Valvetrain

[burell]

Pat, I guess I was too eager to ask burning questions after becoming a member (searching on keywords is a member-only feature). Anyway I have found enough info on accumulators in relation to turbo lag, thanks.

I would very much appreciate additional input on the subject of valvetrains though. I have always found the setup of a shaft with lobes, rockers and strong springs a strange one. In my way of (naive?) thinking there must be much more elegant and efficient ways of opening and closing valves in a particular order and with particular timing. I know of (radical) different mechanical setups that never made it and I guess the way to go eventually is an ECU controlled system of individual valves like they use in F1.

What do the experienced and professional members of this forum think of the classic valvetrain? Do you consider it quite OK or is it just something that has historically grown on us and thus live with?

I also have a less philosofical question: if valve float wasn’t an issue would it be possible to gain more usable power from a typical automotive engine and would this be something engine designers/builders would be interested in?

I really appreciate your input,

Jochen

 

[patprimmer]

Jochen

I am pleased you found your answers without a whole lot of people having to do it all again.

The poppet valve and springs evolved from an array of systems being tried in the late 19th century and early part of the 20th century. These included atmospheric inlet valves, sleeve valves, desmodromic valves, rotary valves and I suspect reed valves, piston ports, peripheral ports and possibly others I have not seen.

Atmospheric valves have limited control and limited rpm potential. rotary and sleeve valves had friction and seal issues. The poppet valve very neatly self seals with a little help.

Desmodromic cams were used on looms to weave cloth for many years before the introduction of the IC engine, so the concept was at least known.

The biggest limitation I see with shafts with lumps operating the valves is that they could only be optimised for a narrow set of operating parameters. "V" tech has broadened that a bit, but in the Honda format, it only optimises 2 sets of operating parameters.

A better performing system would be cams with 3 dimensional profiles, like used to metre fuel in some timed, mechanical, fuel injection system made by Bosch/Kuglefisher and used on some BMWs in the 60s and 70s, but I shudder to think of the cost and maintenance. Even F1 didn't go that way, at least as far as I know.

Computer controls and pneumatics and electromagnets have opened a whole new world. It will be interesting to see if it will ever become economical, durable and reliable enough to be commercialised.

A significant amount of the force used to compress a valve spring is recovered on the closing side of the cam, as the spring assists the cam in turning.

There are some losses in friction and inertia at the various points, and some lost as heat generated in the springs due to hysteresis.

There is also small losses in the the camshaft drive gears, belts, chains or whatever, and in the the extra capacity in the oil pump to supply the cam bearings.

The best valve train I ever saw for a 4 stroke IC had no moving parts. It was a ring shaped cylinder head (shaped like a slice cut from a hollow sphere, so the centre of the slice was the equator). The cylinders, in similar configuration to a radial aircraft motor, rotated inside the head. The ports were peripheral and the size of the bore. The port timing was the length of the slot in the head.

Great mechanical and volumetric efficiency. Lousy thermal efficiency due to the large surface area of the combustion chamber.

Regards
pat pprimmer@acay.com.au
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.

 

[GraviMan]

"A better performing system would be cams with 3 dimensional profiles, like used to metre fuel in some timed, mechanical, fuel injection system made by Bosch/Kuglefisher and used on some BMWs in the 60s and 70s, but I shudder to think of the cost and maintenance. Even F1 didn't go that way, at least as far as I know."

Now there is a concept! In a world where CVT is developing in major strides, I'm suprised some manufacturer hasn't had another serios look at this. This would outperform even the servo-valve systems being developed.

I gather there is some work being done on rotary valves, but with the entire cylinder liner rotating. Can't remember who was looking at this. I do remember that there were wear advantages at TDC, since the piston/liner contact never becomes stationary (most wear occurs at TDC apparently).

Must admit, Jochen, I usually find myself wondering why CFD design techniques hasn't replaced all 4-strokes with turbo 2-strokes! Renault are looking at a 6-stroke engine (for emissions) - YUK!!!

Mart

 

[notnats]

Sleeve valves were fashionable in the 1930's and Ricardo wrote very favourably of them, but it is >30 years since I read his book and can't remember the details.

Jeff

 

[bazeleur]

Actually, there was a pretty interesting discussion on that very same topic over on AtlasF1, you can check it out here:

http://forums.atlasf1.com/showthread.php?s=&threadid=67536

 

[sreid]

F1 racing technology comes up quite often. Questions;

1) I would think that F1 engines still use conventional cam shafts. Do they use roller followers?

2) I believe that every one uses pneumatic valves. My assumption is that an air piston replaces the valve spring to a) reduce weight b) to eliminate valve spring dynamics c) to avoid spring failure due to metal fatigue. Is this correct?

3)How are the cams driven; belts, chains or gears?

 

[tbuelna]

burell:

If your question regarding valve train architecture is specific to production automotive engines, there is really only one design consideration: Can this valve train be produced and sold at a profit? Everything else is secondary.

That is why Corvettes still use pushrods/rocker arms

http://www.chevrolet.com/corvette/standard_optional.htm

and high-end Mercedes-Benzs have gone back to 3-valve (instead of 4-valve) cylinder heads:

http://www.mercedes-benz.com/com/e/home/innovation/recentdevelopments/engines/v12biturbo/index.html

It gets the job done and it's low cost. Automakers exist to generate a profit for their shareholders, not to impress a bunch of gearheads with their technical prowess.

Terry

 

[NormPeterson]

"if valve float wasn’t an issue would it be possible to gain more usable power from a typical automotive engine and would this be something engine designers/builders would be interested in?"

Relatively few people are willing to normally drive in a manner that must use the upper reaches of safe RPM in order to avail themselves of the power that potentially lives up there. This is apparently true even in the relatively hard-core sports car market, where the Honda S2000 now runs a longer stroke version of its VTEC engine for more torque at the expense of specific power and redline RPM.

What the majority of folks want, if they even stop to consider, is sufficient low end and midrange torque to give good response through an automatic transmission (which bleeds off a certain amount of power to operate). On the other hand, peak HP is pretty much meaningless in most vehicles even if you disregard 65 - 70 mph speed limits by as much as 20 mph.

Actually, I see the current gas price trend as an influence to push engine tuning in the direction of scaling back the rpm at which peak torque is reached to correspond more closely to normal driving speeds rather than pushing it "upstairs" for greater specific output.

Re pushrods: - while I'm not particularly a fan of pushrod arrangements, they also permit lighter and less bulky head castings (read: packaging, assembly-line installation and engine CG advantages) and are at least "adequate" for street driving.

Norm

 

[carbonlife]

"if valve float wasn’t an issue would it be possible to gain more usable power from a typical automotive engine and would this be something engine designers/builders would be interested in?"

I'd go as far as saying that since burell is asking about a "typical" engine, that valve float simply isn't an issue - at the design stage, it's certainly not the rpm limiting factor. It's other things which cost a lot more to engineer for high rpm that mean specific output isn't higher. I must say though that Norm is probably very close to the mark about designing high revving engines for the masses - every manufacturer who has done this has usually ended up being criticised in the marketplace and had to re-engineer for more torque.


Pat says:
"The biggest limitation I see with shafts with lumps operating the valves is that they could only be optimised for a narrow set of operating parameters. "V" tech has broadened that a bit, but in the Honda format, it only optimises 2 sets of operating parameters."

Yes, but isn't it amazing how wide those operating parameters can be made to be. I think that quite a few people have thought about infinitely (or large but finitely) variable valve timing but that Honda's real coup was getting most of the advantage by a relatively simple mechanism. I mean, if vtech had a third operating lobe, would that really make the engine more effecient, more drivable, more powerful? The operating parameters are so wide that making them any wider would mean pushing the rpm limit up and this would get expensive in terms of other components........

That's not to say that I like poppet valves.

"rotary and sleeve valves had friction and seal issues."

And I think that you're right to use the past tense here. These problems were generally overcome, weren't they? Even with 1960s materials, the Aspin rotary valve eventually got oil consumption down to an acceptable level. Well, they have their disadvantages but they are great at some things like charge stratification meaning that very lean mixtures could be burnt. And they are very efficient in terms of combustion needing very little ignition advance. Somehow I don't think we'll ever see a return of such devices since development of a cheap thing like poppet valves has come such a long way.