[Viper2000]

Originally Posted by Voyager

Just a fun little nit-pick, that is not necessarily true. When an engine is rated for a specific output that just means that it has been tested for that power output without failure for a minimum percentage of runs. It does not necessarily mean that the engines have actually been characterized as to their true maximum strength.

I can be fairly confident about the (safe)life of a Merlin III at +25 psi, because this was the boost rating cleared for N17's low level speed record attempt, and AFAIK the life it was cleared for was 15 minutes; Morgan & Shacklady provides some interesting details in their book about the Spitfire.

Rolls-Royce were the world experts in breaking their engines, since essentially their entire development strategy was to clear an engine for a given rating, then push it harder and strengthen whatever broke first. They also did the same sort of thing with overhaul life extensions; once more than about 30% of engines reaching maintenance organisations were reaching their full design life they'd go for a life extension; this allowed the life of fighter Merlins to rise from 240 hours in 1939 to over 300 hours (single engine) or 360 hours (twins) in 1944/5.

I agree that there would inevitably be some scatter in the life of an engine, especially when we're talking about overloading and running to failure; that's why I said "15 minutes or so".

Originally Posted by Voyager

I suspect this predict, test loop is why in WWII most engines tended to both be rated in multiples of 50hp, and increase in multiples of 50hp. That's not a natural thing.

Merlin ratings at least tended to be to the nearest 5 bhp - here's a few early engine ratings relevant to CoD:

R.M.1.S.
(Merlin II,III,IV,V)
Combat power:
1310 bhp/3000 rpm/9000' at +12 psi boost
or for Merlin III in Sea Hurricane only:
1440 bhp/3000 rpm/5500' at +16 psi boost
Takeoff power:
880 bhp/3000 rpm/SL at +6½ psi boost
Cruising:
2600 rpm/+4½ psi boost
Climbing:
2600 rpm/+6¼ psi boost

R.M.2.S.
(As R.M.1.S. but with increased takeoff power using 100 octane fuel; superseded by the use of combat ratings on R.M.1.S.*)
Combat power:
1000 bhp/3000 rpm/15500' at +6½ psi boost (provisional rating, not type tested)
Takeoff power:
1000 bhp/3000 rpm/SL at +8¼ psi boost
Cruising:
2600 rpm/+4½ psi boost
Climbing:
2600 rpm/+6¼ psi boost

R.M.3.S.
(Merlin XII)
Combat power:
1280 bhp/3000 rpm/10500' at +12 psi boost
Takeoff power:
1175 bhp/3000 rpm/SL at +12½ psi boost
Cruising:
2650 rpm/+7 psi boost
Climbing:
2850 rpm/+9 psi boost

Harvey-Bailey, A. (1995) The Merlin in Perspective - the combat years (4th edition). Derby: Rolls-Royce Heritage Trust.

After this most ratings were at the nearest integer value of boost in psi, which makes sense because it's quite hard to imagine a pilot trying to set boost to the nearest ¼ or ½ psi in combat.

Meanwhile, on the other side of the pond, Americans were building Merlins under license and fitting them to Mustangs.

The original Merlin Mustang was the Mustang X, built at Hucknall and fitted with the Merlin 65 (basically a Merlin 66 with a 0.42 reduction gear). This was rather an interesting machine; it featured a chin radiator to supplement the Allison Mustang's small belly radiator, and had Spitfire exhausts.

North American subsequently came out with the Mustang III/P-51B/C and fitted it with the American equivalent of the Merlin 65, which was the V-1650-3.

Since the Americans liked to rate their engines in terms of "Hg absolute manifold pressure, somebody obviously got out their slide rule and did a quick conversion.

1 standard atmosphere = 101325 Pa = 14.696 psi = 29.92 "Hg

Therefore, if you convert +18 psi boost into "Hg you get:

29.92*(14.696+18 )/14.696 = 66.57 "Hg.

So they rounded up and called it 67 "Hg, and the Americans got about an extra 0.65% power in MS gear:

R.M.10.S.M.
(Merlin 65, 66, 67, 85, 85B, 86)
Combat power:
1705 bhp/3000 rpm/5750' at +18 psi boost MS gear (5.79:1 ratio)
1580 bhp/3000 rpm/16000' at +18 psi boost FS gear (7.06:1 ratio)
Takeoff power:
1315 bhp/3000 rpm/SL at +12 psi boost
Cruising:
2650 rpm/+7 psi boost
Climbing:
2850 rpm/+9 psi boost

(85 cleared to takeoff at +18, 85B cleared to climb at 2850 rpm in FS gear, 86 has different supercharger gear ratios, all 0.42 reduction gear apart from 66 which has 0.477)

V-1650-7
(Obviously it's American so it doesn't get an MoS rating number)
Combat power:
1720 bhp/3000 rpm/6200' at +18¼ psi boost MS gear (5.80:1 ratio)
1505 bhp/3000 rpm/19300' at +18¼ psi boost FS gear (7.35:1 ratio)
Takeoff power:
1490 bhp/3000 rpm/SL at +15¼ psi boost
Cruising:
2700 rpm at +8 psi boost

(ibid)

Of course, since this is from a Rolls-Royce publication, the American manifold pressures have been converted back to the nearest ¼ psi boost.

[You can't really compare FS gear performance directly because the supercharger gear ratios are too different. The small difference in MS gear ratio is probably just because the Americans used different gear grinding machines. For much the same reason the V-1650-7 has a 0.479 reduction gear ratio vs 0.477 for Spitfire Merlins (almost everything else used 0.42 in order to swing a bigger prop). Anyway, I included the FS gear performance for completeness.]

In peacetime, when developing engines for sale to a customer, obviously the power, FTH and SFC are the main parameters of interest. These would usually be backed up by a contractual performance guarantee of some sort. The boost and rpm required to get to the guaranteed performance are both entirely academic; the customer doesn't care as long as he gets the performance. The engine manufacturer predicts what he can make, knocks off a small margin for safety, and then tries to get a contract to build the engine. Once everybody has signed on the dotted lines, the engine manufacturer then needs to develop his product until it meets the guarantee, or else pay out as demanded by the penalty clauses.

In wartime, things are different. The objective is simply to improve upon the performance currently available. So development doesn't stop when a target performance level is reached, and therefore the power achieved is academic. What matters is how much boost and rpm the engine can handle, the objective being to roll that performance out to as many engines in service as possible via whatever mods are required, whilst meanwhile attempting to develop future engines to handle more.

Pretty much the only reason that the power is even specified in the rating is so that you can put a random sample of engines on the dyno for production or maintenance quality control purposes. Once the engine is in service it's going to be flown according to the boost, rpm and temperature limits; the actual installed power output is unknown.

As for the Reno guys, they certainly push to high boost, but I think much of the extra performance probably comes from extra rpm, cropped supercharger rotors and copious ADI.

The R.M.17.S.M. was actually tested at +36 psi and 3100 rpm in 1944 for about 2620 bhp; that's about 103" Hg. It was also the only "stock" Merlin to get extra power from more aggressive cams.

What the Reno guys seem to do is run at or close to the overspeed limit of 3600 rpm, chuck in vast amounts of ADI and then up the boost until they're on the edge of detonation. Then they can then crop the supercharger rotors until they're riding the edge of detonation when running WOT on a hot day and call the job a good'un (boost varies as rpm^2, so 3600 rpm almost certainly gives them more boost than they can actually use - something like 140" Hg according to a quick back of envelope calculation, though how many of them actually crop their supercharger rotors I don't know; likewise I don't know if they've started playing about with the valve timing).

They certainly often fit V-1710 conrods and various other things to strengthen the engine, and put spraybars in front of the cooling system so that they can undersize the intake for reduced cooling drag.

I'd estimate that they're chucking out something close to 4000 bhp by the time they've finished. One day I'd like to go out there to see and hear the action before they all run out of spares and/or somebody from the American equivalent of the HSE decides that air racing is too dangerous...

*The list of ratings dates from 1948 and therefore shows the final version of R.M.1.S.; R.M.2.S. is a fossilised rating which shows the old 87 octane version of R.M.1.S. but with increased takeoff power before the boost control before +12 psi operations with the boost control cutout mod were cleared.