Which country designed the best engines for WWII?

[Shortround6]

Stanley Hooker's auto biography has a short anecdote about the fit of the Merlin and Ford of England tooling up to make Merlins.
There may have been a bit of exaggeration but basically Hooker claimed that Ford of England figured out how to mass produce Merlins like "shelling peas" and they were very good engines with none failing their tests.
And Ford was doing this before Packard got involved .

Vee's for Victory also has a few short sections on the difference between building luxury car engines of the day and building Allison aircraft engines. Cadillac being a major sub contractor to Allison.
Any high power aircraft engine was built to standards of fit very close to racing cars, not production car engines.
Just about all luxury car makers of the 1930s used large, low stressed engines for flexibility (less gear shifting),
Smoothness (heavy engines also dampened vibration), and quietness.

Aircraft engines leaked oil, however a lot more either went through the rings burned up in the combustion chamber or was blown out through the breathers as the crankcase was pressurized by blowby at times.
A few ounces or a cup of oil can make a large mess. A lot of these aircraft engines could go through several gallons an hour at cruising speeds.

 

[pbehn]

Stanley Hooker's auto biography has a short anecdote about the fit of the Merlin and Ford of England tooling up to make Merlins.
There may have been a bit of exaggeration but basically Hooker claimed that Ford of England figured out how to mass produce Merlins like "shelling peas" and they were very good engines with none failing their tests.
And Ford was doing this before Packard got involved .

Vee's for Victory also has a few short sections on the difference between building luxury car engines of the day and building Allison aircraft engines. Cadillac being a major sub contractor to Allison.
Any high power aircraft engine was built to standards of fit very close to racing cars, not production car engines.
Just about all luxury car makers of the 1930s used large, low stressed engines for flexibility (less gear shifting),
Smoothness (heavy engines also dampened vibration), and quietness.

Aircraft engines leaked oil, however a lot more either went through the rings burned up in the combustion chamber or was blown out through the breathers as the crankcase was pressurized by blowby at times.
A few ounces or a cup of oil can make a large mess. A lot of these aircraft engines could go through several gallons an hour at cruising speeds.

Part of the story of Ford Manchester was importing a special machine tool, I think for the main bearing journals from Switzerland via USA. The first got sunk so they ordered another. They fully understood the need for accuracy and fine tolerances and got the machines to do it. When war was declared a team of machine tool specialists from Germany installing machines in the RR factory were among the first to be "interned for the duration".

 

[Dash119]

Gentlemen,

He was referring to infernal combustion engines, which I believe Rolls Royce did assemble by hand...

The internal combustion engines were produced on an assembly line.

 

[pbehn]

Gentlemen,

He was referring to infernal combustion engines, which I believe Rolls Royce did assemble by hand...

The internal combustion engines were produced on an assembly line.

But the Schneider trophy was won with infernal combustion engines? Only 19 R type engines were made in two years, these were made by hand and were the engines Lovesey was involved with, probably why he had no concept of production engineering and the reason he was put where he was by RR to learn about it.

 

[Shortround6]

Part of the story of Ford Manchester was importing a special machine tool, I think for the main bearing journals from Switzerland via USA. The first got sunk so they ordered another. They fully understood the need for accuracy and fine tolerances and got the machines to do it. When war was declared a team of machine tool specialists from Germany installing machines in the RR factory were among the first to be "interned for the duration".

Part of Hooker's story was he was sitting in his office (shared with another RR official) when one of the men from Ford came in and said they could not build engines with the tolerances listed on drawings. Hooker asked if that was because Ford could not maintain the tolerances. (a bit being a snob about it as he explains in his auto-biography) The Ford man said no, it was because Ford needed much tighter tolerances in order to assemble engines using interchange parts.
You don't build cheap cars in quantity using semi-skilled labor and stay in business otherwise.

That said the Merlin needed pretty tight tolerances for it to work. Any high performance aircraft engine did. But the tight tolerances needed to be specified on the drawings to being with and maintained through manufacture of the parts and through dozens of inspections so that when the parts reach the assemblers 99% (or better) of the parts will fit together with no fitting or selecting. Selecting pistons and con rods gets a bit tricky as each assembly and each group of 12 pistons/con rods has a certain weight tolerance. You don't sit there and "file to fit" it the piston/con rod is too light or heavy it gets put aside and another one is weighed and if it "fits" with the group it is used. When you get enough light (or heavy) piston-con rod assemblies to make a complete engine you can assemble a light or heavy crankshaft assembly and put that in an engine. All the piston-conrod assemblies had a certain weight range they had to fit in in relation to the 12 cylinders used. However the total range of possible weights was larger.

A complete pair of assembled rods, pistons, pins and rings in an engine had a permissible variation was 1 oz. for Merlin II.
Permissible variation of pistons selected for one engine was 1/2 oz.
Permissible variation between any two pairs of connecting rods, pistons, pins and rings fitted to one engine was 1 oz.


There was a considerable amount of fitting/selecting even so, but the stories of "filing to fit" are not true.
The better the manufacturing and in a lot of cases that means better inspection, the more parts that make it to the fitters are in tolerance, and require less selection.

Tolerances are from "Aircraft Handbook" by Fred Colvin 1942 edition section III.

People that tell tales about loose tolerances or how engines were assembled need to look at handbooks that describe actual tolerances and how the engines were torn apart and reassembled. The book has 37 pages on overhauling a Merlin II had is pretty much "guidance" for a man who has gone to school to learn all the basic techniques.
Even with the book I am no more qualified to rebuild a Merlin engine than I am to launch a moon rocket.

 

[pbehn]

Part of Hooker's story was he was sitting in his office (shared with another RR official) when one of the men from Ford came in and said they could not build engines with the tolerances listed on drawings. Hooker asked if that was because Ford could not maintain the tolerances. (a bit being a snob about it as he explains in his auto-biography) The Ford man said no, it was because Ford needed much tighter tolerances in order to assemble engines using interchange parts.
You don't build cheap cars in quantity using semi-skilled labor and stay in business otherwise.

I know the story but as I remember it is was Cyril Lovesey that had that conversation or exchange. mentioned in the book by Hooker "Not much of an engineer"

 

[Shortround6]

Perhaps Hooker inserted himself into the story, I belief they shared the office.
The Ford man was most probably there to talk to Lovesey.

 

[pbehn]

Perhaps Hooker inserted himself into the story, I belief they shared the office.
The Ford man was most probably there to talk to Lovesey.

It was a remark that as I understand it they both were amused at, showing how much they had to learn at that time, but it was way back in the history of the Merlin. The article these stories about Packard come from use that remark and are deliberately ambivalent about who it was said to and when, the reader gets the impression it was made to a Packard employee in 1941 which was certainly not the case. I havnt read the book, but I have read elsewhere that the first discussions with Ford were actually Ford from France with a view to license building the Merlin there. There were many discussions and remarks, strange to build a whole engineering thesis and history on one.

 

[MiTasol]

Part of Hooker's story was he was sitting in his office (shared with another RR official) when one of the men from Ford came in and said they could not build engines with the tolerances listed on drawings. Hooker asked if that was because Ford could not maintain the tolerances. (a bit being a snob about it as he explains in his auto-biography) The Ford man said no, it was because Ford needed much tighter tolerances in order to assemble engines using interchange parts.
You don't build cheap cars in quantity using semi-skilled labor and stay in business otherwise.

That said the Merlin needed pretty tight tolerances for it to work. Any high performance aircraft engine did. But the tight tolerances needed to be specified on the drawings to being with and maintained through manufacture of the parts and through dozens of inspections so that when the parts reach the assemblers 99% (or better) of the parts will fit together with no fitting or selecting. Selecting pistons and con rods gets a bit tricky as each assembly and each group of 12 pistons/con rods has a certain weight tolerance. You don't sit there and "file to fit" it the piston/con rod is too light or heavy it gets put aside and another one is weighed and if it "fits" with the group it is used. When you get enough light (or heavy) piston-con rod assemblies to make a complete engine you can assemble a light or heavy crankshaft assembly and put that in an engine. All the piston-conrod assemblies had a certain weight range they had to fit in in relation to the 12 cylinders used. However the total range of possible weights was larger.

A complete pair of assembled rods, pistons, pins and rings in an engine had a permissible variation was 1 oz. for Merlin II.
Permissible variation of pistons selected for one engine was 1/2 oz.
Permissible variation between any two pairs of connecting rods, pistons, pins and rings fitted to one engine was 1 oz.


There was a considerable amount of fitting/selecting even so, but the stories of "filing to fit" are not true.
The better the manufacturing and in a lot of cases that means better inspection, the more parts that make it to the fitters are in tolerance, and require less selection.

Tolerances are from "Aircraft Handbook" by Fred Colvin 1942 edition section III.

People that tell tales about loose tolerances or how engines were assembled need to look at handbooks that describe actual tolerances and how the engines were torn apart and reassembled. The book has 37 pages on overhauling a Merlin II had is pretty much "guidance" for a man who has gone to school to learn all the basic techniques.
Even with the book I am no more qualified to rebuild a Merlin engine than I am to launch a moon rocket.

Well put.

All I would add is that automotive tolerances and clearances usually appear tighter than aviation ones, especially with radials, because of the much larger size of the components and the wider range of operating conditions.

A .01 tolerance on a small 50mm piston is exactly the same percentage as .03 on a 150mm piston. Add to that the operating conditions where the pilot reduces power on a radial. The heat going into the cylinder is reduced. It takes a relatively long time for the piston temperature to drop because it is only cooled by oil splash but the cylinder barrel itself will cool much quicker so, unless the clearances are wide enough the barrel will shrink and seize the piston. This is less a problem on a liquid cooled engine but other, more complex, cooling factors a apply there.

As you say having the book is only a part of the knowledge required but by reading the book you will find all the critical non-standard information you need such as, to use your example, the the maximum weight difference between the lightest and heaviest piston or connecting rod. One of hundreds of standard practices you learn before that is to match the heaviest rod with the lightest piston etc in order to balance out the individual component weights as far as is practical.

 

[Deleted member 68059]

"balancing" is one of the most misunderstood topics of engine design, and is sadly subject to some truly terrible myths.

One very good example of this is balancing things like connecting rods to within some miniscule mass, on the grounds
that this "must" make it a "better balanced" engine.

In truth, it is entirely dependant on what it is you are trying to achieve.

For perspective, a typical bearing loading might approach 100kN or 10 tons, which in turn is accompanied by
second order vibrations which may (depending on the bank angle) produce pure sideways forces of equivalent to perhaps
1 ton.

The insanity of sitting filing down the last 1/4 of a gram of weight from the parts is readily apparent here from the
perspective of the bearing loads involved.

This changes somewhat in very modern engines, which may be described as high-speed, which is to say, that the dynamic forces
from the motion of the components exhibit a very high portion of the bearing loads.

Conversely, low speed engines (certainly anything below 3000rpm), tend to be what we could call "gas-load dominated",
which is to say, that the loads produced by the gas pressure acting on the piston totally overshadow the loads from the
rotating and reciprocating parts. (gas loads remain fairly constant with rpm, but dynamic loads obviously
can rise to multiples of their magnitude from idle speed to maximum crank speed.)

This is most clearly demonstrated by very large diesels, which are often seen to have no
crankshaft counterweights what-so-ever, because the speeds are so low, the forces
from the motion of the piston are so tiny compared to those from the gas load
that they are essentially a waste of time.

This is complicated somewhat because bearing loads are not quite the same as "balance",
as, technically speaking the gas loads do not really impact engine balance as they act
equally in all directions.

This is all my diplomatic way of saying that in engines like those, it is an embarrasing waste of time
playing with a couple of grams here and there, when thousands of kilos equivalent of force are flying about all over the
place.

Sitting polishing rods down to within a 10th of a gram end-to-end is something amateur tuning shops
in the 1980`s did to convince people that they must know what they`re doing because they were doing
something "more exact", which all stemmed from the notion that "blueprinting" was something to be
applied to all areas of the engine, until people actually sat down and said, hang on... why ?

Much the same thing as polishing ports to a mirror finish "because shiny", until it was mentioned to them
by gas turbine experts, that the lowest surface drag actually required a very carefully controlled, and actually
pretty rough satin finish after all... but that didnt stop generations of hot-rod owners driving off with
polished ports and pistons balanced to within a 10th of a gram - feeling that nice warm feeling that
they had a "professionally hotted up" engine.

 

[swampyankee]

"balancing" is one of the most misunderstood topics of engine design, and is sadly subject to some truly terrible myths.

One very good example of this is balancing things like connecting rods to within some miniscule mass, on the grounds
that this "must" make it a "better balanced" engine.

In truth, it is entirely dependant on what it is you are trying to achieve.

For perspective, a typical bearing loading might approach 100kN or 10 tons, which in turn is accompanied by
second order vibrations which may (depending on the bank angle) produce pure sideways forces of equivalent to perhaps
1 ton.

The insanity of sitting filing down the last 1/4 of a gram of weight from the parts is readily apparent here from the
perspective of the bearing loads involved.

This changes somewhat in very modern engines, which may be described as high-speed, which is to say, that the dynamic forces
from the motion of the components exhibit a very high portion of the bearing loads.

Conversely, low speed engines (certainly anything below 3000rpm), tend to be what we could call "gas-load dominated",
which is to say, that the loads produced by the gas pressure acting on the piston totally overshadow the loads from the
rotating and reciprocating parts. (gas loads remain fairly constant with rpm, but dynamic loads obviously
can rise to multiples of their magnitude from idle speed to maximum crank speed.)

This is most clearly demonstrated by very large diesels, which are often seen to have no
crankshaft counterweights what-so-ever, because the speeds are so low, the forces
from the motion of the piston are so tiny compared to those from the gas load
that they are essentially a waste of time.

This is complicated somewhat because bearing loads are not quite the same as "balance",
as, technically speaking the gas loads do not really impact engine balance as they act
equally in all directions.

This is all my diplomatic way of saying that in engines like those, it is an embarrasing waste of time
playing with a couple of grams here and there, when thousands of kilos equivalent of force are flying about all over the
place.

Sitting polishing rods down to within a 10th of a gram end-to-end is something amateur tuning shops
in the 1980`s did to convince people that they must know what they`re doing because they were doing
something "more exact", which all stemmed from the notion that "blueprinting" was something to be
applied to all areas of the engine, until people actually sat down and said, hang on... why ?

Much the same thing as polishing ports to a mirror finish "because shiny", until it was mentioned to them
by gas turbine experts, that the lowest surface drag actually required a very carefully controlled, and actually
pretty rough satin finish after all... but that didnt stop generations of hot-rod owners driving off with
polished ports and pistons balanced to within a 10th of a gram - feeling that nice warm feeling that
they had a "professionally hotted up" engine.

One other issue is that a minute scratch on that perfect mirror finish is likely to seriously degrade fatigue life, much worse than a similar scratch on a normally finished surface.

 

[Shortround6]

Depending on the engines involved there seem to have been several critical vibration bands, this is just from looking at which engines changed what in their history so I could be way off.
There may be overlap going on. As Calum has said, gas pressures acting on the pistons counts for an awful lot and some parts, like crankshafts, are subject to a variety of vibrations from different sources at the same time.

The Hispano-Suiza Y series engines with their 150mm X 170mm cylinders either had no counter balancing or very little when they were running at 2400rpm. They introduced counter balancing (or more of it?) when they went to 2500rpm and any of the variations by the Russians, Swiss and later French (the 12Z) that ran even faster introduced a lot more counter balancing and/or heavier crankshafts. But they all were operating at higher pressures in the cylinders with greater gas pressure loads on the pistons, the reciprocating speeds went up, and the torsional vibrations ( crankshaft trying to wind and unwind as the cylinders fired) went up and us armchair engine designers can only make the vaguest of thoughts about beam strength and/or distance between bore spacing and size of the bearings (both diameter and length).

The Merlin ran at the same speed for it's production life except for a few competition engines. The Allison gained 27lbs (?) when it went from the 6 counter weight crank to the 12 counter weight crank that was supposed to allow the engine to run at 3400rpm instead of 3000rpm. Of course the cylinder pressures may have gone up in the mean time as the production engines were only cleared for 3200rpm? But the production engines were supposed to use 115-145 or 150 fuel instead of the 100/130 fuel they may have been working on at the start of the new crankshaft design? The 12 counterweight crank did have much lower bearing loads than the 6 counterweight crankshaft and had a much longer fatigue life.

Knowing that your crankshaft broke because of vibration/fatigue is one thing, knowing which vibration pattern or source is the reason is something else. Note also that sometimes you can dampen out certain vibrations by making the crankcase stronger and reducing the flex in the crankshaft.

I have brought it up many times before, Aircraft engines of WW II were built to near (or exceeding) race car standards. Look at the power to weight ratio of the engines and look at how long they lasted.
Then look at how long many of the race car engines of the late 30s lasted or the number of DNFs out of the number of starters for a four hour race or longer.

 

[pbehn]

Depending on the engines involved there seem to have been several critical vibration bands, this is just from looking at which engines changed what in their history so I could be way off.
There may be overlap going on. As Calum has said, gas pressures acting on the pistons counts for an awful lot and some parts, like crankshafts, are subject to a variety of vibrations from different sources at the same time.

The Hispano-Suiza Y series engines with their 150mm X 170mm cylinders either had no counter balancing or very little when they were running at 2400rpm. They introduced counter balancing (or more of it?) when they went to 2500rpm and any of the variations by the Russians, Swiss and later French (the 12Z) that ran even faster introduced a lot more counter balancing and/or heavier crankshafts. But they all were operating at higher pressures in the cylinders with greater gas pressure loads on the pistons, the reciprocating speeds went up, and the torsional vibrations ( crankshaft trying to wind and unwind as the cylinders fired) went up and us armchair engine designers can only make the vaguest of thoughts about beam strength and/or distance between bore spacing and size of the bearings (both diameter and length).

The Merlin ran at the same speed for it's production life except for a few competition engines. The Allison gained 27lbs (?) when it went from the 6 counter weight crank to the 12 counter weight crank that was supposed to allow the engine to run at 3400rpm instead of 3000rpm. Of course the cylinder pressures may have gone up in the mean time as the production engines were only cleared for 3200rpm? But the production engines were supposed to use 115-145 or 150 fuel instead of the 100/130 fuel they may have been working on at the start of the new crankshaft design? The 12 counterweight crank did have much lower bearing loads than the 6 counterweight crankshaft and had a much longer fatigue life.

Knowing that your crankshaft broke because of vibration/fatigue is one thing, knowing which vibration pattern or source is the reason is something else. Note also that sometimes you can dampen out certain vibrations by making the crankcase stronger and reducing the flex in the crankshaft.

I have brought it up many times before, Aircraft engines of WW II were built to near (or exceeding) race car standards. Look at the power to weight ratio of the engines and look at how long they lasted.
Then look at how long many of the race car engines of the late 30s lasted or the number of DNFs out of the number of starters for a four hour race or longer.

That is true, but tuners have to sell parts. I bought two stroke tuning of barrels heads reed valves etc in the early 80s. Everyone compared their various "bits" to see which ones most "looked the part" as if you could tell good gas flow by looking. It was almost as important as the actual power they produced, we were all very partisan about "our" tuner, all the others were of course underhand cheats.

 

[MiTasol]

"balancing" is one of the most misunderstood topics of engine design, and is sadly subject to some truly terrible myths.

One very good example of this is balancing things like connecting rods to within some miniscule mass, on the grounds
that this "must" make it a "better balanced" engine.

In truth, it is entirely dependant on what it is you are trying to achieve.

For perspective, a typical bearing loading might approach 100kN or 10 tons, which in turn is accompanied by
second order vibrations which may (depending on the bank angle) produce pure sideways forces of equivalent to perhaps
1 ton.

The insanity of sitting filing down the last 1/4 of a gram of weight from the parts is readily apparent here from the
perspective of the bearing loads involved.

This changes somewhat in very modern engines, which may be described as high-speed, which is to say, that the dynamic forces
from the motion of the components exhibit a very high portion of the bearing loads.

Conversely, low speed engines (certainly anything below 3000rpm), tend to be what we could call "gas-load dominated",
which is to say, that the loads produced by the gas pressure acting on the piston totally overshadow the loads from the
rotating and reciprocating parts. (gas loads remain fairly constant with rpm, but dynamic loads obviously
can rise to multiples of their magnitude from idle speed to maximum crank speed.)

This is most clearly demonstrated by very large diesels, which are often seen to have no
crankshaft counterweights what-so-ever, because the speeds are so low, the forces
from the motion of the piston are so tiny compared to those from the gas load
that they are essentially a waste of time.

This is complicated somewhat because bearing loads are not quite the same as "balance",
as, technically speaking the gas loads do not really impact engine balance as they act
equally in all directions.

This is all my diplomatic way of saying that in engines like those, it is an embarrasing waste of time
playing with a couple of grams here and there, when thousands of kilos equivalent of force are flying about all over the
place.

Sitting polishing rods down to within a 10th of a gram end-to-end is something amateur tuning shops
in the 1980`s did to convince people that they must know what they`re doing because they were doing
something "more exact", which all stemmed from the notion that "blueprinting" was something to be
applied to all areas of the engine, until people actually sat down and said, hang on... why ?

Much the same thing as polishing ports to a mirror finish "because shiny", until it was mentioned to them
by gas turbine experts, that the lowest surface drag actually required a very carefully controlled, and actually
pretty rough satin finish after all... but that didnt stop generations of hot-rod owners driving off with
polished ports and pistons balanced to within a 10th of a gram - feeling that nice warm feeling that
they had a "professionally hotted up" engine.

True, but at the time this was not known so the manuals of the day are very specific about the maximum difference between the weights of pistons, pins and rods and the standards of the day required matching the heaviest with the lightest to even out the static imbalance. Dynamic balancing was not attempted in those times.

 

[MiTasol]

Depending on the engines involved there seem to have been several critical vibration bands, this is just from looking at which engines changed what in their history so I could be way off.

Many of the engines of the time had restrictions on operating within specific rpm ranges (or specific rpm+boost conditions) because of vibration issues.

 

[Shortround6]

Many of the engines of the time had restrictions on operating within specific rpm ranges (or specific rpm+boost conditions) because of vibration issues.

Hmmm. My old BMW R-90/6 had a restriction (by me) about riding it at around 55-60mph. Under 55 and over 65 or so it was very smooth. However, despite what the advertisements said that 55- 60 something speed range was very tiring to ride at with the engine shaking back and forth.

 

[pbehn]

Were there any other nations other than Great Britain that got a H-block engine into mass production and widescale usage? As far as I know, the Napier Sabre was the only one to make it into production and it had one of the highest compression ratios of any engine of the war. IMO, it was a greater technological marvel than any of the WW2 jet engines. I gotta believe that Great Britain was the leader in technical sophistication when it came to IC engine development and design.

There was also the Rolls Royce Eagle, used i the Westland Wyvern in lieu of turbo props Rolls-Royce Eagle (1944) - Wikipedia

 

[fastmongrel]

That is true, but tuners have to sell parts. I bought two stroke tuning of barrels heads reed valves etc in the early 80s. Everyone compared their various "bits" to see which ones most "looked the part" as if you could tell good gas flow by looking. It was almost as important as the actual power they produced, we were all very partisan about "our" tuner, all the others were of course underhand cheats.

I was always being beaten by Stan Stephens tuned bikes. Now I know why he was an underhand cheat. I want all the cheap chrome plated plastic crap trophies that were stolen from me.

 

[pbehn]

I was always being beaten by Stan Stephens tuned bikes. Now I know why he was an underhand cheat. I want all the cheap chrome plated plastic crap trophies that were stolen from me.

Dunno when you raced but there was one obvious Stan Steven tuned cheat when I was racing Chris Wyatt, but no evidence that Stan Stevens had anything to do with it, I know when I am racing against a 350 when I am on a 250. That year (1982) the ACU championship was won by Mick Crick who's bike was tuned by Barcol and was no faster than mine, however his dad was a chassis expert on sidecars Windle Crick or Windricks as they were known, so he knew how to set up a bike and had ridden all sorts of competitions since he was a kid. In the last race of the championship at Silverstone all the bikes of the top riders were the same, you could pull out of the slipstream and get alongside but not pass, this resulted in us going into the corners at the end of straights 6 or 8 abreast. I finished 8th, 1.8 seconds behind the winner, but I was in 3rd place with three corners to go, it was complete mayhem. I had people demanding to open up my engine, it was a bog standard Beckett tune, I told them to knock themselves out, just follow me home with a new gasket set. The top boys on Stan Stevens bikes like Gordon Allott, Eddie Boldizar and Kurt Langhan were no faster than mine, the rider and circuit knowledge decided who won.

 

[Reluctant Poster]

Ground crews prefered the Packard ones - they didn't leak like sieves

Urban Legend. Early Merlins leaked glycol before the change to pressurized water cooling. Packard Merlims were all built after the change.