Build the perfect air-cooled engine

[Readie]

Most Herefords did not see combat (if any actually did?)

Some did apparently in 1940 -41.

Handley Page H.P.53 Hereford - bomber

Albeit with little success.
Cheers
John

 

[Shortround6]

When all is said and done the ultimate air-cooled engine is probable going to be an 18 cylinder 2 row radial of about 53-57 litres displacement. supercharger to suit employment and Sleeve or poppet valve to suit initial cost or maintenance requirement.

A VERY interesting article can be found here: http://www.flightglobal.com/pdfarchive/view/1937/1937 - 0507.html?search=Fedden

While everything in it may not be 'gospel' it does show what at least one great engine designer was thinking and why at the time. It also shows that multiple layouts were at least being considered although perhaps with a view of justifying their own position?

 

[Trilisser]

A most interesting article! BTW, I wonder how would it pan out if the ultimate air cooled radial was a CI engine with pure turbocharging...

 

[basil]

Ad ultimate aircooled aircraft piston engine - a very interesting project was the Junkers O-32 (P 130) project which was designed 1945/1946 (already under soviet occupation) by Junkers engineers under the leadership of Dr. Cordes and Dr. Scheibe. It was an aircooled two stroke compression ignition radial in two versions - one 10 cylinder double row radial (34 liter) and one 12 cylinder double row radial (36 liter), each with about 4.000 hp. It would have powered a fan (derived from the Jumo 012 turbojet) instead of a propeller without reduction gearing at a speed of 6.600 rpm - resultig in a thrust of 2.000 kp (19,6 kN) and was proposed to have a fuel consumption of less than a third compared to the Jumo 012. Work was terminated in 1947 because of work on turboprops which finally resulted in the Kusnezow NK-12.

 

[wuzak]

34l/36l, 4000hp, 6600rpm Diesels seem a bit far fetched for WW2 time frame.

 

[merlin]

The Deerhound is an interesting engine, though by making it a three-row you'd have thought 'cooling problems' could have been anticipated!
Having said that the performance of the Mk I, is that different to what happened to the Hercules - similar power target, with the initial one being lower than expected - what happened - it took till Mk III to get there!

Might it have helped focus their minds if the Air Ministry had stipulated either engine. Imagine Bristol's loss of 'face' if they were forced to have Deerhounds in a Bristol aircraft instead of an under performing Hercules.

 

[tomo pauk]

Wonder how well the Homare would've been doing, built maintained by Western standards, while using 'good' fuel. And then install some auxiliary supercharger stage, maybe a fan to cool it while the cowling is made tight...

 

[davebender]

How can any piston engine be considered "perfect" if it arrives too late to matter?

An acceptable air cooled engine will be reliable and in service NLT 1942 (i.e. WWII half over). A perfect air cooled engine will be reliable and in service during 1939.

 

[Shortround6]

"A perfect air cooled engine will be reliable and in service during 1939"

And will have been designed to run on 87 octane fuel, no water injection and using 1938/39 manufacturing techniques. In other words, totally obsolete by 1944.

The Wright R-1820 was giving about 1100hp for take-off in 1939. By 1945 it was good for 1425hp for take-off and post war went to 1525hp. Not only did the fuel change but so did just about everything else EXCEPT the bore and stroke. Different crankcase and crankshaft, different pistons, different cylinder construction, different cylinder head construction, different number of cylinder hold down bolts, and so on. Engine weight changed by around 200lbs.

Versions of this late model air cooled engine powered an awful lot of helicopters (several thousand) but I guess it was far from perfect because the jet engine was already on the scene?????

Piston engines powered the majority of the worlds transport planes for a good decade after WW II ended. Last Lockheed Constellation was rolled in 1958, last scheduled passenger airline flight ( not charter or cargo) was in 1967 as just one example.

 

[tomo pauk]

Perfect engine means that timing is right, among other stuff. Maybe Dave was having that in mind? Even the engines that were in service later than 1939 (R-2800, BMW-801, Hercules) received plenty of upgrades, if not complete redesigns, by 1945.

 

[Shortround6]

That they did but Mr Bender seems to be saying that any aircooled aircraft engine after 1939 was less than "perfect" because it won't be in "time". For "what" he isn't saying.

Nobody was using the air cooled engines in fighters in 1944/45 that they were using in 1939.

And while the jet engine was an indicator of the future it was far from perfect in 1944/45 for fighters and far, far, far, from perfect for many other aircraft propulsion duties.

 

[riacrato]

There's no such thing as one perfect engine, radial or inline- It all depends on the applications you're going for. A 14 cylinder with ~40l displacement might be good for a lot of applications but will run into limitations and is inefficient when the plane gets really big. And an 18 cylinder or even larger engine can be overkill for a smaller aircraft.

 

[wuzak]

The Deerhound is an interesting engine, though by making it a three-row you'd have thought 'cooling problems' could have been anticipated!
Having said that the performance of the Mk I, is that different to what happened to the Hercules - similar power target, with the initial one being lower than expected - what happened - it took till Mk III to get there!

Might it have helped focus their minds if the Air Ministry had stipulated either engine. Imagine Bristol's loss of 'face' if they were forced to have Deerhounds in a Bristol aircraft instead of an under performing Hercules.

Armstrong-Siddeley wanted the Deerhound to be liquid cooled. The Air Ministry wanted an air-cooled radial manufacturer to compete with Bristol, so insisted that the Deerhound be air cooled.

Radials with more than 2 rows faced cooling problems, regardless of whether the rows were staggered or not. The solution used for the R-4360 was shrouds to guide air around cylinders, and that should work for the Deerhound. The solution Armstrong-Siddeley chose was reverse flow cooling.

 

[tomo pauk]

There's no such thing as one perfect engine, radial or inline- It all depends on the applications you're going for. A 14 cylinder with ~40l displacement might be good for a lot of applications but will run into limitations and is inefficient when the plane gets really big. And an 18 cylinder or even larger engine can be overkill for a smaller aircraft.

I agree that a big engine can make the job harder for the airframe designers; hard, but not impossible to build a not-so-big airplane. The renown Bearcat was an example of big engine in a modestly sized airframe. But even the Fw-190 was carrying around some big powerplant, on a wing area smaller than of P-51, let alone Spitfire. We also know the other small airframes with big engines, eg. La series (900 kg dry weight of engine), or maybe Yak-3U, or that 109 prototype with BMW 801.
The ~40L engine can still provide plenty of power, if it has 2 stage spercharger, while using good fuel. Even more power should make 18 cyl engine of such a displacement, smaller pistons mean more RPM.

As for the airplanes that are getting really big, how big would that be? The Lanc, Stirling Halifax were flying with radials displacing under 39 liters, the engines being humble single stage ones. Many German US planes were using 41 or 42 liter radials, again single stagers, to a good effect. If those engines were running into limitations, that was only at high altitude, because none of them were using multi-stage supercharging.

 

[wuzak]

As for the airplanes that are getting really big, how big would that be? The Lanc, Stirling Halifax were flying with radials displacing under 39 liters, the engines being humble single stage ones. Many German US planes were using 41 or 42 liter radials, again single stagers, to a good effect. If those engines were running into limitations, that was only at high altitude, because none of them were using multi-stage supercharging.

The "under 39l" engine would be the 2360ci Hercules.
The "42l" radial would be the R-2600.

Note that the B-17 and B-24 both used radials under 30l - the R-1820 and R-1830. These, too, were single stage units but equipped with turbochargers.

 

[wuzak]

How useful would the Deerhound IIi have been?

About 1800hp, 44" (1118mm) in diameter, around 41l (2500ci). Basically the same power class as an R-2600, but much smaller frontal area. Maybe a bit heavier, though.

 

[riacrato]

I agree that a big engine can make the job harder for the airframe designers; hard, but not impossible to build a not-so-big airplane. The renown Bearcat was an example of big engine in a modestly sized airframe. But even the Fw-190 was carrying around some big powerplant, on a wing area smaller than of P-51, let alone Spitfire. We also know the other small airframes with big engines, eg. La series (900 kg dry weight of engine), or maybe Yak-3U, or that 109 prototype with BMW 801.
The ~40L engine can still provide plenty of power, if it has 2 stage spercharger, while using good fuel. Even more power should make 18 cyl engine of such a displacement, smaller pistons mean more RPM.

As for the airplanes that are getting really big, how big would that be? The Lanc, Stirling Halifax were flying with radials displacing under 39 liters, the engines being humble single stage ones. Many German US planes were using 41 or 42 liter radials, again single stagers, to a good effect. If those engines were running into limitations, that was only at high altitude, because none of them were using multi-stage supercharging.

When I said limitations I was thinking about B-29s, Me 323s and the likes.

Also, while a big radial can be fit into a smaller airframe, why would I power my C-47-class aircraft with a high-class and expensive R-2800 when in the end an inexpensive R-1830 will do just as fine. Granted the engine costs is not all in that decision, but look at C-46 vs C-47.

 

[tomo pauk]

The "under 39l" engine would be the 2360ci Hercules.
The "42l" radial would be the R-2600.

Indeed

Note that the B-17 and B-24 both used radials under 30l - the R-1820 and R-1830. These, too, were single stage units but equipped with turbochargers.

That would technically qualify them as a two stage engines - engine stage + auxiliary (turbo, in this case) stage.

When I said limitations I was thinking about B-29s, Me 323s and the likes.

Okay, I see what you mean.
If, for example, US have had the 2600 cu in, 18 cyl engine, it would've been, in turbo installations, a 1800 HP in early war, 2000 HP in 1943, and maybe up to 2500 HP in early 1945. More than enough to carry the Grand Slam, let alone the nuke, in a 4 engined plane. The engine being small enough to power a fighter of modest size?

Also, while a big radial can be fit into a smaller airframe, why would I power my C-47-class aircraft with a high-class and expensive R-2800 when in the end an inexpensive R-1830 will do just as fine. Granted the engine costs is not all in that decision, but look at C-46 vs C-47.

Maybe your transport aircraft with 2 big engines would be able to carry more payload at greater distances, with less pilots needed per ton-mile?

 

[Shortround6]

C-46 could carry about double the payload and carry it further.

 

[wuzak]

If, for example, US have had the 2600 cu in, 18 cyl engine, it would've been, in turbo installations, a 1800 HP in early war, 2000 HP in 1943, and maybe up to 2500 HP in early 1945. More than enough to carry the Grand Slam, let alone the nuke, in a 4 engined plane. The engine being small enough to power a fighter of modest size?

Not sure that is a safe assumption Tomo.

R-2600 only got up to around 1900hp (ok, so it is a 14 cylinder engine). The R-3350 really only got to 2200hp by the end of the war.

In contrast, some special versions of the R-2800 were able to get 2800hp by the end of the war, but that was really stretching things.