Liberty Ship

By 1941 Britain was losing merchant ships faster than the entire British shipbuilding industry could replace them, and the arithmetic pointed one way: an island that imports its food and fuel loses a tonnage war long before it loses a battle. The following year would be the worst of the whole campaign — Allied shipping losses in 1942 ran to roughly 7.8 million tons — and what was needed was not a good ship but an enormous number of adequate ones, built somewhere the U-boats could not reach.

The design came from Britain — a simple, slow, coal-burning tramp steamer from J. L. Thompson & Sons of Sunderland, chosen precisely because it was old-fashioned and undemanding. The Americans adapted it as the EC2-S-C1, switched it to oil firing, and set about building it in yards that did not yet exist, with workers who had never built a ship.

The first was launched on 27 September 1941, on a day the administration named Liberty Fleet Day, and she was called SS Patrick Henry after the revolutionary who had said “give me liberty, or give me death.” Roosevelt used the phrase in his speech, and the whole class took its name from it.

He was also candid about how they looked. The President reportedly called the ship a “dreadful looking object”, and the press settled on ugly ducklings. Nobody disputed it: the Liberty was a slab-sided, blunt-bowed cargo hull with a single funnel and no pretensions whatsoever. The Patrick Henry took 245 days to build. Within two years that figure would look absurd.

The transformation was the work of Henry J. Kaiser, a construction magnate who had built dams and roads and had never built a ship in his life — which turned out to be the qualification that mattered. His yards treated a ship as a building site: sections were prefabricated elsewhere, brought in and assembled, and above all they were welded rather than riveted, some 600,000 feet of welded joint per hull.

Welding could be taught in weeks to workers with no trade at all, which was the decisive point: Kaiser’s yards were staffed by farmhands, shop assistants and, in very large numbers, by women who had never held an industrial job before the war. The results were extraordinary: average construction time fell from 197 days to about 14, and in November 1942 the SS Robert E. Peary was completed 4 days, 15 hours and 29 minutes after her keel was laid — a stunt, openly so, but one that only an industry which had already made the ordinary case routine could even attempt. In all, 2,710 were built by 16 shipyards.

The ship itself was deliberately unremarkable: about 134.6 m long, carrying roughly 10,800 tonnes of cargo, driven by a triple-expansion steam engine at 11 knots. That engine choice is the same logic that produced the Flower-class corvette escorting her — obsolete technology, chosen because it could be built and repaired by people who were not specialists, anywhere in the world.

A single Liberty could carry the equivalent of some 2,840 jeeps, or 440 tanks, or 230 million rounds of ammunition. Her armament was purely defensive — a gun on the stern and light anti-aircraft weapons, manned by a US Navy Armed Guard detachment berthed among the merchant crew — because at 11 knots, slower than a surfaced U-boat, she had no other answer than the escorts and the convoy around her.

Then they began to break. Liberty ships developed cracks — around 1,200 of them suffered fractures during the war, roughly 30% of the class, and three split completely in two. The failures happened in cold water and often without warning, sometimes at anchor, and for a time nobody could explain them.

The explanation, found by the Cambridge metallurgist Constance Tipper, was that the steel underwent a ductile-to-brittle transition below a critical temperature: in North Atlantic cold, plate that behaved normally in a test shop became glass. And the very innovation that had made the ships buildable made the consequence catastrophic — in a welded hull a crack can run the length of the ship, where a riveted hull would have stopped it at the next seam. Tipper’s work produced the Tipper Test for transition temperature and stands at the origin of modern fracture mechanics.

They were civilian ships with civilian crews, and that is easy to forget. Merchant seamen signed on for wages, were not subject to military discipline, received none of the protections of uniformed service — and died at a rate that was, proportionally, among the heaviest of any American service in the war.

The reason is simple: a Liberty in convoy could not manoeuvre and could not outrun anything — convoys made 7 to 9 knots, and she had barely more — and if she was carrying ammunition or fuel there was frequently nothing left to rescue. What she could do was arrive, and enough of them arrived. By 1943 the Allies were launching merchant tonnage faster than the U-boat arm could sink it, and that crossover — a graph rather than a battle — is where the Atlantic was decided.

The Liberty ship is the least glamorous vessel of the war and arguably the most consequential. Every campaign documented here depended on cargo crossing an ocean: the tanks in the desert, the aircraft in Britain, the fuel at Normandy, the entire American war in the Pacific. The Liberty is the machine that moved all of it, and it won by being simple enough to be built by anyone, anywhere, in numbers that made losses survivable.

Two are still afloat as museum ships — the SS Jeremiah O’Brien and the SS John W. Brown — out of 2,710. The design assumed heavy losses and was drawn to make them survivable: about 1,200 hulls developed cracks and three broke in two, while average build time fell from 197 days to roughly 14.

• Frederic C. Lane, Ships for Victory: A History of Shipbuilding under the U.S. Maritime Commission in World War II (Johns Hopkins University Press).
• John Gorley Bunker, Liberty Ships: The Ugly Ducklings of World War II (Naval Institute Press).
• Roger Chesneau (ed.), Conway’s All the World’s Fighting Ships 1922–1946 (Conway Maritime Press).