Sunday, June 12, 2011

Let's Argue About Aircraft Carriers, Again!

Edit: Oh, Good Lord, according to Google, people are reading this cut-and-pasted monster. So I'm going to edit it up a bit.

Now, I was originally going to blog about K. Maria D. Lane's  Geographies of Mars. That's because I didn't have a copy, so, really, I could say whatever the heck I wanted to say about it. Once I had it, I figured, I could read it, and then I could live up to my obligations to the world of letters with a Very Serious Review.  Such, however, are the wonders of Amazon and UPS that I already have a copy of a book I ordered two days ago. (Maybe the Internet loves me; maybe I checked the wrong box on the delivery options page). So where's that Very Serious Review? Well, I've sort of got to read it, now, before I can say anything scholarly. That doesn't exclude my writing about the subject in a non-scholarly way, but I have a place for such things, where it will help space out the (more blatant) self indulgence.

Speaking of self-indulgence, while I don't follow Lawyers, Guns, and Money closely, when I checked in today, Robert Farley was trying to jump into a conversation about aircraft carriers. The argument is that they're large and they're boats, so they can sink if they hit a rock or an ICBM, and that they're really nothing more than an expensive stage for telling America to itself.*
(Congratulations on the embedding restrictions, SME. I can't imagine a better way of promoting Blood and Chrome! That's sarcasm, BTW.)

There are worse uses for money than settings for hopeful love stories, but my  personal opinion (which is to say, my rehash of Norman Friedman's useful opinion) is that the advent of the aircraft carrier really marks the transition from the weight-limited naval warship  to the volume-limited. The actual weapon system matters far less. It's just a thing that right now that happens to be manned aircraft. The size of the ship comes out of naval architectural logic, and really doesn't dictate costs compared with other factors. So supercarriers it is.

Controversy! I court it! And I have some more substantive things to say about the "vulnerability" of supercarriers.

So, first, about Naval aviation. It begins early.

 We all know that there were naval aircraft in the First World War. We've probably heard about the 1918 carrier raid. But what was the first air-sea battle in history? How about a Jutland that never happened? R. D. Layman, (Naval Aviation in the First World War: Its Impact and Influence (London: Chatham, [1996]) tries to get the details right. (185--200). On 18 August 1916 Scheer sortied the High Seas Fleet, intending to “get Jutland right.” Once again scouting lines of submarines and zeppelins were deployed to shadow the Grand Fleet. Once again, British electronic reconnaissance gave adequate forewarning of the German sortie and was thus out in force. Briefly tracked by the submarines (the zeppelins were neutralised by a low haze), Jellicoe inadvertently broke contact, and briefly the Grand Fleet and High Seas Fleet were converging towards disaster (that is, for the Germans.) At this point a zeppelin signalled the presence of a British battlefleet to the south. Believing this to be the British Channel Fleet and eager to pick off an exposed target, Scheer turned south away from Jellicoe. It was, in fact, the misidentified Harwich Force, and contact was never made. In short, people had to begin thinking about aircraft and naval battles in the immediate aftermath of Jutland. It was not something buzzing around the minds of overstuffed admirals retiring to the club after lunch during the stultifying postwar years when no weapon could be approved unless it could be used to blow up Waziris. (How far we have progressed!) 

This matters because it's often argued by my old buddy, Professor Strawman, that the Admiralty, or possibly the Air Office, starved the Fleet Air Arm between the wars. No money, no thinking. The carriers that Britain actually fought the war with just sort of came out of nothing.

Well, what what are the facts at hand? Rather different, as it happens. It's a good thing that facts are for the tedious.

Fleet Air Arm Votes

Extracts of Vote 4 (grant-in-aid to the RAF for the costs of the Fleet Air Arm, prices in £000,000)

Or consider total aircraft. At the outbreak of World War II, we are often told that the Fleet Air Arm numbered less than 150 aircraft or so. And this is, of course, true. Another interesting way to look at it is to note that several types have acknowledged FAA holdings on the outbreak of war: 51 Fairey Seafoxes, 162 Supermarine Walrus; 165 Blackburn Sharks; 492 Fairey Swordfish, 150 Skuas, or 1020 a/c on charge, not counting remaining earlier types, Gladiators, or Rocs, for a frontline strength of 232. 1939 aircraft strength for the USN/USMC are on file. In short, in 1939, total FAA holdings were 2098 aircraft, including 1,316 combat, 150 transport/utility/262 training, 370 misc. "Complement" is different from planes on hand.

I raise the point because aircraft totals are sometimes used to demonstrate that the FAA was neglected in comparison to the American naval air arm. But you can count the number of aircraft ordered for the USN: Fighters: Boeing FB (43), F2B, F3B (106), F4B (136), Curtiss F6C single seater (75), F8C, also referred to as O2C, in fact a 2 seater fighter/observation/dive bombing type (90), F9C trapeze fighter (9), F11C also BFC, BF2C, 2 seater FO (81), Grumman FF and SF 2S fighter (60), F2F, F3F, in service 1935–41 (189); bombers, dive bombers, torpedo bombers Curtiss CS TB (43), SBC Helldiver to outbreak of war (c. 108), Douglas DT-1 (67), T2D or P2D, a twin-engine bomber with a 55ft span, flown off Langley once (13), TBD Devastator (130), Great Lakes BG-1 DB (60), Martin TBM/T4M 2S torpedo bomber (277), Martin BM 2S DB (32), Northrop BT 2S TB (54); Observers and Observer/fighters Vought O2U (159),O3U/SU (159), SBU, designed as 2S fighter, accepted as a SB (124); Scouts of interest, Curtiss SOC biplane (258); Amphibians Grumman JF, J2F, (c. 110).
So the Navy Department bought, during the entire interwar, only a few more aircraft than the Fleet Air Arm had on hand in 1939. Or 2386 a/c, plus 550 flying boats and numerous trainers.  (Gordon Swanborough, US Navy Aircraft since 1911, passim; and I do mean passim. Notice that the USN has plenty of "fabric-winged biplanes" and flying boats in 1939.

All of this only proves that no-one cares about planes or money. Let's try carriers, instead. One can argue that the Admiralty ordered too few carriers before the outbreak of World War II, but not that it didn't order a great many by contemporary standards. The question of British neglect of naval aviation thus comes down, not to numbers of planes or of ships, but to the fact that the "armoured deck carriers" of the Royal Navy didn't carry enough planes. The crux of the argument becomes that all of this armour cost planes, and that the Admiralty either falsely or correctly thought that carriers were too "vulnerable," the very argument being made today.

Let's by all means, then, rethink this.

In 1931, having just completed the last "large light cruiser" conversion, British planners began working on a brand new aircraft carrier, the design that became Ark Royal. As is his special strength, Norman Friedman gives a full account.

Planners were given a 22,000t displacement limit. The point was aircraft, so it was designed to accommodate the maximum number of a/c aboard, with a barrier so that they could be parked on the deck, at the insistence of Charles Forbes, then Third Sea Lord, and Home Fleet commander during the Norway campaign. (Thus, the man who sent Warspite into Narvik Fjord.Arrestor gear was also approved requiring the development of a feedback-controlled hydraulic "gear." Consequent to the barrier, the Deck Landing Control Officer was introduced in 1937). Speed and range were specified to be high, and that meant carrying a great deal of fuel and other stores, and thus a deep-girder ship with lots of space to put things. 

That meant that the landing deck should be the strength deck. This would maximise internal volume for fuel and such, and structural strength (versus pounding, amongst other issues). But it would reduce both deck area and freeboard. (Although deck length was reduced from 900ft to 800 for navigational reasons, so there were other factors at work.) And it also implied accepting partial destruction of the deck part and thus meant narrower lifts and running additional structural members up through the walls of the hangar, reducing the free volume for planes. Although on the bright side, provision for compromised structural integrity also added robustness for suspended fire curtains, overhead sprinkler systems, and deck fuel delivery pipes. Eagle was chosen as deck park test bed ship, because its capacity went from 18 to 30 a/c if so employed. Arrestors were first installed on Courageous and Glorious by January 1934. In the same month, the Director of Naval Construction calculated that Eagle and Hermes could each take 40 Ospreys (20 above, 20 below), a design much favoured within the Navy's air staff at the timeFurious could take 14 above, 18 each for Courageous and Glorious. “These figures were never used to calculate the full capacity of the British fleet,” Friedman remarks, but there's no particular reason that it should have been. 

What this means is that we often see calculations in terms of numbers of aircraft. Thus, the Japanese had 600 naval aircraft at the outbreak of WWII in the Pacific, America had 400, the British that 165 figure again. Naval air neglect! The thing is, though, while, the Osprey buy was 133 a/c, and there were related 87 Nimrods (Mason, 229–30), the design was an adapted Hawker Hart. At least in theory, the Fleet could ask the RAF for some squadrons of Harts, and the RAF's reserve of carrier-trained pilots would allow them to fly off the carriers from their improvised deck parks and carry out one heck of a "special operation" against Taranto or Kure. And, yes, the records do imply that the Admiralty had a plan for a pre-emptive air attack on the IJN at anchor.

Nevertheless,  the complement question continued to rage in the prewar years. A new perspective came with regime change, as the first postwar Rear Admiral, Aircraft Carriers, reached the Controllership in 1937. Henderson's patron, Ernle Chatfield, had been aboard Lion at Jutland, and was armour mad and happy about new technologies now only dimly perceived in the mind's eye, having gone back under the cloak of nuclear secrecy as quickly as they came out of the old one of "gun versus armour." Whereas the navigators focussed on two-seat fighters vectored in on reconnaissance aircraft with new and existing radio technologies and shooting them down before the fleet could even be located, Henderson and Chatfield focussed on death from above. How many a/c could a CV operate in wartime? Henderson pitched the number low, perhaps for rhetorical reasons, perhaps because he was considering a more pessimistic maintenance scenario than emerged in the Pacific in 1944/45.  Ark Royal was supposed to carry 72, but, he suggested, only 60 (42 Albacores and 18 Skuas [44/22 “at the outside”]) in practice. Then he pointed to the need for light deck armour and light AA gainst fighter attack to argue for 40/12, and then 40, and finally 36, on the grounds that it  would be weird and creepy to fly planes in more or less than their natural squadron strength of 12.

With his ground work prepared, Henderson became  Controller. All sorts of CV proposals were on his desk. There was a "trade protection carrier," linked to the Mobile Naval Base Defence Organisation  concept. Although at some points this project resembled an escort carrier, or the passenger liner/carriers with which the Japanese had so little success, it was probably moving towards a flying boat tender. There was a fine design already in construction, and insistent request for a small, c. 10,000 "expendable" design. And there were other projects, too. 

And Forbes, Chatfield, and Henderson killed them. They wanted plenty of planes, to be sure, but also a large CV force. There would be 8 ships operating 350a/c --for now.  (Not counting either float planes or legacy carriers, so that the First Lord could project a 600 aircraft FAA in the 1938--39 Estimates, presumably forecasting out to a day when the Royal Navy would have no less than 15 modern carriers of the Forbes/Chatfield/Henderson concept). Theis concept would be ships would be large (23,000t), and have armoured decks.

 So this is the substantive, alleged mistake, and the question to be investigated, at long, meandering last.

 The question gets endless play, but we should bear in mind that we are in fact talking about armoured hangars, not armoured ships. That is, we're comparing heavily armoured American carriers that had their armour at the hangar deck level with heavily armoured British carriers that had their armour at deck level. 

The issue, therefore, is operational losses due to damage at the hangar deck level. For although the US lost no Essex class carriers (not that these 27,000 t. sd carriers are directly comparable to the 23,000 t sd Illustrious types), they did suffer heavy casualties due primarily to slanting hits that did all or most of their damage on the hangar level. Given the factors that went into making an individual hit deadly or not, it is also worth looking at damage in smaller classes. So I spent some time with Morison’s 14 volume official history of US naval operations in the Pacific. Note that Morison does not always give complete casualty returns, but I return KIA/WIA where that information is given.

Here's what, historically, happened to an actual carrier force in operations.

 24 October 1944. After a generally excellent interception, a leaking “Judy” (capacity for 1x550lb bomb in bomb bay, 2x66lb underwing) bombs Princeton. The single bomb penetrates three decks and blast sets fires at hangar deck level that penetrate a torpedo mantlet, causing loss of ship, no details of incidental losses, which were heavy and probably implicated losses on a CA alongside.
12:357ff In the wake of Leyte the fast carriers raid the Luzon airfields in November 1944. Intrepid is hit by 1 kamikaze (hereafter k) that detonates bomb at hangar deck level. A k hits Cabot with light damage, 36/16. (Other sources give moderate to severe fragment damage, presumably from a near miss. Take not, those who think that armour had no place on non-battleships. Shrapnel has no friends.) Essex is hit by 1 k with 15 KIA. Intrepid hit again by k within the hour, damage reaching the hangar deck where a fire causes superficial damage and a net 69/35. Lexington takes a k. on the island with minor damage. Essex takes 1 k with minor damage.
Off Lingayen Gulf  Omaney Bay, a CVE, is sunk by single k that feeds penetrating bomb to hangar deck level, fatal fire.  In a strike against Formosa the fast carriers get in the way of a raid on 21 January. Langley takes 2 small bombs from an SE intruder. 2 explosions on flight deck, 3/11. Ticonderoga takes k, 550lb bomb detonates above hangar deck at 1208; at 1250 1 more hits island and avgas sets fires. Total losses 143/202, 36 a/c lost. CV retires from theatre. Intrepid is earning her "Evil Eye" reputation, bombing herself with a 550lb bomb that drops from harness on a taxi-ing Avenger, punching through the wooden flight deck and detonating on the hangar deck, for 52/105.
Off Iwo Jima, an experimental operation in which Saratoga is to give day and night advanced fighter cover ends badly when, with cloud at 3500ft Saratoga is attacked at 1659 by 2 ks that hit short, “hurling their bombs into the ship” and themselves hitting the ship at waterline while a 3rd a/c bombed. 42 a/c are lost aboard, 123/192. Ship retires from theatre.
(54–5) With aircraft cover stripped away, the Bismarck Sea CVE is sunk by 1 a/c. Fire reaches avgas stores, ship abandoned with 218 KIA or WIA.
On 18–31 March the fast carriers (now TF 58) are sent in to attack Kyushu for air superiority preparatory to Okinawa. Oops, on 18 March the TF is rumbled, TG 58.4 attacked shortly after dawn by no more than 50 a/c. Enterprise is hit by a dud. Intrepid takes 2/43 to fragments that start fire on hangar deck. 3 DBs attack Yorktown, 1 500lb bomb passes through 1 deck and “explodes near ship’s side,” holing hull and 5/26. Listed as minor damage.
On 19 March TG 58.2 is again attacked, shortly after dawn. Wasp is bombed by weapon that penetrates the hangar deck, the level where the Essex-class carries its substantial vertical armour that is supposed to exclude enemy weapons from the citadel. Fortunately, the bomb bursts in no 3 deck galley. Fires on 5 decks controlled in 15min while a kamikaze detonates alongside. Casualties are 101/269, but this is listed as moderate damage. I think that this one of the cases in which the entire on duty portion of the carrier air group is killed by smoke inhalation. Franklin is hit by 2 bombs dropped by an intruder not even detected before it bombs. One explodes on hangar deck, one above it. Internal explosions and enormous fire result. Ship not underway until 0300 20 March. 724/265. Covering the withdrawal of Franklin during the day of the 20th, TF 58 comes under heavy, but ineffectual air attack. A k. misses Hancock, which suffers minor damage. Fires due to AA splinters break out on Enterprise flight deck, but the best known aspect of this battle was the intervention of the strategic arm, with a squadron-strength attack by the closest thing the Japanese have to a heavy bomber, the  (Mitsubishi G4M, operating manned glide bombs that were easily dealt with by Hellcats. This might have been a bit more effective. A total of 3 CV retire from theatre per Morison, making the Kyushu raid is one of those embarrassing defeats that so contradict the historiography that is the job of us military historians to make them to have never happened. battles that never happened.  as see below. 

Having learned his lesson, Mitscher took his carriers south of Okinawa during the Ten-Go offensive, struck the planes below, and drained the gas tanks of all but the fighters (Morison 14:209), while  massed k. attacks fell on the surface fleet, and in particular the picket destroyers and amphibious shipping. Nevertheless, some aircraft push south, in spite of not having darkness to mask their approach. The major attack begins at 1:30. Enterprise is first missed alongside , Hancock taking some damage and 72/82 notwithstanding the fact that the bomb is reported so light that it rolls along flight deck before penetrating. (210) On 11 April 2 a near miss k put Enterprise out for 48h, while one to Essex inflicts 33/33 and puts the ship out of action due to machinery damage. T5 58 is again attacked 16 April after a near miss on 14 April when the inbound raid stops to attack the radar screen of picket destroyers. Intrepid is crashed once, missed once, and with fires and structural damage finally withdraws from the theatre 10/87. On 11 May TF 58 is again found. Bunker Hill crashed by a Zero. The bomb (which would be 250lb if identification correct) detonates in the AA tubs while a D4Y either bombs or crashed on board. 396/264. Ship retires from theatre. On 12–13 May Mitscher decides to draw the sting of the Japanese attacks by closing Kyushu and attacking the airfields again. Once again, he is caught retreating on  14 May. Enterprise is crashed once. Bomb penetrates “deep” into the ship.” 13/68. Enterprise retires from theatre, eighth CV to do so, 7th of Okinawa campaign out of 12 that started the campaign. There have now been 3451 American casualties in the fast carrier ops, including 1734 dead, exclusive of Princeton.

This is all very interesting, but one might argue that these were leakers getting through massive defensive screens during equally massive battles. Except that there is the case of the Battle of the Philippine Sea, where everything went massively right, and an American carrier group was still bounced with its pants down. (See Morison, 8: 263ff.) The circumstances perhaps make for a defence, in that the Japanese elected for a complex battle plant hat depended heavily on the Japanese attack being made at extreme range. Although much emphasis is made in the sources on the poor state of his pilots’ training, the failure of the complete, undisrupted raid of 82 a/c to have any effect on the carriers of TF 58.2 is probably due to fuel starvation, although Morison credits (unrealistically) AA. (272). In nigh ideal conditions, with the radar upon which Henderson could not rely, the Japanese put an 82 a/c raid over half the committed major American CVs. However much "Seagull" (probably John D. Cunningham) may protest the superior value of advanced interception over Henderson's small air groups under heavy armour, the RAA had a point in 1937.  

Here is, by contrast, what happened to some heavily armoured carrier subject to very heavy air attack.

On 10 Jan 41, the Med Fleet, with Illustrious in company in the body of the fleet over the standing objections of RAdm Lyster, Adm Aircraft Carriers, who believed that this made the CV overly vulnerable, was headed into Malta, escorting 4 freighters and the crippled DD Gallant. At 10:06, 5 Fulmars ("Seagull's" hasty Osprey replacement) were flown off for CAP. At 11:06 2SM 79 high level bombers were shot down, and 3 Fulmars landed on. At 11:26 2 low level TBs made a pass at the Illustrious, and 2 Fulmars chased them off. At 12:28, a Fulmar “subflight” was ready to take off, and Lyster apparently considered immediate launch, but he required explicit permission from Flag to turn into the wind, and this was not scheduled until 12:35. At 12:34 the Fleet received radar warning of an incoming raid. At 12:35 the first Fulmar was launched. At 12:36 40 German Ju 87 dive bombers are spotted, with 2 Fulmars on flight deck landing on. At 12:38 DB attack begins. 118) 30 Ju 87s attack Illustrious, diving from 12,000 feet to 7000, recovering, then making their final dives. 119–21) A bomb, assessed at 1000 lbs, passes through a portside 2 pdr S1 mount and into the sea, where it detonates and splinters strike the ship. ii) hits the flight deck at the bow and passes through into the paint room at the bow of the hangar, starting a fire. iii) A third hits S2 pom pom and kills entire crew. These three bombs were probably not 1000 lbs, as the Germans DBs did not employ a 1000 lb GP, and these were clearly high capacity weapons intended to attack the AA, probably 550lb HE. iv) a real 500kg AP struck the elevator as it was lofting a Fulmar to the flight deck. The elevator collapsed to the hangar deck and the bomb exploded there, setting the Fulmar on the elevator on fire as well as 9 armed Swordfish and 4 Fulmars on the hangar deck, and damaging the rear 4.5"batteries with splinters. Although the elevator can be closed off from the hangar with an armoured door, it is open at this time. A second 500kg AP (fifth hit) also strikes the aft elevator and again detonates in the hangar deck, or below it in the wardroom flat according to Robert Jackson. The third 500kg AP hits at the bow and penetrates. The fourth hits just forward of the aft elevator, penetrating the fractured deck and detonating on the hangar deck, or according to Jackson in “the compartment below” where it damages the steering lines. At this point the climbing Fulmars just launched reach the diving Stukas and interrupt their aim. By 14:00 Illustrious is underway at fleet speed of 21kts. Loss of steam steering ensues and Illustrious turns out of control in the formation, but steering is recovered by 14:30. At this time it is discovered that hydraulic oil dripping from punctured lines in the steam steering system is feeding uncontrolled fires in the hangar deck. This leak is stopped and the fires brought under control. The spray system was not effective, but direct hoses proved adequate. By 1600 Illustrious was maintaining 15kts and steering. At this time there was another DB attack, and a 500kg AP penetrated the deck just forward of the aft elevator. Once again the fractured armoured deck gives way and the bomb detonates below, I’m not sure where. At 17:30 the ship suddenly heels due to the free surface effect of the large amount of water now shipped in the hangar deck. This is controlled and corrected by heroic damage control. (128) Illustrious left Malta on 24 Jan despite being hit one more time and near missed.
The interesting thing about this narrative is that it reveals that there was not one straightforward penetration of the flight deck in its fully armoured area. The ship’s survival certainly depended on the capability of the armour under the hangar deck to withstand splinters and generalised damage control, but much more clearly on the limited effects of the fires on the hangar, almost certainly thanks to the surviving fire curtains and water mains. The American assessment in 1941 was that similar damage to the just-completing Essex-class carriers would cause their loss, which was the consensus opinion until the revisionists appeared. 

Just to make it clear that this was no fluke, consider some other cases. Formidable’s also suffered campaign-ending damage during an air raid, in which  12 a/c of Stuka Geschwader 2 on an antishipping patrol delivered 3 bombs, of which 1 detonates on the flight deck and 2 cut through starboard side. (87–8) Victorious was much more lightly damaged during “Harpoon,” by two Re 2000 fighter-bombers that plant two bombs on the deck at low level. One breaks up on the flight deck a “dud” the other is rejected and rolls into the sea, where it detonates to no effect.  (Rather lazily, I take this from a secondary source, specifically, Richard Woodman, Malta Convoys 1940–43 [London: J. Murray, 2000], 118ff; with additional information from Robert Jackson, Strike from the Sea: A Survey of British Naval Operations, 1909–69(London: A. Baker, [1970], 80ff, 87–8, 112.

Of course, all this relative invulnerability came at a price; no planes!

We have here two main lines of development. Essexes carried a great many aircraft, and suffered heavy operational losses to relatively light damage. Illustrious and modified Illustrious-types carrier many fewer aircraft, and suffered comparatively fewer operational setbacks from much heavier damage.Thus, neither Armoured flight decks nor their absence were a bad idea. What was a bad idea was restricting aircraft carriers to 23,000, or, for that matter, 27,000 tons displacement. Yes, the enemy had to cause more instances of carrier damage than if there had been, say, half the number of 54,000 ton carriers operating in WWII as 27,000 ton Essexes. But that's not it would have gone. 54,000 tons was far too large given the size of aircraft at the time. This, or this, represents war experience, and are the designs that the admirals would have liked to have had off Crete in the middle of 1941, or in the Pacific. Making carriers small didn't make them less vulnerable by making them cheaper. It made them more vulnerable by making them cheaper. 

Which is not, as far as I can see, that counter-intuitive a conclusion. 

*Really, any opportunity to get Grace Park into this blog is an opportunity worth taking.

No comments:

Post a Comment