Merlin vs. DB601

[GregP]

The Merlin has 11,000 parts. The Allison, also a V-12 of similar displacement and performance, has 7,000.

4,000 is a LOT of screws and bolts, especially considering that the Allison is also held together with screws and bolts! And Allisons don't leak anymore than Merlins do.

The seal between the top and bottom crankcase on an Allison is metal to metal with one small silk thread all the way around with overlapping ends as the entire gasket. The rest is just tightseal, which can be used on ANY engine. Nothing wrong with the Merlin design, but one or two people can get an Allison mostly apart in about a day and a half or less. Ditto for reassembly. The hard part is prepping the parts for proper fit, just as in ANY piston engine. Correct prep is everything.

You have to torque the cylinders on a Merlin every 25 hours or so. Allison cylinders are torqued once, upon assembly, and are never a problam after that. Of course, the cylinders are of Different design and Allison cylinder nuts require 2,200 foot-pounds of torque!

And as a matter of information, I looked at our Merlin for the Hiapano and was wrong. It is a Merlin 224, not a 228. Being a 20-series, it has a 2-speed supercharger and the gearshit is at the bottom of the supercharger case on the left side. Once I was looking for it, the lever was quite obvious and was confirmned by Steve Hinton. Sorry, no pics. The camera was 1/2 mile away at the time.

 

[fastmongrel]

Until someone gets 2 engines an Allison V1710 and a RR Merlin in equivalent condition with the same accesories then strips them down to the very last washer and enters the details into a spreadsheet no one is going to know how many parts there are in each engine. We dont even know which engines supposedly had 11,000 or 7,000 parts it might be comparing a 1930s prototype V1710 and a Merlin 131 with the reversible prop.

Is the number of parts for a V1710 a bare engine is the parts number for a "Merlin Power Egg" an engine with everything bolted on from the prop shaft to the little link that moves the widget that pushes the doofer that starts the waffle pump that turns the flange that spins the dickey that makes the engine go BRMMMM.

I have never worked on a RR aero engine but I have worked on older RR car engines and every single nut, bolt or screw had a washer, every single cable, hose and pipe was clamped in at least 1 place sometimes 2, with a fibre pad under the clamp and a washer on the bolt, nut or screw. Oh how I wish modern cars were put together the same way rather than a thousand brittle plastic clips that require special tools to open and a box of spare clips to replace the half dozen you break every time.

 

[Shortround6]

I certainly don't have the experience to tell where all the "extra" parts are but lets be logical. Like I said earlier, a Merlin, an Allison and a DB 600 series all use the same number of pistons, connecting rods, piston pins, valves, etc, so were do the "thousands" of extra pieces go? Nuts, bolts, washers (fiber or metal) clamps? they all used shaft drives to the cams, not gear towers.

I can certainly understand how, with extra parts, the Merlin could be more of a pain (need more man hours) to tear down and re-assemble. That does NOT mean it would be less reliable or break down more often. That would assume that ALL the parts were made of the same materials, were manufactured the same way, and had to handle identical stresses. I don't know how many changes were made to an item like the Merlin or DB crankshaft. We do know that the Allison went through at least 4 changes in crankshaft and that just shot peening the crankshaft without any other changes in dimensions or material added significantly to it's fatigue life.
There is too much we don't know about the engines to make the assumption that the relative number of parts had anything to do with either reliability or longevity. Assuming the engines were assembled with due care.

For all I know the R-R use of smaller but closer spaced fasteners allowed for better sealing (anyone with experience with most old English gaskets can appreciate this) or lighter cam covers or other access plates due to less 'bending' near the fasteners.

 

[RCAFson]

How many Allison engines had two stage, two speed SCs? Are we comparing apples to oranges?

 

[swampyankee]

Well, power wise both contemporary engines were rather close. IMO the DBs offered some advantages.
Inverted V-design offered better view over the nose,
fuel injection ensured supply at all flight situations,
hydraulic supercharger drive was more efficient,
side mounted supercharger allowed an engine cannon.
cimmex

The hydraulic supercharger drive was not more efficient; its variable speed was by deliberate slippage in a fluid coupling. This may have resulted in improved system performance, but it was not "more efficient."

The inverted-V may have helped visibility in some installations, but not when an annular radiator was used.

The side-mounted supercharger may have permitted an engine cannon, but it also forced some amount of asymmetry in the combustion conditions between the left and right banks of the engine.

Fuel injection was an advantage, especially over the float-type carburetors preferred by Rolls-Royce. The pressure-type carbs used by the US companies did not have that problem.

 

[davebender]

inverted-V may have helped visibility in some installations, but not when an annular radiator was used

1930s German fighter aircraft designs did not employ an annular radiator. So the inverted V12 design helped fighter aircraft visibility in all circumstances prior to fall 1944 Fw-190D9.

 

[cimmex]

The inverted-V may have helped visibility in some installations, but not when an annular radiator was used.

Here I disagree. Only the lower Prop axis of an inverted V-engine allowed the use of an annual radiator. I don't know any plane with an upright V-engine which had an annual radiator
cimmex

 

[GregP]

Hi Shortround,

In your post above you state higher CR makes for better fuel consumption. I believe that is for naturally aspirated engines. Once you start adding boost, the CR is not as important as the final cylinder pressure, and that is controlled almost exclusively by how you employ the boost. If you are running a CR of 5.0 and are running 140 inches of MAP up near the limits of detonation, the cylinder pressure is equivalent to a CR of 7.0 and some reduced boost at the same detonation limit. Do you agree?

Now if you are running at cruise, then you'd have to adjust the boost to give equivalent cylinder pressure to get equivalent fuel consumption. The performance might not be quite the same since the gearing also might not be, but the fuel consumption should be very close at identical internal cylinder pressures.

As for the parts count, I worked at a shop specializing in Allisons and ONLY Allisons. I think we know how many parts it has since we were assembling them. We never built or have seen a "C" engine inside the shop (DO have some parts), though we could certainly have added a "C" type nose case to any existing -89 or above is anyone really wanted a "C" nosecase (if anyone can find one ... try South America in a museum or barn ... that's where the last 2 -3 of them were found). The parts count is for a typical E, F, or G engine. Most of the E and F engines we built were of the -89 / -91 series and up. The difference in parts count between an "E" and an "F" is miniscule. Both have a nosecase and internal bearings, the "F" has internal gears. The difference is that the "F" has a shaft designed to mount a propeller and the "E" has a mounting flange designed to interfce with a U-joint for driveshaft operation of a remote gearbox such as in a P-39 or P-63. The same nose case can be used in a PT boat or a tank, but is usually replaced with a case employing a flywheel weight for smoother non-aviation operation. The parts count delta might be 5 - 8 parts between an "E" and an "F". Within a count of 7,000 it makes zero difference.

The "G" is the last of the higher-power Allisons and has the same parts count as an "E" or "F" but some of the parts are stronger. All the front-running Merlins at Reno are running Allison "G" series rods since the Merlin rods will not live at the power levels they are producing today.

The 11,000 count for the Merlin is for a single stage, single speed engine, but the count is almost identical for a 2-stage engine, no matter the speeds. Adding an impeller, 3 - 4 gears and 2 - 4 bearings makes almost NO diference in parts count at a level of 11,000 parts, and the case is a unit with minimal parts count increase (some extrra bolts). The total delta between a single-stage and a 2-stage might by as high as 50 or so parts out of an 11,000 count. The percent parts difference is well less than 1%, so it makes zero diffrence in an analysis of parts count.

As for the 2-stage Allisons, I am not interested in anything bedore the -7 engine since they were all pretty much experimental before that. If you are interested, you can research them. Starting with the V-1710-7 there were 75 models through the V-1710-149. Of these, 12 models had auxilliary supercharger stages, These were the -93, -97, -103, -109,-109A, -111, -113, -117, -119, -121, and -123.

The V-1710-57 had a single-stage, 3-speed supercharger.

If the Government had funded an integral 2-stage unit, it might have been nice. Unfortunately, it didn't happen.

 

[swampyankee]

Here I disagree. Only the lower Prop axis of an inverted V-engine allowed the use of an annual radiator. I don't know any plane with an upright V-engine which had an annual radiator
cimmex

There were; they were used on US aircraft in the 1920s.

 

[cimmex]

never heard, what type?

 

[fastmongrel]

As for the parts count, I worked at a shop specializing in Allisons and ONLY Allisons. I think we know how many parts it has since we were assembling them

Did an accountant with a spreadsheet and a complete parts list down to the last widget check every part. Is it an about number or an exact number that is being bandied about, people forget how many small parts there are in an engine.

Take a Conrod is it one part or is it 1 x rod 1 x end cap 1 x gudgeon pin 2 x circlips 1 x gudgeon pin sleeve 1 x small end roller bearing or journal bearing 2 x rod bolts 2 x rod nuts 2 x end cap bolt washers plus depending on design maybe some kind of bolt retainer 2 x big end shells and some old engines also have 2 x gudgeon pin end float control washers. Thats about 15 parts for one rod but is it one part because a lot of people call it one part.

I stripped, refurbished and rebuilt a 32cc Lawn Mower engine just before Xmas from 2 boxes of parts and I couldnt tell you how many parts are in the engine I know the carb was about 20 parts because I rebuilt the thing about 5 times till I found out I had the main jet upside down

 

[Aozora]

Fuel injection was an advantage, especially over the float-type carburetors preferred by Rolls-Royce. The pressure-type carbs used by the US companies did not have that problem.

Flight magazine had several pre-war articles debating the merits of carburettors vs fuel injection; one of them is from March 1939:

Aviation History from 1939. Browse historical aircraft from 1939
fuel injection | carburetter system 1939 | 0839

A description of the Stromberg pressure-injection carburettor starts here:

Aviation History from 1941. Browse historical aircraft from 1941

Plus an article on the different systems, starting here:

Aviation History from 1941. Browse historical aircraft from 1941

 

[Aozora]

I stripped, refurbished and rebuilt a 32cc Lawn Mower engine just before Xmas from 2 boxes of parts and I couldnt tell you how many parts are in the engine I know the carb was about 20 parts because I rebuilt the thing about 5 times till I found out I had the main jet upside down

A couple of years ago I helped my grandad strip down and overhaul the gearbox/engine unit of his ancient, but immaculate, Morris 1100. How many parts? Dunno. How easy was it to pull apart and rebuild? I never realised that such a small engine could be so !!!!!! Not the best example of British engineering...

 

[GregP]

I think you are arguing just to argue fastmongrel. The exact method arrived at for the count is easy if you (and anyone else who is interested) know how.

I decline to participate going forward, but please feel free to make a count on your own.

If you elect to do so, you'll come out VERY close unless you miscount. The diference is not worth the effort unless you have it all on computer and can simply look at the total count when you're finished right at the bottom of the database.

Go for it and good luck.

 

[Balljoint]

Compression ratio is basically the volume at TDC vs. volume at BDC, though blow down requirements affect the ideal. If, for a given fuel mixture charge, the mixture is confined to a smaller volume at TDC, the fuel charge in the smaller volume will be at greater pressure and produce a greater force on the top of the piston for the given fuel charge. The greater force will persist but relatively diminish as the piston moves toward EVO. Thus for a given fuel charge, more power and efficiency result from a higher CR until the higher pressure and temperature result in fuel detonation or pre-ignition. Too high CR cause the oxygen to become hyper reactive (changing the electron spin to a higher state) in response to raising temperature and pressure. Octane rating is a measure of fuel susceptibility to this failing.

Supercharging lowers mean thermal efficiency by cramming more fuel mixture into a given volume, i.e. more than atmospheric pressure would. Thus the combustion products have less relative expansion volume so more energy is wasted during blow down opening of the exhaust valve.

 

[fastmongrel]

A couple of years ago I helped my grandad strip down and overhaul the gearbox/engine unit of his ancient, but immaculate, Morris 1100. How many parts? Dunno. How easy was it to pull apart and rebuild? I never realised that such a small engine could be so !!!!!! Not the best example of British engineering...

My dad had a Austin 1100 (same car different badge) and it was a pretty poor example, we used to joke and say it did 5 miles per hour. That is for every 5 miles Dad drove it he spent an hour fixing it

 

[tomo pauk]

Oh, boy - you people mean this? picture

Rest assured - my dad's 1st car, bought in 1971 (or was it in 1972?). He drove it maybe 7-8 years, once some mechanical issues arose it became a challenge to keep it running. Body panels (ie. sheet metal) were, by his words, 'as good as in the Mercedes', on the other hand.

 

[fastmongrel]

I think you are arguing just to argue fastmongrel. The exact method arrived at for the count is easy if you (and anyone else who is interested) know how.

I decline to participate going forward, but please feel free to make a count on your own.

If you elect to do so, you'll come out VERY close unless you miscount. The diference is not worth the effort unless you have it all on computer and can simply look at the total count when you're finished right at the bottom of the database.

Go for it and good luck.

I am not arguing for arguings sake people have stated the Allison and the DB had 7,000 parts and the Merlin had 11,000. Where did these numbers come from are they genuine GM DB and RR numbers or are they someones guesstimate. We can spend whole threads discussing a 1,000 feet of altitude or 10 pounds of fuel but somehow an unsupported bunch of numbers becomes fact with no discussion allowed and if you know an easy method of counting engine parts please link to this method of counting engine parts I am genuinely interested.

 

[stona]

I am not arguing for arguings sake people have stated the Allison and the DB had 7,000 parts and the Merlin had 11,000. Where did these numbers come from are they genuine GM DB and RR numbers

That's a fair point. The provenance of the numbers IS relevant. Also, there may well have been different standards or methods used by different companies, never mind different nationalities, for ascertaining these figures.
I neither know nor care which engine had more parts, I don't see any relationship between that and reliability or ease of maintenance, but just stating unreferenced numbers, arrived at by who knows what method, as comparative facts seems to me a very risky business.
Cheers
Steve

 

[wuzak]

I certainly don't have the experience to tell where all the "extra" parts are but lets be logical. Like I said earlier, a Merlin, an Allison and a DB 600 series all use the same number of pistons, connecting rods, piston pins, valves, etc, so were do the "thousands" of extra pieces go? Nuts, bolts, washers (fiber or metal) clamps? they all used shaft drives to the cams, not gear towers.

I can certainly understand how, with extra parts, the Merlin could be more of a pain (need more man hours) to tear down and re-assemble. That does NOT mean it would be less reliable or break down more often. That would assume that ALL the parts were made of the same materials, were manufactured the same way, and had to handle identical stresses. I don't know how many changes were made to an item like the Merlin or DB crankshaft. We do know that the Allison went through at least 4 changes in crankshaft and that just shot peening the crankshaft without any other changes in dimensions or material added significantly to it's fatigue life.
There is too much we don't know about the engines to make the assumption that the relative number of parts had anything to do with either reliability or longevity. Assuming the engines were assembled with due care.

For all I know the R-R use of smaller but closer spaced fasteners allowed for better sealing (anyone with experience with most old English gaskets can appreciate this) or lighter cam covers or other access plates due to less 'bending' near the fasteners.

I think most of the extra parts of the Merlin are screws and bolts.

On 2 piece Merlin blocks, for instance, there are an extra set of bolts to clamp the cylinder head to the block. I don't think the Allison uses them. There are also ferrules to transfer coolant between the block and head (4 per cylinder) and, IIRC, o-ring seals for each end of those ferrules.