An interesting read about the Packard built Merlin engine. (1 Viewer)

Ad: This forum contains affiliate links to products on Amazon and eBay. More information in Terms and rules

That is a short little piece but I am impressed that it actually quotes from Sir Hooker's book about the automotive parts having to be closer tolerance than the RR parts. Note that RR never did change that practice. The Allison TF-41 engine was a American built RR Spey and AFLC was distressed to find that the American version still followed the RR practice of match drilling bolt holes through assemblies rather than making them all as close to the same dimensions as possible. This wrecked the AFLC approach of taking all engines parts, throwing the parts into inspection, cleaning, and repair and then having parts from different serial numbered engines be assembled into an engine. The USAF had huge problems with the TF-41 and I rather suspect that at least some of them came from the fact that OC-ALC had effectively sabotaged the whole fleet. I was sent to MBAFB to get their A-7D's back in the air because the engines were being removed and replaced so often that the guys in the field had come up with a labor saving method that warped the bleed air ducts.

A friend of mine also told me that years ago he was working for a US airline that used RR engines. He went over to the RR plant and watched as they modified a fan blade assembly right on the production line and then gave it a specific unique part number that reflected the fact it was different than the rest.
 
Note that RR never did change that practice.

Please provide a source for that.

RR made more Merlins than anyone else (over half the global number), its clearly utterly preposterous to claim they did so filing down each bit to fit each engine, I`ve read that article - and its nothing more than a junk-news recyling of badly understood rumours with absolutely zero archival evidence whatsoever to back up any of it. The famous comment from Ford about tolerances has very little to do with the realities of Merlin production.

RR set up a factory in Glasgow which I can see the site of from my bedroom window, to build Merlins, staffed by totally semi-skilled labour (40% female) who had never so much as assembled a lawnmower engine before. Its ludicrous to imagine they did that with hand-fettled parts once it had got going seriously.

Nobody doubts that Packard did an even better job of mass producing the Merlin, in terms of streamlining the production process but that article is utter junk.

"Rolls-Royce had never implemented mass-production techniques to their assembly lines"

Yet..... Glasgow factory which started production before war had even broken out produced almost 24 thousand Merlins, and started well before Ford or Packard had gone near it;
Presumably all hand filed by inept British bumpkins ? :rolleyes: who somehow "hand built" each engine whilst being bombed....I have absolutely zero reason to "support" RR, but I have a serious bugbear about junk "history" online written by people with absolutely no idea what they`re talking about.
 
Last edited:
My "source" was the USAF experience with the TF-41 as well as my friend's experience with RR jet engines. In addition a professor in college told me he flew P-51's in WWII and his had a "real" RR Merlin, not a Packard, saying there was a big difference, the RR Merlins having their pistons individually selected so that they were as close as possible to the same weight. The professor said it made a big difference in how smooth the Merlins ran.

So, in WWII they were using the craftsman approach to build Merlins and in the 1960's were using a similar approach with jet engines. Does not sound like much of a change in philosophy.

Now, I have also read that the Spitfire XVI used Packard Merlins and that they had a lot more reliability problems with them. This does not seem to match the P-51 experience with the same engine. Admittedly the Spitfire used a much smaller propeller and I have no idea what that would make a difference.

By the way, did RR use the Shadow Factory approach with Merlins, with pieces built by cottage manufacturers?
 
The comment was made by Lovesay who was on the development of the Schneider trophy engines, learned to fly and set up a flight testing centre at Hucknall. He worked on the Merlin development programme and when speaking to some guy from Ford was probably just trying to be polite and get to the bottom of the issue.
 
A friend of mine also told me that years ago he was working for a US airline that used RR engines. He went over to the RR plant and watched as they modified a fan blade assembly right on the production line and then gave it a specific unique part number that reflected the fact it was different than the rest.
Maybe it was for an engine that was different from the rest.
 
I do think and i stress think, you are quite right. But....thinking is not facts in this case.
And you do not bring them. Logical yes. Facts no.

Please provide a source thats supports your thinking.

The Merlin production figures for each plant are so well known they`re all neatly listed even on Wikipedia.

Rolls-Royce Merlin - Wikipedia

As for my confidence in my assertions, I have been to Rolls-Royce Derby archives, Rolls-Royce corporate archives, Rolls-Royce Bristol archives, and
read all the personal correspondance of Ernest Hives himself - and in addition been invited to lecture at Rolls-Royce about the German
opinion of the Merlin engine.

So I dont have any time for nonsense history like that article.

The notion of each engine from Britain being some sort of cobbled together junk-heap with each bit fettled to fit is dispelled by a simple mathematics
excersize in looking at the number of engines, and the time in which they were built. Everything in Britain from about 1940 onwards was being mass produced,
if it hadn`t been, there`d have been no Merlins, during the Battle-of-Britain every single Merlin came from one production line at Derby, as the output
from the other factories (also RR run) hadnt come online yet.

I would imagine that a lot of hand fettling happened in Britain, as we were getting bombed, and the Merlin needed a lot of early modifications to cure
all sorts of defects, so changes would have been happening constantly. But this is all totally misrepresented by one quote about tolerances, which
has been used to create a historical nonsense. In addition most repairs and modifications were made by RR to each engine when they came back
to the factory for a rebuild. The whole system was exceptionally well organized and RR had total control, when the RAF tried to have their people
take care of the engines, Hives overruled the Air Ministry and said he would only supply engines if RR had full control of field servicing by RR
personell. These people imagining the British at the time were bumbling about like simpletons have no idea. RR had an utterly ruthless grip on aero
engine design, manufacture repair and servicing from start to finish - to an extent which in some respects makes the Germans appear amateurish
(that doesnt meant I`m suggesting the Americans were less professional, I`m simply stating that saying Packard were wonderfully professional
and did a marvellous job DOESNT mean the Brits were floundering about like blind lambs.

The number of sub-contractors supplying Merlin bits for all the RR run factories in England also shows the "hand fitting" story to be nonsense,
all these shadow plants and suppliers would never have time to get hundreds of rejects sent back, and would have been fined and lost their
contracts. They managed to do all this because they all had proper drawings whereby people with no aero engine experience whatsoever
could make Merlin parts. How it is imagined that this all happened by any other means is a total mystery to me.

I also happen to have all the Merlin reliabilty data, and it shows he RR built engines to be excellent, again, making a nonsense of how
else 80,000 Merlins were somehow all "hand-made" by magic, when Packard needed a perfect production plant operating like a swiss
watch to make 55,000. Quite obviously, both firms did a world-class job of mass producing their variants, the Packard effort being
impressive as they made many improvements and did so without the expereince of RR staff at as close a hand as they were in Britain.

As for the tolerances, Available at Kew National Archives in London
AIR-10/2175 "Merlin-I Fits and Clearances October 1938"

Merlin I Aero-Engine: Part II Schedule of Fits, Clearance and Repair Tolerances | The National Archives

Even in 1938 the Merlin crank (for example) had a limit on the journal diameter of +/-0.00025 of an inch,
and the tolerance stack on the clearance was 0.75 thousandths.

This is not far off modern automotive steet tuning best practice.

https://www.mahle-aftermarket.com/media/local-media-north-america/pdfs-&-thumbnails/cl77-1-205r.pdf (Mahle Powertrain / Clevite are one of the worlds largest
manufacturers of automotive crank bearings today)

"Let's pick some typical manufacturing tolerances and look at the potential clearance range that results. A tolerance range (from min. to max. sizes) of .0010" is typical for most crankshaft journals"

================================================================================================================
So let me condense that down. A Merlin Crank from 1938 in Britain, before the production even really got started had a crank journal diameter grind limit 50% tighter than a current automotive production car engine (mahle document dated 2005, so: 67 years later.....we`re half as good 1938 Merlin.).
================================================================================================================

Tols.png


A page from the Merlin XX "fits and tolerances" manual dated May 1945.... its the same. So apparently no improvement was needed either....although the clearance has been cut a bit
all the tolerances remain identical from the Merlin-1 in 1938 to 1945.

Source: Merlin XX, 21, 22, 22A, 23, 23A, 24 and 25 Aero-Engines | The National Archives

1945.png


I`m not suggesting that Packard didnt change any tolerances to suit themselves, but I am proving that its utter nonsense to suggest that even the Merlin-I wasnt toleranced to a level necessary for mass production, even by todays standards.

We can also compare to Daimler-Benz DB600 series, with the German reputation for prescision, the DB605 crankshaft drawing for example shows
the crank main bearing journal grind tolerance at +/- 0.0004"

So the 1938 Merlin-1 crank journal grind limit is also not only better than a modern engine but tighter than a Daimler-Benz.

crank.png

Before anyone says "yea but wartime Germany..." thats the same tolerance as the Swedish licence drawings for the DB605 dated November 1945 (which
are not DB drawings, but entirely redrawn).

(Kurbelwelle = Crankshaft)

"D" is the crank journal diameter, and its 100mm g6, which you can look up in the table at the end, its in "mm" so you`ll have to convert.... -0.034 -0.012mm = 0.022mm delta
= 0.022/25.4 = 0.00086" = +/- 0.00043"

2019-12-27 23_36_15-A4029.jpg - Windows Photo Viewer.png


Tols.png



Tols.png
 

Attachments

  • Tols.png
    Tols.png
    115.9 KB · Views: 334
Last edited:

To nail the lid firmly shut on the quality of this article

"They even tightened critical bolts by trained feel, rather than with calibrated torque wrenches. In effect, each Rolls-Royce-manufactured Merlin was a hand-built engine"

Meanwhile: Actual Merlin assembly manual page:

Not only is it full of torque wrench numbers, but even tells you which EXACT wrench to use for each bolt. :rolleyes: I`m not trying to open up another
"us vs. them" argument about Rolls-Royce and Packard, I`m just saying, if people write articles about this, and refuse to even do a
a modicum of research first - its not very useful to anyone.


Torque_2.png
 
Last edited:
They even tightened critical bolts by trained feel, rather than with calibrated torque wrenches. In effect, each Rolls-Royce-manufactured Merlin was a hand-built engine"

I remember years ago at the Motorcycle show at Birmingham NEC a Torque Wrench Manufacturer had a demo stand where anyone could try and torque up a series of bolts by feel. The guy running the demo stand said no one had ever managed to get 2 bolts to the same torque and even experienced mechanics regulary overtightened a bolt. I had a go and was shocked to be told I had applied double the required torque.

Getting a bolt to 300 inch pounds (which iirc is about 25 foot pounds) dozens of times a day 6 days a week would be impossible without a regulary calibrated Torque Wrench.
 
Sometimes the Americans and the British measured things a bit different (See measuring the bore or pistons on the Allison and the Merlin, one measures along the wrist pin, the other measures at 90 degrees to wrist pin, doesn't mean one is slopper than the other.
In some cases a spot of filing might have been done to a piston. However that would be for weight and not dimensions.

Overhaul directions for a Merlin say "The permissible variation in weight for one assembled pair of rods, pistons, pins and rings, in an engine, is 1 oz."

In some cases they may have drilles a few blind holes in the skirt or bottom of the piston to reduce the weight.
There is an almost two page chart on "Pistons and connecting Rods-Fits and Clearances" with 33 ref. numbers/entries. With new and allowable wear tolerances.
 
The following video shows Ford of England making V8s in 1936.

All the nuts and bolts are banged on with air wrenches, not a torque wrench in sight. A lot of hand work I would say. Compare this video to the Rolls Royce video I posted previously.

C E Harrison was a Rolls Royce employee who worked in the engine testing section. He went to the US for pilot training but ended up at Packard joining them in March 1942. The following quotes are from his 1984 lecture to the Royal Aeronautical Society as reprinted in the fall 2004 issue of Torque Meter:
"...the Americans were used to coarse, heavy threads whacked up with air guns, so we had to establish torques on all nuts and bolts and issue torque wrenches to all engine builders. Packard really made every effort to do a good job and made instant dismissal the penalty for anyone found with a hammer or file in a tool box."
"There were some some features of their engines that were magnificent, one being the finish of bevel gears, but in general it was hard to distinguish between a Rolls Royce part and the Packard equivalent."
 
Sometimes the Americans and the British measured things a bit different (See measuring the bore or pistons on the Allison and the Merlin, one measures along the wrist pin, the other measures at 90 degrees to wrist pin, doesn't mean one is slopper than the other.
In some cases a spot of filing might have been done to a piston. However that would be for weight and not dimensions.

Overhaul directions for a Merlin say "The permissible variation in weight for one assembled pair of rods, pistons, pins and rings, in an engine, is 1 oz."

In some cases they may have drilles a few blind holes in the skirt or bottom of the piston to reduce the weight.
There is an almost two page chart on "Pistons and connecting Rods-Fits and Clearances" with 33 ref. numbers/entries. With new and allowable wear tolerances.
There is a misconception that Rolls Royce was doing something unusual in weighing and matching pistons. At the 18.00 mark of the following video you can see that Ford weighted and matched the pistons and con-rods for the V8s they were mass producing at the River Rouge plant.
 
There is a misconception that Rolls Royce was doing something unusual in weighing and matching pistons. At the 18.00 mark of the following video you can see that Ford weighted and matched the pistons and con-rods for the V8s they were mass producing at the River Rouge plant.

The first thing after market tuners used to do was lightening and balancing.
 
Sorry update to issue, someone elsewhere pointed out that the Torque wrench part number was post-war, and it seems that booklet was re-printed in Nov 1945. So here is another Merlin manual page from 1944 to confirm wartime usage.
1578176441724.png
 

Users who are viewing this thread

Back