On valves

[Red Admiral]

With engine speeds rising to over 3000rpm the traditional poppet valves couldn't keep up as the springs used would not close quickly enough. Thats not the entire story as the springs start resonating, but that wasn't known or possible to find out.

What do you guys think of the possibility of using desmodromic valves to increase engine speeds and so also power? They're similar to poppet valves but with two cams which both open and close the valve. Google gives a nice animation of how they work.

[ChrisMcD]

Hi Red Admiral,

As I understand it, improvements in spring metalurgy in the late 40's allowed engine speeds up to at least 8,000rpm - and were relatively cheap as well as being a logical progression of existing technology (ie big fat American radials).

That is why the sleeve valve systems have never really made much progress after the Centaurus.

Desmodromic valves are expensive and high maintenance, so they have never made the grade either - apart from Italian motor bikes and Mercedes sports cars - which rather makes the case.

How about electomagnetic valve actuation? No camshafts to go wrong and complete computer control.

http://www.grandprix.com/ft/ftpw012.html

Trust the French to take a logical idea to extremes!

[Wuzak]

Electromagnetic valve actuation isn't practical for high speed engines yet, though it could be used for aircraft engines at their lower rpm.

In the '40s, of course, this was far beyond the realm of possibility. Desmodromic valve actuation wasn't. I believe Mercedes Benz had used desmodromic valves prior to WW2 in their GP cars.

Another F1 technology that may have been useful for aero engines is the pneumatic valve spring. Renaulyt first used that in 1986. Basically it is a closed system which uses air pressure to return the valve to its closed position. In F1 they have a small bottle, and over time their is inevitably some leaks. If the leak is too severe the engine will run out of air before the end of the race, but it can be quickly recharged in the pits.

On an aero engine this could be bleed air from the supercharger, or from an auxiliary supercharger, so that the air pressure is maintained. Only question is, were the seals of the 1940s up to the task?

There were other advantages of sleeves over poppet valves. The former allow for a more compact engine, the distance from the crank centreline to the top of the cylinder/valve cover.

[Tony Williams]

quote:Originally posted by ChrisMcD

That is why the sleeve valve systems have never really made much progress after the Centaurus.
I don't think so. Most of the planned big piston engines of the late 1940s were switching to sleeve valves. What killed them off was the advent of the gas turbine, not improvements to poppet valves.

Tony Williams
Military gun and ammunition website: http://www.quarry.nildram.co.uk

[ChrisMcD]

Hi Tony,

I know that the theory was that jets were simpler and cheaper to build ( although Hives' made the famous comment about soon doing something about that!).

The main problems with big sleeve valves were sealing the sleeves against oil loss and heat transfer - neither of which has really been solved for traditional 'Bristol style' sleeves as far as I know.

The old joke about the Beverley was that with four Centarus the pilots never got lost because they could always follow their smoke trail back the way they had come from.

Improved valve springs meant that the big American radials could be given another boost of power to keep them dominant before the jets took over - which I think is your point about the fact that apart from the Centarus, all the other late 40's big sleeve valve engines remained 'planned'

[Montana]

Hi,

as I can see it, this really is an expert's thread. The contributions so far contain a REAL lot of truth. So there's only a little which I can contribute, too:

The desmodromic valve gear has other advantages than resistance against valve-spring-flutter, too. In a desmodromic valve drive, you are free to give the valves MUCH higher acceleration/deceleration during the opening/closing-cycle, than with a standard cam/spring-design. Thus, the effective maximum open time for a valve can be up to 1/3 BIGGER than with springs.

Sleeve valves: Hmm - do I really have to tell you about my fondness of these...? Sleeve-valves are GREAT. but they inevitably are extreme oil-throwers, too, as you can't design proper gaskets to work on the cylindric surface, which should keep the oil-film between liner and cylinder from entering the exhaust-ports. The flat-disc-rotary valve, on the other hand, can be build with proper gaskets, reducing oil-throwing close to nil...:D

Materials: Today, poppet-valves mustn't be made from steel, only, but may also be made from Titanium or, even Ceramics. (You wouldn't think that, almost every modern Mercedes-Benz-car you can buy nowadays, is equipped with ceramic valves, would you...?)

The lighter the valves, the better. And the heat-resistance-capabilities of Titanium and Ceramics are legendary. ANd both are still capable of being sodium-cooled, too, of course!

Elsewhere in this forum, I have lectured you about the optimum piston-engine, already. But, using modern techniques and materials, classic aircraft of ANY make and kind, radial or inline, aircooled or watercooled, today may also be driven to new limits, yet.

With lighter valves, being made from new materials, with new valve-spring-design, with computer-calculated cam-design, with high effichiency-superchargers, intercooling, with port-fuel-injection and ADI-injection, with newly designed pistons and, conrods made from titanium, rather than steel: Even with standard AVGAS LL100-fuel, an Allison V-1710, for example, may be re-designed to put out approx. 2430 HP at 4500 rpm nowadays. And you could boost this power output even further to about 3000 HP, if this engine was running on pure Methanole or, Ethanole (YEAH, MAN: IT'S A GREEN ENGINE THEN!!!)

As Chris mentioned, modern spring-metallurgy allows for increasing the revs up to 4500 rpm - and higher! - without fuss anymore. So even the "classic" valve-train may be retained.

One indication: in the 1940ties, an average piston-speed of 15 meters per second was considered maximum permissible. (Allison, RR Merlin)

As of today, engineers feel totally secure to have an engine running at average piston speeds of 25 meters per second constantly and routinely for HUNDREDS and THOUSANDS of ours.

Coincidentally, the TBO for engines over the last 60 years has been increased from 400 hours (1940ties) over 800 hours (standard-Allison and Merlin, today refurbished with modern methods) over to 2500+ hours in modern for freshly designed engines.

There are other, "optimum" options, though, which would allow yet another 1/3 increase in power output. But then you had to divert from the classiv V-12-design, and had to resort to completely new solutions.

Cheers!

Montana

[Wuzak]

Hi Montana, I wondered how long it would be before you threw in the possibility of disc valves.....


You're right about the problems of sealing sleeve valves - and this is probably a reason they weren't adopted for car engines. Today's materials and manufacturing capabilities should minimise the oil sealing problem, as the materials can be more stable at the required temperatures and tolerences can be made tighter.

As an aside, it would be interesting to know if the Sabre was better or worse than the Bristol sleeve valve engines for slinging oil. In theory the liquid cooled engine should have better temperature control and the clearences will be less, especially when the engine hasn't fully warmed up.

[Montana]

Wuzak,

try this link: http://www.rcvengines.com/

This company claims to have overcome the sealing problem.

I have doubts about the suitability of this technique for aircraft engine.

Reasons:

1. The size of the ports is highly restricted, due to the approach they chose to porting.

2. The seals appear to me like kind of a "designed wear"-type. It's like a chunk of something, pressed against a grinding-wheel (the rotating sleeve).

This kind of sealing shure DOES work - but only for a short time, until the seal has been completely ground off.

See and check yourself.

By the way: Being from the motorcycle business myself, I haven't heard anything about a bike with such an engine being mass-produced, yet - although this angine is claimed to be around for a good couple of years, already.

Seems like, somebody knows - or: has found out - something that's not contained in this website...

Besides: WHO NEEDS ball bearings to center the sleeve in the cylinder? I mean: Bristol, Napier and, RR didn't. And their sleeves were so big - you could raise chicken inside...!

Cheers!

Montana

[Wuzak]

Ball bearings? That we the installation bulkier, surely?

[Montana]

Yeah,

only reasonable reason ( :D ) I can think of, is reducing friction, which might be worth the effort in small engines with small outputs.

Every other reason is - err: beyond reason, I am afraid...

[:o)]

Cheers!

Montana

P.S.: BTW - have you ever heard of THESE? You won't believe your eyes!

[8)]

http://www.me.berkeley.edu/mrcl/mini.html

http://www.me.berkeley.edu/mrcl/micro.html

Cheers! [8D]

[Wuzak]

Yes, when I posted that I hadn't looked at the site and realised that these are model engines.

At that size sleeve friction would be a very significant factor.

A company I used to work for had made a few 2 stroke model engines. The prototype had a roller bearing crank. It ran beautifully. Someone decided to change that to a plain bearing, which could barely be started.

Plain bearings need oil pressure - either supplied or hydrostatic pressure - to work properly and reduce friction. I suspect that in the model engine it just wouldn't get enough pressure/oil.

[Red Admiral]

http://www.modelenginenews.org/gallery/croft/eagle/index.html

Model Rolls-Royce Eagle 22 with sleeve valves. It does actually work. I met one of the writers for Model Engineer last year and we got talking about this sort of thing. The sleeve vales for the model were extremely tricky to manufacture. Unfortunately it doesn't run properly on standard petrol. Runs ok on methanol however. Its quite an achievement.

Thanks for the comments on the valves.

[Wuzak]

Even though it is a model engine, it is much larger than the ones <ontana mentioned.