Such a liquid space could, however, do no more than slow down the destructive force of a torpedo explosion, without stopping it. In fact, it was deemed necessary to back up the oil space with a torpedo bulkhead of thin armour. Following the trials with HMS Belleisle, the Dreadnought of 1906 was provided with armoured vertical screens covering the magazines. In order to protect her against two torpedo hits, her main bulkheads, unpierced by watertight doors, were carried up to 9ft (2.75m) above the normal waterline, to avoid capsize through flooding of side compartments below the armour deck. For the same reason, she had a large metacentric height, which meant more rapid rolling and made her a slightly less stable gun platform, but it was considered a worthwhile trade-off.
From HMS Bellerophon of 1909, the magazine screens of the type fitted in Dreadnought were replaced by internal armoured bulkheads up to 1in (25.4mm) thick, extending from the forward magazines to the aft magazines. This followed reports from the Russo-Japanese War that ships torpedoed or mined had only been lost if their magazines exploded. The armoured anti-torpedo bulkhead was intended to stop plating and torpedo fragments from reaching the magazines. Even so, the shock of a torpedo explosion was likely to send fragments of the torpedo and the ship’s outer hull through the oil, and through this inner armoured bulkhead as well. So designers kept adding more and more layers to the side torpedo defence system. Of course, there was a physical limit to this progression, and space was at a premium.
There were two solutions: to increase the beam of the ship, keeping the explosion as far from the vitals as possible; or find more internal space by replacing boilers and machinery with more compact modern versions. The latter solution explains why, when the Royal Navy was rebuilding its Queen Elizabeth-class battleships in the late 1930s, instead of going for more powerful power plants, as the Japanese were doing, they opted instead for more compact machinery.
The best TDS turned out to be the American 1915 design built into the Tennessee-class super-dreadnoughts. They benefited from multiple layers of vertical bulkheads, voids and liquid-loaded spaces in the following order:
— outermost, a void, which carried the risk of asymmetrical flooding if pierced, but which could easily be rectified by counter-flooding the voids on the opposite side of the ship;
— next, a plain bulkhead;
— then a liquid-loaded layer;
— bounded by a first elastic armoured bulkhead;
— a second liquid-loaded layer;
— bounded by a second elastic armoured bulkhead;
— a third liquid-loaded layer;
— a third elastic armoured bulkhead;
— an inner void;
— an unarmoured holding bulkhead.
Each armoured bulkhead was sufficiently spaced from the next inboard to avoid touching the latter under the expanding force of the explosion. This system worked exactly as planned when USS California was twice torpedoed at Pearl Harbor. Her inner holding bulkhead was not perforated. She still sank, however, because of a lack of preparedness which compromised her watertight integrity.
In ships where it was desirable to increase the distance between the outer plating and the vitals, a simple remedy was to graft an external bulge or blister onto the outside of the hull. Bulges were first applied to coast-bombardment monitors before their general application to capital ships. In the RN Renown class, Courageous, Glorious and Furious, and later HMS Hood, this bulge was made an integral part of the hull design.
As an example of a TDS which was modernised and extended, illustrated here is the scheme as originally built into the USS Arizona in 1915, and her modernised protection as reconstructed in 1929–31. Note in the hold plans that the multiple layer bulkhead protection was at its widest abreast the boiler rooms. If the ships had been refitted with the more compact but heavier turbo-electric drive as suggested, they would have saved sufficient internal space to have extended this system to abreast the engine rooms.
Examples of ineffective TDS were many and varied. HMS Hood’s bulges were filled with sealed hollow steel tubes, designed to absorb the effects of a torpedo explosion by being crushed. A similar arrangement was fitted inside the bulges added to the ‘R’-class battleships. They appear to have been completely in effective. Possibly the worst TDS system of all was that conceived by Italian designer Pugliese, and fitted to the Littorio class as well as the rebuilt Cavour and Andrea Doria classes. Pugliese placed his faith in a large horizontal cylinder running the length of the TDS. Inside the cylinder were the ineffective sealed steel tubes. The main disadvantage of this system was that the explosive force, finding its way around the sealed tubes, met up with the inner holding bulkhead which, because of the large cylinder outboard of it, was concave in design, in other words the very opposite of what was required to resist the explosion. The results were plain to see at Taranto.
The TDS did nothing to protect the narrower ends of the ship. A torpedo hit on the unprotected bow area could reduce her speed and seaworthiness. A hit on the stern could damage or even destroy the rudders and propellers. One feature which American designers favoured was the skeg, an extension of the hull structure to surround and protect the propeller shaft, fitted in place of the more usual external struts. These could have a serious effect on handling, but if correctly designed could, as a minimum, reduce the risk of damage to the inner shafts. A side effect of the use of skegs was the ability to increase the width of the stern area, thus providing increased protection to the aft magazines.
Here, once again, size did matter. The more complete the watertight integrity of individual compartments, the better were the chances of a ship struck by a torpedo surviving. Late British designs such as HMS Vanguard largely did away with horizontal openings in lateral bulkheads, many compartments being accessed only from above.
Providing a longitudinal bulkhead down the centreline of the boiler and engine rooms seemed initially to guarantee against losing all power to a torpedo hit on one side. Many examples of loss through capsize were needed before this dangerous arrangement was superseded. On the other hand, dividing compartments by means of lateral bulkheads running across the ship from one side to the other posed the risk of the free water effect seriously reducing the ship’s stability, also leading to capsize. Combining the two into multiple compartments was the best solution.