Why such a low RPM for the Griffon?

[SeaWorldOrBust]

It's always struck me that the ultimate developments of the Rolls-Royce Griffon (at least that I'm aware of) were not massively more powerful than the later Merlin derivatives, owing to the Merlin's typically greater boost pressures and higher RPM. While increasing boost on the Griffon was probably a non-starter given it's relatively large cylinders (hence the eventual development of the Eagle H24), I've never understood why the max RPM was so low, at only 2750 (slightly lower than the earlier .

The 1930s R, with identical bore and stroke as the Griffon, operated at the same 3000 RPM as the Merlin, and while being a racing engine, it obviously didn't need to have the same longevity as the military Griffon, improvements in metallurgy and construction techniques in the intervening decade would seem to have made it possible to run a 37 litre engine reliably at high RPMs. The Jumo 213 entered service at around the same time, was around the same weight, and had around the same capacity, yet was operating reliably at 3250 RPM.

I would assume there was some technical reason for the limitation but I've never heard what it is. Anyone have any knowledge on the subject, or ideas as to why this might be?

 

[Shortround6]

Most engine designers didn't like to exceed 3000fpm of piston speed. There are several reasons relating to force on the connecting rods and crankpins and such.
This was not a hard a fast rule but it was a generally accepted guide line.

Exceeding the 3000fpm guide needed special attention and/or accepting shorter overhaul life.

Also most aircraft engines used fixed ignition timing. The ignition system didn't move the ignition timing earlier in the piston travel to compensate for higher rpm.
Flame travel in the cylinder is pretty much constant and most engineers want the combustion to be pretty much finished by the time the piston is at about 20 degrees past top dead center. They were getting variable timing late in the war or they sometimes had two positions. Starting/idle and everything else.

going back to the Griffon, RR used better gas to increase the boost pressure which increased the mean pressure in the Cylinder. The BMEP went from about 180psi in a prewar Merlin running 6lbs of boost to around 290-300 in a late war Merlin or Griffon. Griffons and post war Jumo 213s (built in France) were actually pretty close in the power per sq in of piston area.
Griffon used higher cylinder pressure to make up for the lower RPM. You need to be a much better armchair engineer than I am to figure out the actual differences. More but smaller booms per minute and the strain of keeping everything together.

I will also note that most people (writers about engines) figure that piston rings/skirts make up about 80% of the friction in an unsupercharged engine (sometimes power for supercharger gets figured into the same loss catagory as friction) and since friction goes up with the square of the speed increasing RPM needs to be looked at carefully.
Remember that these high output aircraft engines were equaling or beating the power output of the best car racing engines of their day on per pound basis.

 

[SeaWorldOrBust]

The Jumo 213 did receive the sort of special attention you refer to though. With a ~6.5 in stroke (very comparable to the 6.6 of the Griffon) at 3250 rpm, it had a piston speed just over 3500 fpm — almost identical to the R, (which had a longer stroke by 0.1 in, but operated 50 RPM lower) which obviously opted for option two — accepting a shorter engine life. However the R also ran at a much higher +21 lbs boost — the same as the dimensionally identical, but lower revving Griffon running on the 150 octane fuel you alluded to (later versions ran at +25 llbs, but +21 that was the initial limit on the 65) and as a result had about a 300-350 hp advantage over the post war Arsenal 12H or the wartime Griffon running 150 Octane.

It seems to me that had Rolls-Royce put more effort into reliable operation of the Griffon at higher RPM, they could have had a genuine competitor to the Napier Sabre, capable of similar power in a smaller, simpler, and more practical package, as well as handily outperforming any planned/likely developments of the 213 or DB 603 thanks to the availability of superior fuel. They already had experience with an engine operating under those conditions with the R, I guess it seems strange to me that more effort wasn't made to match that performance in a more robust engine, especially when the Griffon is already a more heavily-built engine than the R (at least if the weight difference vs the presumably comparable single-stage Griffon is indicative of anything).

 

[tomo pauk]

It seems to me that had Rolls-Royce put more effort into reliable operation of the Griffon at higher RPM, they could have had a genuine competitor to the Napier Sabre, capable of similar power in a smaller, simpler, and more practical package, as well as handily outperforming any planned/likely developments of the 213 or DB 603 thanks to the availability of superior fuel.

From 1944 on, Griffon was a better engine than the Sabre in any metric bar the engine length? Yes, the Sabre 7 was supposed to do 3000+ HP with water injection, while being heavier than the latest Griffons, and without the hi-alt abilities of these Griffons.
Between the 3-speed 2-stage supercharged 100 series Griffon, 150 grade fuel operation, jet engines and the RR Eagle 24 introduction, I'd say that RR have had the bases well covered.

 

[Simon Thomas]

From Lumsden.
These were the engines in the Tempest Vs and Spitfire XIVs chasing the V-1's in mid 44.

 

[SeaWorldOrBust]

The Sabre II came into production well before the 60 series Griffons and had a pretty clear advantage in horsepower until the introduction of 150 octane fuel for the Griffon put it about on par with the IIA in the Typhoon, though still well behind the IIB in the Tempest which entered service around the same time, and was about on par with the much later 100 series Griffons at low altitudes. By the time the two stage Griffon was cleared for 21 lbs boost, the Sabre V was making 2600 hp reliably, though it didn't actually enter production afaik until after the war in the Tempest VI. Sabre VIIs could likely have entered service around the same time as the 100 series Griffons.

The Griffon was still a more compact, simpler, and cheaper engine, easier to maintain and held its performance much better at high altitude, but comparing contemporary iteratiions of each, the Sabre, for all its flaws, still typically had about a 200-400 hp advantage at each engines optimal altitude. I suspect a higher revving Griffon could have bridged that gap whilst maintaining most of it's various other advantages, apart from perhaps weight (the R made very similar power to the Sabre V), and I can't imagine it taking much more work than it took Napier to get the Sabre running semi-reliably. The Germans managed to optimise the 213 for high RPM operation despite presumably greater challenges in material qualities, I suspect Rolls Royce could have managed it. It's curious to me that they didn't pursue this avenue even originally, but especially when the boost limiations of the Griffon's large cylinders became apparent and they opted to instead develop the Eagle, despite the obvious flaws of the similar Sabre.

 

[Simon Thomas]

"the Mark V which was introduced in Autumn 1944 with an output of 2600 h.p. at 2500 ft. and 2,300 h.p. at 12,750 ft. and a boost pressure of +15 lbs./sq.in."

Quote from the information in this post:

Napier Sabre WW2 overview.

Reciprocating aero engines: design and development AVIA 46/160. First page is missing. Neil

ww2aircraft.net

 

[SeaWorldOrBust]

The Griffon 65 was cleared for use with 150 octane fuel (initially allowing +21 lbs boost for ~2200 hp) in November of 1944 so that's more or less the relevant comparison for the 2600 hp Sabre V time wise. Technically the Tempest VI equipped with that engine didn't go into production until later (I believe late 1945 but someone can correct me), however that had more to do with Hawker and the RAF than with the maturity of the engine. Originally the RAF wanted the Tempest I with the Sabre V and wing root radiators, however that order was changed to a simpler, Tempest V-based Mk VI with the traditional chin radiator, I believe because by that point the Tempest was mainly being used for ground attack, and it was perceived that the chin radiator would be less vulnerable to ground fire.

 

[tomo pauk]

What weight increase you can see for the Griffon turning much greater RPM than it was historically so? What kind of propeller for it? Suitable aircraft for such the Griffon?

 

[pbehn]

Using the same engineering and metallurgy as engine stroke increases the max RPM goes down, limited by many factors but mainly maximum piston speed. When I was racing motorcycles, it was just after Mike Hailwood made his comeback on the twin cylinder desmo Ducati. Desmo valve actuation and some fancy coatings of the cylinders allowed a brief change to the norm and it beat four cylinder four strokes. Then Honda decided to enter GP racing with an oval piston eight valve per cylinder 500 cc motorbike. It revved to over 20,000 RPM but as I read from an engineer at the time. it is one thing getting an engine to rev to 20,000 RPM it is another thing to get it to produce more power at the time. Pic below, it was actually a 8 cylinder engine with 4 combustion chambers.

 

[Engineman]

Also most aircraft engines used fixed ignition timing. The ignition system didn't move the ignition timing earlier in the piston travel to compensate for higher rpm.
Flame travel in the cylinder is pretty much constant and most engineers want the combustion to be pretty much finished by the time the piston is at about 20 degrees past top dead center. They were getting variable timing late in the war or they sometimes had two positions. Starting/idle and everything else.

As this question is posed for one of the largest capacity V-12 engines, I guess your reply "Also most aircraft engines used fixed ignition timing." applies to them? In that case, I must point out that almost all contemporary large WW2 Aero-engines actually did have sophisticated ignition timing adjustment. Generally, the timing would achieve around a maximum 40 to 45degrees of advance BTDC at about Climb power, often mechanically linked to the throttle position, throttle position itself usually also implied certain RPM and Boost (or MAP) power setting on CS props. Usually, an increase to full throttle, War Emergency or T/O power would also see a small reduction in the advance setting with such high Boost, and lower throttle setting would see a progressive reduction in advance, with the lowest advance at closed throttle. However, the ability of Magneto's to perform timing adjustment depended upon the mechanical system used to change the timing. If the timing adjustment mechanism physically adjusted the magneto driveshaft, the adjustment range may have been 45degrees relative to crankshaft. But, if the timing adjustment was achieved by the simpler system of rotating the points backplate, the range of timing adjustment would be less, because the points would be opening further and further away from the optimum flux generated in the magneto armature, so making a weaker and weaker spark.
In some cases the fully retarded starting sparks were generated from a starting magneto, or a "shower of sparks" vibrating contact and coil, that fed an independent conducting track and distribution contact on the magneto distributor rotor. If the minimum advance of the magneto was say 15degrees BTDC, the starting contact was about 15 degrees behind the normal contact, so giving starting timing (well, by dint of distribution) at about TDC.

Eng

 

[Snautzer01]

As this question is posed for one of the largest capacity V-12 engines, I guess your reply "Also most aircraft engines used fixed ignition timing." applies to them? In that case, I must point out that almost all contemporary large WW2 Aero-engines actually did have sophisticated ignition timing adjustment. Generally, the timing would achieve around a maximum 40 to 45degrees of advance BTDC at about Climb power, often mechanically linked to the throttle position, throttle position itself usually also implied certain RPM and Boost (or MAP) power setting on CS props. Usually, an increase to full throttle, War Emergency or T/O power would also see a small reduction in the advance setting with such high Boost, and lower throttle setting would see a progressive reduction in advance, with the lowest advance at closed throttle. However, the ability of Magneto's to perform timing adjustment depended upon the mechanical system used to change the timing. If the timing adjustment mechanism physically adjusted the magneto driveshaft, the adjustment range may have been 45degrees relative to crankshaft. But, if the timing adjustment was achieved by the simpler system of rotating the points backplate, the range of timing adjustment would be less, because the points would be opening further and further away from the optimum flux generated in the magneto armature, so making a weaker and weaker spark.
In some cases the fully retarded starting sparks were generated from a starting magneto, or a "shower of sparks" vibrating contact and coil, that fed an independent conducting track and distribution contact on the magneto distributor rotor. If the minimum advance of the magneto was say 15degrees BTDC, the starting contact was about 15 degrees behind the normal contact, so giving starting timing (well, by dint of distribution) at about TDC.

Eng

And i am in school again. Thank you.
I like that a lot.

 

[pbehn]

I think a quote from Lovesey gives an important insight to this ""The impression still prevails that the static capacity known as the swept volume is the basis of comparison of the possible power output for different types of engine, but this is not the case because the output of the engine depends solely on the mass of air it can be made to consume efficiently, and in this respect the supercharger plays the most important role." The Merlin, Griffon and R2800 were slightly "long stroke" motors with the stroke slightly larger than the bore. As Lovesey said RPM and static swept volume are not the deciding factors in engines where the "boost" can be varied and exotic fuels prevent detonation. The "trick" is to start combustion off before TDC with a mixture made up of droplets of a certain size. As the engine passes TDC the mixture is burning rapidly the droplets are vaporising which has a needed cooling and slowing effect, and then as the piston accelerates down the bore the expanding flame front exerts pressure on the piston, with an ideal of maximum pressure at 90 degrees rotation for max torque. It is all very clever stuff, way above my pay grade, I understood the principles when I read about it but it is high end science.


Edit: quote taken from wiki article on the Griffon Rolls-Royce Griffon - Wikipedia

 

[RichardSuhkoi]

Most engine designers didn't like to exceed 3000fpm of piston speed. There are several reasons relating to force on the connecting rods and crankpins and such.
This was not a hard a fast rule but it was a generally accepted guide line.

Exceeding the 3000fpm guide needed special attention and/or accepting shorter overhaul life.

Also most aircraft engines used fixed ignition timing. The ignition system didn't move the ignition timing earlier in the piston travel to compensate for higher rpm.
Flame travel in the cylinder is pretty much constant and most engineers want the combustion to be pretty much finished by the time the piston is at about 20 degrees past top dead center. They were getting variable timing late in the war or they sometimes had two positions. Starting/idle and everything else.

going back to the Griffon, RR used better gas to increase the boost pressure which increased the mean pressure in the Cylinder. The BMEP went from about 180psi in a prewar Merlin running 6lbs of boost to around 290-300 in a late war Merlin or Griffon. Griffons and post war Jumo 213s (built in France) were actually pretty close in the power per sq in of piston area.
Griffon used higher cylinder pressure to make up for the lower RPM. You need to be a much better armchair engineer than I am to figure out the actual differences. More but smaller booms per minute and the strain of keeping everything together.

I will also note that most people (writers about engines) figure that piston rings/skirts make up about 80% of the friction in an unsupercharged engine (sometimes power for supercharger gets figured into the same loss catagory as friction) and since friction goes up with the square of the speed increasing RPM needs to be looked at carefully.
Remember that these high output aircraft engines were equaling or beating the power output of the best car racing engines of their day on per pound basis.

I always wondered about piston velocity losses. How much was ring to bore vs. Oil losses? Since iron rings slip much better on chrome, if the actual metal contact friction was significant then a change would be apparent with a change to chrome. But if instead the losses were more from work done on moving the oil away and in impacting an under cylinder piston top cooling jet, then chrome would have a negligible benefit.
New VW GTI select on or off under piston oil cooling to avoid the losses except when at high power. What is the reality of this?

 

[SeaWorldOrBust]

I think a quote from Lovesey gives an important insight to this ""The impression still prevails that the static capacity known as the swept volume is the basis of comparison of the possible power output for different types of engine, but this is not the case because the output of the engine depends solely on the mass of air it can be made to consume efficiently, and in this respect the supercharger plays the most important role." The Merlin, Griffon and R2800 were slightly "long stroke" motors with the stroke slightly larger than the bore. As Lovesey said RPM and static swept volume are not the deciding factors in engines where the "boost" can be varied and exotic fuels prevent detonation. The "trick" is to start combustion off before TDC with a mixture made up of droplets of a certain size. As the engine passes TDC the mixture is burning rapidly the droplets are vaporising which has a needed cooling and slowing effect, and then as the piston accelerates down the bore the expanding flame front exerts pressure on the piston, with an ideal of maximum pressure at 90 degrees rotation for max torque. It is all very clever stuff, way above my pay grade, I understood the principles when I read about it but it is high end science.


Edit: quote taken from wiki article on the Griffon Rolls-Royce Griffon - Wikipedia

The Lovesey quote seems to be more concerned with the question of swept volume (I suspect the context is a comparison of the Merlin with the larger but less heavily boosted Daimler Benz engines). He doesn't mention RPM, an increase in which will also result in a more or less proportional increase in the mass of air an engine can consume in a given interval (and thus more power), and crucially, will do so without increasing the susceptibility to knock (though of course higher piston speeds can also cause issues). This was the issue Rolls Royce ultimately ran into with the Griffon, as the relatively large capacity of the individual cylinders meant the engine was less resistant to knock than the smaller Merlin (something that can still be observed in contemporary Reno racers powered by the Merlin vs the Griffon), hence the eventual decision to switch development to the 24-cylinder Eagle instead of moving to a larger V12. In addition to being larger in swept volume, the Eagle revved to a much higher 3500 RPM, and in theory was meant to be capable of higher boost pressures (+28 lbs) due to the smaller individual cylinders, though I suspect this would have caused problems with the engine's sleeve valves. The Eagle was an over-square motor, but again the Jumo 213 and Rolls Royce R were both undersquare, yet managed 3250 and 3200 RPM respectively, with piston speeds of around 3520 fps if memory serve.

I could potentially see it being difficult to calibrate the combustion timing you refer to in a higher revving motor, but I can't imagine it would have been more difficult that solving the boost limitations of a sleeve valve designt that they would have inevitably run into with the Eagle. It seems a higher revving Griffon would have been a more promising way forward. Perhaps not suitable for the very-high power applications the Eagle was intended for (3000-3500 hp) but probably would have given an appreciable boost in performance to the later Spitfire marks, the Spiteful, and potentially would have provided a higher altitude (and potentially lighter) alternative to the Sabre for the Tempest/Fury, both of which were tested with Griffons but ultimately never produced with them.

 

[Shortround6]

There is a formula to try to adjust the piston speed for bore to stroke ratio. This is an estimate as pistons do vary in weight in reality.
A smaller bore than stroke gets a bit of benefit compared to an over-square engine as this formula is concerned with the loads on the crankshaft, not piston ring problems.

mean piston speed divided by the square root of the stroke/bore ratio.

That is stoke/bore and not the common bore/stroke ratio.

A bigger diameter piston moving at the same speed is going to exert more force.

This was figured out by Frederick Lanchester, English engineer 1868-1946

 

[pbehn]

The Lovesey quote seems to be more concerned with the question of swept volume (I suspect the context is a comparison of the Merlin with the larger but less heavily boosted Daimler Benz engines). He doesn't mention RPM, an increase in which will also result in a more or less proportional increase in the mass of air an engine can consume in a given interval (and thus more power), and crucially, will do so without increasing the susceptibility to knock (though of course higher piston speeds can also cause issues). This was the issue Rolls Royce ultimately ran into with the Griffon, as the relatively large capacity of the individual cylinders meant the engine was less resistant to knock than the smaller Merlin (something that can still be observed in contemporary Reno racers powered by the Merlin vs the Griffon), hence the eventual decision to switch development to the 24-cylinder Eagle instead of moving to a larger V12. In addition to being larger in swept volume, the Eagle revved to a much higher 3500 RPM, and in theory was meant to be capable of higher boost pressures (+28 lbs) due to the smaller individual cylinders, though I suspect this would have caused problems with the engine's sleeve valves. The Eagle was an over-square motor, but again the Jumo 213 and Rolls Royce R were both undersquare, yet managed 3250 and 3200 RPM respectively, with piston speeds of around 3520 fps if memory serve.

I could potentially see it being difficult to calibrate the combustion timing you refer to in a higher revving motor, but I can't imagine it would have been more difficult that solving the boost limitations of a sleeve valve designt that they would have inevitably run into with the Eagle. It seems a higher revving Griffon would have been a more promising way forward. Perhaps not suitable for the very-high power applications the Eagle was intended for (3000-3500 hp) but probably would have given an appreciable boost in performance to the later Spitfire marks, the Spiteful, and potentially would have provided a higher altitude (and potentially lighter) alternative to the Sabre for the Tempest/Fury, both of which were tested with Griffons but ultimately never produced with them.

Lovesey was, if anything speaking about the Vulture and Sabre, which were due to replace the Merlin and would have if they could have been made reliable and higher boost pressure was not possible. The Vulture was in essence 2 x V 12 Peregrine engines working of the same crank, so it had almost twice the machining time and costs as a Merlin was heavier a and bigger AND due to various issues with it when the Vulture was abandoned the Merlin was producing similar outputs on bench tests. RPM is obviously linked to swept volume. The static swept volume x RPM /2 is the theoretical volume of air per minute consumed by a four stroke. Do not treat RPM as just a number that can be easily increased, it cant. The stroke of an Eagle engine was 130mm, a Merlin 152.4mm and a Griffon 167.6. To find the max speed of the piston at 3000 RPM it is Pi x stroke x60 to give mm/hour if the con rods are infinitely long.. BUT every revolution the piston stops at TDC and BDC. The longer the stroke the faster the piston must go at max speed but it still comes to rest twice per revolution, the increase in forces is exponential and the Griffon not only had a longer stroke but a bigger bore so heavier pistons too. Not only do the forces increase exponentially the maximum piston speed gets to a level that it is faster than the gas is expanding so does not achieve more power.