An interesting read about the Packard built Merlin engine.

[Shortround6]

At the factory they may have been pre matched and sent out in sets, to the assembly area/s.

I do not have the manufacturing tolerances. Like at what point a piston is judged to heavy or too light to try and put into a set.
However you still have the complete rod assembly weight requirement.

You do want the engine assemblers actually assembling engines. However that does not mean that each assembler, or small team, had to install each and every piece. Sub assemblies from another group, dept, could be used and would probably be more efficient. Like crankshaft assemblies showing up on the production line with pistons, rods and bearings already in place. Cylinder heads likewise might show up ready to bolt on rather than assembly team stopping after putting the cylinder blocks on and grabbing a pair of heads and putting the valves, valve springs, retainers, camshaft, rocker arms, etc into the heads before they put the heads on the engine.

 

[wuzak]

I frequently see anecdotes saying that P-51 pilots could tell a Rolls Engine from a Packard, how many P-51s were fitted with Rolls Royce engines?

Maybe 2 or 3, but these weren't production aircraft.

The obvious one is the Mustang X, converted to the Merlin by Rolls-Royce in the UK.

Maybe the XP-51B, because Packard production wasn't yet up to speed, and the XP-51F - fitted with Merlin 100 series?

 

[wuzak]

At the factory they may have been pre matched and sent out in sets, to the assembly area/s.

I do not have the manufacturing tolerances. Like at what point a piston is judged to heavy or too light to try and put into a set.
However you still have the complete rod assembly weight requirement.

The weight would have been checked as part of the quality procedures, as well as the physical dimensions of the piston. From there it shouldn't be too difficult to set them in sets that meet the criteria. Similarly with the rods.

 

[RW Mk. III]

If all pistons were matched for weight by the worker how were they matched. Your going to need hundreds of pistons to choose from and a very accurate set of scales for each worker. To assemble one engine how many pistons do you have to go through to find 12 that match and how lucky would you need to be for the first 12 you picked up to match. Instead of the hundreds of Merlins per day that British factories churned out the RAF would be lucky if they got one a day.

The system you are suggesting would suit the narrative of RR having lovingly handcrafted the engines. But I suspect was not the methodology that would have been applied. I am by no means a production specialist but I do know the value of a process to reduce a man-hours to product ratio. Having many workers with many scales sorting hundreds of pistons is a very haphazard way to manage things imo. One or two workers with one scale, sorting the pistons by weight and dimensions as they become available and inventorying, labelling and creating matched sets to be sent downstream would be an appealing approach in a mass production environment.

Now if I apply a bit of lateral thinking I can streamline the task of these two workers further. -this is where I point out that I cannot find any detailed information about how this sorting and matching was actually accomplished as a procedure by Allison or RR. It is more a thought experiment about the likelihood of the craftsman approach in an industrial nation at war-

Consider US Patent US73280334A.

"This invention relates to machines for testing and grading eggs and particularly to that type of machine wherein the eggs are moved over a source of light by means of an endless conveyor and graded as to size by determination of weight."

This patent was filed 06/28/1934. The device itself was not the first of its type nor particularly groundbreaking for 1934. For example: a mail sorting machine was first used in service in 1927 in the Netherlands, coin sorting machines quite before that but I cannot determine exactly where they came about first.

I realize it seems silly but it should dispel any (I apologize if I misinterpreted the thrust of the quoted post) notion that the weight matching would be in any way onerous in an industrial setting. I chose this patent partly because I like this image of all those eggs working through these things and partly because it demonstrates that the pistons could be rapidly sorted by weight and dimensions automatically in a manner not detrimental to precision parts. Eggs require a gentle touch as I understand it. This approach would have the advantage of saving time/personnel while introducing a QC safeguard, a machine-derived measurement. A parameter generated by a machine and check-measured by a human will always be preferable to two human measurements. At least for my money.

I apologize for introducing agricultural industrialization into this.

 

[Shortround6]

People seem to think that the Piston dept simply sent a bin full of almost but not quite finished piston over to the assembly area for the engine assemblers to pick through and sort out and finish up to fit with crocus cloth or something.

you put a nick in a rod or piston skirt and you are setting up a potential stress fracture starting point. These parts were going to be finished to a very high degree before they ever get to the assembly area/s.

 

[buffnut453]

This all gets down to the core assertion that RR tolerances were crap until the US automotive industry fixed the problem. It's absolute nonsense. You can't hand-pick components in a construction line and still maintain throughput.

The criteria for a sub-assembly would be a QC check but you don't check every sub-assembly on a production line, you check a percentage. To achieve a 1oz margin for the entire assembly of crankshaft, pistons and con rods, then it suggests upstream processes would output components such that any 12 pistons and con rods, when connected to the crankshaft, could be expected to fall within limits with a very high degree of repeatability. The alternatives are a slowed production line and either wasted production or go back to highly-skilled individuals to fettle out-of-spec sub-assemblies back into spec. The problem with the latter option is that it doesn't help detect upstream problems that caused the out of tolerance issue in the first place...so you're back to your upstream tolerances.

I'm afraid this whole issue of RR not having robust production line processes for Merlins prior to US automotive companies getting involved is absolute nonsense and the more it gets repeated, the angrier it makes me. It smacks of nationalistic willy-waving rather than actual technical study and reasoned analysis. Not one of the "hand crafted by elves" stories has EVER come up with a document that shows how poor RR specs were compared to the U.S. automotive industry, indeed most are of the "I knew a guy once who worked [insert location here] who always said that...." That's not proof or evidence, it's hearsay or folklore.

 

[buffnut453]

Here's a pretty balanced, albeit not especially detailed, study of whether Packard Merlins were any better than RR units (or vice versa). It includes a number of observations already made in this thread, notably that RR in the UK produced more Merlins than anyone else, and that it is inconceivable for such a number to be achieved using only skilled tradesmen. RR had effective production lines that largely used unskilled or semi-skilled workers to churn out engines by the thousand.

Rolls-Royce Vs. Packard: Who Built a Better Merlin?

Just to put a final nail in the nationalistic coffin, we owe an incalculable debt to the countless workers, on both sides of the Atlantic and elsewhere, who provided the tools so that Allied military forces could get the job done. Yes, America was the Arsenal of Democracy but it was very much a team effort, with bright ideas emerging on both sides of the Atlantic, to the benefit of the entire world.

 

[Reluctant Poster]

I previously posted a video of Ford manufacturing V8 engines pre war. They were weighting the pistons so I don't see why this was an impossible task for Rolls Royce.

 

[fastmongrel]

The system you are suggesting would suit the narrative of RR having lovingly handcrafted the engines. But I suspect was not the methodology that would have been applied.

I wasnt suggesting this was how it was or should have been done I was replying to the myth that Rolls Royce workers would hand pick and measure parts before assembly then file to fit each part. This is so stupid I cant understand why this myth persists there is no way the four Merlin factories in Britain could have churned out 100,000 or so engines if each part was hand fitted.

The only way that so many Merlins could be assembled was if every part was manufactured to exacting and repeatable tolerances by machines.

The main thing that annoys me so much about the myth is not that its so obviously untrue but that the "Us clever Americans had to fix the engine then teach the dumbass Limeys how to build it" meme is so utterly disrespectful to the men and women who worked long hours (72 hours a week at the Glasgow factory) under difficult conditions of rationing, blackouts and the chance a 250 kg bomb would land on your head at any time.

US industry did so many amazing things in the period 1940 to 45 so why do some people feel the need to steal the honour belonging to others.

 

[RW Mk. III]

I wasnt suggesting this was how it was or should have been done I was replying to the myth that Rolls Royce workers would hand pick and measure parts before assembly

I'm sorry I mismeasured the intent of your initial post. We are obviously in agreement.

 

[pbehn]

It is easy to forget in the discussion how quickly the Merlin went from just another engine to almost the only game in town. Without the war being declared there would probably just been a few thousand examples made. It was only the specified engine in the Hurricane Spitfire Defiant and Battle to start with. If RR knew they were required to make 160,000 engines they would almost certainly have designed a different engine.

 

[Reluctant Poster]

Here's a pretty balanced, albeit not especially detailed, study of whether Packard Merlins were any better than RR units (or vice versa). It includes a number of observations already made in this thread, notably that RR in the UK produced more Merlins than anyone else, and that it is inconceivable for such a number to be achieved using only skilled tradesmen. RR had effective production lines that largely used unskilled or semi-skilled workers to churn out engines by the thousand.

Rolls-Royce Vs. Packard: Wh

Here's a pretty balanced, albeit not especially detailed, study of whether Packard Merlins were any better than RR units (or vice versa). It includes a number of observations already made in this thread, notably that RR in the UK produced more Merlins than anyone else, and that it is inconceivable for such a number to be achieved using only skilled tradesmen. RR had effective production lines that largely used unskilled or semi-skilled workers to churn out engines by the thousand.

Rolls-Royce Vs. Packard: Who Built a Better Merlin?

Just to put a final nail in the nationalistic coffin, we owe an incalculable debt to the countless workers, on both sides of the Atlantic and elsewhere, who provided the tools so that Allied military forces could get the job done. Yes, America was the Arsenal of Democracy but it was very much a team effort, with bright ideas emerging on both sides of the Atlantic, to the benefit of the entire world.

o Built a Better Merlin?

Just to put a final nail in the nationalistic coffin, we owe an incalculable debt to the countless workers, on both sides of the Atlantic and elsewhere, who provided the tools so that Allied military forces could get the job done. Yes, America was the Arsenal of Democracy but it was very much a team effort, with bright ideas emerging on both sides of the Atlantic, to the benefit of the entire world.

I have read that article before and agree that it is better than most

Here's a pretty balanced, albeit not especially detailed, study of whether Packard Merlins were any better than RR units (or vice versa). It includes a number of observations already made in this thread, notably that RR in the UK produced more Merlins than anyone else, and that it is inconceivable for such a number to be achieved using only skilled tradesmen. RR had effective production lines that largely used unskilled or semi-skilled workers to churn out engines by the thousand.

Rolls-Royce Vs. Packard: Who Built a Better Merlin?

Just to put a final nail in the nationalistic coffin, we owe an incalculable debt to the countless workers, on both sides of the Atlantic and elsewhere, who provided the tools so that Allied military forces could get the job done. Yes, America was the Arsenal of Democracy but it was very much a team effort, with bright ideas emerging on both sides of the Atlantic, to the benefit of the entire world.

I have read this article in the past. The author makes some good points, but I have to disagree with his characterization of Packard "operating at a scale that dwarfed Rolls Royce.". As I have noted in the past Packard had far more in common with Rolls Royce than with Ford or Chevrolet. In fact, it was a bit of a mutual admiration society. A Packard ad from 1934 opens with "Only Packard and Rolls Royce" and goes on to describe how both companies manufacture super silent gears.

Because of the depression, sales of luxury automobiles fell dramatically, Packard reaching a low of 2407 units in 1933. Obviously, this was untenable and in fact all of Packard's rivals, except for Cadillac and Lincoln, disappeared. Rolls Royce's solution was to become an aircraft engine manufacturer with a side line in luxury automobiles. Packard took a different tact and moved down market into the near luxury niche dominated by Buick. The 120 (a straight 8) was introduced in 1935, followed by the Six in 1937. By 1940 the Six (rebadged as the 110) accounted for 2/3 of sales with most for the rest being 120s.

Meanwhile, in the same timeframe, Rolls Royce was rapidly expanding engine production producing more than 7100 Merlins in 1940.

There is no comparison in the level of complexity between a high-power aircraft engine and an automobile engine, even a V-12 The following video showing the rebuild of a Packard V12 illustrates this.



Here's a description of the differences between aircraft and automotive practice, which I believe was based on Packard data:

The point is that producing 7,000 plus Merlins represents an enormous effort fully comparable to Packard producing 98,000 automobile engines.

Rolls Royce had 12,500 employees in 1939 rising to 38,600 in 1941. Packard had slightly fewer employees.

As a special bonus attached is Packard's 40th Anniversary book. Note that it shows Packard did in fact weigh and match automobile pistons.

 

[fastmongrel]

I would add the Whitley to the list of specified engine pre war. The Armstrong Siddeley Tiger engined production was over quite quickly as soon as the Merlin X was available

 

[pbehn]

I would add the Whitley to the list of specified engine pre war. The Armstrong Siddeley Tiger engined production was over quite quickly as soon as the Merlin X was available

It sort of illustrates the point. The Whitley started with Tiger engines and was changed pre war with some Merlin engined planes in service. The Halifax had already been changed and the prototype flew in Oct 39, the Manchester/Lancaster would follow. Before the extra RR factories were producing it was obvious that they werent enough.

 

[tomo pauk]

D Deleted member 68059 - a member on the another forum posted this:

According to Volume 1 of 'the magic of the name' states the following regarding the two-piece block used on Packard engines, 'their early engines, Merlin 28's were similar to the Merlin XX, but had a two-piece block designed in the USA.

(my bold)

I've read in your book that 2-piece blocks were slated for Merlin 60 production at 1st; care to add something that might clear the picture about who designed the 2-piece block for Merlin?
FWIW, the book 'The Merlin in perspective' also notes that there were both American- and British-type of the 2-piece blocks.

 

[wuzak]

D Deleted member 68059 - a member on the another forum posted this:

According to Volume 1 of 'the magic of the name' states the following regarding the two-piece block used on Packard engines, 'their early engines, Merlin 28's were similar to the Merlin XX, but had a two-piece block designed in the USA.

(my bold)

I've read in your book that 2-piece blocks were slated for Merlin 60 production at 1st; care to add something that might clear the picture about who designed the 2-piece block for Merlin?
FWIW, the book 'The Merlin in perspective' also notes that there were both American- and British-type of the 2-piece blocks.

Rolls-Royce designed the two piece block, but Packard put it into production first since Rolls-Royce was a bit too busy churning out engines for the changeover.

The change to the Merlin 61 allowed Rolls-Royce to change to the 2 piece block.

I believe the Packard initially used a different solution to connecting the water passages in the block and head, but this was later changed to the Rolls-Royce design.

 

[Crimea_River]

.......I believe the Packard initially used a different solution to connecting the water passages in the block and head, but this was later changed to the Rolls-Royce design.

Water passages between the cylinder block and head on the Packard Merlin 29 shown below. 54 individual brass connectors with 2 or 3 O-rings each fit into tapered holes in the block and head. Getting proper alignment of each of these while installing the head is a MAJOR pain in the butt.

 

[Deleted member 68059]

D Deleted member 68059 - a member on the another forum posted this:

According to Volume 1 of 'the magic of the name' states the following regarding the two-piece block used on Packard engines, 'their early engines, Merlin 28's were similar to the Merlin XX, but had a two-piece block designed in the USA.

(my bold)

I've read in your book that 2-piece blocks were slated for Merlin 60 production at 1st; care to add something that might clear the picture about who designed the 2-piece block for Merlin?
FWIW, the book 'The Merlin in perspective' also notes that there were both American- and British-type of the 2-piece blocks.

Naughty Tomo, thats on page 409 of my book, you`ve not read it yet ?

(this is a quote from a Packard document)

This is reference #21 of Chapter 1945:

 

[pbehn]

This all gets down to the core assertion that RR tolerances were crap until the US automotive industry fixed the problem. It's absolute nonsense. You can't hand-pick components in a construction line and still maintain throughput.

The criteria for a sub-assembly would be a QC check but you don't check every sub-assembly on a production line, you check a percentage. To achieve a 1oz margin for the entire assembly of crankshaft, pistons and con rods, then it suggests upstream processes would output components such that any 12 pistons and con rods, when connected to the crankshaft, could be expected to fall within limits with a very high degree of repeatability. The alternatives are a slowed production line and either wasted production or go back to highly-skilled individuals to fettle out-of-spec sub-assemblies back into spec. The problem with the latter option is that it doesn't help detect upstream problems that caused the out of tolerance issue in the first place...so you're back to your upstream tolerances.

I'm afraid this whole issue of RR not having robust production line processes for Merlins prior to US automotive companies getting involved is absolute nonsense and the more it gets repeated, the angrier it makes me. It smacks of nationalistic willy-waving rather than actual technical study and reasoned analysis. Not one of the "hand crafted by elves" stories has EVER come up with a document that shows how poor RR specs were compared to the U.S. automotive industry, indeed most are of the "I knew a guy once who worked [insert location here] who always said that...." That's not proof or evidence, it's hearsay or folklore.

I just read in a book on the Lancaster that the Ford shadow factory in UK required a special machine made in Switzerland to achieve the tolerances they wanted, it could machine to 1 millionth of an inch. They eventually imported it via USA, it took three attempts, the first two are somewhere at the bottom of the Atlantic.

 

[brianwnz]

Well, a little bit off the Merlin topic, but still on tolerances - here's an old Allison document regarding piston balance to within 1/4 oz. for the set (of six) for the Allison Navy Diesel in the late 1920's.
Now before you assume this is better than Ford or Merlin, recall that they talk of the reciprocating mass of the piston, rod, piston pin & bearing, which if all had a 1/4 oz. on the plus side, could add up to 1 oz. as an assembly.

It does highlight an example of some US factory tolerances (for a one-off engine) from Allison in the 1920's/30's..... and I just wanted to share the drawing
.....If anyone has any earlier/later Navy Diesel piston drawings, I'd appreciate a copy!

Thanks to John and RRHT-Allison Branch for the drawing ......

And sorry to move off the Merlin engine if anyone is annoyed - there's some great information been shared here by others.