Did Sleeve Valves Have an Advantage at Very High RPM?

[ThomasP]

re sleeve valve engines used in automobiles:

I read an industry periodical article a number of years ago (maybe 30 or so?) and it gave 2 primary reasons why the sleeve valve did not gain more popularity post-WWII. The conclusion was that the initial purchase cost was significantly higher (high enough that it put the majority of people off buying them), and that the different mechanism involved and knowledge required also discouraged the automotive service industry. These 2 factors kept the vehicles equipped with sleeve valve engines in the well-to-do and niche market. The article also made the point that this was in spite of significantly improved reliability and fuel efficiency. They were also supposed to be significantly quieter. I can not comment on the accuracy of these statements, but it was a technology based periodical and not an advertising type.


re modern sleeve valve engine being practical in use:

When I worked at Chrysler on the M1 Alternate Power Plant Project, the diesel engine guys investigated the sleeve valve system re air-cooled diesel engines and concluded that it could work quite well and have some potentially significant advantages over the air-cooled poppet valve diesel engine. But there was no interest in developing what would effectively be an almost entirely new engine, with the associated costs and a questionable future market.


Also, although I do not disagree with Snowygrouch's above statements in general, I feel it is only fair to point out that the 20,000 rpm Renault 3.5L V10 only has a 2" stroke (maybe slightly less?) and is intended to last only a few hours (6-7?) operational between major inspection and possible repair or replacement. The cost of the engines works out to ~$1,000,000 purchase price per hour of operation , although I am sure the price would come down if they were produced in larger quantities . The (then current)* rules that govern the engine designs used in Formula 1 encourage high rpm and short stroke. They get their high volumetric efficiency by using a large piston diameter:stroke ratio, allowing 4 unusually large valves per cylinder on relatively small displacements. The F1 engine designs are not at all suitable for use in turning large aircraft propellers - the weight of the reduction gear box alone would probably be several times the engine weight. (The M1 Abrams' X-1100 transmission/gear-box weighed ~2x what the turbine engine weighed, and the turbine engine weighed ~5x what the helicopter turbine (that the M1's turbine is derived from) weighed.)

*edited from "current" to "then current"

 

[wuzak]

Also, although I do not disagree with Snowygrouch's above statements in general, I feel it is only fair to point out that the 20,000 rpm Renault 3.5L V10 only has a 2" stroke (maybe slightly less?) and is intended to last only a few hours (6-7?) operational between major inspection and possible repair or replacement.

Around 40mm.

The cost of the engines works out to ~$1,000,000 purchase price per hour of operation , although I am sure the price would come down if they were produced in larger quantities .

They went from unlimited engines (using a different one for qualifying and the race) to a single engine per weekend (and later to two weekends per engine, but I can't recall if that was in place for the V10s, or when the V8s came in).

Each team would have two cars, 16-17 races per season, plus pre-season test sessions (one car only) of 10-15 days. So you were looking at ~40-50 engines per team plus spares.

The (current) rules that govern the engine designs used in Formula 1 encourage high rpm and short stroke. They get their high volumetric efficiency by using a large piston diameter:stroke ratio, allowing 4 unusually large valves per cylinder on relatively small displacements.

Current engines have bore fixed at 80mm, capacity at 1,600cc and number of cylinders 6. Stroke is 53mm.

They are turbocharged, with a motor/generator, so that helps the VE.

The rpm used is determined, essentially, but the fuel flow regulations, which permits a maximum flow rate of 100kg/h, but only at 10,500rpm and above. Below 10,500rpm the maximum fuel flow is proportional to the rpm.

 

[ThomasP]

Hey wuzak,

Thanks for the info, I had not realized how much the rules had changed.

 

[Deleted member 68059]

Hmm where did I advocate using an F1 engine in an aircraft ?

I was purely pointing out that a sleeve cannot possibly be of any advantage at high engine speeds from a breathing perspective
as a poppet valve engine without even having forced induction can still overfill the cylinder at 20,000rpm.

This is not because its oversquare (althought it obviously helps), thats done to lower the piston speed and stop it falling to bits, the
100%+ filling is due to resonance tuning of the induction and exhaust pipe lengths for the most part.

 

[ThomasP]

Hey Snowygrouch,

I know you did not mean that the F1 engine would make a good aircraft engine. I just figured it was worth pointing out that super high rpm engines had problems all of there own - I suspect that if the same F1 design concepts were used in a full size aircraft engine, the piston speeds and rpms would be about the same as for a modern design sleeve valve engine of the same power.

I would think that the tuning of the induction and exhaust systems for volumetric efficiency can be done for either type of engine. I touched on the tuning subject a few years ago in another thread on this forum, starting here: "Why so few planes that fired thorugh the propeller hub?". Have you run across any tests that indicate it would be easier/more efficient for the poppet valve than for the sleeve valve engine in the Merlin/DB605 range? I am curious.

 

[Deleted member 68059]

Hey Snowygrouch,

I know you did not mean that the F1 engine would make a good aircraft engine. I just figured it was worth pointing out that super high rpm engines had problems all of there own - I suspect that if the same F1 design concepts were used in a full size aircraft engine, the piston speeds and rpms would be about the same as for a modern design sleeve valve engine of the same power.

I would think that the tuning of the induction and exhaust systems for volumetric efficiency can be done for either type of engine. I touched on the tuning subject a few years ago in another thread on this forum, starting here: "Why so few planes that fired thorugh the propeller hub?". Have you run across any tests that indicate it would be easier/more efficient for the poppet valve than for the sleeve valve engine in the Merlin/DB605 range? I am curious.

Difficult to say, I "suspect" that the transition of the standing wave from a displacement node to anti-node might happen in a faster and more defined
manner with a poppet as the seat/valve interface releases around the whole periphery at once. But I`m not convinced that would make a whole lot of
difference to the practical tuning possibilities, as they`re principally defined by "open-closed" organ pipe resonance theory anyway. (as opposed to
open-open).

The trouble with all this stuff is that "interesting" valvetrains with lots of "potential", usually just dont work because its a lot more complicated that just "bigger windows"
to let air in.

A prime example is the five valve poppet head, in theory, more flow area and a considerable boost in peak power, in practise, it varies from mediocre, to disasterous.
The reason being that the angle of the middle port tends to screw up the nice tumble flow from the two outer ports, and thus combustion is worse, and power worse.
(see the ill fated Yamaha designed five valve Formula One heads)

Oil consumption is another major no-no for todays engines, and I am exceptionally doubtful that a sleeve would suceed in that area either (not outside of
special lab engines anyway, with some incredibly expensive sealing systems.

Ultimately, the liner needing to go up into the junk head, is always going to be a very severe downer (because you get a very tortous path of either air or
water to try to cool it) on combustion chamber metal temperatures, and I am
fairly certain this is the primary reason for the very low peak boost levels achived in WW2 by Bristol and Napier sleeve systems. It could probably be
considerably improved these days, but the trouble is, so have poppet engines, so I think the "delta" between them will remain similar in terms of knock
limit.

 

[Trilisser]

using a completely different engine to say sleeves are better whilst saying "don't compare it with THAT" one.

I think you've discovered the hole in your methodology !

Richardo are still going, why do YOU think they have not made one in three quarters of a century?

You've also ignored the data driven point about boost levels, still waiting for you to post a single piece of actual data to support your assertion?

Why didn't the Centaurus or sabre exceed 15psi boost in the whole war? Why didn't the crecy work?

We already know, and it's all in the book, with references.

Sleeves give you a considerably lower knock limit. This is not a discussion point I'm afraid.

If you think with the pressure today to develop better engines that the fact nobody is making them quote "means nothing" then I'm not sure I can help you any further since your proclivity for the technology is obviously ideological in root,

Well, how much boost would the Merlin have "taken" with a CR of 7.2:1? Second, shouldn't air-cooled radials be compared to air-cooled radials of similar size? Thus the R-3350 is by far the closest comparison.

And why should the Sabre need to exceed +15 lbs. when even at that it far exceeds the output of the its closest comparison (Griffon)? How many German engines exceeded +15 lbs.? Neither does the R-3350 exceed it.

Again, what current makers actually do is of little relevance because their main interest is not technical excellence but the bottom line. You yourself wrote elsewhere how cost-conscious car business is. I have no doubt that sleeve-valves are more expensive to make. It is also a fact that turbo-compounding offers substantial benefits in fuel-efficiency, yet where are cars with turbo-compound engines? Again: technical excellence vs. bottom line.

And aren't sleeve-valves actually banned in major car racing classes?

 

[Shortround6]

You can answer some of your own questions.
It is not always about technical excellence.

Aircraft designer doesn't care about the HP per cu in.
They care about HP per pound of weight.

Choices the engine designers make reflect in the weight of the engines/power plant.

If engine A makes 10% more power than Engine B but weighs 15% more it is not a good choice for an aircraft engine.
If engine A costs significant more to buy than Engine B it is not good choice.
If engine A costs more to maintain than engine B it is not a good choice.

The company making the airplanes and their customers have to consider all the factors. Some don't always line up with either engine A or B.

But deciding on Engine A or B based on max boost tolerated or max power per cu. in. is not a way to decide the usefulness of engine.

Try telling us the weights of the engines involved and also trying telling us the altitudes at which the power was achieved or what power could be used in a high speed cruise at a certain altitude.
Engines that use two stage superchargers have a bit of extra weight for the extra supercharger stage and also are using up more power to drive the superchargers leaving less power for the prop.

Answer questions like this can lead to answers that help explain why the sleeve valve engines were not a success.
They didn't offer anything that the poppet valve engines couldn't match when a number of factors were taken into account.

 

[Simon Thomas]

I would venture to say that the Bristol Hercules sleeve valve was a success.
Real production started in 1939 and continued until Rolls-Royce refused to make any more in 1967.
Over 57,000 of them were built.
The final TBO was 3500 hours.

 

[Shortround6]

I would venture to say that the Bristol Hercules sleeve valve was a success.
Real production started in 1939 and continued until Rolls-Royce refused to make any more in 1967.
Over 57,000 of them were built.
The final TBO was 3500 hours.

Was it?
Try to fit a 1960s Hercules into an early 40s airplane.
Try to fit the parts out of a 1960s Hercules into an early 40s engines.

The same can be said of the R-2800 to be fair.

What was the final TBO of an R-2800?
It was probably lower but how much lower?
What was the first cost?
What was the overhaul cost?
What was the budgeted cost per hour of operation?

The R-2800 went through several complete revisions.
So did the Hercules.
the Hercules gained about 450lbs from the MK VI to the MK 773.

Let's all be sure we are comparing apples to apples.

 

[Engineman]

I would venture to say that the Bristol Hercules sleeve valve was a success.
Real production started in 1939 and continued until Rolls-Royce refused to make any more in 1967.
Over 57,000 of them were built.
The final TBO was 3500 hours.

Interesting, the Bristol Hercules. Lots of positive points, when it was mature. But, I think it benefitted from some fortunate circumstances. It was able to suit many UK transport aircraft after WW2 and particularly Military transports. This kept things going. Also, was attractive as a British product, not a USA engine in UK aircraft. It just fitted-in, although it wasn't cutting edge.

Eng

 

[Deleted member 68059]

The Hercules was a very very good engine indeed, but the effort expended making it so, was immense. Nobody other than an obsessive
genius like Fedden could (or would) have done it.

However, Fedden could easily have made a radial engine of similar output with poppet valves instead, and probably would have cost Bristol about
5x less to develop.

To give you some idea of what that means, Bristol had the Firth-Brown firm do the metallurgical development for the sleeves, it cost them £ 2 Million GBP.
(roughly enough to have bought you THIRTEEN complete squadons of Spitfire MK1`s - there were about 21 Squadrons of them in the entire RAF
at the peak of the Battle of Britain).

It does not take a lot of imagination to wonder how much better a supercharger you could have had for that £2 million quid, if it hadnt been spent
developing a centrifually cast high-expansion nitrided sleeve technology (which was an incredible, but also wholly unecessary investment in science)

 

[Shortround6]

Interesting, the Bristol Hercules. Lots of positive points, when it was mature. But, I think it benefitted from some fortunate circumstances. It was able to suit many UK transport aircraft after WW2 and particularly Military transports. This kept things going. Also, was attractive as a British product, not a USA engine in UK aircraft. It just fitted-in, although it wasn't cutting edge.

Eng

For the 1940s (peacetime) and some of the 1950s the Hercules was a very, very, Very attractive as a British product as there was a whopping import duty on American made products to help get the balance of trade anywhere near back in balance. This applied to the commonwealth as well. Which helps to explain the Canadair North Star with Merlin engines.

British airframes using British engines in that time period wasn't just a matter of being "suitable" it was close to a matter of necessity.
Unfortunately it might have resulted in fewer airframe orders as the British airframe makers were "locked" into British engines and couldn't offer options to foreign air lines.

Again, a simple data sheet doesn't show some of the decisions that were made in order to get to a finished product.

 

[wuzak]

Real production started in 1939 and continued until Rolls-Royce refused to make any more in 1967.

What was the market for new Bristol Centaurus in 1967?

Griffon production stopped ~1953, despite the RAF still using the Griffon powered Shackleton until the 1990s.

 

[Simon Thomas]

My comment was based on the paper "The Bristol Sleeve Valve Aero Engines" Patrick Hassell Rolls-Royce Heritage Trust.
The latest dataplate I have seen for a Hercules engine had a date of 1968.
The Hercules were possibly wanted by Straits Air Freight Express (who used Bristol Freighters until 1986) or Wardair (who used Bristol Freighters until 1977). Just a WAG.
I have no data on the end of production for the Centaurus.

 

[MiTasol]

This kept things going. Also, was attractive as a British product, not a USA engine in UK aircraft. It just fitted-in, although it wasn't cutting edge.

Eng

And the British were completely anal on that.

Back in 1964/65 the New Zealand National Airways Corporation, NAC, which later merged with TEAL to become Air New Zealand, was looking for a Viscount replacement.

At the time NAC were the worlds largest F-27 operator so had lots of experience with Rolls Royce who made the Dart engines for both aircraft. As a result of this experience they considered Rolls Royce support a nightmare and a major handbrake on operational reliability. Also at that time both Boeing and Douglas offered Brit or US engines on all aircraft.

The aircraft considered included the DC-9, Caravelle and either the BAC-111 or Trident, I can't remember which but more likely the BAC-111 (check Flight or the NZ Herald archives of period). The 737 was looked at but was just a project with none flying and not even a commitment to build as the minimum order level had not been reached.

The Brit aircraft came out on top by a significant margin so NAC proposed an order for x aircraft fitted with Pratt and Whitney JT-8D engines and said they would cough for the conversion costs. The Brits refused, saying we only fit RR engines, so NAC bought the 737 off the drawing board. This raised the number ordered to above the minimum Boeing needed to commit to production and a star was born.

Had the Brits fitted the Pratts to their aircraft as NAC asked it is likely the 737 program would have died as Boeing were already voicing this probability. This would have made the Brit aircraft much more attractive to those other operators who had had a gut-full of Rolls Royce product support and it may well have been that aircraft that sold over 15,000 examples to date and is still in production. Instead production of both Brit aircraft died soon after.

In 1971 Rolls Royce went into receivership - one very small step from bankruptcy, but that is a whole different story. Customer support dissatisfaction was just one factor. And not just from Bristol engine owners and NAC.

 

[spiralcopter]

Thanks all, especially Calum. I had read the enginehistory article on sleeve valves vs. poppet valves, and thought that their assessment was fairly scathing. They concluded that sleeve valves had categorically worse power density, but conceded a slight edge in knock limit thanks to greater swirl due to the intake configuration. It sounds like even that advantage evaporated fairly quickly, since the Merlin had such a high boost limit relative to the sleeve valve engines.

Speaking of boost and sleeve valves, are they limited by the sleeves being bulged outwards, or by their lower knock limit?

Edit: Nevermind, I see that this was already answered, just missed it before.