With the brilliant South Dakota class, American designers built the finest of the treaty battleships. However, the clouds of war were gathering on the horizon, and rumors were circulating of Japanese battleships that dwarfed the ships being built under the Second London Treaty. In an attempt to prevent this from happening, the authors of that treaty had included an "escalator clause", which allowed the treaty powers to raise the treaty limits if ships that breached them were built by other powers. Public opinion was still largely pacifist, so it wasn't until March 31st, 1938 that the US, Britain and France exchanged notes invoking this clause, ultimately signing a protocol raising the battleship tonnage limit by 10,000 tons. This laid the foundation for the greatest of all battleships, the Iowa class.
Iowa in 1945
The extra 10,000 tons gave designers several options, and studies began almost immediately. One possibility was to fit another triple 16" turret on a stretched South Dakota hull, with the tonnage going into a longer citadel and bigger engines to maintain 27 kts. Another was to upgrade the guns to 18",1 which kept the length of the citadel down and allowed the armor to be upgraded. Both of these would have been entirely in keeping with the traditional American policy of prioritizing firepower and protection over speed. But some in Preliminary Design suggested a radical alternative, a very fast ship intended not to stand in the line of battle but to hunt down Japanese cruisers.2 The first sketches were of a ship armed with 12" guns, capable of cruising for 20,000 nautical miles at 15 kts,3 and with a top speed of 35 kts. To hold displacement down to even 50,000 tons would have meant that the ship was armored against 8" weapons, with an immune zone of 10,000 to 30,000 yards.
New Jersey in 1943, with French battleship Richelieu
This design was obviously too big and of rather dubious effectiveness, but the idea of a really fast ship appealed to the General Board. Even the North Carolina and South Dakota classes, much faster than the pre-treaty American battleships to allow them to operate with the carriers, were still slower than the flattops. A sketch was done of a 30-knot SoDak, which appeared to displace around 37,600 tons, prompting a request for a 33-knot design, which came in at 39,930 tons. However, a faster ship would need more freeboard forward to stay dry and a deeper armored belt, which would push up weight by about 1,375 tons directly. However, each ton of direct weight would require more structure and machinery to maintain speed, so the total impact of the changes was actually 2,400 tons. Suddenly, the design was pushing the new 45,000 ton limit for a gain of only 6 kts.4
One of Iowa's turrets under construction
This seemed an inadequate improvement over the SoDak, and so the General Board directed that the study include some of the 16"/50 guns that had been built for the battleships and battlecruisers cancelled under the Washington Treaty in place of the 16"/45s used on the previous two classes.5 It also allowed draft to be increased despite the potential to complicate the use of harbors and docks. This increased propulsive efficiency, which let them save weight. More structural weight was saved by the unusual hull form, which concentrated buoyancy amidships, where the heavy weights were. However, the lack of buoyancy near the bow made the ships very wet forward. Even with these innovations, the cost of the 16"/50 guns was too high. Besides the direct increase in turret weight, about 400 tons per turret, the barbette diameter would need to be increased from 37'3" to 39'4", and once the structural and propulsive changes were taken into account, the total cost of the new weapons was 2,000 tons, pushing the total tonnage of the sketch to 46,551 tons.
Missouri at Ulithi, 1945
While some tonnage could be saved by minor tweaks to the design, such as cuts to splinter protection and the underwater belt, it was obvious that the only option was to reduce the weight of the turrets. The Bureau of Construction and Repair (C&R) found a smaller turret in one of the tables sent to them by the Bureau of Ordnance (BuOrd), with a barbette diameter of 37'3". This would let them shave 785 tons, which when coupled with 435 tons in minor weight reductions and the consequent reductions in indirect weight, let them cut the 2,000 tons required to bring the design in under the treaty limit. This triumph of design was presented to the General Board on June 2nd, 1938.
An external and internal profile of Missouri, 1944
Unfortunately, they forgot to tell BuOrd that they'd made this decision. The turret in question was nothing more than a sketch, and one that ignored the past few decades of development in heavy gun mounts. It would have been significantly slower-firing than the turrets on the North Carolina and SoDak, and unable to elevate to 45°. BuOrd instead worked on a 39' diameter turret, a scaled-up version of the mounts developed for the previous battleships. It wasn't until November, when both designs were nearly finished, that the two groups figured out that the turret wouldn't fit into the barbette.
Wisconsin in the Pacific, January 1945
Changing the design to fit the turret would have caused the ship to breach the treaty limit. The biggest limit was the strength of the upper deck, which would be compromised if both barbettes were enlarged in their existing positions. To keep the amount of deck between the turrets the same, Turret 1 would have had to be moved forward, into a region where the weather deck was rising rapidly to keep the ship dry at high speed. This in turn meant that Turret 2 and the conning tower would have had to be raised to clear it. The only obvious solution was to go back to the 16"/45 gun, but BuOrd was able to save the day by developing a new, lighter, weapon, the superb 16"/50 Mk 7 gun, which could fit into the 37'3" barbette.6
Iowa in drydock showing her unusual underwater hull form
Resolving the problem with the guns cleared the last major roadblock to production, although a few changes were made while the ships were under detail design. The biggest was subdividing the engine spaces, a decision made in November 1939. Originally, the ships had been planned with unit machinery on the lines of the South Dakotas, with four rooms each containing boilers and a turbine set. However, it was realized that while the length of the rooms on the SoDak, with only 131,000 SHP, were acceptable, the Iowas, with 212,000 SHP, would require 64' rooms, meaning that a torpedo hit on the bulkhead dividing two rooms would flood 128'. This was too much, and in November of 1939, the scheme was changed so that engines and boilers were in separate rooms, cutting floodable length in half. This did require extra weight, but Britain and France had suspended the treaty on the outbreak of war. To keep draft from increasing further, a foot of beam was added, resulting in the 108'6"-wide Iowas being the widest vessels ever to transit the Panama Canal. Underwater, the vessels had twin skegs on the inner shafts for hydrodynamic reasons, and a bulbous bow to reduce resistance at high speeds.
Iowa from above, showing her unique hull form
The first two units, Iowa and New Jersey, were ordered on July 1st, 1939, from the New York and Philadelphia Navy Yards respectively under the provisions of the Second Vinson Act, followed by Missouri and Wisconsin from the same yards the next June. The lapse of the treaty limits allowed the transverse bulkheads closing the fore and aft ends of the armored box to be strengthened, but armor for Iowa and New Jersey had already been ordered, and the improvement was limited to the later ships. Four ships was considered sufficient for their role in screening the carriers, and it was expected that the next class would revert to the more traditional American priorities of firepower and protection. The Fall of France, almost simultaneous with the ordering of the second pair of ships, changed things. Congress passed the Two-Ocean Navy Act, which authorized two more ships, Illinois and Kentucky, along with five ships of the Montana class, which ultimately came from the studies of a 27-knot ship with 4 turrets. Unfortunately, none of these ships were completed, as the events of the war shifted building priorities to carriers, escorts, and landing craft.
Iowa moored in San Pedro
As Norman Friedman said, "If these four were among the most elegant capital ships of the Second World War, they were also among the most puzzling." Their high speed, radically at odds with the conventions of US battleship design, made them ideal carrier escorts for the war that actually happened, but could have potentially compromised them in a surface action. The demands of extremely high speed, which in a seaway often made them faster than 38-kt destroyers, meant that in terms of surface combat, they were little better than the preceding South Dakotas. But that basis was good enough, when coupled with the incredible 16"/50 Mk 7, to make the Iowas by far the best battleships ever built. Not only would they have been able to defeat the Yamatos in battle, but they also gave excellent service as escorts in WWII, as bombardment vessels off Korea, and were versatile enough to be brought back into service to fight modern wars. Ultimately, any design where every completed ship is in the service of a front-line navy on the 50th anniversary of being laid down has to be pretty special, and the Iowas were even better than that.
1 This work was done before the 16" limit was reaffirmed. ⇑
2 There are obvious parallels between this concept and some of the thinking behind the early battlecruisers, and it later came to fruition in the Alaska class. ⇑
3 Previous American battleships had only been capable of 15,000 nm at this speed. ⇑
4 It bears pointing out that some of the tonnage increase went into buying back the less obvious sacrifices made on the SoDaks. Corridors on Iowa are wider, there's more space in the cabins, and so on. ⇑
5 Sumrall suggests that this was driven by the development of the 16" superheavy shell. Range for the 16"/45 would fall from 40,600 yds for the 2,240 lb shell to 36,900 yd for the 2,700 lb shell, and the General Board wanted to keep a range of at least 40,000 yds. ⇑
6 Astonishingly, there were no reprisals against BuOrd for this screw-up, probably because they were able to develop a new gun in record time. ⇑
Comments
What is that round platform on the very tip of the bow?
There were a couple of 20mm AA guns mounted there.
Conventional wisdom is that battleships and battlecruisers were regarded as the primary naval combatants and aircraft carriers a secondary fleet element used for scouting, harassment, etc in advance of the main fleet engagement, until the destruction of the US battle line at Pearl Harbor forced a reevaluation (on both sides). The various naval arms control treaties of the interwar period, through 1935, seem to reflect this belief - aircraft carriers are discussed, but explicitly not considered capital ships and the primary concern seems to be that nobody try to sneak in a quasi-battlecruiser by sticking heavy guns on an aircraft carrier.
If the United States was explicitly designing a class of battleships to keep up with aircraft carriers in 1938, devoting ten thousand tons of displacement to speed that could not be used in the context of a 27-knot battle line, that suggests that the consensus within the USN had shifted to aircraft carriers as the primary fleet elements going forward, well before Pearl Harbor. In particular, if the Japanese are rumored to be building super-battleships and the USN gets one round of naval building to counter them, it seems significant that the focus is on "make sure we can still protect our aircraft carriers" rather than "make sure we can match the Yamatos in terms of guns and armor".
The conventional view of naval thinking before Pearl Harbor is obviously wrong. This isn't particularly controversial if you read the better histories, which often emphasize how air-minded the USN and RN were. I'd even point out that the rapid development of air power meant that the situation at the time of the Second London Treaty was very different from the earlier treaties, but the 1936 treaty didn't have number caps on any type. In theory there was no particular reason not to build carriers that violated the treaty limits and declare them as capital ships. All that said, I see a couple of mistakes in your analysis.
First, the USN didn't know it had one round of naval building to create a counter. There were always parallel studies for a 27 kt ship with more guns and heavier armor, and they were intended to be the main thrust of ship development in the long term. If the US had gotten drawn into the war two years earlier, the Iowas might well be the kind of curiosities that the Montanas and Lions are today. If the war had been delayed a year or two, we might have seen Montanas completed, and that changes the narrative. As it was, the General Board thought they needed carrier escorts more than they needed the extra guns a 45,000-ton, 27 kt ship would have brought to the table.
Second, the USN didn't know very much about the Yamatos at the time. There was general agreement among the naval attaches that they were going to build a ship with 16" guns and a displacement around 45,000 tons. This is obviously a close match to Iowa, and if you feel a strong need to protect your carriers from marauding Kongos, then the fast battleships make perfect sense.
That source says the 1938 estimates were pegging the Yamato at twelve 16" guns. Also 45,000 tons minumum, and if they weren't expecting 30+ knots on that displacement then they presumably were expecting more than a 12" main belt. So the Iowa as planned would have been outclassed in a gunnery duel by the Yamato as anticipated.
It's mostly the timeline I am interested in. The Washington and First London naval treaties were clearly negotiated by people who considered the aircraft carrier a supporting element. This is also reflected in the early versions of War Plan Orange, and in the design of the USS Ranger. All signs point in the same direction in 1930.
Second London (1936) is less clear due to its limited scope, and I need to track down a copy of the 1935 revision to War Plan Orange, but it still doesn't look like naval aviation is considered even comparable in importance to battleship gunnery.
In 1938, Japan has already started what will be later known as "World War II", and nobody in what will later be known as "The Pentagon" is confident that the United States is going to have time to build out the Iowas and Montanas both before being dragged into that conflict.
Putting the top priority on a ship that will certainly be outgunned by the new Japanese battleships, specifically because it will be able to screen the carriers, suggests a remarkably fast shift from aircraft carriers as a supporting arm to aircraft carriers as the preferred combat arm.
I think that you're modeling the USN as thinking rather more like the denizens of a second-rate warship forum than they actually were. While Iowa vs Yamato is the headline case, and it's certainly not something you're going to neglect in your analysis of a new design, it's not the only thing you're going to take into account either. It's a trade of 25% of the ship's firepower in exchange for being able to stay with the carriers. The loss of firepower is annoying but probably not disastrous if you run into Yamato, particularly if you build two (or five) ships to the Japanese one. But having a battleship in the screen if the carriers run into a Kongo, or even some cruisers? That's going to make a lot more than a 25% difference in that action.
But it was already being correctly reported in January 1938 that the Japanese were building two Yamatos, with plans for at least one more. And whatever ships the United States builds, get split 50/50 between the Atlantic and Pacific fleets. So even assuming the war waits for the US to build out the Iowas, that gives you rough parity in super-battleships for the decisive battle, not 2-5:1
Also, even the Pacific Fleet Iowas can't be in two places at the same time.
If the US had built Montanas and deployed them with the battle line, yes, the carriers would have had to flee and maybe be sunk if they ran into Japanese battlecruisers. That's very bad.
If the US builds Iowas and uses them to screen the carriers, then the US battle line gets to flee or maybe be sunk, if it encounters the Japanese battle line at full strength. That's also very bad
And if the US builds Iowas and matches them against the Yamatos, that's the worst of both worlds. The Iowas are with the battle line, so the carriers are still vulnerable to battlecruisers. And on the battle line, the Iowas' 33-knot speed can't be effectively used and the USN has wasted 40,000 tons of steel and 300,000 hp of machinery to make the tactical equivalent of a SoDak and match it against a ship at least 33% bigger and more powerful.
None of these plans make sense unless the expectation is that it's OK to match a SoDak against a Yamato because that entire contest is secondary to the part where the Japanese fleet is crippled by air attack first. Which, fine, is what actually happened. I'm just surprised to see such a strong commitment to that plan as early as 1938.
In 1938, it was reasonable to assume the bulk of the US fleet would be in the Pacific. War hadn't even broken out in Europe, and the fall of France radically changed the naval balance in that theater. And there's no reason to discount the SoDaks and even the North Carolinas in an engagement with the modern Japanese battleships (assuming the two sides have evenly-matched old battle lines) so you are looking at a substantial numerical superiority for the USN. In that context, yes, it makes sense to focus on making your battleships more useful instead of more powerful. (Also, none of the 45,000 ton slow designs were particularly compelling. The Montana only started to work when the treaty was suspended.)
As for carrier operations, if the carriers were operating near the battle fleet (as was doctrine at the time) it probably would have been possible to concentrate their escorting battleships in time for the fleet engagement. That was what the US planned to do with TF 34 during most of the war as fought.
Also, it's worth pointing out that there's valuable context in the North Carolinas. They were originally intended as fast battleships for escorting the carriers, and the standard speed was only increased to 27 kts when intelligence of the increased speed of the Japanese battle line came in. So for both the first class of new battleship after the treaty and the first class after the limit was raised were supposed to be fast. They really were worried about the carriers facing surface attack, and the example of Glorious suggest they weren't really wrong about this.