could the Allison engine have done what the Rolls Royce Merlin did?

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

I found this

In October 1939, City of Flint was carrying a cargo of tractors, grain and fruit to Britain. On 9 October, the German pocket battleship Deutschland seized the City of Flint, declaring her cargo to be contraband and the ship a prize of war. A German prize crew was put on board the ship to sail her back to Germany.[13][14]

At 01.50 hours on 29 Oct 1939, U-34 fired two torpedoes at two steamers and one destroyer in convoy HX-5A about 180 miles west of Lands End and claimed two hits. In fact, only the ship of convoy commodore, the Malabar (Master Henry Herbert Armstrong), was hit and sunk. Five crew members were lost. The master, the commodore (Rear Admiral G.W. Taylor, RNR), two naval staff members and 66 crew members were picked up by HMS Grafton (H 89) (Cdr M.S. Thomas, RN) and landed at Plymouth.
 
The RR files I have say that the Merlin 1 engine was never properly tooled up, I.e. it wasn't mass produced in general principle. So we can see that sometime between then and about 1940/1941, RR was in some transition phase between knocking up a few merins at derby and managing a series of large factories mass producing them.

So the simple answer to the source of this silly myth is that the notion of how RR made engines came from anecdotes from this transitionary period.
 
I read somewhere on the story of the Ford Manchester plant that the first discussions between RR and Ford were with Ford France. The Trafford Park factory in Manchester started work to become a shadow factory in 1938. Mass production is a science and a skill, it would have been foolish of RR NOT to pick the brains of Ford engineers if they had the chance.
 
I have alway assumed that the comment/story re Packard tightening up tolerances was referencing the need for Packard to tighten its own tolerances. The tolerances for the Packard automobile engines of the time were not as tight as the tolerances on the Merlin engine.
 
Totally agree re: New Factory design for specific application. Under any circumstances, Ford would start with the Process Plans for each part, carefully consider the machine tools required to make them, to world class process plan(s) desired for each class of parts. The next consideration is to match as well as possible the flow of parts from one set of machine requirement to the Next and plan for common batteries if possible to minimize movement (and tracking) of parts - with an eye to complete as much of a complex part as possibe in one machining/process center as possible. Quality and throughput time drive costs.

Part classification systems arose in 70s' to aggragate families of parts as designed to combine with processes to provide insight on better planning for throughput. Battery environments such as vehicles are less difficult than bespoke manufacturers that design and build 'common' theme parts in many variations (Wellsite drill tools for example). Airframe companies are notorious for what 'looks like battery' but are more like bespoke because of the complex shapes of machine parts from one aircraft to another - often better served by subcontracting than purchasing five axis mill machines for deep pocket forgings as embedded in F-111 swing wing installations.

One goal of Group Technologies is to match as clsely as possible families of parts with certain design attributes (shaft versus aforementioned swing wing support) to process plans which can be grouped into one concentrated batch of machines - and be completed in that group, thereby eliminating all together the cherished tradiions of parts chasing from one end of the factory to another.
 
ThomaP - while I don't discount the possibility that Packard needed to 'tighten' its own tolerances, I don't think it can be based on facts unless access to RR drawings and comparable Packard drawing can be compared.

An industry (American automotive/engine) that thrives on reliability supported by global distribution of interchangeble parts strongly suggests that neither Packard nor Ford needed instruction on tolerances to support that vision?

That said, I don't have that history and knowledge that youand others have of the R-R, Ford and Packard deliberations and outcomes in the engineering dimensional changes to argue the point.
 
I do not think it is a case of Packard needing instructions on tolerances so much as it is that the required tolerances for the relatively small in3 and low HP/in3 automobile engines of the time would have been significantly different than for relatively large and high HP/in3 engine like the Merlin/V-1710/DB601/etc.

Using the crankshaft as an example:

As pointed out by Snowygrouch in his post "An interesting read about the Packard built Merlin engine." the Merlin crankshaft tolerance of +/-0.00025" was ~twice as tight as modern crankshaft tolerances for automobile engines. This was an amazingly tight tolerance for 1940, and was presumably a necessary tolerance.

It was not until the early-1990s that holding diameter tolerances of +/-0.00025" became relatively common on machined parts. This was due to a combination of increased capability of modern computer controlled turning and machining centers and the associated reduced cost of achieving the tolerance. Prior to the early-1990s a tolerance of +/-0.0005 was the best commonly achieved. During WWII +/-0.001 was the tightest commonly achievable tolerance.

(A general rule of thumb in machining is that for every time you reduce the tolerance by half you double the time it takes to make that particular feature.)

The cost of a Merlin engine (I think) was somewhere around $6000 in 1941. The cost of the most expensive Packard sedan was ~$5800.

There is no reason to believe that the Packard or Ford automobile engine of 1940 was held to tolerances as tight as modern high performance automobile engines. It would not have been economical or necessary within the industry needs of the time.

I know that if I were approached to manufacture a scaled up engine (such as from a 356 in3 V8 to a 1650 in3 V12) one of the first things to go through my head would be that - due to the size alone - tolerances would have to be tighter in order to achieve the required cumulative runout/ straightness/ flatness (to manage vibration and sealing problems) and to maintain fit/interchangeability of parts.
 
Last edited:
Tolerances dont always scale up, you have to work to different tolerances or have different ways of doing things.
 
To further complicate or colour the debate, the RR Trafford Park factory was set up to be near the Vickers Avro Manchester factory, but the Manchester used Vulture engines. Lost in the sands of time are the discussions of what the Trafford Park factory would make Vulture or Merlin, we know when the Vulture was cancelled, but prior to that it must have been increasingly "on the back burner". Also I read that they imported a special precision machine tool from Switzerland for the main bearing journals, the first was lost when a ship was sunk (maybe one of those Callum posted about) so they got another, ordered from the Swiss but shipped via USA due to "hostilities".
 
True - but debate between Packard and R-R is simply cut off by R-R in the context "we are the customer, giving you a contract to meet our specifications depend on a.) Willingness to do so, and b.) capability to do so, and c.) adhere to the QA/QC demands contained theerin".
Not to belabor the point, but didn't Packard deliver the required tolerances required by R-R? As a tangent, the tolerances achieved dimensionally and spherically exceeded that in the 1945-1950 timeframe at Lawrence Livermore, Hanford, Oak Ridge and Savannah River nuclear weapons designs and fabrication. The tolerance issues were achieved by combination of computer controls and polishing techniques. Pause - I know this anecdotally, not by hands on observation. The increased tolerance achivements were requiredto improve efficiency of nuclear chain reactions
The point is well taken and illustrates the imbedded direct costs and overheads associated with producing the Merlin and the Packard. There is no question that either engine required significantly more investment in machine tools and processes than associated with a Pacakrd (or Ford) vehicle of the 1940s. But recognize that tooling is depreciated from the Balance Sheet while overhead is baked into Indirect Costs - who knows what the unit cost (not price) was unless the books are available in detail. I assume (bad word) that the Packard contract was let as Cost Plus with exquisite details contained in R-R/Packard Agreement
There is no reason to believe that the Packard or Ford automobile engine of 1940 was held to tolerances as tight as modern high performance automobile engines. It would not have been economical or necessary within the industry needs of the time.
Agreed again - but the question being asked is 'Could They', and what extra ordinary steps, if any, would have been required to a.) acquire the tooling, b.) firm up the process plans to machne, inspect and approve the 'extreme' (greater than auto, but within reach technically) tolerances required for criical parts, c.) acquire and install the upgraded tools (i.e including making in-house, d.) train for the upgraded processes and equipment. All this would have led to pricing and contractural framework to be Paid.
Absolutely agreed. I will bow to your hands on experience in the related field, noting that even at Bell Helicopter and Vought and Lockheed were uch considerations and design attributes requiring 'flatness' and 'sphericity' beyond standard manufacturing strike zone.

I was Program Manager for 18 mo for GE contribution to AFCAM and the one customer requiring Q Clearance was mean little prick with last name Rickover - whom I really respect and would have possibly feared had I ever worked in his chain of command. He had a ost interesting approach to make on 'uncomfortable' during his personal interview. If you failed it - you were out - and Nixon couldn't over-ride.

General Dynamcs was a major component of the design and manufacturing corps we looked at. That included all major contractors that specifically had huge investments in tooling ranging from Jeweller's lathes to extreme tolerance capable 5-axis bad boys. There is a story behind this, but the nuclear sub business (and associated Contractors located in CA and WA and N and SC) were a major focus.
 
I am not saying in anyway that Packard or Ford were not able to meet the specifications. I am only addressing the possible origin of the statement(s) (I have read accounts of the statement as coming from the Ford rep or the Packard rep) that 'they would have to tighten the tolerances for mass production'. Such a statement would make sense if it was addressing a need for Packard or Ford to adopt tighter tolerances than they usually used for their mass produced automobile engines. It would not make any sense otherwise.
 
I think you are getting the wrong end of the discussion. The whole "thing" stems from a discussion between Cyril Lovesey of RR and someone from Ford in the early days of Ford being involved in RR production. To paraphrase the guy from Ford said "we cannot manufacture to these tolerances" to which Lovesey replied, are they too demanding" the Ford guy said "no, to make car engines we use much tighter tolerances because every piston has to fit in every cylinder". However at the time Lovesey had no experience of production engineering, his main work was on the racing engines, where you probably do machine the bores and then machine pistons to fit.
 
I admit I got a little down into the weeds but the point I was trying to make was that GM was investing very large sums of money in all sorts of things (appliances, air conditioning, leaded gasoline) in the 20s and 30s and they continued to do so even in the depression. For example, they bought Opel for 33.36 million in 1931 and built a new factory for it as well. Other examples include buying Vauxhall and Holden and launching a new truck brand in the UK (Bedford).
Perhaps a better way to state my case was that GM became a billion-dollar corporation in terms of market valuation in 1926. In 1929 they were neck and neck with Standard Oil (New Jersy) for the title of world largest corporation with a revenue of $1.5 billion. Obviously, those numbers dropped over the next few years, but GM was always profitable throughout the depression. Even if you cut that revenue in half the $500,000 still qualifies as petty cash to a company the size of GM. What I draw from all this is that GM had little faith in Allison. That is why I asked how much GM invested in North American. The move from the east coast to California and the establishment of a new factory was not driven by military contracts and must have been paid for by GM.
 
Hey pbehn,

By the time that discussion took place (1940?), several thousand Merlins had been produced. By that point in time RR and associates would have sorted out the process such that at least 90% of the pistons and 90% of the cylinder bores would meet full interchangeability standards, and probably more like 95% and 95% - at least in terms of machined dimensions. The casting weights might have tossed a problem into the mix, such that the pistons might have been sorted by weight? I have seen Merlin pistons that were machined into weight specifications by removing material from the inside-bottom of the casting but I do not know if this was the normal method during the war, or if they were supplied in matched sets.
 
And despite the size of GM (per your numbers which I wont dispute) $500,000 in the late 30s is about 12 million in today's dollars. I don't know if you actually worked in the aviation manufacturing industry, but I seen managers canned for not making profit margins, let alone losing a fraction of that amount.
 

Posts like this are why I love this forum. There's so much to learn!
 
I dont think it is stated when the discussion took place because it makes a better story if it isnt, it also isnt stated that the discussion was with Ford not Packard. By 1940 RR were already mass producing Merlins in Crewe and the Manchester and Glasgow factories were either in production or tooling up. My guess is the discussion was much earlier, when Ford were first starting to work on the Trafford Park factory so 1938-39.
 
Actually the move from Dundalk MD to Inglewood Was driven by the BT-9 and then the BC-1 follow up funded the primary expansion at Mines Field, Inglewood. Just prior to that NAA got its foot into the AAC world with the XO-47 (Kindelberer/Schmued design).

GMC bought NAA in 1928, then bought Fokker Aircraft and formed GMAC to operateboth NAA and Fokker - then acquired Berliner-Joyce Aircraft. After the Knute Rockne crash in a Fokker F-10 in 1931, the Teterboro plant was closed and remaining folks moved to Berline-Joyce plant at Baltimore airport and Dundalk, MD. About this time, the GMC Board exec Ernie Breech hired Kindelberger. The YO-47 contract was let as the first external contract award obtained by NAA.

GMC funded $125,000 toward General Order GA-15 (XO-47) for competition in the three seat observation competiion. Shortly afterwards GMAC took the name North American Aviation.

When NAA decided to compete for the two seat monoplane trainer, (future BT-9), GMC invested another $16,000 for the NA-16 proposal effort. I believe that GMC might have financed several loans related to production assets after that but have found no records to substantiate.
 
The contract/s for the 0-47 involved 239 aircraft. The last contract was for 74 aircraft placed in 1938 so NA was either making money or repaying loans.
Please note that Curtiss only got orders for 215 P-36s from the US government (and about 700 export orders) So NA was doing pretty well compared to other US Airplane makers.
Everybody had eyes on the future.
GM's problem with Allison was that until the order came for the P40 engines Allison was pretty small potatoes in the engine field.
In 1938 they delivered 12 engines, in 1939 they delivered 48 ( for YP-37s, Airacudas and prototypes)
Wright built 1800 R-1820s in 1938, not counting the smaller Whirlwinds (249) and 33 about R-2600s
P & W built 695 R-1830s in 1938 and was building R-985s, R-1340s, R-1535s, R-1690s. They had been working on the R-2800 since 1936.
The US Airlines and commercial aviation showed no real interest in liquid cooled engines and the Navy, once airships were gone, were only showing polite interest. (just making sure they didn't miss something) but that left the Army as pretty much the only customer and the Army had their own oar in the water with the Continental hyper engine (largely designed by Army officers/employees) . And Lycoming was chasing the same market.

GM In other areas could see a way to dominate a market or at least be a major player. It was one thing to loose money in the short term to reach a new market or to create a new market. But the market was looking pretty iffy over the winter of 1938-39 for Allison. Wright and P & W were getting overseas sales in huge quantities.
Allison had no real production facilities.
 
If I may throw some auto data into this mix, Packard was a high end builder. There were no cheap Packards. Their competition was Cadillac (GMC),Lincoln (Ford), Pierce Arrow and some semi-custom companies such as Duesenberg and Auburn. Packard top engine in 1939 was their V-12 of 473 cubic inches. Packard engines were well built and reliable with close tolerances in the automotive world. They were the most logical choice to build Merlins.
 

Users who are viewing this thread