De Havilland Mosquito (Wood vs. Metal)

[Deleted member 68059]

There is a "Parts Interchange" manual for Packard/RR Merlins so that service people know which part number goes with which corresponding part number.

Obviously some bits are not like the supercharger drive gearing. It is correct that the British/US partnership on the Merlin was a two way thing and
Packard had people over in England to visit and talk over all the latest mods and so on, just as RR had people at Packard.

 

[swampyankee]

There was a lot of industrial engineering and metrology information interchange before WWII, which included a mutual redefinition of the inch so UK and US inches were actually the same.

 

[buffnut453]

I'm still struggling to understand the relevance to Merlin production of the U.S. having to convert metric measurements. Were ANY RR Merlin measurements in metric units? I'd be surprised (but happy to be proven wrong).

 

[MiTasol]

I'm still struggling to understand the relevance to Merlin production of the U.S. having to convert metric measurements. Were ANY RR Merlin measurements in metric units? I'd be surprised (but happy to be proven wrong).

You are correct - the Merlin was actually all in inches BUT all hardware and threads were British (BA, BSF, BSW), all materials were British (DTD specs instead of AMS specs) and material sizes were British gauges instead of US gauges. This is an even bigger cluster phuque as you can see below (two of six pages) because there are so many different US and UK metal and wire gauges so Packard had to know which of the many British wire gauges that RR used for their valve springs etc.

 

[buffnut453]

You are correct - the Merlin was actually all in inches BUT all hardware and threads were British (BA, BSF, BSW), all materials were British (DTD specs instead of AMS specs) and material sizes were British gauges instead of US gauges. This is an even bigger cluster phuque as you can see below (two of six pages) because there are so many different US and UK metal and wire gauges so Packard had to know which of the many British wire gauges that RR used for their valve springs etc.

View attachment 701803

I get the difference between British and US specs/standards,,,and it was a huge challenge that is too-often overlooked. I was just completely confused by repeated mention in this thread of having to convert metric to US imperial because it has no relevance to the Merlin or, indeed, to any British-manufactured aircraft in the 1940s.

 

[MiTasol]

Except all those with de Havilland Gypsy engines.

The de Havilland use of metric goes back to just after WW1 when Geoffrey de Havilland and Frank Halford bought a massive number of war surplus French V8 aero engines, designed a new four cylinder crankcase and sold the result as a light aircraft engine, the Cirrus. All dH piston engines were derived from that and kept the metric hardware.

 

[WARSPITER]

Packard Merlins only had different threads on carburettor and fuel pumps as they were different makes.

All other fasteners were imperial. Packard even made their own taps and dies for the purpose. None
of the Merlin fasteners were metric. UNF and UNC came later and more modern Merlin rebuilds can
include those threads but not in the war years.

 

[GregP]

The only "bad press" that came out of the Rockne incident were the limitations of wood aircraft in commercial airlines operations.

OK

That lies with some thinking in the Air Ministry. Sometimes you have some individuals who want to be conservative in their approach to an engineering solution, my guess is they were looking at producibility and cost.

Actually, I thought and have heard both in print and from people who were contemporary at the time, that the US population more or less declined fly in Fokkers after the Knute Rockne incident and that aircraft with primary wood structures were popularly seen as unsafe. That's what my grandfather said, too.

 

[FLYBOYJ]

Actually, I thought and have heard both in print and from people who were contemporary at the time, that the US population more or less declined fly in Fokkers after the Knute Rockne incident and that aircraft with primary wood structures were popularly seen as unsafe. That's what my grandfather said, too.

I believe the Curtiss T-32 Condor was also effected by this as well

 

[MiTasol]

And the Stinson trimotor

 

[pbehn]

I get the difference between British and US specs/standards,,,and it was a huge challenge that is too-often overlooked. I was just completely confused by repeated mention in this thread of having to convert metric to US imperial because it has no relevance to the Merlin or, indeed, to any British-manufactured aircraft in the 1940s.

I think the confusion starts with the origins of the standard metre. Both the USA and the UK standardised their inch measure on the standard metre, this is hardly surprising since that was precisely why they were involved in the whole enterprise to begin with. The standard prototypes were actually made in London by Johnson Matthey 30 bars were produced and the USA received bar No 27. History of the metre - Wikipedia

 

[Howard Gibson]

But then the spring would not be interchangeable with the RR spring as it would not fit the keeper and base ring perfectly and I pretty sure the requirement was that Packard and RR parts must be totally interchangeable.

Yes it would be. The Americans at Packard will have no problems interpreting the inch dimensions on Rolls Royce drawings. They will have problems with obscure British material specifications. The elastic modulus of steel, which determines the spring rate, is not affected by heat treatment. We are concerned with strength and fatigue. If a Merlin engine must run 100 hours between overhauls and the Merlin runs at 3000rpm, we know how many cycles the spring must survive. Problem not solved, but definitely it is solvable.

 

[Howard Gibson]

There was a lot of industrial engineering and metrology information interchange before WWII, which included a mutual redefinition of the inch so UK and US inches were actually the same.

I have posted some pages with units and conversions of the 1835 Mechanic's calculator on my website.

Weights and Measures MDCCCXXXV

In 1835, 1mm = 0,03937".

That is 1/25.4 to the accuracy of the log tables they must have used to do the math back then.

 

[pbehn]

Yes it would be. The Americans at Packard will have no problems interpreting the inch dimensions on Rolls Royce drawings. They will have problems with obscure British material specifications. The elastic modulus of steel, which determines the spring rate, is not affected by heat treatment. We are concerned with strength and fatigue. If a Merlin engine must run 100 hours between overhauls and the Merlin runs at 3000rpm, we know how many cycles the spring must survive. Problem not solved, but definitely it is solvable.

Are you sure? Heat Treatment of Spring Steels | Metallurgy

 

[MiTasol]

Yes it would be. The Americans at Packard will have no problems interpreting the inch dimensions on Rolls Royce drawings. They will have problems with obscure British material specifications. The elastic modulus of steel, which determines the spring rate, is not affected by heat treatment. We are concerned with strength and fatigue. If a Merlin engine must run 100 hours between overhauls and the Merlin runs at 3000rpm, we know how many cycles the spring must survive. Problem not solved, but definitely it is solvable.

You must remember that there were many errors made in the US, and no doubt a lot of other countries, by assuming that what they were looking at was how they understood it. So long as Packard knew which gauge a piece of material was in there was no problem unless the person reading the drawing read it as the "usual" gauge to which they were accustomed to working with. As the six pages I posted on metal and wire gauges shows that can produce massive errors.

Another major issue was that US (and Asian) drawings are in third angle projection whereas the British and most continental countries use first angle projection. When the US started making Hispano cannons they did not realize this so the end result was that the US Hispanos were a mirror image of the English and European Hispanos and did not fit the British aircraft that they were ordered for.

 

[Howard Gibson]

Are you sure? Heat Treatment of Spring Steels | Metallurgy

Yes, I am sure. I am a mechanical engineering technologist with a three year diploma, and I have been doing mechanical design for over forty years. I took metallurgy, and six semesters of mechanics of materials. Steel has an elastic modulus of around 29 million psi, or 210GPa (GN/m^2). The yield and ultimate strength of steel is due to carbon content and heat treatment. Alloying elements affect the heat treatment, and possibly, corrosion resistance.

On a coil spring, I need to know the free length, the diameter of the coil, the diameter of the wire, and the number of free coils. All of this can be can be worked out by reading Rolls Royce drawings, or by examining a coil spring with Vernier calipres. I can work out the forces on the spring as it opens and closes valves, and the resulting stresses. The worst case is that the British use some nut-case wire gauge (the term "British Standard" is an oxymoron). A flakey wire gauge is a problem if I am making one Merlin engine. If for some reason, I am making perhaps 55,000 engines, I can amortize some tooling. I also know what space the spring fits in. I can work out the performance of the Rolls Royce spring, and I can design a new spring that uses American gauges, and fits in the Rolls Royce space.

If I design trucks, tractors and combine harvesters, I am going to have problems re-engineering guns.

 

[ThomasP]

Howard Gibson is correct.

When referring to metals and similar materials, Elastic Modulus is a property of the metal that is dependent on the chemical composition - not the microstructure. Heat treatment affects the microstructure but not the chemical composition (if you do not count things like carbon vaporization/loss, outgassing, etc, which changes the chemical composition during the heat treatment process) and therefore will not affect the Elastic Modulus (aka Young's modulus).

What this means in an application such as a leaf spring: At some point as more weight/force is applied to a non heat treated spring it will eventually exceed its Yield Strength at a given flex/distance/travel. If after the spring is heat treated the same weight/force is applied to the spring it will flex the same distance, but depending on its heat treated properties it may be able to flex further with the addition of more weight/force before reaching its Yield Strength.

For those of you not familiar with Yield Strength, this is the point at which the flex of the spring (in this case) will exceed the ability of the inherent strength of the material, such that it will not return to its original form. IE a leaf spring that is flat/straight before the weight/force is applied will not return to flat/straight after the weight/force is removed.

edited for clarity(?)

 

[MiTasol]

A flakey wire gauge is a problem if I am making one Merlin engine. If for some reason, I am making perhaps 55,000 engines, I can amortize some tooling. I also know what space the spring fits in. I can work out the performance of the Rolls Royce spring, and I can design a new spring that uses American gauges, and fits in the Rolls Royce space.

Your spring will now only be interchangeable if the mating parts are changed as an assembly and then there is the problem in the overhaul shop how do they work out (without fine tolerance measuring each spring) if it is an brand A or brand X spring, keeper and base ring that they need to replace. From your background you know the problems that will quickly arise if a fractionally oversize valve spring is fitted to standard size mating parts (or standard spring to undersize mating parts). And remember that they must be interchangeable with every other part number x spring used on that series engine. EG the Rolls Mk 24 and Packard 224 (and a couple of other models) have maximum interchangeability).

If I design trucks, tractors and combine harvesters, I am going to have problems re-engineering guns.

The American contractors did not re-engineer the Hispanos. They just built them straight off the first angle Brit drawings using third angle interpretation which is why then ended up mirror image.

**********************************************
​

I agree that British Standard is an Oxymoron as is the concept that the US were any better at that time. AC, AN, NAS, Parker hydraulic fitting were in use in aircraft prewar and early war and those have a wide range of thread types and mating angles etc etc etc and interchangeability is almost non existent without changing the pipe/hose simultaneously. And the component end of those fittings can have short or long tapered or parallel threads with makes life even more "interesting" for the poor suffering repair man. And the US had many different gauges for sheet and wire.

Likewise Metric was a mess pre ISO.

I refer to the BS and US standards as using inFerial measurements from joining inferior with imperial.

On the nuts and bolts side of things, with one exception in the common sizes, the US National Course and Brit Whitworth thread have identical threads per inch but different thread angles and thread forms. This made it possible to fit US nuts to UK bolts and vice versa but that naturally caused hardware failures. The US and UK worked hard on solving that problem and created Unified thread using the best features of each thread. Whitworth was declared inactive for new design in November 1946 and disappeared from British mass production in the very early 50s. I do not know the US dates but for them the change was easier as Unified used the American head sizes and tools.

Australia, which has officially been metric for decades, not only still uses large amounts of Whitworth but recently goods have begun appearing which have what is called a 13mm bolt. Naturally, when you measure it up, it is nothing more than a rebadged 1/2 Whitless. No respected tool manufacturer has made Whitworth spanners for many years so that means you must use a loose fitting real metric or inch spanner or the good old Aussie UNF spanner - NO, not a Unified National Fine spanner but a Universal Nut F...er as shown below.

 

[WARSPITER]

NO, not a Unified National Fine spanner but a Universal Nut F...er as shown below.

This is the other Nut F.....g tool we use.

 

[pbehn]

Yes it would be. The Americans at Packard will have no problems interpreting the inch dimensions on Rolls Royce drawings. They will have problems with obscure British material specifications. The elastic modulus of steel, which determines the spring rate, is not affected by heat treatment. We are concerned with strength and fatigue. If a Merlin engine must run 100 hours between overhauls and the Merlin runs at 3000rpm, we know how many cycles the spring must survive. Problem not solved, but definitely it is solvable.

Yes, I am sure. I am a mechanical engineering technologist with a three year diploma, and I have been doing mechanical design for over forty years. I took metallurgy, and six semesters of mechanics of materials. Steel has an elastic modulus of around 29 million psi, or 210GPa (GN/m^2). The yield and ultimate strength of steel is due to carbon content and heat treatment. Alloying elements affect the heat treatment, and possibly, corrosion resistance.

The two statements in bold are completely contradictory.