Engineers of Victory: The Problem Solvers Who Turned the Tide in the Second World War (6 page)

And that they did very well. As the number of U-boats available to Doenitz rose steadily during 1942, their crews also grew in experience, their detection equipment became more reliable, their range was increased by the introduction of “milk cow” refueling subs, and they were masterfully coordinated by Doenitz. America’s entry into the conflict gave them fabulous opportunities against a new enemy and his merchant fleet, almost completely unprepared for this type of warfare, all the way along the still floodlit eastern coast. Longer-range submarines were sent farther afield to pick off rich targets near Sierra Leone, in the eastern Caribbean, off Buenos Aires, and off the Cape. But Doenitz never stopped hammering away on the crucial North Atlantic routes.

Total Allied shipping losses had jumped from about 750,000 deadweight tons in 1939 to an awful 3.9 million tons in the chaotic year of
1940, increased in 1941 to 4.3 million tons, and then soared again in 1942, as we have seen, to a colossal 7.8 million tons. Of course, there were heavy losses in other regions—off Dunkirk, in the South Atlantic, in the Mediterranean, and by 1941–42 in the Far East—but the heaviest losses (for example, 5.4 million tons of the 1942 grand total) occurred in the North Atlantic. By comparison, Doenitz’s U-boat losses were moderate in those years: around twelve in 1939 and around thirty-five in 1941, increasing to eighty-seven in 1942. These U-boat losses were completely sustainable; the merchant ship losses—and equally the losses of their experienced crews—were much less so.
⁶

It is therefore not surprising that the shipping losses of March 1943 scared Churchill and the Admiralty. If Doenitz’s wolf packs could inflict that much damage during dark and stormy conditions—the main attacks ceased only after March 20, when already rough waters were joined by what was essentially a massive Atlantic hurricane—the planners worried how their Allied convoys could survive in lighter and calmer times, with the moon shining across the waters. Would the loss rate double again by May and June? And were the submarines becoming harder and harder to trace and to sink? The jubilant U-boat crews and their determined admiral must have hoped so.

And yet the one-sided results of the March battles were never repeated again. In fact, they turned out to be the high point of the submarine offensive against Allied shipping, a momentary peak that then fell away so precipitously that each side was stunned by the transformation. It is hard to think of any other change in the fortunes of war that was both so swift and so decisive in its longer-term implications.
⁷

Precisely because it was so, it is important to take a closer look at the epic convoy battles of March to May 1943, when the balance of advantage swung so decisively from U-boat triumph to U-boat disaster. Fortunately, the sources for this story are excellent, down to the hour-by-hour tracking of virtually every submarine’s movements and the turn of every convoy.
⁸

The month of March began badly for the Allies. While the American, British, and Canadian naval authorities were at their Atlantic Convoy Conference, hammering out decisions regarding zones of control, reinforcements, and the rest, a confident Doenitz was dispatching more and more submarines to join each of the four large wolf packs he had
established in the central Atlantic, usually two in the center and one each on the northern and southern flanks. Moreover, at this stage in the intelligence/decryption conflict the Germans very much possessed the upper hand; B-Dienst was providing its chief with extraordinarily complete descriptions of the time and course of the Allied convoys, sometimes even before they left harbor. By contrast, the code breakers at Bletchley Park and at the Admiralty were having difficulty reading German messages days after they were sent. In sum, the shepherds, though gallant, were more than normally disadvantaged, weaker than usual, groping in the dark. The wolf packs were ready to pounce.

Thus their slaughter of Convoy SC 121, which sailed from New York to various British ports on March 5, 1943. Despite the weather being perfectly foul, some U-boats not picking up signals, and a late rush of a few Allied escort reinforcements, the odds were overwhelmingly in Doenitz’s favor. The great expert on these March convoy battles, Juergen Rohwer, provides us with a meticulous order of battle:

So there were fifty-nine vulnerable and slow merchantmen, with initially only five escorts, against twenty-six U-boats, and with no air cover for the convoy until the third day of the fight—and what was air cover anyway if the submarines attacked chiefly at night? The result was an ordeal from March 7 until March 10, when Doenitz called off his boats. Thirteen merchant ships totaling 62,000 tons had been sunk, but not a single submarine had been lost. It was perhaps the most disproportionate, one-sided battle of the entire war—and was deeply satisfying to Hitler, to whom Doenitz regularly reported.

There were, however, another couple of early March convoys across the Atlantic that also command attention. Convoy ON 170, for example, was ably directed away from all of these deadly mid-Atlantic battlefields and thus steamed across the northern waters without a loss and without (so far as we can tell) an encounter with a U-boat. Here was the case for shepherds and sheep simply taking the high Alpine passes and avoiding the wolf-strewn valleys below. Many an Allied convoy, in fact, survived the crossing unscathed, either because of clever routing or simply because Doenitz had directed all his boats to go after a different one.

A more mixed story is that of Convoy HX 228, which fought its way across the Atlantic between March 7 and 14 in a craziness and confusion that might remind naval historians of Nelson’s entanglements with the French fleet at the Battle of the Nile (1798). At one stage in this battle the destroyer HMS
Harvester,
having rammed U 444, got its propeller shaft entangled in the latter’s rudder and was only released by the French frigate
Aconit
ramming and sinking the submarine. The
Harvester
was torpedoed the next day, but the
Aconit
promptly sank the submarine that had done so, U 432. At the end of it all, HX 228 lost only four merchantmen plus the destroyer, while the escorts—joined, limpingly, by the first escort carrier, USS
Bogue,
a harbinger of things to come—kept a good account of themselves throughout. The U-boat crews were composed of formidable and intrepid men, but the British, American, and Canadian sailors—a small number of old hands and a vast recruitment of new officers and crew to their navies and merchant marines—showed themselves on this occasion to be equally resourceful.

However suggestive of a possible Allied recovery at sea, these hints were swept away by the achievements of the U-boats against convoys HX 229 and SC 122 between March 16 and 20. This was the most frightening moment, and not just for the fate of those two groups of merchantmen but for the overall convoy strategy as well.

Unlike a classic land battle (between Greeks and Spartans, or Wellington and Napoleon), where each opponent was roughly similar in composition, the two sides’ forces in the Atlantic struggle were very different. Doenitz’s U-boats were, roughly, all the same; the captain and crew of an older Type VII submarine were no doubt envious of those in
the faster, larger, and better-equipped Type IXs—unnecessarily, as it turned out—but all of them could reach out far into the Atlantic, fire their deadly torpedoes, and dive fast, away from counterattack.
¹⁰
By contrast, the Allied convoys contained a motley assembly of ancient tramp steamers, ore carriers, oil tankers, mail and passenger ships, and refrigerated ships.
^b
The cargoes they carried were equally heterogeneous—grain, linseed, meat, army supplies, aircraft fuel, sugar, bauxite (for aluminum), steel, tobacco, “African produce” (probably vegetable oil and hardwoods), and everything else needed to keep a nation of forty million people at war. British and American merchant ships were reinforced by boats flying the Panamanian, Norwegian, Greek, Polish, and Dutch flags. One of the unintended consequences of Hitler’s aggressions was that the island state’s limited fighting resources were boosted by considerable numbers of foreign merchant ships and crews, foreign fighter and bomber pilots, and foreign infantrymen—and Britain was happy to take them all.

The story of convoys HX 229 and SC 122 confirmed that the Royal Navy faced one of the greatest logistical challenges in all of military history. There were thousands of Allied merchant vessels on the high seas at any given time—probably up to twenty convoys, plus hundreds of independently sailing boats. From Trinidad to New Jersey, and from Adelaide to the Cape, the lines stretched out, with most of this produce ultimately destined for the critical North Atlantic passageway. As the maritime war unfolded, the convoys would necessarily become larger and larger, which was no bad thing in itself. During the Casablanca discussions on the shipping crisis, P. M. S. Blackett, the Admiralty’s chief of operational research, had impressed listeners with an analysis showing that a convoy of sixty or even ninety ships was a more efficient way of getting goods across the Atlantic than a convoy of thirty; the number of escorts remained roughly the same, limited more by shipbuilding production and other duties (Operation Torch) than
anything else, and the U-boats only had a limited number of days and hours in which to attack, and a limited number of torpedoes, too. That mathematical analysis reinforced the planners’ conviction that the convoy system was the best one to pursue, but it still left a practical problem: how on earth did one get such a large and heterogeneous bunch of merchant ships from one side of the ocean to the other, especially when the Allied warships were themselves such a mixed bag of destroyers, frigates, corvettes, cutters, trawlers, and others?

One response, which went back to almost the beginning of the war, was to make a simple division between “fast convoys” and “slow convoys.” Much flowed from this, including the different nomenclature (“SC” was a slow convoy, “HX” a faster one, the latter usually coming out of the great harbor of Halifax, but also from New York itself). These convoys could leave from separate ports and be timed to arrive in the United Kingdom (or, on their return journeys, into East Coast harbors) on different days. Slower escorts such as sloops and armed trawlers were assigned more often to the slower convoys. The schedule of aerial patrols could be arranged accordingly. Allied warships dispatched to join one convoy in midocean might be instructed to help another one if it came under heavier attack. To be sure, and to the fury of every escort commander, all convoys, whether fast and slow, would have their stragglers—how could it not be, with forty, fifty, or sixty oddly assorted ships in a single group? Overall, dispatching large convoys and dividing the merchantmen into fast and slow concentrations made a lot of sense. Yet what would happen if the number of U-boats was simply too great?

Already by March 13, 1943, B-Dienst had evidence that the slow convoy SC 122 (fifty-one assorted vessels, with four or five close escorts) had set off from New York. It was to be followed a few days later by the fast convoy HX 229 (forty-one vessels, with four escorts), from the same port. The latter fact was not clear to the Germans at this time, but the intelligence nevertheless gave Doenitz plenty of time to order his western patrol group to ready for an attack, while also instructing additional U-boat groups to move westward, toward the key zone in the middle of the ocean. An examination of the route charts, losses, and reports of the convoys (there actually was a further fast convoy of twenty-five ships, HX 229A, well to the north, off Greenland, at this time), and especially of Rohwer’s reconstruction of the maneuvers of
the individual U-boats, leaves the reader with a sense not only of the complexity of this contest but also of its enormous size. Correlli Barnett has it right: “It could be said that for the first time an encounter in submarine warfare attained the scale and decisive character of the great fleet battles of the past.”
¹¹
In fact, one would probably have to go back to the grim multiday fights of the mid-seventeenth century between the Dutch and British navies to find a good historical equivalent.

The German attack upon these two convoys seemed to unfold like clockwork, although serendipity played a role, too. Because of engine trouble, U 653 was slowly heading westward to a relay point when it saw the approaching convoy HX 229 on the horizon, steaming toward British harbors. Its captain, the bemused though resourceful Lieutenant Commander Feiler, dived under for a long while as the entire convoy passed over it. When he resurfaced, the ocean was clear; the convoy had steamed on, and U 653 could send the critical message to Doenitz’s headquarters, which then took immediate action. Twenty-one submarines responded to that news, a clear testimony to the way that electronic communication was changing the art of war.
¹²

The seas were terribly rough, but the wolf packs pressed in, sensing that this was a great opportunity. The clear, moonlit night of March 16–17 truly was a night of the hunter, for which many merchantmen were to pay the price. The attackers were also advantaged by the fact that on that same night the HX 229 escort commanders decided to slow the convoy in order to pick up stragglers and, by doing so, unwittingly bumped into the first cluster of U-boats. Had that decision to slow down not been taken, the submarines “would certainly have passed by to the stern.”
¹³
But once the convoy had been detected, there was little that its escort commander, Lieutenant Commander Luther in the destroyer HMS
Volunteer,
could have done to lead his flock to safety. His direction finder had located two U-boats 20 miles away and closing, so he had dispatched another escort to drive them away. But if one group of U-boats had missed their target, others surely would not. By this time, both the American and British admiralties were picking up U-boat signals from all around the convoy, so alterations in course to avoid one cluster of attackers simply brought HX 229 closer to another.