In retrospect, were the BMW radial engine developments a mistake?

[z42]

V-16s were generally considered to be heavy for the power developed.

Well, what are the alternatives? Upthread you were lamenting that cylinder volumes are getting awfully big as you're trying to increase the size of a V-12. As sufficient quantities of very high octane aviation gasolines weren't available to Germany, cranking up the boost a la Merlin wasn't an option for them. So the remaining options were to increase volume and/or rpm. A H-24 is perhaps a "natural" design point, but will almost certainly be heavier than a V-16 of equivalent volume, though might compensate by running the smaller pistons at a higher rpm, producing more power. X layouts, well that looks quite risky just looking at the troubled development history of them. I'm suggesting a V-16 could have been a reasonable and technically relatively safe option.

As quick snap shot of this, in 1951 Alfa Romeo was running a 1.5 liter straight 8. They had stated work in 1938 so there was a lot of development. An awful lot of development.
At the end of the 1951 season they were getting 420hp/9,300rpm from a 363lb engine running 98.5% Methanol fuel and 3.10 Atm manifold pressure.
BRM was trying to sort out their 1.5 liter V-16 engine which finally gave them 430 hp/11,000rpm from a 525lb engine. They were using 4.85 Atm pressure.

Now consider that the aircraft engines had economy of scale. They also were running on gasoline and not Methanol. The BRM engine was one of the ones that used the central gear tower and separated the engine into two 750cc V-8s (with a 135 degree angle between the Vs) and again, power was taken from the middle of the engine.

Another racing V-16 were the pre-war Auto Union V-16 race cars: Auto Union racing cars - Wikipedia
This were fairly successful in the pre-war Grand Prix races in Europe. From the table on that wiki page, the final V-16 was a 6L affair producing 620 hp at 5000 rpm and a modest 0.95 bar boost. No idea what fuel was used, except an anecdotal note that drivers were coughing up blood due to inhaling exhaust fumes from the cars on the track.

 

[Shortround6]

Some of the disadvantages of the V-16 were known at the time. They expected a V-16 to need a heavier crankshaft and not just by the amount of the extra length. You need larger diameter main bearings at least because of the torsional stress and vibration. You also need a stronger crankcase for the same reason. This was part of the reason that that some of the V-16s were built as two V-8s. The extra length and weight that resulted from that was thought to be less than trying to build the V-16 the 'normal' way. A 'normal' V-16 would use 9 main bearings. The 'split' V-16s used 10 main bearings. Other engineers may have disagreed with them but the engineers that went with the 'split' thought their way would be lighter.
Now if you are getting 33% more displacement for even 40% more weight and you can't build a big V-12 of the desired power any lighter perhaps the V-16 is a solution.

There is a much older thread on this subject or on the DB 609 16 cylinder engine. The DB 609 was supposed to use the same size cylinders as the DB 603 and not the DB 601/605.

Opportunity lost - DB 16 cyl

It is historic fact that Damlier Benz worked on a 16 cylinder development of the DB603 line of engines under the DB609 type number. Development of this engine was abandonned in April of 1943. From what little I have been able to find out, it was a conventional V16 design with a crankshaft...

ww2aircraft.net

The Auto Union engines were very interesting engines. But they also illustrate some of the trade offs. They also used some solutions that might not scale up very well.

They used one camshaft for the entire engine using pushrods and rocker arms for exhaust valves. On the last engines they were using 75mm pistons and 85mm stroke and two valves in each cylinder. Item 3 in the drawing was the intake passage, no separate intake manifold as we know it. Of interest is that the engine's crankshaft was only partially counterweighted. They got away with it for several reasons however in 1938 when they had to change to a 3 liter engine they responded by going back to the just about the original engine dimensions and cutting out 4 cylinders making a V-12. Which they ran about 2000rpm faster and at the higher rpm they had to use a fully counterweighted crankshaft.
Many large Aircraft engines were not fully counterweighted because the rpm that they ran at didn't require it at the power levels/rpm they were operating at. Sometimes changing the RPM by 100-200 rpm required major changes to the counterweight set up.

BTW the fuel blend used in 1938-39 was 86.0% methanol, 4.4% nitrobenzol, 8.8% acetone and 0.8% sulphuric ether.

 

[don4331]

Depends on what engine type is to be built. If BMW is to make a V12 that is 30-35% or greater displacement than the DB 601 or Jumo 211, and can offer 25-30% more power, then it is worthwhile to shell monies to BMW.



You can note that nobody is suggesting outrageously big cylinders here.



Twin row radial will be operating with just two valves per cylinder, historcally it took more time to get good and efficient supercharges on ww2 radials than on ww2 V12s, and the radial will make greater drag than a decent V12 installation. BMW managed to make fresh air intake either inefficient (that is for 'normal' installations) or draggy (for outside intakes).
BMW 801 consumed more fuel than DB 603 or Jumo 213, let alone DB 605; what it consumed was pricy and not something Germany was awash.

Larger cylinders are a double edged sword. Using RR engines as example, Griffon bore is ~10% larger than Merlin = area ~20% larger. So, I need to make liners 20% thicker to retain cylinder pressure. Which means I have decreased the cooling ability by 20%. So, if engine is limited by how well we can cool it, we haven't gained anything. Which is why Griffon 65 is only making ~15% more power (basically difference in stroke) over Merlin 66 (2,300 vs 2,000hp on 150 octane) while weighing ~250kg more (1 ton vs 750kg).

Then you get into details like does your larger engine react as quickly to throttle position changes, critical in fighters.

 

[tomo pauk]

Larger cylinders are a double edged sword. Using RR engines as example, Griffon bore is ~10% larger than Merlin = area ~20% larger. So, I need to make liners 20% thicker to retain cylinder pressure. Which means I have decreased the cooling ability by 20%. So, if engine is limited by how well we can cool it, we haven't gained anything. Which is why Griffon 65 is only making ~15% more power (basically difference in stroke) over Merlin 66 (2,300 vs 2,000hp on 150 octane) while weighing ~250kg more (1 ton vs 750kg).

Then you get into details like does your larger engine react as quickly to throttle position changes, critical in fighters.

150 grade fuel was a luxury that ww2 Germany was ill able to afford. We can also note that Germany decided that big engines were the way to go, and that RAF was favoring the heavy Griffon over the lighter Merlin for the late Spitfires and Spiteful. Probably had a lot to do with Griffon-powered Spitfires having a 30-40 mph speed advantage over the Merlin-powered Spitfires.

 

[BarnOwlLover]

One thing to note about the V-16s. When Vittorio Jano designed the Alfa Romeo straight 8 engines (basically half a V-16), he designed them as two four bangers joined by a gear train. This was to prevent crank and camshaft flex issues.

 

[z42]

Larger cylinders are a double edged sword. Using RR engines as example, Griffon bore is ~10% larger than Merlin = area ~20% larger. So, I need to make liners 20% thicker to retain cylinder pressure. Which means I have decreased the cooling ability by 20%. So, if engine is limited by how well we can cool it, we haven't gained anything.

Yes, I'm sure we can all agree that engine design is all about compromises and tradeoffs. As an aside, I think the limitation with big cylinders is more about piston speed and potentially flame propagation speed, at least as long as we're talking about liquid cooling. For air cooled, yes, cooling is certainly a very critical issue that gets way worse with bigger cylinders.

So bigger volume has problems in keeping a good power/weight ratio, be it due to above issues with bigger cylinders, or requiring heavier-than-V-12 construction with more cylinders and otherwise keeping it the same as a "traditional" WWII aero V-12.

And availability of very high high octane aviation fuel limited how much Germany was able to increase the effective volume by increasing boost. As an aside, how come current F1 engines can run at apparently up to 500 kPa (5 bar) boost with very close to RON 95 gasoline you can find at any gas station in most of the world (or AKI around 90 for you Americans)?

So lets say BMW, in this hypothetical scenario that is the topic of this thread, wants to make a liquid cooled engine a step more powerful than DB 601/605 or Jumo 211 to be introduced into service around, say, 1940 or so. Target power could be around 1200 kW (1600 hp) for the initial version, with some headroom for more powerful versions later in the war (say around 1700 kW / 2300 hp). So what about going in a radical way for that third way of increasing power, namely increasing rpm. Yes, this is totally crazy but lets do some napkin math for fun. So a typical mid-late war engine (DB 605) produced around 35 kW/L at 2800 rpm. Now something like 7000 rpm wasn't unheard of in pre-war auto racing engines. So lets aim for 6000 rpm. 6000/2800*35 = 75 kW/L. Aiming for 1200 kW we need a capacity of a mere 16L! However at 6000 rpm I'm quite sure we need each cylinder to be quite small, can't use those dinner plate size pistons of typical aero engines. So let's use a, say, H-24 configuration. And if we slightly increase the volume to 18L to give some headroom we get 0.75L/cylinder. Which is slightly on the large side compared to current day auto racing engines, but not completely outlandish either.

If we assume a square engine, 0.75L cylinder gives us a bore and stroke of 98.5mm. At 6000 rpm that gives us a mean piston speed of 19.6 m/s (3860 fpm). High but not completely out of the question either (Jumo 213J at 3700 rpm had a MPS of 20.35 m/s or 4006 fpm).

Ok, I'm slightly worried about the MPS. So lets iterate a bit and make the engine slightly oversquare, say a stroke/bore ratio of 0.8. That gives us a bore of 106 mm and a stroke of 85 mm, and hence a MPS of 17 m/s (3350 fpm).

 

[Shortround6]

The German designers of the 1930s were not dumb. They took a certain path (large displacement and low revs) for what they considered good reasons.
Please note that the aircraft designers didn't care what displacement the engines had. What they cared about was the weight and the physical size (length/height/ width) of the engines.
Strangely enough (or not so strange) the early Allison (C-15), the Merlin II, the DB 601A and the Jumo 211 all weighed within 42lbs from heaviest to lightest. They tended to trade places form take off power to power at altitude and the FTH altitude varied by 3000ft from highest to lowest. I think it is safe to say that they were actually very close to one another in capability. Things changed latter but often not as much as it is sometimes thought. They were not on quite the same schedule though.
The Germans knew that a larger but slower revving engine could be built for the same weight. There was also the question of bearings and/or lubrication and piston speed ( piston rings) Engine design is not just the actual design of the engine, it is the ability to actually make the engine in quantity at a decent price. This goes back to the support each engine maker could get from all the support industries, like casting and forging. Or bearing suppliers. For example Allison found a new way of casting the cylinder blocks pretty much by accident. Somebody put them in touch with a husband and wife team of sculptors that were making cast aluminum sculptures using their own casting process. This allowed for either stronger for the same weight or lighter for the same strength cylinder blocks/crankcases. Alcoa had failed several attempts at coming up with better castings.
Allison was making much more money selling bearings to major engine companies during the 20s and 30s than they were making trying to sell engines.
Somebody once said " you use roller bearings when you don't trust your plain bearings".
Bristol was using forged cylinder heads on radials when P & W was using castings, P & W changed to forgings later as the power went up but at some point in the 30s the US foundries could make better castings than the British could which in turn affected the choices that P & W and Wright could make in materials/design.

The path to high rpm had several problems.
One was the piston speed.
Another was the simple fact that the friction went up the square of the speed. Double the speed of the sliding surface (piston ring or surface of a main bearing) and you had 4 times the friction.
Going right along with that was that the stresses did the same thing. All the rotating and reciprocating parts would have 4 times the stress at double the speed.

Yes you can use smaller parts to keep the actual speed down while using high rpm.

 

[don4331]

I used 150 octane fuel as it was the end of the line for development for all intents.

My point was a larger cylinder might wind up making absolutely no extra power because: flame propagation/cooling/piston speed/material limits might simply result in larger/heavier package.

The fuel in F1 might have same RON number but it isn't same stuff you buy at the pump; it is a very carefully created blend with much greater tolerance to pre-ignition. But the biggest factor allowing high boost is direct injection - there is no fuel in the engine to pre-ignite.

How many pre-war racing engines lasted 100 hours at full power? Continuing your F1 example, engines aren't running 75% open throttle over course of a race. In order to make 100% power, you must be at rpm for peak power, throttle wide open and load equal to power. The engines last 7 races? @ <2 hours race (maybe 3 per weekend including qualifying and practise) = 20 hours. Which is long ways from what your airplane needs.

 

[Shortround6]

One big difference (among many) is that most WW II aircraft engines used fixed ignition timing. The spark plugs fired the point if the engine was idling or if it running at full speed.
Some engines did use a starting delay. Once the engine started and was idling the engine went to full advance. Some people were fooling around with variable ignition timing at the end of the war.

There is a lot of things going on in background. This is a selected sample (cherry picked if you will)
The 1938-39 Auto Union 3 liter V-12 used a piston speed (corrected) of 3,207fpm.
The Bristol Pegasus XVIII 28.7 radial 9 used a piston speed (un-corrected) of 3,250fpm.
Jumo 213 at 3250rpm had a piston speed ((un-corrected) of 3,519fpm.
Most of the engines in world were running at 3000fpm or under (un-corrected)
The RR Vulture at 3000rpm was running at 2750 piston speed ((un-corrected)

Corrected piston speed is a way of factoring in over square cylinder dimensions and under square cylinder dimensions. Every engine listed used under square dimensions but the Pegasus used the most under square of all, a bore 0.768 times the length of the stoke.

 

[don4331]

Given most WWII engines operated behind a constant speed propellers so operated at basically same speed all the time, fixed ignition timing isn't big deal.

 

[Shortround6]

Given most WWII engines operated behind a constant speed propellers so operated at basically same speed all the time, fixed ignition timing isn't big deal.

109E could run at 1300rpm for long range cruise and 2400rpm for full power. Could run at 2500rpm emergency power at times.
Most other engines showed similar speed ranges.
SBD-3 idled at 600-800 rpm, could cruise at 1400-1600rpm and full power at 2300rpm.

Fixed ignition timing was a compromise. It worked but was not ideal. Also a difference between aircraft engines and car engines.

 

[Engineman]

One big difference (among many) is that most WW II aircraft engines used fixed ignition timing. The spark plugs fired the point if the engine was idling or if it running at full speed.
Some engines did use a starting delay. Once the engine started and was idling the engine went to full advance. Some people were fooling around with variable ignition timing at the end of the war.

Hi,
Many German combat engines used variable ignition timing in the power range, not just for starting.
Cheers

Eng

 

[Engineman]

Given most WWII engines operated behind a constant speed propellers so operated at basically same speed all the time, fixed ignition timing isn't big deal.

Hi,
Just to note that the German combat engines had critical variable ignition timing.
Cheers

Eng

 

[z42]

The German designers of the 1930s were not dumb. They took a certain path (large displacement and low revs) for what they considered good reasons.

Where am I claiming they were dumb? I'm sure they were brilliant, and knew more about engine design than I ever will (in case it's not clear, I'm not a professional engine designer). My only advantage is the benefit of 80 years of hindsight, and comparatively easy access to mountains of information written about engine design since, both in print and online.

In this thread I'm merely engaging in some light-hearted speculation about alternative history paths, had BMW not been forced to pivot to radials.

Please note that the aircraft designers didn't care what displacement the engines had. What they cared about was the weight and the physical size (length/height/ width) of the engines.

I'm sure we're all aware there were no London and Washington aeronautical treaties limiting aero engine displacement.

In my above napkin math calculations the volume was merely used as a tool to characterize a hypothetical high rpm engine targeted at a specific power. Not as a goal in itself.

The path to high rpm had several problems.
One was the piston speed.
Another was the simple fact that the friction went up the square of the speed. Double the speed of the sliding surface (piston ring or surface of a main bearing) and you had 4 times the friction.
Going right along with that was that the stresses did the same thing. All the rotating and reciprocating parts would have 4 times the stress at double the speed.

Yes you can use smaller parts to keep the actual speed down while using high rpm.

Yes, there are challenges, I'm not denying that. But partially the increased stresses and friction from increasing rpm are compensated by the higher rpm allowing the engine to be smaller with a shorter stroke, which reduces the piston speed and centrifugal stresses.

In a way, one can see the Sabre as a development in the high rpm direction, though obviously not nearly as radical as my napkin proposal above. Sabre had about the same volume as the Griffon, but ran 1000 rpm faster, I suspect largely due to having pistons half the size and shorter stroke.

For a modern example, see the Rotax 900 series engines for light aircraft. 5800 rpm takeoff, 5500 rpm max continuous, 2000h TBO. Fuel consumption slightly better than the slow revving direct drive large displacement Lycoming /Continental they compete with. Keep in mind these Rotax engines, while current day, are designed to be cheap and mass produced, not pushing the envelope like a WWII military engine or a racing engine.

The latest and brand spanking new Rotax 916iS BTW produces about 85 kW/L, slightly higher than the 75 kW/L my above napkin proposal has.

 

[wuzak]

Another factor in moving to lower displacement, higher rpm engines is the size of the reduction gears required to get the output speed to suit the propeller.

There must come a point where the reduction in engine size is offset by the increase in gearbox size/weight.

 

[tomo pauk]

I used 150 octane fuel as it was the end of the line for development for all intents.

Roger that.
However, this thread is about what can be done by 1940-43 in the ww2 Germany, not what can be done in the UK or USA by 1944-45.

My point was a larger cylinder might wind up making absolutely no extra power because: flame propagation/cooling/piston speed/material limits might simply result in larger/heavier package

Reality was that bigger cylinders were making better power in ww2, provided same technology level, fuel and ADI (if used), engine layout and manufacturer.

 

[Deleted member 68059]

One big difference (among many) is that most WW II aircraft engines used fixed ignition timing. The spark plugs fired the point if the engine was idling or if it running at full speed.
Some engines did use a starting delay. Once the engine started and was idling the engine went to full advance. Some people were fooling around with variable ignition timing at the end of the war.

There is a lot of things going on in background. This is a selected sample (cherry picked if you will)
The 1938-39 Auto Union 3 liter V-12 used a piston speed (corrected) of 3,207fpm.
The Bristol Pegasus XVIII 28.7 radial 9 used a piston speed (un-corrected) of 3,250fpm.
Jumo 213 at 3250rpm had a piston speed ((un-corrected) of 3,519fpm.
Most of the engines in world were running at 3000fpm or under (un-corrected)
The RR Vulture at 3000rpm was running at 2750 piston speed ((un-corrected)

Corrected piston speed is a way of factoring in over square cylinder dimensions and under square cylinder dimensions. Every engine listed used under square dimensions but the Pegasus used the most under square of all, a bore 0.768 times the length of the stoke.

This is incorrect regarding the timing, even Jumo210 has variable ignition timing. Usually it was done by oil pressure. (210 G manual)

Early timing is 40 degrees before TDC

Late timing is 10 degrees after TDC

Jumo 211 manual diagram below to show it more easily>

Merlin >

Centaurus >

 

[Simon Thomas]

Taurus and Hercules also have variable timing.

 

[tomo pauk]

BMW taking up the P & W engine seems hard to pin down. P & W shipped an engine to BMW in May 1933 but that certainly doesn't prove a license agreement. It appears that the Ju 52 used a succession of BMW engines that started as licensed copies. Aside from the JU-52s the BMW 132 doesn't seem to have gone into much until 1936-37.

(guess this is a better location to discuss BMW engine stuff)

Licence deal was signed on January 3rd 1928, both for Wasp and Hornet, to be produced by BMW and sold across Europe (bar the UK). The deal was not supported by state funds, and it was not received well by the government; BMW managed to pull this due to their well-off financial situation due to export of their engines.
The BMW Hornet passe the type test by March 28th 1930.
BMW bought 100 of Hornet engines at P&W (1.7 millions RM total price) in order to supply them to the airframe manufacturers until their own production started.
By 1930, however, BMW was not exporting engines as good as before (eg. Soviet Union - previously the best costumer for BMW - showed no interest in the Hornet), and they have had no radial engines to sell abroad yet, so the licence agreement was cancelled by 1931. 'Replacement' deal was made, to cover just the German market.

Source: "BMW Flugtriebwerke" book, pg. 93-97

The deal to outfit the Ju 52 with BMW 132s was a life line for BMW.

 

[Shortround6]

In my above napkin math calculations the volume was merely used as a tool to characterize a hypothetical high rpm engine targeted at a specific power. Not as a goal in itself.

See Napier and a few others.

In a way, one can see the Sabre as a development in the high rpm direction, though obviously not nearly as radical as my napkin proposal above. Sabre had about the same volume as the Griffon, but ran 1000 rpm faster, I suspect largely due to having pistons half the size and shorter stroke.

It weighed as much as a Merlin, made 1000hp at 8,750ft. the 8,000ft difference is important. Granted the air cooling did limit the power.
But look at the Sabre (same designer) again. The Sabre VA went 2500lbs while the Griffon VI was 1790lbs. The engines were remarkably close to each other in BMEP, Identical in power to weight and nearly identical in piston speed (uncorrected).
You will notice I keep sticking corrected and uncorrected piston speed in. Corrected piston speed is figured by dividing the mean piston speed by the square root of the stroke/bore ratio. This approximates the difference between over square and under square engines. Over square engines may still be better, just not as much as just plain piston speed would leave us to believe. Without actually weighing pistons and rods this was an approximation. Fat wide bore pistons weigh more than skinny pistons and put more stress on things.

For a modern example, see the Rotax 900 series engines for light aircraft. 5800 rpm takeoff, 5500 rpm max continuous, 2000h TBO. Fuel consumption slightly better than the slow revving direct drive large displacement Lycoming /Continental they compete with. Keep in mind these Rotax engines, while current day, are designed to be cheap and mass produced, not pushing the envelope like a WWII military engine or a racing engine.

It it tough comparing engines of different time periods. Metallurgy, manufacturing processes and oil/lubrication changed tremendously even in 10 years let alone 60 years.
Valve springs made a large advances from the 1920s to 1940, they made more advances ever since. Changes in springs, push rods, rocker arms has kept pushing the upper limits of valve train technology. We do have to make sure that we are comparing the same things. In the late 60s US V-8 engines were reaching rpm 40-50% higher than what they were doing in the mid 50s. however drag racing rpm limits were often 1000-1500rpm higher than the same engine would use in a 3-4 hour road race. Changes in cam ramp profiles also helped even with the same total lift and duration.
From the car racing world in 1956 Maserati was using 60 weight oil in their 2 1/2 liter Grand Prix engine, BP produced a 40 weight oil for the engine but added anti scuff compound to handle the high loadings between the cam lobs and finger followers. The main bearings had taken care of by quadrupling the the size of the oil gallery feeding the mains to improve flow and provide a reserve of cool oil near the mains.

In the late 40s and though a lot of the 1950s Lycoming was making high rpm engines, relatively speaking. They were using gear reduction on flat fours and superchargers on some the sixes and the eight. Then they just started making the cylinders bigger and the high revving engines went away. The reduction gears went away and the superchargers went way.
The engines lasted longer and they sometimes cost 1/2 as much to overhaul and they could run on lower cost fuel. The 0-480 six became an 0-540 six.
Turbos began to replace gear and belt dive superchargers by the early 60s.

Using modern (or even not so modern) examples to show what they could have done in WW II often gives a distorted picture.