You have done a nice job of going through some of pathways. Some of which are nearly incomprehensible to some of us today.
I'm glad you find it worth replying to, but I know what you mean about "nearly incomprehensible"!!
Fedden and his fixation on the sleeve valve seem to had had a very large hand in this. Apparently there is no surviving accounting of the cost of the various engines in either pounds Sterling or man-hours. Fedden had nearly bankrupted the company by the time the sleeve valve engines were a viable product. Please note there were two phases of this so some accounts are a little unclear. First phase was just getting the Perseus into production.
View attachment 717640
Visually this was a look into the future for British radial engines, at least it used a long cord cowl. However it seems the Pegasus was not quite ready for prime time. This has to be seen in context as the the other main contender, the Armstrong Siddeley Tiger was sliding into obsolescence.
From Wiki so...............
"The Mark IV introduced the much more powerful
825 hp (615 kW)
Bristol Perseus sleeve valve radial engine enclosed in a
NACA cowling which significantly improved performance, increasing maximum speed to 156 mph (251 km/h) and rate of climb to 840 ft/min (4.3 m/s).
[7] In this version, the Perseus had
overheating problems and was deemed unsuitable for tropical service[8] with production limited to 18 aircraft, all of which served with the home based squadrons"
When I was trying to find out if there's any information on the original narrow-body fuselage design for the Botha (there isn't!), google bounced me up a remark saying that the Perseus struggled with dust as well...
The Perseus went through 4 or 5 different cylinders with increasing fin area from 1932 to 1940.
The 2nd phase was 1939-40 with the Perseus in large scale production and the Hercules (using the same cylinders) also going into large scale production. This is when they discovered that they could not make the sleeves stay round in mass production and the engines (mostly Hercules?) were using excessive oil in as little as 20 hours of operation.
The main demand for the new designs down to 1940 seem to have been for the Persus, principally the Botha, which used nearly 1200 engines in total, but I've not found when their production-run was complete; the only other significant early number was for the Lysander, with about 500 Perseus-engined airframes down to August 1940, with the Skua and its turret-fighter variant taking another 300-odd in the same timeframe; by contrast, the Hercules seems to have only flown in
very limited numbers at first, with the four-engined short S.26 flying-boat, of which just three were built, and about twenty of the twin-engined Saunders-Roe Lerwick, which between them account for around fifty motors...
The first really serious Hercules production is for the Beaufighter, which starts deliveries in July/August 1940, but even this needs no more than 200-odd engines in 1940 (the hundredth Beau completed in December '40 and that presumably includes some early Mk. IIs with Merlins), and rather less than 2000 engines by the end of 1941 (there are less than 1000 Beau Mk. Is in total, though I can't find precise dates, and some percentage of them are likely later); the Short Stirling only runs to a handful of machines in 1940, and around a hundred and fifty in 1941: total Hercules numbers across the two types from mid-1940 to end 1941, a maximum of about 2500, don't seem out-of-step with the 2000-odd Perseus motors in a comparable period 1939-1940...
Hercules machines only seem to really kick in around 1942, and then they
really kick in, with the Beaufighter Mk. VI, the Armstrong-Whitworth Albermarle, and the engine swapped onto the Wellington, the Halifax, and even a few hundred Lancasters, with a total of maybe 40k motors on these types by 1945, which makes even Mercury production in '37-'41 look rather modest, and then after that it's on a whole range of British and French cargo types...
I wouldn't be completely surprised if there was a hard-swap around August 1940 from serial production of the Perseus to the Hercules, but for whatever reason, it seems to be 1942 when things get going...
They were also realizing that 900hp engines were not actually going to power war winning aircraft. Why this was so surprising is a little hard to figure out. The Sleeve valve program at Bristol was always a little behind the curve. The backwards Americans were building 30 liter engines that made 900hp in the mid 30s when the Pegasus was struggling to make 800hp. By 1940 the American engines were making 1000-1100hp (the American 1200hp was pretty much sea level for take-off). Less said about the Aquila the better, The Americans were using engines of that size to power trainers in tier below the AT-6.
The British were forced to keep making Mercuries because they didn't have much else. They needed the Pegasus cylinder production to make Hercules engines. They had only made about 4-5 of Aquila engines despite all the press. And the Taurus was another lost cause. 5/6th size of twin wasp, over heating problems as it was and the sleeve valves didn't like high boost (high cylinder pressures) left few, if any, options.
Basically the engines didn't make enough power and no amount of fooling around with skinny fuselages was going to solve that. They were building Blenheims by the hundreds and they should have had a pretty good idea of what kind of power they needed to get the desired performance from a plane of that size. Clip the wing, take the turret off, fair the wheels over. See what it actually took instead of assuming that sleeve valves were going magically sprinkle unicorn poop on the aircraft they were dreaming up and take them to next level of performance.
I'm prone to defending faintly absurd causes like the viability of a narrow-body Blackburn Botha (obviously, removing that appalling dorsal-fuselage drag-brake with the Lewis guns in it would help too; "would be significantly improved by the turret off the Defiant/Roc" is not something you can say about that many aircraft), and more generally I have an amateurish sense that smallish radials might make more sense in pairs than they do individually, but you're not wrong; the question is, why were
they wrong?
1. Presumably they thought drag was prohibitive on a big blunt radial - it's quite puzzling reading narratives surrounding the Short S.30 that emphasise the
minimal diameter of the NACA cowling on the Perseus, which suggests they didn't really understand what it could do...
2. Was there some perceived advantage in cruising range? The RAF seems to have emphasised an economical trog at 10,000ft, and they thought - rightly or wrongly - that the Perseus would be more able to heft freight cargo across the Atlantic than the Pegasus...
This is probably the best data on Iowa speeds that exists
http://www.navweaps.com/index_tech/tech-003.php#:~:text=Iowa's%20Speed,to%20perform%20measured%20mile%20tests.
During Operation Hailstone (the raid on Truk in Feb 1944) Iowa and New Jersey were reported as running at 30 knots while chasing escaping Japanese ships.
www.navweaps.com/index_tech/tech-029.php
Note how big an increase in shp is required to raise the speed of the vessel from around 25-26 knots to about 30 knots. Providing that extra power alone means a bigger ship.
Subsequent trials in the 1980s seem to verify that the Iowas were capable of 32.5 knots at 212,000shp at near full load displacement of 56,900 tons.
www.navweaps.com/index_tech/tech-104.php
The thing that is usually forgotten is that in WW2 ALL 10 US fast Battleships spent the majority of their time from mid-1943 onwards as escorts to the carrier groups. From mid-1943 new procedures were laid down covering the operation of these groups. Operating speeds were in the range 15-25 knots and the whole task group would generally manoeuvre as a single entity. If a carrier needed to operate its aircraft outwith a normal planned operating sequence, either the whole group would alter course and speed or the individual carrier would manoeuvre as far downwind as possible while remaining within the screen of escorting destroyers. It would then operate as fast as necessary to launch/recover aircraft before taking up its designated position in the TG again, all the while remaining within the destroyer screen.
For that reason an Iowa was bound to be operating at the same speed as a North Carolina or South Dakota for the vast majority of its time at sea.
Aye, I'm aware it takes an awful lot of extra horsepower to brute-force even a small speed increase out of a big ship that's trying to power through the water it's piling up in front of it - the approximate equivalent of transonic drag in aviation; but while some ships did get something out of lean proportions and/or big boilers, I've not seen any evidence the
Iowa class was one of them...
The claim of over thirty knots in 1944 was based on the bridge speedometer - an unreliable instrument, especially at high speeds; this is the equivalent of IAS on an aircraft...
In 1985, as the article notes, what was
actually achieved on trials was just under 29½ knots at a little over 140,000 shp; this was combined with an (unreliable, unspecified) bridge speedometer reading from an earlier engine-power trial which had reached over 185,000 shp, which was then adjusted, based on the (questionable) assumption that you could calibrate it by scaling 140,000 shp to 29½ knots, to give a (theoretical) speed of 31 knots at around 185,000 shp; the numbers thus produced were then adjusted
again for a (theoretical) maximum output of 212,000 shp; this is an old-established technique for estimating
theoretical maximum speeds for capital ships. The only
real number here is 29½ knots.
A good example, because it's entirely explicit, is provided by the 1916 "large light cruisers", which happen to be favourite ships of mine;
Courageous reported a speed on her sea trials of just below 31 knots, perhaps measured by the ludicrously unreliable old-fashioned method of throwing a log on a string off the back of the ship and seeing how fast the rope payed out; this unreliable result was achieved off what was said to be just over her notional 90,000 shp design power, and more certainly at a little over 22000t displacement and 25'8"ft draft, and on this basis she was allowed to be capable of 32 knots off 110,000 shp overload power at her design parameters of 19000t and 22'8", in order to say that the engines met the official performance requirements - though these were all unrealistic figures for the ship as-built; without derailing the discussion with more detail, the only clearly reliable figure for the type I've seen is a fraction above 30kts on the measured mile, and with a length:beam ratio of 9:1 and a deep-load draught of scarcely 25ft, these are the most innately streamlined capital ships to ever swim...
Similar examples can be multiplied - typical giveaways are the quoting of design displacement/draft figures rather than as-built ones, which imply that the speed has also been adjusted, or results based on power trials rather than speed trials, where it's sometimes obvious that design power is being artificially
assumed to produce design speed and maximum overload is being attributed a scaled-up rate of knots...
In practice, there seems to be a fairly consistent "wall" for real speeds for battleship-shaped ships around 29½ knots, as that's where practical trials reports for pretty much all the WW2 types max-out regardless of length/beam ratio, horsepower and draught; I'm prepared to believe that one or two
might have done a little better, more from hull configuration than power (e.g. the Italian
Littorio class at their very light trials draught/displacement, and HMS
Vanguard with her big forecastle and transom stern); but even those numbers I'm wary of, as they simply lack
obvious artifice, and I've not seen anything reliable for the
Iowa class...
This isn't to say I'm an expert! As with anything I say here, you should consider me as towing a large gunnery target...
Ordered as an Amphibian Boat Reconnaissance aircraft for the FAA, Supermarine couldn't build them in 1940 due to demand for Spitfires. Production had to wait until new contracts were placed with Saunders Roe in Jan 1942 with the first production aircraft flying in Jan 1943.
Wikipedia mentions an earlier contract with Blackburn that was issued in 1940 and cancelled in 1941, information it says comes from a 1981 history of Supermarine; supposedly that was to do with airframe capacity, but the intended Perseus engine wasn't in production after August 1940 for anything except
perhaps the Blackburn Botha (and I say
perhaps because I'm simply not sure when production was done there); or was it the other way up, and the Sea Otter was an attempt to find a
use for Perseus engine-manufacturing capacity?
But their first operational service was with the RAF in the Air Sea Rescue role from Nov 1943, alongside many aircraft types including the Walrus. When the FAA began to form squadrons on the Sea Otter in Nov 1944 it was again in the Air Sea Rescue role. Immediately postwar one squadron used them minespotting in the clear waters of the Med.
Postwar each RN carrier generally had one, again for Air Sea Rescue duties. That continued until the Korean War when they were replaced by Westland Dragonfly helicopters.
Fascinating. ASR makes sense as a main role at that date... I simply hadn't really realised how much the Sea Otter was used like that, and of course it does make a little flying-boat relevant; but I'm pretty sure it wasn't
designed primarily for that role...
Incidentally, the FAA began removing the catapult flights from its cruisers from late 1942 / early 1943. As a measure of the speed with which that happened 700 squadron, responsible for training crews and managing individual flights, reduced from 63 aircraft at its peak in June 1942 to just 20 in July 1943. Amongst the very last catapult flights were the Curtiss Kingfisher flights on Emerald and various AMCs in spring 1944. Blame radar and the fact that the role of the cruiser changed from patrolling the sea lanes (something taken over by long range flying boats and land based aircraft) to escorting the fleet. The hangar space freed up was used for a lot of purposes, not just cinemas, on pre-war designed ships whose crews had increased dramatically with all those new electronic devices and much heavier AA armaments.
Oh, definitely because of radar. The cinema on HMS
Renown is just a symbol of the way the whole midships floatplane aparatus on RN capital ships was suddenly no longer needed...
Though the cruisers did continue to be used on patrol until 1944 - with several of the Italian co-belligerents taking up the South Atlantic slack; I'd guess that
Emerald - which still had an old-fashioned charthouse and topmast, not a tower bridge - simply didn't have much useful space for radar, her Kingfisher presumably being because she wasn't really designed for anything as hefty as a Walrus or an Otter (she seems to have used a Seafox with the rest of the smaller RN cruisers until they went out of style)...
The USN retained its floatplanes on cruisers and destroyers right to the end of the war and beyond (around 1948 IIRC). In June 1942 it issued a requirement for a new "observation seaplane". That became the Curtiss SC-1 Seahawk. A production order was issued in June 1943 and the first flights went aboard the USS Guam in Oct 1944.
Compare the IJN's active enthusiasm for replacing half the guns on battleships and heavy cruisers with a seaplane deck, which makes more sense when you realise they've become a search-radar platform, and the reduction in main armament is practical because the fire-control radar removes the need for large randomized salvos...
The RN's contrasting rejection of the floatplane is, therefore, the more curious decision; obviously they'd found masthead radar more useful than the Walrus in the
Bismarck chase in North Atlantic conditions, and as you say they'd expended the effort on convoy shadowing so they generally had shore-based overflight, but carriers also seem to be more integrated with their capital ships (to the extent that Cunningham literally had
Formidable in the battle line at Cape Matapan), so perhaps they don't perceive the need to give a battleship force a large screen of seaplane radar pickets
sans carrier in quite the same way...
... because their carriers are providing the battle squadron radar pickets...
And (thinking out loud) this might suddenly make sense of the peculiarly large two-seater carrier-fighters the Fleet Air Arm is getting. The Fulmar is, perhaps, designed in anticipation of
ASV, though in the end, of course, it inevitably ended up being flown by the Swordfish instead...
Yes. Aircraft would lose lots of speed in dogfights, and/or lose altitude to convert potential energy into kinetic energy (speed).
Aircraft coming out of dogfights were vulnerable because they were slow. Not only were they easy targets for slashing attacks, the maximum Gs that you can pull depends on your speed (and air density, and is limited by structural strength or the pilot blacking out, etc., etc.).
That certainly explains the limits of the Skua, if the airframe simply wasn't able to stay flying at the speeds it would need to drop to during combat, and was going to suddenly fall sideways out of the sky mid-manoeuvre...
That was the advantage of the two American engines. They didn't have to be 'tricky'. P&W had make an R-1535 (after they made the first R-1830s) because the navy wanted a small dimeter engine for good pilot view. Turns out the commercial market wasn't interested in it. It was too big for single engine private/commercial planes and two small for twins where they didn't care as much about the view from the cockpit as the engines were out on the wings and they wanted more power for planes like the DC-2/3 anyway. Army didn't want it either, they weren't lading on carrier decks.
I'd never thought that the Skua's nose configuration might be seen as an advantage for the cockpit view. I'm not sure I
want to think that it was designed that way deliberately...
