more common.

Hybridizing allows better performance with a right-sized engine, but it's not the only solution.

Turbocharging systems can be designed for economy applications if they're designed for it (most up to now have just been designed for power) by allowing the engine to be downsized. Very good intercooling is important to allow the compression ratio to be maintained. The turbo controls have to be designed to minimize exhaust back-pressure when running at part load (most turbo control systems don't do this).

You can probably gather that I'm not entirely a fan of cylinder shutoff systems. I think they're a way to try to get a last gasp out of today's oversized engines.

 

[patprimmer]

Some of my opinions in this are based on a friends recent experience. It is anecdotal and is not back to back, but it does support my thoughts on the subject.

He had a 2005 model Nissan Maxima V6. He was very happy.

H traded it on a 2008 Honda Accord V6 with cylinder deactivation expecting a significant improvement in economy.

In fact he got very similar performance and economy from both cars. They are similar weight, similar size and aero shape at least to the eye and similar engine size and layout. The guy is meticulous in observation and record keeping

The cylinder deactivation in this not very scientific test seems to have yielded no gains.

I currently have no data foe gear ratios and "normal" change points for either car, but I expect this guy would notice and comment if the rpm t change point as significantly different.

My daughter has a 4 cylinder Honda Accord. It's performance is more than adequate for daily driving and it gives very good MPG for a car it's size, but the whole car is smaller and lighter than the V6 version.

My conclusion from that collection of anecdotal and not back to back data is that the cylinder deactivation is a marketing exercise because it sounds intuitively like it should reduce fuel consumption, that is until you think a bit deeper

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[thundair]

After reading this thread the only cylinder deactivation I can think of is to;
Use 3 BMW boxer twins with planetary gears in between and as you accelerate release the band on the ring for the front engine with a shut down, and between the second engine with a shut down, and between the third engine and final drive.
Wouldn't that be a 3 speed?

Cheers

I don't know anything but the people that do.

 

[izzmus]

"Most auto companies in North America have been selling people oversized engines, and the media hype this up by complaining about anything taking longer than 8 sec to get to 60 mph as being "slow" (meanwhile, I can keep up with traffic on my 12 hp 125cc motorcycle everywhere except steep uphills and in strong headwinds!)"

Based on observation, I am of the mind that if GM sold trucks with 2 liter engines, and recalibrated the electric throttles to give full throttle at 15-20% pedal travel, most people wouldn't notice the difference.

Certainly I have no trouble keeping up with traffic with a fully-overloaded telco truck, which typically are the heaviest duty trucks available with a six cylinder. Sometimes one-ton, sometimes only a 3/4-ton, but the GVWR is just a number and not actually taken seriously.

 

[patprimmer]

In the late 80s, early 90s they did that trick with te throttle of their 90 deg V6. The throttle opened substantial faster off idle than through he second half of the movement, so it felt like it had so much power you just needed o touch the throttle, well at least they did on the Aussie Holden Commodore model.

Also GM or their associates do produce and successfully market 2 litre 4 cylinder 1 Tonne utilities or light trucks in most markets outside North America. This tells me it is not GM ideology, but their perception of different markets that influences their local model range.



Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[DanEE]

from Pat's post......
"My conclusion from that collection of anecdotal and not back to back data is that the cylinder deactivation is a marketing exercise because it sounds intuitively like it should reduce fuel consumption, that is until you think a bit deeper"

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

I've thought the same thing...

Has anybody found or know of any published engineering/research info quantifying gains from cylinder deactivation (which to me is old tech) as opposed to other methods such as Variable Valve Timing (with intake and exhaust independently variable from each other)?

I would think that VVT can not only widen the power band, but I believe can improve economy do a better job of minimizing pumping losses at light throttle at the same time.

I believe BMW has a throttle-less (no butterfly throttle) engine under development that uses variable lift intake valves as the mechanism to modulate injected fuel/air quantity (power levels). Seems like this could be the ultimate method off transitioning from no throttle (closed valve, efficient air spring mode) to part throttle to full throttle overall efficiency.

 

[patprimmer]

The BMW system simply uses the valves as a throttle.

I expect a good VVT can get more power and a wider power band out of a smaller engine and the option of accurately controlling valve timing and lift to optimise fuel efficiency at part throttle and a controlled compromise between power and economy at WOT.

I also expect the smaller engine can have cascading effects on chassis and drive train weight further enhancing economy

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[GregLocock]

There's quite a few SAE papers

http://www.sae.org/technical/papers/2007-01-1292

looks directly relevant.

Heywood suggests that the pumping work in an engine could be as much as 40 kPa, so halving that would be roughly the same as eliminating ring friction. There's an additional benefit from eliminating some of the throttle 'work' as well.





Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GregLocock]

http://www.sae.org/technical/papers/2001-01-3591

as well

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GregLocock]

The first paper gives an /estimated/ improvement of 5.5-7.5% on city cycle. It is rather long on mechanical implementation, rather short on the actual performance beenfits. The second paper is more (over?) optimistic, more theoretical, and more informative.

I built a crude model that suggests the reduction in fmep is around 20-40 kPA at best. There are some thermodynamic effects on top of this, positive ones in that the firing cylinders are probably more efficient, negative in that the 'gas spring' cylinders are not adiabatic.

There is also a reduction in the mechanical power to drive the valvetrain, I have not put this in yet. That is quite significant.

Incidentally I was wrong above there is no particular change in throttle work for a given torque output, at a first order level. Similarly you are pumping twice as much air through half as many valves - does that really sound like a recipe for success?



Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[cibachrome]

My 2008 Chevy Tahoe has the Eaton Active Fuel Management system (internally called "Displacement on Demand"). After 24k miles, and using the ecm provided instant fuel economy, I can see the change from 18 to 26 mpg on a flat road at 55 mph very day, both directions. This includes trailer towing(boat and hay delivery). The ratio is the same when using 85% ethanol. Best not to use cruise control because the power demand on hills cancels some of the gains. The name of the game is to keep the V4 symbol from jumping to V8. The sound and vibration increase is readily evident. GM managment wanted a 'transparent' system that would minimize the vibration. Some powertrains still need development on this because of oxygen migration back to the exhaust valves.

I wish there was a switch to select fuel economy over smoothness. Even at 80mph out on I96, the economy shift is pretty interestin to play with. A constant throttle position strategy works best for me with this system. My V6 company car uses the same system and the percents are nearly identical. Truck drafting is also pretty interesting to experiment with...

 

[pontiacjack]

Wow- 44% increase in fuel economy?!