A battle conducted on the basis of replacing losses faster than they can be inflicted is, however, a sterile and dispiriting business. The Admiralty rightly strove, from the outset of the U-boat war, to hit back and to give better than it got. While its overriding policy was to assure “the safe and timely arrival of the convoy,” a policy pursued from the war’s first day to the last, and its first requirement of the escorts was that they should drive attacking U-boats out of contact, it also sought their destruction. At first, given the paucity of escorts, convoy protection was all that could be attempted. Moreover, escort crews and captains were inexperienced, weapons inadequate and means of detection primitive. Britain entered the war with about 180 destroyers, 60 surviving from the Great War, the rest dating from 1927 onwards. Many, however, were required to work with the fleet, while none was strictly adapted to anti-submarine duty. They had torpedo tubes, which were redundant, too many guns and, paradoxically, were too fast; their large engines consumed too much fuel on long Atlantic crossings. What was needed, as Churchill insisted from the outset, were smaller ships, of greater endurance, mounting fewer surface weapons but also able to carry more depth charges. The first expedient were the Hunt-class destroyers, of which eighty-six were eventually built, but they proved to need too frequent refuelling and were unstable in North Atlantic storms. Adapted American destroyers, of which fifty were swapped for the use of Caribbean bases, were more successful, though not as much liked by the Royal Navy as the ten loaned U.S. Coast Guard cutters, roomy vessels of excellent seakeeping qualities. A hasty expedient was the corvette, developed from Antarctic whalers, good seaboats, though violent rollers, but too slow, small and weakly armed to provide the backbone of the escort fleet. That was eventually supplied by the new classes of sloops and frigates, in effect small destroyers, possessing enough speed to outrun U-boats on the surface but economic in fuel consumption.27
While the Royal Navy was building up an adequate escort fleet, and the Royal Canadian Navy was embarking on a twenty-fold expansion to become the third largest in the world, the British were also developing and refining their anti-submarine methods. At the heart of the system lay the Operational Intelligence Centre (OIC) in the London Admiralty, divided into four sections, of which the U-boat Tracking Room was dominant; it was commanded from 1941 onwards by Rodger Winn, a recovered polio victim, who quickly established an intimate relationship with the naval team at Bletchley and proved an inspired interpreter of Enigma decrypts. The OIC was linked by secure teletype and telephone to Western Approaches Command, initially located at Plymouth, later at Liverpool, the principal receiving port for North Atlantic convoys. Western Approaches, commanded by the dramatically extrovert Admiral Sir Max Horton, was the headquarters that fought the Battle of the Atlantic from the European side. Its control room was dominated by an enormous wall chart which showed a constantly updated display of the positions of convoys, and U-boat patrol lines and wolfpacks. “All Americans who visited British military agencies in 1941 were impressed by the degree of unification which had been achieved in the Battle of the Atlantic. From the War Cabinet to the Admiralty and Air Ministry, to Bletchley Park and the OIC and Derby House [Western Approaches], all hands worked with an extraordinary singleness of purpose.”28 There was no such efficiency on the German side. As in so many other fields of warmaking where direct comparisons can be made, British democracy proved more efficient than German dictatorship in fighting the Battle of the Atlantic. Direction of the U-boat war was characterised by suspicion and rivalry, particularly in seeking the ear of the Führer. While the British, in conditions of the strictest secrecy, included all who needed to know in the struggle, Dönitz confined his side’s direction to a tiny group, located at Kerneval, in Brittany, until a commando raid alarmed him into ordering its transfer to Berlin. The result of this self-imposed exclusion from the wider German war effort was that the navy remained frozen within the technology and strategy with which it began the war—using U-boats scarcely superior to those of the Great War to conduct group attacks on the surface against oceanic convoys—while its opponents were mobilising every sort of anti-U-boat measure to defeat its Atlantic offensive.
Dönitz cannot be accused of neglecting the training of his crews. Even as the prospect of Atlantic victory slipped away from him, newly built U-boats and their fresh crews were spending as much as a year in the sheltered waters of the Baltic “working up” for battle on the convoy routes. U-boat tactics, however, remained static. Dönitz clung to the “single idea” he had conceived as a young officer in the post–Versailles Treaty navy: of using submarines as surfaced torpedo boats to hover at the limit of visibility on the flanks of an intercepted convoy and then to attack in swarms as darkness fell. He made little allowance for the ability of his opponents to vary their defensive technique. That was a grievous mistake, for the British, assisted as the Battle of the Atlantic drew out by the Canadians and Americans, varied their technique in many ways, invoked more new technologies than the Germans developed in response, and ultimately triumphed by an ingenuity of which Dönitz proved incapable.
At the outset, the British convoy escorts, weak in number, often no more than four warships for a lumbering assembly of as many as sixty merchantmen, responded to attack by running down individual U-boat contacts. They were sometimes able to drive the enemy down and sustain the attack by Asdic—the echo-sounding device developed by the Allied Submarine Detection Investigation Committee in 1918, led by the Canadian scientist R. W. Boyle and later known as sonar—but were usually forced to break contact when the convoy drew away and required their return to the protective screen. Escort successes in the early days were dispiritingly few. In 1940 they destroyed only twelve U-boats, for a loss of 133 merchantmen inbound in convoy to the British Isles.
As the convoy battles intensified, the British developed a new technique. In 1941, the newly founded escort groups—prefixed B for British, C for Canadian, later also A for American, each with a distinguishing number and consisting of 6 to 8 warships kept together under a designated commander—began to react aggressively. The British and Canadians, as they were later to instruct the Americans, knew by 1941 that the U-boats were trained to attack on the surface at night. They also knew that the worst losses, whether or not there had been an Ultra intelligence warning, usually occurred on the first night of attack. When a torpedo struck, therefore, the escorts forming the convoy screen, knowing that the attacking U-boats would seek to roam surfaced within the convoy’s columns, turned inwards and fired star shell (“snowflake”) to illuminate the scene. If U-boats were identified, they were engaged with gunfire or set up for ramming; if they submerged, the escorts proceeded to establish sonar contact and to drop and fire depth-charges. The purpose, besides that of hoping to hit the U-boat, was to keep it down, at slow or negative speed, thus allowing the convoy to sail out of contact, usually by making an emergency turn.
When many U-boats were in contact, however, turns could not solve the problem. The convoy simply ran into others and the battle had to be fought out. Daylight forced all U-boats to submerge, thus reducing their speed below that of the target and allowing a well-handled convoy to disappear if luck was with it. The first night of a convoy battle was notoriously the worst. As the number of escorts increased, moreover, and it became possible to form support groups, held in readiness to reinforce a hard-pressed convoy, subsequent nights also became harder for the U-boats.
As the number of escorts grew, and the skills of their crews increased, the odds began to turn their way. By the end of 1943, some escort groups had become very skilful indeed, none more so than the 2nd Escort Group, commanded by Captain F. J. Walker. He had developed a “creeping” technique, designed to overcome the loss of sonar contact which always occurred in the last 100 to 200 yards of an attack. As the distinctive “pinging” of the sonar beam on his submerged hull stopped when the attacking vessel passed ahead, a cool U-boat captain would take the opportunity for violent evasive action, often successfully. Walker perceived that by using two ships, one to keep up the sonar contact from a distance, the other to creep up silently on the target until it was overhead, the U-boat could be caught by a surprise pattern of depth-charges. The moment was signalled either by light, flag or the new TBS (talk-between-ships) radio, an American invention of which Walker made copious use.
Between 31 January and 19 February 1944, Walker’s five sloops, Woodpecker, Wild Goose, Magpie, Kite and Starling (his own ship), sank six U-boats by the “creeping” method in the Western Approaches, to which Dönitz had again begun to send his boats after their withdrawal from the central North Atlantic six months earlier. The crew of the last victim, U-264, got out and were saved; the other five disappeared with all hands.
Walker’s group got the chance it did because of an abundance of inward-bound convoys, which drew the U-boats into contact. Success in anti-submarine warfare almost always came in the vicinity of a convoy. Simply searching for U-boats with “hunter-killer” groups, the method favoured by Churchill in 1939–40 and by Admiral King in 1942, resulted in pointless quartering of an empty ocean. No amount of intelligence could guide anti-submarine vessels on to targets if they remained submerged, or dived when an attacker was sighted, as U-boats always did when searching for prey. The only exception to that pattern came in the middle of the war, when Dönitz began to send U-boat tankers to refuel attack boats in mid-ocean. The American escort carriers were particularly successful in destroying tankers and their clients found in groups. On 4 October 1943, aircraft from the USS Card found U-460 refuelling three Type VII U-boats, sank U-460 immediately and later sank U-422. Again, however, Card was in company with a convoy, bound for Gibraltar.
A potent source of technical intelligence was high-frequency direction-finding (Huff-Duff), a pre-war development but perfected and widely distributed only in 1943, which permitted a bearing to be taken even on the briefest U-boat transmissions and displayed visually on a screen. Since it was a passive device, giving no warning to the target of the interception, it increasingly allowed U-boats to be engaged without warning, particularly in conditions of poor visibility. It was extremely accurate over short distances. Huff-Duff, however, had teething troubles and was slow to win the confidence of escort commanders. They initially preferred high-frequency centimetric radar, which appeared at almost the same time. The various models of the 271 radar set gave clear definition up to 8,000 yards and had the additional advantage of defeating German detection devices. Because they themselves had not yet developed a centimetric system, the Germans discounted the likelihood that the British had done so, with lamentable results for U-boat survival.29
In the final battles around the convoys sailing the central North Atlantic route in 1943, U-boat losses reached insupportable levels. Despite forming wolfpacks as many as fifty strong, the U-boats could not break through. In the attacks on ON206 and ON520 in mid-October, only one merchantman, and that a straggler outside the escort screen, was sunk, for the loss of six U-boats in two days.