Armored protection against gunfire was initially a simple affair of interposing sufficiently thick pieces of armor on a ship's vertical surfaces to stop incoming, short-ranged, low-trajectory gunfire. To keep the ship's armament intact, guns were enclosed in armored barbettes or gunhouses, and in casemates or redoubts. To keep the ship afloat, the waterline received light armored plating along its length, both above and below the surface. Forward, up to medium armor protected the ship's bow, often to a height greater than most of the protective waterline plating. Aft, abreast the steering gear, the armor again thickened to medium thickness. Abreast the highly explosive magazines and the machinery essential to the vessel's mobility, the waterline plating was thickened into a heavy belt. Above this heavy midships belt another medium thickness belt was intended to keep light and medium caliber shells out of the casemate batteries. Deck armor was minimal and largely intended only to keep out the fragments of shells detonating above protected spaces. Outboard, near the heavy midships belt, the armored deck was often sloped downward at about 45 degrees to meet the bottom of the belt, again to assist in keeping out fragments, this time of shells detonating just after piercing the heavy armor. This armor scheme developed along with the pre-dreadnought battleship, which typically sported four heavy turreted guns, eight to twelve medium caliber turreted or casemated guns, and twelve to twenty-four unprotected or - at best - splinter shielded light guns. In 1905, the Battle of Tsushima revealed long range fire to be a significant hazard, it became necessary to armor the horizontal surfaces more heavily. Armored turret tops and conning tower roofs appeared and some casemates received overhead protection as well. The initially relatively low trajectories did not require more than about double the previous splinter protection. This was the standard the original long range gunnery ships, the early dreadnoughts, were built to. It would prove tragically inadequate at Jutland as ranges increased further and shells fell at ever steeper trajectories. Unbeknownst to the antagonists of Jutland, the problem of adequate protection against long-range gunfire has already been developed across the Atlantic.
As early as 1908, the United States Navy began work on its "1912 Battleship" design, intended to succeed the still-building New York Class. The General Board of senior and retired admirals, displeased with the wasteful disposition of the five- and six- turret Delawares, Floridas, Wyomings, and New Yorks, opted for triple turrets to save length and make more tonnage available for protection. When even this did not yield adequate protection against plunging (steep trajectory) gunfire, the decision was taken to adopt the "all or nothing" scheme of armoring.
The best description of this concept comes from Norman Friedman:
The logic of 'all or nothing' protection was that at very long ranges, ships would be attacked primarily with AP shells since hits might be anywhere on a ship, and HE would be useless against thick belt or deck armor. In consequence only the heaviest armor (or no armor at all) was worth using: anything in between would serve only as a burster. By way of contrast, the Royal Navy concluded at about this time that heavy HE shells would be extremely effective against unarmored portions of ships, and used considerable amounts of medium armor, which would resist HE fire, in its dreadnoughts. Only after World War One did the Royal Navy adopt "all or nothing" protection, in the abortive 1921 battleships and battlecruisers, and in the Nelsons, all of which were expected to fight at longer ranges than those envisioned for the earlier British Dreadnoughts. To the extent that the U.S. battleships, then, were designed specifically to fight at extreme ranges, they were well ahead of their time.
The result was the armor scheme of the USS Nevada, which eschewed any medium protection in favor of a thicker belt, massive turret faces, and thicker turret tops and decks than contemporary practice. It is interesting, as Friedman notes, that the USN took this step before the actions at Dogger Bank and Jutland made it clear that long-range engagements would necessitate better horizontal protection. Part and parcel with the armor scheme went the USN concept of the "raft body." Since the "all or nothing" scheme eliminated much of the light to medium armor intended to protect a ship's buoyancy fore and aft, it was necessary to ensure sufficient buoyancy was enclosed within the heavy armored box to ensure the ship could survive if its ends were riddled and flooded. This provision was sometimes missed by designers in other countries, resulting in designs as vulnerable to dangerous trims from flooded ends as the pre-"all or nothing" SMS Lutzow at Jutland.
The "all or nothing" scheme dominated the post-Jutland era of battleship design undergoing refinements once battleship construction resumed after the Washington Treaty. Side armor was canted outward at the top, and decks increased in thickness as ranges increased and shells fell at steeper angles. Armored belts were moved inboard from the ship's side to accommodate and / or interact with the torpedo protection below and / or outboard of the belt. Some designs added a "decapping" surface of light armor or thick hull plating outboard of the main armored belt. Many different schemes of horizontal protection were tried including, most prominently a single thick deck, and a thick armored deck with a thin "splinter" deck below for catching fragments. The threat from aircraft also saw the addition, in some designs, of thin armor on the weather decks to keep out HE bombs, and to prematurely initiate fuze action in heavy AP bombs before they could reach and penetrate the main armored deck. This concept seemingly reverses the priorities of the "all or nothing" scheme until one considers the fact that a vertically falling bomb enjoys significantly greater penetrating power than a shell. Also, a shell passing more or less horizontally through the unarmored portions of the ship is less likely to connect with something vital on the way through. A bomb falling onto an armored portion of an "all or nothing" ship is going to connect with a vital space. If the bomb is heavy enough and dropped from high enough, no practical amount of armor will stop it. Thus it becomes necessary to detonate the weapon as soon and as high in the hull as possible. Despite the fact that horizontal protection was often spread over several layers, it remained concentrated over the vital spaces and thus was in keeping with the "all or nothing" idea. A bomb striking the unprotected ends of the ship could easily pass out the bottom without detonating. It's important to remember that the term "all of nothing" refers to devoting the armoring effort entirely to protecting certain vital spaces in the ship, not to placing the armor into single thicknesses only. In Second World War action, the "all or nothing" scheme proved successful. Incremental armoring did not, although available test cases were few. HMS Hood, featuring the older incremental armor was lost most probably to a shell passing through a thinner upper belt and into a magazine. DKM Bismarck also featured incremental armoring, as was largely laid waste during its final battle above its armored deck by AP shells detonated by a medium thickness upper belt. HIJMS Hiei and Kirishima, both with the older protective scheme, suffered heavily from fires started in their casemates by AP shells. These fire contributed along with other damage to their untenability and eventual loss after damage off Guadalcanal. Details of DKM Scharnhorst's loss are lacking, but probably resemble Bismarck. In contrast USS South Dakota came off fairly well when a power failure and an ill-considered turn silhouetted the ship to Japanese fire off Guadalcanal. The ship was extensively hit by common, HE, and AP shells in the superstructure, most of which passed through while causing relatively little damage.
According to Dulin and Garzke:
Many small fires in the superstructure were quickly put out... Despite 27 shell hits, the ship's hull strength, buoyancy, and general stability were not materially affected.
- "Battleships and Battle Cruisers, 1905-1970" by Breyer
- "Battleships: United States Battleships in World War II," by Dulin and Garzke
- "Battleships: Allied Battleships in World War II," by Dulin and Garzke
- "Battleships: Axis and Neutral Battleships in World War II," by Dulin and Garzke
- "U.S. Battleships: An Illustrated Design History" by Friedman
- "The Naval Battle of Guadalcanal" by Grace
- "The First Team and the Guadalcanal Campaign," by Lundstrom
- "History of U.S. Naval Operations in World War II, Volume V, 'The Struggle for Guadalcanal'," by Morison
- 1 May 2000