Realistic limits of the small 12 cyl engines, 1935-45

[tomo pauk]

... and by small 12 cyl engines in think of engines that are a good deal smaller then the small, but firmly mainstream V12 engine - the RR Merlin. So engines like the Kestrel/Peregrine, Jumo 210, small Fiat, HS and Curtiss engines (for the sake of discussion let's assume the siblings of these are still being developed in this time frame).

Yes, we will not be getting the equivalent of the 2-stage RR Griffon or the Jumo 213E here.

As-is, RR gotten the most from this, with Peregrine making around 1000 HP with the 100 oct fuel. Some of the engines, like the Jumo 210, will need to get to 3000, if not to 3200 rpm in order to compete. Being small, and looking at what Jumo were doing with the 213 series, even 3300 rpm does not look outlandish come mid-ww2.
Better S/Cs will also be needed, especially if the RPM range remains modest, talk 3000. Dry engine weight will easily approach 500 kg here.

 

[GrauGeist]

Are we talking small like the Ranger V-770 or the Packard V-1410?

 

[don4331]

I was thinking about this in conjunction with what would be possible with modern material/modern analysis:

Sticking with RR Peregrine size: 5.0" bore/5.5" stroke = V-1295 (in American terms): 960 hp @ 3,000 rpm, ?9? psi boost:

Torque (hp*5252/rpm) = 1,680 ft.lbs works out to average BMEP of ~90 psi​

​

Merlin 71: 5.4" bore/6.0" stroke = V-1650: 1710 hp @ 3,000 rpm, 18 psi boost:

Torque = 2,990 ft.lbs works out to average BMEP of ~130 psi​

If RR had applied all the improvements to the Peregrine which were done to the Merlin

Back figuring the 130 psi BMEP = 1,460 hp...not too shabby.​

But wait, those boys at Junkers got the Jumo 213E up to 3,250 rpm; for the 213J to 3,700 rpm

If RR can hold the BMEP while adding 250/700 rpm respectively, the "hyper" Peregrine will have 1,590/1,810 hp​

We will note the stroke of the Peregrine is ~9% shorter than Merlin, so 250 rpm increase (8%) - basically same max piston speed (there are vibration issues that need to be analyzed and addressed.)​

This version of the Peregrine isn't going to be substantially lighter than a Merlin - it needs basically the same supercharger/intercooler and same reduction gear, and it needs material in the block/heads/etc to survive that power. Now, where did the tooling for the Whirlwind go?

​

If you could go to the same piston speed as the Jumo 213J with its 165mm stroke, you could spin your Peregrine to 4,370 rpm! Then, same BMEP of the Merlin 130 25psi boost = 160 psi BMEP. And it would be making >3,100hp - taking names and kick @$$.

 

[tomo pauk]

Are we talking small like the Ranger V-770 or the Packard V-1410?

Probably the smallest being the Jumo 210 (19L), and the biggest being the Fiat A.30 lineage (24.5 L, or 1464 cu in).

 

[Shortround6]

I was thinking about this in conjunction with what would be possible with modern material/modern analysis:

Sticking with RR Peregrine size: 5.0" bore/5.5" stroke = V-1295 (in American terms): 960 hp @ 3,000 rpm, ?9? psi boost:

Torque (hp*5252/rpm) = 1,680 ft.lbs works out to average BMEP of ~90 psi
​

Merlin 71: 5.4" bore/6.0" stroke = V-1650: 1710 hp @ 3,000 rpm, 18 psi boost:

Torque = 2,990 ft.lbs works out to average BMEP of ~130 psi​

If RR had applied all the improvements to the Peregrine which were done to the Merlin

Back figuring the 130 psi BMEP = 1,460 hp...not too shabby.​

But wait, those boys at Junkers got the Jumo 213E up to 3,250 rpm; for the 213J to 3,700 rpm

If RR can hold the BMEP while adding 250/700 rpm respectively, the "hyper" Peregrine will have 1,590/1,810 hp
We will note the stroke of the Peregrine is ~9% shorter than Merlin, so 250 rpm increase (8%) - basically same max piston speed (there are vibration issues that need to be analyzed and addressed.)​

This version of the Peregrine isn't going to be substantially lighter than a Merlin - it needs basically the same supercharger/intercooler and same reduction gear, and it needs material in the block/heads/etc to survive that power. Now, where did the tooling for the Whirlwind go?

​

If you could go to the same piston speed as the Jumo 213J with its 165mm stroke, you could spin your Peregrine to 4,370 rpm! Then, same BMEP of the Merlin 130 25psi boost = 160 psi BMEP. And it would be making >3,100hp - taking names and kick @$$.

I think there is a mistake in your math?
Something seems to be off by a factor of 2?
Peregrine at 885hp is listed in old book/s at 180 BMEP.
Merlin X 1130hp (?) is 181.
Jumo 210 Da 700hp is 170.

Watch the increases in rpm. Stresses go up with the square of the speed.

 

[don4331]

I think there is a mistake in your math?
Something seems to be off by a factor of 2?
Peregrine at 885hp is listed in old book/s at 180 BMEP.
Merlin X 1130hp (?) is 181.
Jumo 210 Da 700hp is 170.

Watch the increases in rpm. Stresses go up with the square of the speed.

I might be: 4 cycle, not 2 cycle oops; but it is is consistent, so the power numbers remain same.

I'm aware of the stresses, RR would need to redesign connecting rods as per the Jumo 213J. tomo pauk asked for the limit, not what was easy.

 

[Shortround6]

... and by small 12 cyl engines in think of engines that are a good deal smaller then the small, but firmly mainstream V12 engine - the RR Merlin. So engines like the Kestrel/Peregrine, Jumo 210, small Fiat, HS and Curtiss engines (for the sake of discussion let's assume the siblings of these are still being developed in this time frame).

Yes, we will not be getting the equivalent of the 2-stage RR Griffon or the Jumo 213E here.

As-is, RR gotten the most from this, with Peregrine making around 1000 HP with the 100 oct fuel. Some of the engines, like the Jumo 210, will need to get to 3000, if not to 3200 rpm in order to compete. Being small, and looking at what Jumo were doing with the 213 series, even 3300 rpm does not look outlandish come mid-ww2.
Better S/Cs will also be needed, especially if the RPM range remains modest, talk 3000. Dry engine weight will easily approach 500 kg here.

We can look at few other options/guides.
Like the Lycoming O-1230 (20.2 liters) and the Continental XI-1430. We might also consider taking a power saw to the Napier Sabre and cutting down to a 12 cylinder engine.

Many new engines were designed as single cylinder or 2 cylinder test rigs. 2 cylinder test rig allows for testing of the connecting rod/s and crank journals. I am not concered about if these were flat engine or Vs or which direction the V pointed in, Once you have a running engine then bending it was not that hard.

I am not saying that these engines were successful. They did exist at the time and were designed for high output per unit of displacement. Achieving that was a bit harder (sarcasm).
But they also bring up some problems.

Single cylinder from the O-1230. The studs for fastening on the cam box/rocker cover/s are the top. Using single/paired test cylinders cut development costs and did make things a bit cheaper making a prototype 12 cylinder engine. However it also pointed out some problems the older V-12 engines had. There was a lack of rigidity with 12 separate cylinders instead of 2 blocks of 6 cylinders. The crankcase and crankshafts were longer to fit the separate water jackets (which added to the vibration problems). Also meant that the crankcase and crankshaft was heavier than for a Monoblock engine. We can debate about if it reached the designed 1200hp at 3400rpm and a 50 hour test is not a 150 hour standard US test.
What is harder to gloss over is the 1325-1342lb weight for level supercharged engine and the need for a turbo and intercooler to meet the listed 1200hp at 25,000ft.
The engine was offering better power per liter than the much larger Allison but it was not offering any better power to weight. And it was longer (CG problems? in existing designs)

I am critical of the H-S Y engines but compared to a Peregrine the differences are minor in effective performance. Peregrine 885hp at 15,000ft (4550 meters) for 502kg vs 12Y-31 = 860hp at 3250metes for 468 kg or the 12Y-45 that offered 920hp at 4200 meters for under 500kg.

Small engines can often offer better power per unit of displacement (liter). The Real trick if they can offer better power kilogram.

 

[tomo pauk]

Single cylinder from the O-1230. The studs for fastening on the cam box/rocker cover/s are the top. Using single/paired test cylinders cut development costs and did make things a bit cheaper making a prototype 12 cylinder engine. However it also pointed out some problems the older V-12 engines had. There was a lack of rigidity with 12 separate cylinders instead of 2 blocks of 6 cylinders. The crankcase and crankshafts were longer to fit the separate water jackets (which added to the vibration problems). Also meant that the crankcase and crankshaft was heavier than for a Monoblock engine.

Thank you.
For the purposes of this thread, we will assume that the 12 cyl engines are of monobloc construction.

I am critical of the H-S Y engines but compared to a Peregrine the differences are minor in effective performance. Peregrine 885hp at 15,000ft (4550 meters) for 502kg vs 12Y-31 = 860hp at 3250metes for 468 kg or the 12Y-45 that offered 920hp at 4200 meters for under 500kg.

Small engines can often offer better power per unit of displacement (liter). The Real trick if they can offer better power kilogram.

The smaller cubic capacity will usually mean that engine is smaller. A smaller engine can allow for the cooling system to be installed just under the engine on a 1-engined fighter without paying the drag penalty like a bigger engine with the same radiator set-up. Total weight of the closely-coupled liquid cooling system should also be lighter, and the target surface for enemy bullets is smaller. Smaller engine should should also allow for a better over-the-nose view.
On a 2-engine aircraft, smaller stature = lower frontal area = lower drag penalty. Also less blocking of the view for the crew.

As for Peregrine vs. HS12Y - Peregrine is just starting out. It can be safely over-boosted with the high octane fuel, unlike the 12Y. If the S/C is of the similar quality as what the Merlin XX gotten, we'd be looking at ~900 HP at 18-19kft, and ~1000 HP at 12kft, plus another 15-20% more HP with moderate over-boosting, albeit on the lower altitudes. The HS 12engine capable for such a power will need to pork up to the VK-105PF standard, that is basically a new engine weighting now 600 kg (? ; we save some weight due to the 1-speed S/C on the HS 12Y on steroids) that will have the repercussions both on the engines' production line as well as to re-balance the aircraft - okay, a 100-130 kg increase in the bare weight is not the end of the world, but still.
The Peregrine will need just the transplantation of a supercharger section (picking up perhaps 30 kg?).

I don't know whether the Peregrine was with 1-piece block, or 2-piece block construction.

 

[Shortround6]

Some of the Fiat engines used separate cylinders, I believe the Packards did too.
Or the they used a hybrid construction. Separate cylinders but one piece heads not just cam boxes.
But this does show the problems of updating old engines. They were never designed to operate at either the rpm or the cylinder pressures of a "modern" engine and modern could mean late 1930s.

I am using a sort of short hand on the Peregrine construction. They had two blocks of 6 cylinders each bolted/fastened to the crankcase (3rd piece ) and the crankcase was actually two pieces. Upper and lower. The old engines often used a separate cooling jacket for each cylinder (12 for a V-12) with a lot of "flexibility".

"A smaller engine can allow for the cooling system to be installed just under the engine on a 1-engined fighter without paying the drag penalty like a bigger engine with the same radiator set-up."
Some this is dependent on the cooling load, not the size/displacement of the engine. Radiator needs to get rid of waste heat from 900hp (?). It doesn't really matter how big the the engine blocks/passages are.................except.................What kind of coolant is being used? Plain water, or pressurized water (with a little anti-freeze),or Ethelene Glycol (100%) or water/Ethelene Glycol mix? You might need 3 if not four different radiator sizes to dissipate the same amount of heat. Some this depends on how deep the engine company or air force dipped into the Ethelene Glycol well
US dove in deep and tried to use 300 degrees F coolant temperature. Turns out that 1, Ethelene Glycol doesn't transfer the same amount of heat per KG of coolant passing through the engine per unit of time. 2, the engine/s transferred more of the heat load to the oil system so much of the smaller radiator gain was lost to larger oil coolers. Turns out that a 30% Ethelene Glycol 70% Water mix is about the the most efficient for transferring heat. Some engines required no increase in water passage size (flows might have to be adjusted) to handle higher powers (within reason ) when they switched from pure Ethelene Glycol.

Please note that both Allison and RR were using pure Ethelene Glycol in the late 30s and 1940 and into 1941(?) so using legacy engineering from 1935-1938/39 was doing them no favors.

 

[don4331]

Thank you.
For the purposes of this thread, we will assume that the 12 cyl engines are of monobloc construction.

The smaller cubic capacity will usually mean that engine is smaller. A smaller engine can allow for the cooling system to be installed just under the engine on a 1-engined fighter without paying the drag penalty like a bigger engine with the same radiator set-up. Total weight of the closely-coupled liquid cooling system should also be lighter, and the target surface for enemy bullets is smaller. Smaller engine should should also allow for a better over-the-nose view.
On a 2-engine aircraft, smaller stature = lower frontal area = lower drag penalty. Also less blocking of the view for the crew.

Yes, the engine is smaller, but not its not 1:1 (displacement is cubic, area is square), hence why the Griffon only had 0.4 ft^2 more frontal area than the Merlin (7.9 vs 7.5) despite of the 36% displacement increase. (Exact increase gets tricky as packaging gets better with later designs)

US dove in deep and tried to use 300 degrees F coolant temperature. Turns out that 1, Ethelene Glycol doesn't transfer the same amount of heat per KG of coolant passing through the engine per unit of time. 2, the engine/s transferred more of the heat load to the oil system so much of the smaller radiator gain was lost to larger oil coolers. Turns out that a 30% Ethelene Glycol 70% Water mix is about the the most efficient for transferring heat.

Ethylene glycol has only about 1/2 the specific heat capacity of water, so you either allow it to heat 2X (50° -> 150°C rather than 50° -> 100°C) or you increase the flow rate 2X. The issue with going to 150°C as noted; the heat goes into the oil...and oil doesn't like that temperature (In addition to requiring larger oil coolers*). Increasing flow rate resulted in fluid moving to fast to pickup the heat.

*There is also the issue that cold oil gets thick and stops flowing in the extremities of the cooler, so only a fraction of the cooler functions.

Or just one radiator if you do like Junkers and use oil to water coolers (Liquid to liquid cooling is very efficient and Ethylene glycol + water doesn't have the same coring issue).

The Allison actually has significant baffling in the cylinder block to ensure more cooling medium at the top of the cylinder where it is need more and less at the bottom.

Coming up with a crankcase/cylinder block/head solution which was both strong enough to seal boost/combustion pressure and cooling medium and be manufacturable was challenging (tolerance stack up).

Kestrel/early Merlin combine cylinder block & head into single casting, bolting cylinder sleeves to the head face.​

Starting with Merlin XX the head is separated from the cylinder block to ease manufacture (so you have upper/lower crankcase, 2 cylinder blocks and 2 heads).​

 

[Engineman]

Coming up with a crankcase/cylinder block/head solution which was both strong enough to seal boost/combustion pressure and cooling medium and be manufacturable was challenging (tolerance stack up).

Kestrel/early Merlin combine cylinder block & head into single casting, bolting cylinder sleeves to the head face.
Starting with Merlin XX the head is separated from the cylinder block to ease manufacture (so you have upper/lower crankcase, 2 cylinder blocks and 2 heads).​

The very early PV12 and Merlin B had the large crankcase with integral cylinders and detachable heads, changing to separate monobloc cylinders and ramp heads for the Merlin I and then single-plane one-piece blocks for Merlin II onwards .
Lumsden lists the Merlin 61 as the first Derby Merlin type with 2-piece cylinders. However, Packard built their V-1650-1 and Merlin 28 with 2-piece from the start, later with the R-R design.
In UK, later Merlin XX and similar marks were upgraded with 2-piece blocks.

Eng

 

[yulzari]

Due to supplier issues the OTL Whirlwind had less than ideal radiators. Better ones could well be able to cope with as much power as one could feasibly make for a Peregrine.

The second stage supercharger on a Peregrine would intrude into the wheel well space behind the engine so some changes would have to be made to fit and cope with changes in the CofG. However Petter did inform the Air Ministry at a late stage that he found how he could fit Merlins (single stage) so it may not be an insuperable problem for the Whirlwind to carry a two stage Peregrine.

 

[tomo pauk]

Just having the up-to-date 1-stage S/C on the Peregrine would've much improved it. Eg. the Merlin XX was making about 1220 HP at 16520 ft (similar goes for the Merlin 45), where the Merlin III was making 1030. So, conservatively, if the new S/C adds just 15% power at the 'normal' altitudes, we might've seen the "Peregrine Mk.II" making almost 1020 HP at 15000 instead of 'just' 885. With some over-boosting, it might've gotten at ~1150 HP at ~10000 ft?

An engine similar to the original Peregrine by Fiat would've also been interesting.

 

[Shortround6]

The Merlin XX 'technology' was production ready in Aug/Sept of 1940. Two stage supercharger was 1 to 1 1/2 years away. IF Petter had actually figured out how to put a Merlin in a Whirlwind without a substantial 'stretch' then a few extra inches for a Merlin XX style supercharger on a Peregrine should be easy. Then fix the damn air intake

An engine similar to the original Peregrine by Fiat would've also been interesting.

Or a modernized/updated A.30??
FIAT A.30 RA - 800 HP engine?

Keep the bottom end with perhaps a few tweaks (a bit better balance?) keep the cylinder head design but figure out how to make one piece cylinder blocks (or two piece, 3 cylinders in each block) stick a supercharger on it. Peak power at 3500-4000 meters.
24 liters at 2900rpm should give about 8-9% more power than the Peregrine at altitudes were the Italian supercharger is effective?

Fix the Italian supercharger??
feeding the air from outside of the impeller area (very low) is not getting high efficiency from the supercharger.

 

[Bretoal2]

For me, a "small" engine is an engine with small cylinders.

In this sense (or IMHO, if you prefer...) , the Hispano-Suiza 12Y is not really a good example, with a 3.004 liter individual capacity - among the biggest !

I bet a good contender would have been the Napier-Halford Dagger, with a 0.700 liter individual capacity. This engine lacked only two things : a much better supercharger and, obviously, a front fan cooling system such as the BMW 801. But I think that when the Dagger reached 1,000 hp, Halford was too much involved in the Sabre development.

An other detail : the Peregrine cylinder banks were 1-piece, identical to the first Merlin models.

 

[Shortround6]

I bet a good contender would have been the Napier-Halford Dagger, with a 0.700 liter individual capacity. This engine lacked only two things : a much better supercharger and, obviously, a front fan cooling system such as the BMW 801. But I think that when the Dagger reached 1,000 hp, Halford was too much involved in the Sabre development.

I disagree.
The Dagger was 'much to clever' and that is in a bad way.
It makes the French 700hp 14 cylinder radials look good.

24 cylinders and 1390lbs (630kg) is an expensive way to 1000hp at 2650 meters.
Just for engine design it is good in some way and not good in others. A Merlin at 3000rpm has about 81% of the piston ring drag/friction that the Dagger has at 4200rpm.
Piston/ring friction is about 80% of the total drag in many engines not counting the supercharger cost.
Dagger used a 7.5 compression ratio due to the small cylinders, even with 100 octane it may either hit it's limit or need lower compression.
Using more boost pressure puts you back in the heat problem.

Compared to a radial you have less room between the cylinders/heads and the cam boxes block some of the air flow across the cylinder heads.
All the radials that gained power increased their fin area and many in three ways. Making the fins long is one of them. In the inline engines you have problem,
You cannot make the fins much longer along the axis of the engine and you can't make them much wider (across the engines width) as you start getting uneven cooling around the circumference of the cylinder and piston.
A valiant effort but I would pass.

 

[tomo pauk]

Or a modernized/updated A.30??
FIAT A.30 RA - 800 HP engine?
Keep the bottom end with perhaps a few tweaks (a bit better balance?) keep the cylinder head design but figure out how to make one piece cylinder blocks (or two piece, 3 cylinders in each block) stick a supercharger on it. Peak power at 3500-4000 meters.
24 liters at 2900rpm should give about 8-9% more power than the Peregrine at altitudes were the Italian supercharger is effective?
Fix the Italian supercharger??
feeding the air from outside of the impeller area (very low) is not getting high efficiency from the supercharger.

The supercharged sibling of the A.30, that makes ~950 HP at 4000m, puts it almost in the ballpark of the early DB 601A - excellent for the air force that is badly out-powered in any year of ww2 we pick and choose. Even more of a boon when looked from the supply side of the coin, since Fiat was, after all, the most powerful Italian company to make the engines.
Fixing the S/C, perhaps by looking at what the DB 601 had, might've added another 10-12-15% of power.

It does not solve the Italian engine-related problems, but it makes them easier to bear.

 

[z42]

Or just one radiator if you do like Junkers and use oil to water coolers (Liquid to liquid cooling is very efficient and Ethylene glycol + water doesn't have the same coring issue).

IIRC the P-51H also moved to such a design with an oil-water cooler?

 

[yulzari]

I disagree.
The Dagger was 'much to clever' and that is in a bad way.
It makes the French 700hp 14 cylinder radials look good.

24 cylinders and 1390lbs (630kg) is an expensive way to 1000hp at 2650 meters.
Just for engine design it is good in some way and not good in others. A Merlin at 3000rpm has about 81% of the piston ring drag/friction that the Dagger has at 4200rpm.
Piston/ring friction is about 80% of the total drag in many engines not counting the supercharger cost.
Dagger used a 7.5 compression ratio due to the small cylinders, even with 100 octane it may either hit it's limit or need lower compression.
Using more boost pressure puts you back in the heat problem.
View attachment 851654
Compared to a radial you have less room between the cylinders/heads and the cam boxes block some of the air flow across the cylinder heads.
All the radials that gained power increased their fin area and many in three ways. Making the fins long is one of them. In the inline engines you have problem,
You cannot make the fins much longer along the axis of the engine and you can't make them much wider (across the engines width) as you start getting uneven cooling around the circumference of the cylinder and piston.
A valiant effort but I would pass.

I have a soft spot for the Dagger ever since I found out that de Havilland had been tasked with reviewing the cooling and concluded that Napier had been too busy trying to pack more air in the front rather than getting a low pressure at the exit to draw the air through. My informant was of the view that de Havilland would have chosen a reverse flow route. I suppose de Havilland's Albatros would give us some idea of what he meant.

 

[GrauGeist]

Just to add to the mix, what about the Czech engines developed by Praga?

The two I have in mind are the ESV and ESVR series V-12 engines.

Were these dead-ends or did Germany's annexation kill their development?

tomo pauk what are your thoughts.