BMW Valvetronic System

[pikeslime]

Is the BMW valvetronic system, or any other variable valve train system, truly THROTTLE-LESS? They promote it as such.

 

[ivymike]

I assume you mean to restrict the question to gasoline engines that do not utilize direct injection?

It's my understanding that the BMW system is truly throttleless. I haven't taken one apart personally, but I don't see any reason to disbelieve the published material.

 

[aorangi]

Yes, and the BMW Valvetronic are the first production gasoline engines ever to be throttleless. You can find plenty info and sketches about it on the web. Even direct injection gasoline engines need a throttle.

 

[ivymike]

Even direct injection gasoline engines need a throttle

I thought that one of the main advantages of going to direct injection was the elimination of the throttle?

 

[ivymike]

some articles that I found on the subject:

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

http://www3.bc.sympatico.ca/Volvo_Books/specart10.html

"Mitsubishi 1.8 litre Gasoline Direct Injection (GDI) engine ... A direct injection, stratified charge engine need not have a power-robbing thottle plate. "

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

http://www.carenthusiast.com/20020901_roadtest_mitsubishishogunpinin.html

" In essence, the idea is to control the work output by the amount of fuel injected (not by adjusting the throttle, which reduces the engine's efficiency)."

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

http://www.fev-et.com/03eng/01co/e_co_ot_df.html

"In comparison with a well optimized, conventional PFI engine, there is a potential benefit of up to 25% in fuel consumption due to the throttle free and stratified operation."

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

The middle article seems to be a review of a MY2002 throttle-less GDI-powered SUV... is that not correct?

 

[ivymike]

More on the Mitsu GDI -

Several "tech" sites state emphatically that the Mitsu GDI doesn't have a throttle, yet this Mitsu site

http://www.mitsubishi-motors.co.jp/inter/NEWS/9704-09/9709.html

says that at least one of their GDI engines has an electronically controlled throttle valve.

 

[GregLocock]

Well, I'd guess that DI makes it easier to use stratified charge, and hence throttleless operation, but perhaps to get the last bit of part throttle performance the friction loss in a throttle plate is an acceptable compromise sometimes?

Cheers

Greg Locock

 

[evelrod]

YEAH, Isaac!!! What he said.


Rod

 

[aorangi]

Toyota and Renault have abandoned the stratified charge operation of their direct injection engines, so they are always throttled except at full load.

On the 156 JTS (Jet Thrust Stoichiometric) Alfa-Romeo uses the stratified operation up to 1500 rpm only and most of the others up to about 3000 rpm under low load. In the stratified mode, the throttle is wide open, but still a throttle is needed in the homogenous mode at higher loads and revs.

The stratified charge operation brings a load of problems such as NOx emissions control and top ring durability, due to higher combustion temperatures and increased heat transfer to the piston because of the larger surface of its specifically shaped head – at least in the wall control system. To full benefit of a stratified charge engine, sulfur free gas is needed.

On the opposite, control of the engine output by infinitely variable valve lift and duration seems to me a more logical and better approach since it allows optimum cylinders filling under every condition, without the drawbacks of the stratified charge mode.

Cheers
Aorangi

 

[ivymike]

Another point is that there's no power wasted to needlessly lift the valves too high

power wasted to needlessly lift the valves? you mean valvetrain friction?

 

[aorangi]

Yes, I mean that and also that you don't get back the full force of the spring on the closing of the valve since some of its force is used to counterbalance the inertia of the valve and rockers. On the other hand, the valvesprings must be stronger because of the increased inertia of the dual levers system.

 

[GregLocock]

When I worked for an organisation that really cared about fuel consumption we measured the power absorbed by the valvetrain. Since this was a long time ago I can't recall the exact number, but I think it was around 2 hp for a 2 litre engine at 3000 rpm. Doesn't sound much, until you realise that at 50 mph cruise we were only using 18 hp to push the car.

Cheers

Greg Locock

 

[ivymike]

Yes, I mean that and also that you don't get back the full force of the spring on the closing of the valve since some of its force is used to counterbalance the inertia of the valve and rockers

So a significant amount of energy is lost in the system via something other than friction? Where does it go?

can't recall the exact number, but I think it was around 2 hp

2hp sounds a bit high to me, but I'll have to look at some numbers tomorrow before I say anything definite. Even at it's max, valvetrain friction is pretty small compared to some of the other sources of friction within an engine (slider assy, journal bearings, oil pump, etc)...

 

[ivymike]

ok, I have some numbers in front of me. For a single cam from a (particular) dohc, 4v/cyl 2.2L gasoline engine, motored friction power at 3000 engine rpm is about 0.25 HP. Total friction power for both cams at that speed is 0.5 HP. Friction torque increases slightly at lower rpm (but power drops).

 

[GregLocock]

Oh, well that'll teach me to rely on my memory. Nitty gritty question time - would the power taken to drive the valvetrain change if the engine was running? We measured it by strain gauging the spokes on the drive sprocket.

Cheers

Greg Locock

 

[ivymike]

yes, it would change. The most significant factor would be the increase in force to open the exhaust valves. I don't expect that it would double the friction torque (I'd guess 10% change or so).

 

[aorangi]

Found some data about mechanical losses in (-) for diesels. (after Ricardo).

Pistons, rings against cylinders: 44% (50%)
Main and crank journals: 22% (24%)
Gas exchange: 20% (14%)
Valvetrain: 8% (6%)
Oil, water and fuel pumps: 6%

Also some old data about a small 4 cylinder of 1100cc which I think was a Renault. The difference of losses with and without lifters are detailed at different revs. Total mechanical losses at 6000 rpm amounted to 32 hp:

1000 rpm : 0.5 hp
2000 rpm : 0.7 hp
3000 rpm : 1.3 hp
4000 rpm : 2.04 hp
5000 rpm : 3.5 hp
6000 rpm : 8.9 hp

I suppose there's a great difference if the lifters or rockers are flat or roller type. Anyone having up to date data ?

Cheers
Aorangi

 

[ivymike]

There is a difference in effective friction coefficient between flat and roller followers. Whether or not you consider that difference large is a matter of perspective.

The numbers that you mention seem close enough to be correct for a particular engine, but there are significant variations between engines. If I get a chance before turkey day (the US holiday "thanksgiving&quot, I'll look up some more info from testing of modern engines. I won't, of course, be able to share specifics about the engines, but I can at least try to generalize based on what I find (same as above).

 

[GregLocock]

A large automotive V8 DOHC, each camshaft, about 2 Nm, at normal oil temp

Another DOHC, I6, roller followers, very high torque at cranking speeds, say 5 Nm (to be safe) at 20 deg C, falls to 1 Nm/shaft, at 3000 crank rpm, 110 deg C, double that if cold.

Note the gotcha when converting camshaft torque to power.

The second set of results was obtained by motoring the shaft only, so there are no combustion forces involved.

I think the difference between my hazy recollections and Aorangis data, and ivymike's low figures and these test results, is probably the roller followers.

After all we wouldn't have switched to these expensive items unless they saved a measurable amount of friction.



Cheers

Greg Locock

 

[ivymike]

roller followers also offer significant durability advantages, and can sometimes allow you to use a more aggressive cam profile. Too bad they don't fit so well on top of tiny OHC engines.