A 'proper' way to have a 24 cylinder liquid-cooled aero engine for the ww2?

[MiTasol]

DB 606
View attachment 653349
I would be a little leery of using maximum "box" dimensions as they can be sometimes misleading. A lot of engines had a few lumps and bumps that added to the basic structure of the engine and it many cases, were a number of feet back from the front of the engine.

and the Griffon Spitfires they just put streamlined teardrops in the cowl so that the extra width had almost no effect.

 

[daviemax]

You're really asking the wrong question. If you're asking what is the best layout for 1,500hp on 1/Sept/'39, the answer is 14 cylinder radial - It will give you 1,500 hp on 87 octane fuel reliably when war erupts. It will have room to improve 1,900 hp with 100/130 fuel and improvements to supercharging (mechanical and/or turbine) and inter-cooling.

But what goes around comes around, and the question rears its head again late war - what is the best layout for 3,000 hp on 17/Aug/'45. 18 cylinder radials/V-12s are really pushed to make that power. Is the best solution for 3k hp, a double 11 cylinder radial - Wright R-4090; quad 7 cylinder radial - P&W R-4360 or some liquid cooled engine with 24 cylinders - Napier Sabre/RR Eagle, Double V - DB610/Allison V-3420 or Hex Jumo 222.

Wright only made 3 - R-4090s. (OK, they were having issues with the R-3350 at time, and need to focus). And its not a very streamlined engine (58" diameter)

P&W R-4360 was made in qty (almost 20k) but only couple hundred by end of war. I'm not sure you could start in '35 and have a running engine in '39 (P&W R-2800s were busting cranks left, right and center during that period, 4 row radial has even more complex crank) and at 3,700lbs probably too heavy. But can you start with a quad version of the Twin Wasp Junior for 1,500hp in '39 and improve/enlarge it to quad Twin Wasp for 3k hp (on ~2,500 lbs) without completely retooling the factory? (Interesting R-4360 with only 7 cylinders in larger than the 11 cylinder R-4090 at 61" diameter, Twin Wasp is over a foot smaller in diameter).

The hex 24 has all the connecting rod bearing issues of the X engine with 2 more slaves per rod. You also have to deal with primary and secondary vibrations of 6 - 4 cylinder engines which a hex more/less is. Junkers couldn't get them past the prototype stage even by end of war and built fewer than 300. We don't have until '50 when Dobrynin got a couple dozen running.

Vulture is done in more by 100 octane fuel than by the Merlin. In '35, RR is working on engines in 4 classes: Kestrel XXX - 750hp (engine is changed so much it is renamed Peregrine), Merlin - 1,000hp, update R (will be name Griffon) - 1,250hp and Vulture of 1,500hp. With an expectation that the fully developed version with make about 20% more on 87 octane fuel (900, 1,200, 1,500 & 1,800hp respectively). The nail in the coffin from 100 octane fuel - it allows 60%+ increase in power. RR can't figure out a way around the French multi speed patent, so until France falls, you are stuck with single speed engines.

So, the Peregrine needs to be redesigned for 1,200hp, the Merlin for 1,600, the Griffon for 2,000 and the Vulture for 2,400. And there is a war on so pick 2 - RAF needs Merlin for its front line fighter, FAA has funded the Griffon, Avro says they can redesign their bomber around 4 Merlins and it will be better able to take war to Germany.

However, if we never start the Vulture but rather commence with a double V Kestrel, we are sacrificing 2/3 of mid-war Merlin production. Vulture heads and cylinder blocks have same bore spacing as Merlin (even though the Vulture has same bore and stroke of Kestrel), so the manufacturing floor equipment could be converted to making Merlins with relative ease. Also, depending on supercharger setup of the double V Kestrel - i.e. if you keep individual superchargers ala DB605 superchargers on DB610, you don't have the tooling for the 1st stage for 60 series Merlin. Only just over 500 produced, not exactly a roaring success.

Even with those caveats and the weight advantages noted by @swampyankee, I can't put my money on the "X-24". I know the solution for the connection rods is built up crank, but that gets horribly complex (maintaining tolerances). And the engine needs balance "pucks" like a radial as the rotating mass doesn't follow a circular path (its a funky ellipse) so, its still going to be hard on bearing.

DB601/605/603 all had issues with oil foaming. When oil foamed, bearing failed with disastrous consequences. When you are running a double V, the failure of one V, doesn't immediately fail the outer - which has severe consequences - Ronson: lights every time. Messerschmidt and Daimler-Benz fought over what caused the issue (oil tank or oil pump design), in the end, Messerschmidt installed oil/air separators and resolved their issues. I am not aware that Heinkel ever did (Not enough planes made it back to properly diagnose the issue and resolve it?)

Allison V-3420, lost critical head way when V-1720 was prioritized, and then more when engineers fought to make a single supercharger and experimental turbo-supercharger work. Someone in management needed to step in and have them get the engine working with individual superchargers/production turbos. Then work on the single supercharger/fancy turbo.

Germany and USA made very limited numbers (~2,500 & 150 respectively) so not very successful compared to 70k V-1720 and 60k DB601/605s.

I'm not aware of any "flat H" engines with poppet valves in the power range we are talking. I suspect that having poppet valves with their associated springs and cams made the engine too wide for the air frame guys. I've seen the drawings for the Merlin "vertical H" engine, and it is seriously tall.

Which leaves us with the sleeve valve engines for H engines - and sleeve valves were definitely the "fad" in '35 and pretty good solution if you are stuck with 87 octane. You can't replace a 2,000 hp Napier Sabre (87 octane rating) with a hypothetical 1,250hp RR Griffon in '39 (the engine that is being completely redesigned to be a 2k hp engine on 100 octane, while Napier is forecasting well over 2,400hp for Sabre on same fuel). OK, Napier could have used a little assistance from the gov't to pass on how Bristol resolved the sleeve valve issue. (As gov't is paying for the engines, assigning the patent for center-less grinding and bringing in efficiency experts from automotive e.g. Ford would have solved most of Napiers issues). >5k Sabres isn't the 100k for Merlin, Cyclones or Wasps, but given the limited number of planes using (chicken/egg thing?), still more than all the other combined during war.

Based on historical - flat H sleeve valve.

I still think a double Peregrine has merit as you could start double Kestrel in '35 and get 5 years experience in before you are at war with your 1,750hp double Peregrine. Makes for a very different looking Typhoon...

I really appreciate the detailed erudition in this answer. Thanks!

 

[Shortround6]

Even derated to use regular fuel the Fiat would possibly have topped 2500 hp in 1939

Going from a racing engine to a production engine involves 3 things (at least).
1. can you get the power that you want from "production" fuel?
2. can you get the power you want (or even 80% ) for long enough periods of time for engine to be a worthwhile service engine?
Or do you have to "beef" up the engine (and hundreds of pounds) and run hundreds of RPMs slower to have decent engine life.
3. Getting engine power at sea level is a whole lot easier than getting power at 13,000-15,000ft.


Napier got over 1000hp from the Lion at the end of the 1920s
The Merlin was estimated to be worth 1500hp at sea level even using 87 octane fuel if the engine had been set up for sea level.
It took a while (several years) for Merlins to be rated at over 100 hours in service use. Merlins were steadily improved and were worth around 400 hours (depending on fighter or bomber transport engine use) by 1945.

 

[wuzak]

Hmm - let me think. 3100 hp in 1934 beats the 1500 hp in 1939 Vulture hands down.

Therefore you should have said The only engines that historically meets met the '39 date is the 3100hp Fiat AS.6 and 1500hp Vulture.

Even derated to use regular fuel the Fiat would possibly have topped 2500 hp in 1939

3,100hp used very special fuel developed by Rod Banks, who had also developed the fuel for the Rolls-Royce R.

The R was capable of >>2,000hp. From memory 2,300hp in 1931 for the Schneider Trophy race and over 2,500hp for the speed record run. A special sprint version managed over 2,700hp

The R was detuned for use as a regular engine, known as the Griffon I. This was rated at ~1,500hp.

Given that, it would seem that the best an AS.6 would probably have done by 1939 would be around 1,800hp.

 

[wuzak]

And you should add the Italian Fiat AS.6 Fiat AS.6 - Wikipedia to the list of engines that flew.

3100 hp in 1934.

That had as much potential as some of the other types discussed and solved the torsional twisting of crankshaft(s) and camshafts issues that would have eliminated L12 and probably V18 engines.

The AS.6 had two separate crankshafts that were independent of each other.

The front crankshaft drove the rear prop, the rear crankshaft drove the front prop.

The rear crankshaft drove the supercharger and accessories. The carburettor was mounted tot the supercharger.

The reduction drives were mounted between the crankshafts and drove the props via extension shafts in the vee.

IIRC the rear half was started first, then the front section.

Mixture distribution was a big problem with the AS.6, and there were problems with back-firing.

 

[nuuumannn]

The AS.6 had two separate crankshafts that were independent of each other.

The front crankshaft drove the rear prop, the rear crankshaft drove the front prop.

The rear crankshaft drove the supercharger and accessories. The carburettor was mounted tot the supercharger.

The reduction drives were mounted between the crankshafts and drove the props via extension shafts in the vee.

IIRC the rear half was started first, then the front section.

Mixture distribution was a big problem with the AS.6, and there were problems with back-firing.

Good info, wuzak. Here's a rear view out of interest's sake.

Fiat AS.6

 

[MiTasol]

The AS.6 had two separate crankshafts that were independent of each other.

The front crankshaft drove the rear prop, the rear crankshaft drove the front prop.

The rear crankshaft drove the supercharger and accessories. The carburettor was mounted tot the supercharger.

The reduction drives were mounted between the crankshafts and drove the props via extension shafts in the vee.

IIRC the rear half was started first, then the front section.

Mixture distribution was a big problem with the AS.6, and there were problems with back-firing.

The mixture problems were solved in 34 by simulating ram air though fuel injection would have been a far better solution. As a concept this was revolutionary and produced a light strong and powerful engine. Fortunately the Italian powers that be decided liquid cooled engine were a thing of the past and never developed this engine. The much later AS.8 had potential but was way too late to prove itself.

On the plus side the AS.6 engine drove contra-rotating propellers which eliminated a number of flight issues like torque rolls, turning ability and the need for the offset fin which created drag which slowed other aircraft down.

We all tend to forget the Italians because most of their aircraft (except the trophy racers) were designed to compete with the French for how heavily the designer could use an ugly stick and it still fly. AFAICR the Italians never won that competition.

We tend to forget that on their engines they thought way outside the box so developed a lot of interesting engines - many that worked well. If my memory is correct the Italians were the first to build a production engine (as apposed to racing engine) that simultaneously conquered the pound per horsepower and horsepower per cubic inch barriers.

Naturally, like everyone else, they also had their spectacular failures - I seem to remember it was an Italian engine that had the supercharger mounted under the engine instead of behind it thus almost doubling frontal area.

 

[eljefe]

Two things:

1. In all this, is there any preferred cross section for the engine (presumably to reduce drag)? I mean, would we prefer a 90-degree X, or an X with angles of 60-degrees and 120-degrees? If the 60-120 version, would we have the longer dimension upright or lateral (60-degrees on top, or 120-degrees on top)? Same question for an H configuration...sideways like a Sabre or vertical like a Hispanio-Suiza? Etc.

2. On the Fiat, I'm trying to figure out where the shaft passes through the front section of the engine. How can it be co-axial without interfering with the connecting rods as the crankshaft swings round and round?

 

[wuzak]

Two things:

1. In all this, is there any preferred cross section for the engine (presumably to reduce drag)? I mean, would we prefer a 90-degree X, or an X with angles of 60-degrees and 120-degrees? If the 60-120 version, would we have the longer dimension upright or lateral (60-degrees on top, or 120-degrees on top)? Same question for an H configuration...sideways like a Sabre or vertical like a Hispanio-Suiza? Etc.

2. On the Fiat, I'm trying to figure out where the shaft passes through the front section of the engine. How can it be co-axial without interfering with the connecting rods as the crankshaft swings round and round?

The reduction gears are in the middle of the engine.

The extension shafts to the propellers run in the vee of the front half. Similar to motor cannon position on the Daimler-Benz DB 601/605/603. The extension shaft in the rear half runs inside the extension shaft from the front half.

Thus the rear crankshaft drives the front propeller, and the rear crankshaft drives the rear propeller.

https://oldmachinepress.files.wordpress.com/2012/10/fiat-as6-gears-shafts.jpg

FIAT AS.6 Aircraft Engine (for the MC.72)

The FIAT AS.6 engine ultimately produced 3,100 hp (2,312 kW) and was used to propel the MC.72 to 440.682 mph (709.209 km/h). But development of the 24-cylinder engine was plagued with issues and co…

oldmachinepress.com

 

[MiTasol]

Two things:

1. In all this, is there any preferred cross section for the engine (presumably to reduce drag)? I mean, would we prefer a 90-degree X, or an X with angles of 60-degrees and 120-degrees? If the 60-120 version, would we have the longer dimension upright or lateral (60-degrees on top, or 120-degrees on top)? Same question for an H configuration...sideways like a Sabre or vertical like a Hispanio-Suiza? Etc.

2. On the Fiat, I'm trying to figure out where the shaft passes through the front section of the engine. How can it be co-axial without interfering with the connecting rods as the crankshaft swings round and round?

Frontal area is the usual key

For a fighter a narrow engine is desirable as that keeps the fuselage narrow - anything more than the width of the cockpit is waste.

For a bomber probably the same as the flat engine makes a bigger undercarraige bulge but on four engines the horizontal would reduce frontal area..

 

[sotaro]

Seems like the H engine was the most attainable with the least amount of cutting edge technology. Despite the Packard X engine, the gestation of most X engines was long and ended badly. An H engine (with two crankshafts) is more of a packaging and systems development with no cutting edge technology. One crank is better than two, but only if it works as reliably. Given the difficulties most countries/companies faced in securing sufficient engineering staff, limiting the technological hurdles is a way to keep a project on track. As 5 years seems to have been a common development time, trying to meet Tomo's timeline of 1939 requires starting in 1934. A way to save some time is to use the already developed components from the V12, such as the piston, valve, combustion chamber shape, intake and exhaust sizes, timing and valve duration etc. This can be done with both X or H, but the H has the potential to use the head casting general shape, the same bore spacing etc. If independent operation is considered useful, a counter rotating prop hub would be the major technological challenge. Starting with a V12, the VV24 (such as the DB604, 610 or V3420) is definitely easier to develop than an H24. However, the H is a better planform for single engine aircraft than the VV as it is narrower. It is important to keep in mind that the flat 12, as used in an H24 uses a 6 throw crankshaft and is not a boxer engine. Both connecting rods of the paired cylinders share the same journal, just like an automotive V8.

The Chrysler V16 was laid out in the pattern of the Fiat AS.6. It was started way to late to make it into the war, and it suffered from the issues of excessive length. For a single engine aircraft, placing the pilot far back behind the engine and fuel tanks can be a major disadvantage. I enjoyed this discussion.

 

[sotaro]

Another variant not mentioned, nor seen, might have been an inverted V next to an upright V. With the exhaust of open end of the V passing over the crankcase of the other V. Development of the inverted or upright V (depending on what was available in the engines of that country) would be relatively easy, and the rest of the development would be similar to the VV (either upright or inverted). An advantage might be narrowness. A disadvantage might be engine fires given the Luftwaffe experience of the He 177. The crankshafts could be the upper and lower points of the engine rectangle. Less development cost than an H, maybe some challenging packaging. The counter rotating prop would be a benefit, but not a requirement.

 

[MiTasol]

Interesting concept. Could make contra prop easy and height and width would be only marginally larger than the V and H engines. Should be less than the X an W engines. Engine mounts may be an issue though.

 

[wuzak]

Interesting concept. Could make contra prop easy and height and width would be only marginally larger than the V and H engines. Should be less than the X an W engines. Engine mounts may be an issue though.

The width will depend on how much offset there was between the two crankshafts.

The V-3420 had 10-11 inches between its crankshafts. If your upright and inverted vee engines had a similar spacing, side to side, then it would not be wider than an X-engine.

To be narrower this arrangement would need the offset to be vertical, keeping teh width of the V-12, but making the powerplant extremely tall.

 

[wuzak]

The Chrysler V16 was laid out in the pattern of the Fiat AS.6.

It was, in that the power take-off was in the centre of the engine.

However, the Chrysler had a single crank (made of more than one piece), whereas the AS.6 had two independent crankshafts, driving two independent coaxial shafts.

Central power take-off or camshaft drive can be found in many motorcycle designs.

 

[MiTasol]

The width will depend on how much offset there was between the two crankshafts.

The V-3420 had 10-11 inches between its crankshafts. If your upright and inverted vee engines had a similar spacing, side to side, then it would not be wider than an X-engine.

To be narrower this arrangement would need the offset to be vertical, keeping teh width of the V-12, but making the powerplant extremely tall.

Post 112 specified the engine offset so that the exhaust passed over the top of the lower engine (With the exhaust of open end of the V passing over the crankcase of the other V) so something like this if 15 degrees rotation added. Obviously a common crank case would be narrower and by redesigning the oil sump on the top engine and removing it from the bottom engine some height could be removed also. Don't think I would like to work on it but it would probably still be more maintenance friendly than any H or VV.

I should have butchered a cross-section instead of a photo

 

[sotaro]

MiTasol and Wuzak, thank you for your interest in this engine layout idea. I love the photo of the end of the engine rearranged. Here is a diagram I found. I am surprised at how long the counterweights are. The intention is minimal area without excessive development time. I was thinking each V would have a vertical cylinder bank with the lower, inverted V's crankcase above the upper upright V's crankcase, but with clearance for the exhaust. I think that layout would require that none of the crankcase and block castings could be joined for the two engines. If a taller engine were acceptable, a common crankcase could be cast holding the two crankshafts side by side assuring good alignment and saving some weight.

https://www.thetruthaboutcars.com/wp-content/uploads/2014/03/Allison9.jpg

Wuzak, what do you think the weight of the Allison crank, either 6 or 12 counterweights is?

 

[MiTasol]

The length of the counterweights is no greater than the crank throw so the length is not a limiting factor. I do not know what the weight is but it is reasonably large as it is to counterbalance not just the crank throw but also to compensate in part for the rods and pistons.

 

[GrauGeist]

The DB610 had the individual crankshaft centerlines about as close together as you're going to get with coupled/paired engines, regardless of the configuration (positioning of the engines).

Another factor to consider, is the size of the coupler and it's interior components needed to transfer power from the engines to the drive shaft.
In the attached photo of the DB610, you can see the size of the coupler housing in relation to the engines and it's size was relevant to the gear sets as well as drive disconnects for each engine.

 

[sotaro]

The length of the counterweights is no greater than the crank throw so the length is not a limiting factor. I do not know what the weight is but it is reasonably large as it is to counterbalance not just the crank throw but also to compensate in part for the rods and pistons.

I just measured the drawing of the Allison end on. From the outside end of the counterweight to the center of the crankshaft was almost exactly (measuring with a ruler) as the length of the outside of the connecting rod as it is around the crankshaft, to the center of the crankshaft. I was surprised.