Explore US alternatives to British sleeve valve engines (and the ramifications)

[Shortround6]

So, how much faster could the Gloster f.5/34 go with another 220 hp i.e. 26% increase.

According to the cube root law it might have gone 341mph.

Of course this assumes no increase in cooling drag. No increase in induced drag due to greater weight. It also assumes the Gnome Rhone made it's power at the same height as the Mercury, which it didn't. At 14,000ft it would have been down 21-23hp from it's 12,800 rated altitude.

The Gloster always seems TGTBT (To Good To Be True).
French Bloch 152 using the same Gnome Rhone engine as you are proposing.

Max speed was the same 316mph as the Gloster?
2 feet shorter fuselage, over 3 feet less wingspan, 80% of the wing area.

Neither plane had self sealing tanks or armor.

Alternative engines have been suggested in the past which include the P W R1830 weighing in at 1,250 lbs.

Wiki is not quite correct on this. P W R1830s that weighed 1250lbs were early versions. These were limited to 2550-2600rpm and 900-1000hp for take-off and had single speed superchargers with full throttle heights down around 6500ft.
The R-1830 went to over 1400lbs pretty quick, French put one in the Bloch 152 to make the Bloch 153. It weighed 1423lbs and gave 1050hp at 2700rpm for take-off. Versions with 2 speed superchargers went closer to 1475lbs.
The 14 cylinder engines are going to need heavier exhaust systems, heavier cowls, possibly heavier engine mounts, starters and propellers.

 

[merlin]

Interesting the Estimate of 341 mph.
The other parallel comparison though is the Italian Macchi C.200 Power-plant 870 h.p. Fiat A.74 R.C.38 max. speed at 14,750' 312 mph; C.201 Est. max speed with 1,000 h.p. Fiat A.76 R.C.40 341 mph - both were 14-cyl radials.

 

[tomo pauk]

The MC.200 have had a rather small wing - 181 sq ft vs. 230 sq ft for the Gloster F.5/34; wing span shorter by more than 3 feet. It will be faster than the Gloster, until/unless this one does not get the Taurus or R-1830.

 

[pbehn]

Were power losses higher for sleeve valve engines than for a poppet valve engine?

 

[Shortround6]

Were power losses higher for sleeve valve engines than for a poppet valve engine?

Hard to say. The sleeve valve might have more friction, but poppet valve might depend on valve spring tension. Sleeve valves usually used higher compression.

 

[KiwiBiggles]

Hard to say. The sleeve valve might have more friction, but poppet valve might depend on valve spring tension. Sleeve valves usually used higher compression.

As you say, the sleeve valve might have higher friction losses. But the poppet valves will definitely have losses due to hysteresis in the valve return springs, and losses of spring energy when the returning valve tries to 'drive' the camshaft. Further, opening the poppet valves against the higher pressure in the cylinder requires energy, which is also lost.

The usually higher compression ration of sleeve-valve engines gives greater efficiency, which means lower, not higher, losses.

 

[tomo pauk]

I'm not sure that sleeve valve engines have had higher compression ratios, because of the type of valve-train. German engines were all poppet-valve engines, and their compression ratios were pretty high, from 7:1 to 8:1 often being the values.
The low compression ratio will give higher fuel consumption, but will allow for greater boost = more power.

 

[wuzak]

I'm not sure that sleeve valve engines have had higher compression ratios, because of the type of valve-train. German engines were all poppet-valve engines, and their compression ratios were pretty high, from 7:1 to 8:1 often being the values.
The low compression ratio will give higher fuel consumption, but will allow for greater boost = more power.

With the same boost level teh sleeve valve engine would have a higher CR.

 

[Shortround6]

The compression ratio thing was another benefit that sort of fell by the wayside as time went on. The sleeve valve engine projects started when 70-77 octane fuel was common (and getting an engine to run several hundred hours without breaking a valve spring was cause for taking out magazine ads). With such low octane fuel there wasn't a lot of room to swap back and forth between cylinder compression and boost. ANY advantage was welcome. As fuels got better you could do more trading. Better cylinder head design and better finning helped cut down on the exhaust valve hot spot (along with the sodium filled valves). Liquid cooled engines are a bit different than air cooled in this regard and even the Hercules went through quite a number of cylinder head designs as development went on. Late war and
/or post war ones were copper alloy for better heat transfer.
Cylinder size also has something to do with it. Smaller cylinders can use a bit more compression than large cylinders, everything else being equal. One of the reasons for the multi-cylinder engines vs 12 or 14 cylinder engines. Again, better fuels took away some of the need for tricky engine set ups.

 

[Balljoint]

With the same boost level teh sleeve valve engine would have a higher CR.

Because of the hot exhaust valve? The sodium cooler poppetversion should have helped.

 

[tomo pauk]

With the same boost level teh sleeve valve engine would have a higher CR.

Care to shed some light on that?

 

[wuzak]

Care to shed some light on that?

That was the theory.

I don't think that there are directly comparable engines which can be used to illustrate this.

 

[Shortround6]

It also changed over time.
According to some accounts Fedden started working on the sleeve valve aircraft engine in 1927. They got the first "production" Perseus running in 1932. Although in first few years production might have been counted on the fingers of both hands. Perseus used the same bore and stroke as the Hercules (and the same as the Mercury).

Many of the problems with 1926-28 poppet valve engines were done away with during the 30s. Better valve springs, the Sodium cooled valves and better casting/forging techniques allowed for narrower, closer spaced and deeper fins for better cooling (this last also helped the sleeve valve engines). A LOT of engines improved from 1932/33 to 1938/39 let alone from what they were in 1927. Both sides of the argument were aiming at moving targets.
Bristol was finally able to make high powered, very reliable engines. The question as to wither it was worth the cost has never really been answered.

 

[kool kitty89]

For Bristol - the 'twin Mercury' seem to be a recurring idea. Though, it would take a lot of money to go both with sleeve-valve designs and a twin 'classic' engine in the same time.

Yes a lot on that was discussed here: http://www.ww2aircraft.net/forum/engines/bristol-radial-engine-development-14674-2.html

Including the issue with Bristol engineers favoring Sleeve Valves due to apparently not seeing a way of retaining their 4 valve per cylinder configuration in a 2-row set-up. But even so, the 2-valve per cylinder engines seemed to manage quite well as it was. (Alfa Romeo did produce a 2-row Mercury style engine, though Najima produced the 14 cylinder Ha-5 earleir than that, and that design developed into something roughly in the Hercules power range but smaller in diameter)
The 4-valve arrangement may have been more important for the low-octain requirements for both valve temperatures and engine RPM, so for Bristol's requirements, it might have made more sense than dropping to 2 valves. (better power for given displavement should have been possible -as with 2 vs 4 valve inlines, but Japan and American radial engine designs managed to make up for that with larger displacement with similar bulk/weight that Bristol was managing with sleeve valves)

Simply continuing development for higher power versions of the Mercury and Pegasus would have been plenty useful too. Between the two of them they should have filled all the roles the Taurus did (or sometimes failed to) historically. 4VPC might not have been necessary on the whole, but Bristol had experience with those and building on long held experience would make sense.

That said, a major design aspect of Nakajima's Ha-5 was taking the best aspects of early 1930s Bristol designs AND contemporary American radial engines while elliminating unnecessary complexity. (indeed, the dual/mirrored -front/back- cam arrangement on 2-row American designs made it fairly straightforward to 'double' the Mercury/Jupiter valve arrangement ... and I find it hard to believe that Bristol Engineers didn't realize this -the single cam ... mess the likes of the 14K and 14N used would never work for 4VPC, though)


A decision to rely on licence production would have either come too late to be useful over refining Bristol's own designs (perhaps except for the troublesome Taurus) or would have been so early that development of Bristol's own poppet valve based designs would have made more sense. (even if they'd foregone 2-row designs, further developments of the Mercury and Pegasus would have been significant, or possibly other 9-cylinder derivatives using larger bores/displacements)


Edit: it's also worth noting that Nakajima had the Ha-5 prototype running in 1930, 2 years before Bristol's single-row Sleeve Vale Perseus first ran.

Interesting the Estimate of 341 mph.
The other parallel comparison though is the Italian Macchi C.200 Power-plant 870 h.p. Fiat A.74 R.C.38 max. speed at 14,750' 312 mph; C.201 Est. max speed with 1,000 h.p. Fiat A.76 R.C.40 341 mph - both were 14-cyl radials.

The MC.200 have had a rather small wing - 181 sq ft vs. 230 sq ft for the Gloster F.5/34; wing span shorter by more than 3 feet. It will be faster than the Gloster, until/unless this one does not get the Taurus or R-1830.

The P-36 and Re.2000 would be better comparisons ... P-36 less so given the low rated altitudes for most engines, higher weight in some arrangements, and higher power engines in some as well. (I don't believe 2-stage supercharged R-1830s were ever used either, even on the more interesting fan-cooled XP-42)

The Re.2000 is probably the best comparison given with 329 MPH top speed at 13,000 ft. It had a fairly similar wing area (moderately lower at 220 sq ft), and a considerably higher weight AND a fairly heavy engine with lower power output than the 14N with 986 hp at 13,000 ft. (in fact, performance was rather similar to the Pegasus XVIII, though a couple inches narrower and a few hundred pounds heavier)


All that said, a Pegasus or Taurus powered F.5/34 would have been more interesting historically. (more so if Bristol had continued Pegasus development -let alone Mercury ... similar displacement to the Taurus, wider, but lighter and more proven)

 

[fastmongrel]

Is there any chance the Pegasus could have been fitted with bigger bore cylinders to make a more powerful engine of the same diameter as the real life Peggy. The Wright R1820 is roughly similar in size and weight but a bit bigger capacity so a 30 litre Pegasus should have been easily capable of 1000 hp with a decent supercharger in 1939 and with development and 100 octane 1200hp plus by mid war. I can think of several aircraft that would have been better with a big bore Peggy, the Sunderland, Wellington and Skua come to mind.

 

[Shortround6]

You aren't really going to get much without a major re-work. The Pegasus was 1753 cu in vs 1820 so the Wright was only 3.8% bigger. The Pegasus used a narrower bore but a longer stroke.

Make sure you are comparing like to like when comparing the R-1820 as there so many and they varied quite a bit in weight. Wiki says 1184lbs but that is for the 1000hp T.O. version. The 1200hp T.O. versions went closer to 1320lbs and the end of war 1425hp versions went 1360lbs or more. There was a lot more going on than filling the fuel tank with 100/130 and futzing with the boost control

The G-200 (1200hp) R-1820 used a steel crankcase instead of Aluminium and had a LOT more fin area on the cylinder barrels and heads compared to plain "G" version. A few other differences too.

Wright did a number of redesigns on the R-1820 during it's history, sometimes all that was left was the Bore and Stroke

Pegasus actually doesn't get you much over the Mercury as far as a fighter engine goes either. Mercury being good for 840hp at 14,000ft. Even the two speed Pegasus was only good for 885hp at 15,500ft. It weighed about 120lbs more than the Mercury.

The bottom end of the Pegasus may have been the limiting factor. Even with 100/130 it was NOT allowed the same increase in boost pressure the Mercury was.

And if Bristol is doing major redesigns on the Pegasus what else isn't getting worked on? new versions of the Hercules?

edit; Short bit on the History of the cyclone 9: http://www.enginehistory.org/Wright/Cyclone9Facts.pdf

 

[tomo pauk]

Thanks for the link

 

[kool kitty89]

Pegasus actually doesn't get you much over the Mercury as far as a fighter engine goes either. Mercury being good for 840hp at 14,000ft. Even the two speed Pegasus was only good for 885hp at 15,500ft. It weighed about 120lbs more than the Mercury.

The bottom end of the Pegasus may have been the limiting factor. Even with 100/130 it was NOT allowed the same increase in boost pressure the Mercury was.

And if Bristol is doing major redesigns on the Pegasus what else isn't getting worked on? new versions of the Hercules?

I was suggesting more hedging their bets with continued development alongside some of the sleeve valve designs but ... on that note, it'd be a lot more efficient to focus on one family of designs entirely. (ie either go all sleeve or all poppet)

I'd have suggested pulling development resources away from the Taurus (along with all the small displacement sleeve valve radials) in favor of a mercury and/or pegasus followon, but then you could argue in favor of the Perseus too. (fewer teething/reliability problems than Taurus, more design commonality with Hercules and Centaurus and may have been wider but the compact size of the Taurus doesn't justify its other problems or preference for that over the Perseus)


There's the bigger argument over poppet vs sleeve in general, but that's been discussed plenty before and a bit in this thread.



Hmm, though thinking back to the original topic, and possible 'holes' in the British engline lineup that could have been filled, the R-2800 actually might have been pretty interesting given how long it took to get the Centaurus into production. Particularly the higher power -and very different in design- C series of engines with the forged cylinder heads, though some of the earlier models might have been significant enough to be worth looking into earlier too -namely the 2-stage supercharged versions with water injection capability. (most of the single stage R-2800s wouldn't be a big enough leap over the Hercules to be worthwhile)

 

[Shortround6]

A lot depends on timing.

It took Ford about one year to go from breaking ground to rolling R-2800 engines out the door. It took another 3 months to break 200 engines a month and it took another year to work up to over 750 engines a month.

The R-2800 C wasn't production ready until the beginning of 1944. And then it took the P W satellite plant 6 months to go from 9 engines a month to 300 a month.

The Centaurus might have been ready sooner a, if it hand not been side tracked to allow the Hercules to be developed quicker (shortage of engineers) and b, if Fedden hand't been forced out of Bristol.

If you don't have enough engineers/draftsmen to work on the Centaurus do you have enough to set up an R-2800 plant? Change over all the drawings (reverse what Packard had to do for the Merlin). Design/modify jigs and fixtures to suit available machine tools? And many of these engine programs relied on hundreds of subcontractors.

The Taurus was pretty much a lost cause. It traded complexity and cost for small frontal area at a time when most customers were shifting to bigger engines anyway. A problem with long development times. Take too long and the market shifts before the product is ready.

 

[kool kitty89]

The Taurus was pretty much a lost cause. It traded complexity and cost for small frontal area at a time when most customers were shifting to bigger engines anyway. A problem with long development times. Take too long and the market shifts before the product is ready.

Again, it seems odd that as much work was put into it as there was when the Perseus seemed to have plenty of potential with many fewer problems. (plus the higher power/displacement Perseus 100 derivative)

A 14 cylinder Perseus 100 derivative would have been interesting too, but given we're focusing on engineering/manufacturing efficiency here, focusing on the basic Hercules seems like a better bet. (or things like developing higher altitude supercharger configurations)