Hitler’s U-Boat War- The Hunted 1942-45

by Clay Blair

  As Slessor tells it, King was not very helpful. He offered to send Squadron VB 128 (upgraded from Hudsons to Venturas) and Squadron VP 63 (Catalinas with MAD gear)† temporarily to Iceland so that Slessor could shift British aircraft from Iceland to southern England to reinforce the Bay of Biscay offensive. Slessor was not satisfied with his offer since what he really wanted for Musketry-Seaslug-Percussion were very-long-range B-24s with centimetric-wavelength radar. Nonetheless King loaned these two Navy squadrons to Coastal Command, sending the Venturas to Iceland and the Catalinas, which were not wheeled (amphibious), to Wales for the Biscay offensive. The Catalinas arrived the last week of July; the Venturas in August.

  Nor was that all. King, who had been pressing the British to mount an intense Biscay air offensive for almost two years, eagerly approved of Musketry-Seaslug-Percussion. To support the offensive he did, in fact, agree to transfer temporarily to Coastal Command five squadrons of very-long-range B-24s fitted with centimetric-wavelength radar with PPI displays. These units, which comprised about seventy-five aircraft, were:

  • The Army Air Forces ASW Squadrons 4 and 19, based at Gander in Newfoundland. All twenty-four B-24s of these two squadrons arrived safely at St. Eval, thirteen planes on June 30 and eleven on July 7. These squadrons, not yet integrated into Tenth Fleet and commanded by Howard Moore, were designated 479th Group. The group began flying missions over the bay on July 13. Subsequently it moved to a new RAF field at Dunkeswell in adjacent Devonshire.

  • Two more Army Air Forces ASW Squadrons, 6 and 22. These twenty-four very-long-range B-24s arrived at Dunkeswell in August. However, Squadron 6, less its aircraft, was soon reassigned to another command.

  • Two U.S. Navy very-long-range B-24 squadrons, VB 103 and VB 105, comprising advance elements of Fleet Air Wing 7. According to Slessor, VB 103 relieved Army Air Forces Squadron 6 at Dunkeswell, leaving a total of five new American very-long-range ASW B-24 squadrons (about seventy-five aircraft) to reach England between July 7 and September 1, at which time the Army Air Forces Antisubmarine Command was deactivated.*

  By September 1, John Slessor at Coastal Command had about 150 American-built VLR B-24s in ten squadrons. Five were British units (53, 59, 86, 120, 224) and five were American (USAAF 4, 19, 22, and USN VB 103 and VB 105). Over the following two months, another U.S. Navy B-24 squadron (VB 110) arrived in southwest England to join Fleet Air Wing 7.† Thereupon the three former Army Air Forces squadrons (4, 19, 22) comprising the 479th Group, transferred to the England-based American Eighth Air Force. At the same time, Tenth Fleet shifted the (MAD-equipped) Catalina Squadron VP 63 from Wales to Morocco to mount experimental MAD ASW patrols over the approaches to the Strait of Gibraltar.

  Slessor wrote ungraciously that owing to Admiral King’s foot-dragging, most of these American ASW squadrons arrived too late to be of any real help in Operation Derange or Musketry-Seaslug-Percussion. “Actually, as it happened, it did not make much difference because we were able to defeat the U-boat by ourselves,” he wrote.‡ In his attack on the Americans—especially Admiral King—Slessor omitted to recall Churchill’s famous commitment to Washington—”Give us the tools and we’ll finish the job”—or that the two hundred aircraft in Coastal Command’s ten B-24 squadrons, two Catalina squadrons, the Ventura squadron, and the remaining three of the original British “jeep” carriers (Archer, Biter, Tracker) were indeed tools, from the Arsenal of Democracy.

  “ADAM” AND “EVE”

  The U.S. Navy’s crash program to build one hundred four-rotor bombes for breaking four-rotor naval Enigma proceeded at a frenetic pace at the National Cash Register Company in Dayton, Ohio. The goal was daunting: to produce, virtually on an assembly-line basis, the most complicated and fastest electromechanical devices ever created. Not surprisingly, there was constant friction as to method, design, and manufacturing technique among the four principals running this program: Joe Wenger and Howard Engstrom in Washington and Ralph Meader and Joe Desch in Dayton. Furthermore, Engstrom wrote,* OP20G was engaged in a “constant battle” for priority materials, “always hampered by the fact that the purpose of the machines had to be kept secret from the various agencies controlling priorities.”

  By the end of 1997, the NSA still had not publicly released sufficient documents to recount this epic struggle in detail. However, the American historian Colin B. Burke, who was a scholar in residence at the NSA in 1994-95, was allowed private access to the unreleased material and published his version of the struggle.† Burke’s version gives great credit to the genius of electrical engineer Joe Desch of NCR, less so to the genius of mathematician Howard Engstrom, chief of OP20G R&D. This thesis cannot be tested for accuracy and balance without the availability of the actual documents; however, as of this writing, it must do.

  Burke asserts that at the beginning of 1943 the American bombe was “tragically behind schedule” and that the Americans had yet to make a “real contribution” to the Enigma problem. There was no bombe prototype and there was growing concern about the “practicability” of some of the components. “Tragically behind” seems to be an overly harsh assessment; after all, the Naval Computing Machine Laboratory (NCML) at NCR, established on November 11, 1942, was less than two months old. Moreover, the four men running the program had yet to absolutely agree on the final bombe design.

  What made the American bombe situation somewhat embarrassing was, as previously related, that in December 1942 the British codebreakers at Bletchley Park had finally broken into Triton utilizing their three-rotor bombes. As also related, they were able to do this by exploiting the Enigma materials seized from the sinking £7-559 in the Mediterranean. However, the British three-rotor bombes were maddeningly slow and not sufficiently sophisticated to deal with the anticipated German upgrades in the complexity of naval Enigma.

  By March 1943, Burke explains, NCR had produced two “wheezy” prototypes, “Adam” and “Eve.” However, he states, these machines were full of bugs. “The group at NCR could not even tempt the two Bombe prototypes to run for more than a few inadequate minutes.” Even with improvements, Burke says, Adam and Eve “refused to run for more than a few hours without spurting oil or developing incurable cases of faulty electrical contacts.” The latter problem was caused when the fundamental elements of the bombe, the very-high-speed commutators (large rotors) overheated and lost their shapes. As June approached, says Burke, the American bombe “seemed to be too much of a challenge, even for the talented Joe Desch.”

  In his history of the bombe project, Howard Engstrom wrote contrarily that

  The first two experimental models of the American 4-wheel Bombes [Adam and Eve] were put into operation at the U.S. Naval Computing Machine Laboratory in May 1943, and were immediately successful. Secure means of communications were established between Washington and Dayton and the machines were operated there in order to be under observation by the maintenance and design engineers. [Meanwhile] production models started to be put into operation in early June.

  The big high-speed commutators did, indeed, cause problems, Engstrom continued. He and Joe Desch fell into dispute over the method to eliminate this serious defect. Desch urged that greatly improved sets of small wheels like those in Adam and Eve were the solution. However, Engstrom calculated that the use of small-wheel sets would double the number of wheels to be employed to forty thousand and “would present extreme difficulties in maintenance and in rewiring” in case the Germans made changes. “To make the machines workable it was imperative that we use a single size wheel for all four positions in the machine,” Engstrom wrote. Although the Engstrom solution had yet to be tested, he was confident that it could be made to work and rather than “holding up the entire project,” he ordered that production of the bombes proceed forthwith “against the advice of the engineers.”

  What emerged from these debates was an electromechanical marvel, designated “Bombe, Model 530.” These bombes were big: eight feet long, seven feet high, and two feet wide. Each weighe
d five thousand pounds. Each contained thousands of high-tolerance parts, including 1,500 vacuum tubes arranged in what was “very high advanced circuitry,” as Burke put it. Each generated tremendous heat, so much so that powerful air conditioners were required to maintain a reasonably comfortable work environment. Each machine cost $45,000.

  Not without justification, Burke rhapsodized that each bombe

  housed sixteen four-wheel sets of Enigma analogs [i.e., clones] and the Welch-man diagonal board [i.e., the powerful plug-board system conceived for the original Turing British bombe]. It had all the circuits needed to link them together and test the hundreds of thousands of possible combinations of wheel positions against a text crib for inconsistencies. Its sixty-four double Enigma wheel commutators each contained 104 contact points which had to be perfectly aligned when they touched the copper and silver sensing brushes. Such alignment and synchronization were difficult to achieve, especially for the fast wheel which rotated close to two thousand RPM.

  That is only a crude glimpse into the interior of Model 530. Notably, it also contained some “truly distinctive” digital electronics/designed by Joe Desch and his engineering aides. Burke again:

  The sixty-four double commutators on the Bombes could be driven at very high speeds by mechanical means, but motors and shafts and electromechanical counters were unable to track their position …. [Desch] created a digital electronic tracking and control system that amazed the Navy’s engineers. His 1,500-tube system did more than record the position of a hit: It exercised control. It was able to track the wheel positions, signal the motor and the clutches, then reverse the machine’s action until it had returned all the wheels to their hit position. At that point the wheel locations were [automatically] printed and the operator could signal the machine to restart its search for consistencies.

  During May and June 1943, Navy personnel who were to operate and maintain the bombes reported to Dayton. These included a large contingent of Waves who were to program and run the machines twenty-four hours a day on three shifts. The new men and women were astonished at the power and speed of the bombe. Given a “crib” (known or assumed plain text) a bombe could carry out 20,280 “trys” per second. It required merely fifty seconds to carry out a complete short three-rotor “run” and twenty minutes to do a “long” four-rotor ran.

  Inasmuch as the buildings to house the bombes on Nebraska Avenue in Washington* were not yet ready and the bombes were undergoing nearly continuous mechanical revision and improvement, the codebreakers first used them operationally at NCR, Dayton. Mathematician Joseph Eachus, the Engstrom protégé who had served a highly productive tour at Bletchley Park, reported there to help establish a temporary message-analysis center. Washington sent Dayton the Enigma intercepts by secure teletype; Dayton replied with decrypts that were delivered immediately to “Frog” Low and Kenneth Knowles of the Tenth Fleet staff.

  The British, meanwhile, had completed four-rotor bombes of two different designs. They were slower than the American bombe, requiring thirty minutes for a full run compared to twenty minutes. They were also much less reliable.† The British built sixty-nine four-rotor bombes.

  The early erratic yields of the NCR four-rotor bombes at Dayton in May 1943 probably helped confirm the long-held suspicion that the Germans were reading the Allied convoy code. As related, in June 1943, the Allies made emergency alterations to that code that left German codebreakers blind and deaf and later in the summer introduced a new convoy code that the Germans were unable to break.

  ACTION IN THE NORTH ATLANTIC IN MAY: I

  The Allied convoy traffic on the North Atlantic run in May 1943 was about average: fifteen sailings, eight eastbound and seven westbound. There were about seven hundred ships in these fifteen convoys. Three “jeep” carrier support groups (British Archer and Biter; American Bogue) operated in the North Atlantic during the month. U-boats sank only six of the approximately seven hundred ships in the fifteen convoys.‡

  As related, in the opening days of May, two German groups, Amsel and Fink, consisting of about forty boats, were attacking convoy Outbound North (Slow) 5, in the waters between Greenland and Newfoundland. Owing mainly to heavy fog, which favored the radar-equipped air and surface escorts, on May 7 U-boat Control canceled the operation and dissolved Amsel and Fink. As a result, some boats headed for France, some to a U-tanker.

  U-boat Control organized the remaining twenty-nine boats, including some that were damaged, into two new groups named for German rivers. These were Elbe (seventeen boats) and Rhein (twelve boats). Based on B-dienst information, U-boat Control shifted the groups southeastward to intercept two eastbound convoys, Halifax 237 and Slow Convoy 129. The damaged boats, or those low on fuel, were to serve only as convoy spotters. In order to bring greater force to bear, Control directed the five remaining boats of the ill-fated group Drossel, which, as related, had patrolled west of Cape Finisterre, where two pairs of its boats had collided, to go west at highest speed.

  B-dienst provided quite specific and current information on the routes and positions of these two convoys. Acting on this information, a Rhein VII, U-359, commanded by Heinz Förster, age thirty-three, making his second patrol, found Halifax 237 on May 9. It was composed of forty-six ships, guarded by Canadian Escort Group C-2, commanded by E. H. Chavasse, who at first had but five warships: the ex-American four-stack British destroyer Broadway, the new British frigate Lagan, and three corvettes, two Canadian and one British. In due course, another Canadian corvette joined. When Western Approaches deduced from Förster’s contact report that Halifax 237 was in danger, it ordered British Support Group 5, the “jeep” carrier Biter and her screen, to reinforce the convoy.

  At first U-boat Control directed all three groups, Rhein, Elbe, and Drossel, to attack Halifax 237, but almost immediately it modified these orders. The twelve boats of group Rhein and seven of group Drossel (the latter reinforced by the recently sailed VIIs U-221 and U-753) were to attack Halifax 237. Group Elbe, reduced to sixteen boats by the departure of U-266, but reinforced by eleven other boats to twenty-seven (seven of them IXs), was to hunt for and attack Slow Convoy 129. Commanded by Ralf von Jessen, age twenty-six, making his second patrol, U-266 refueled from the tanker U-461 on May 12. While homebound two days later, a Halifax of British Squadron 58, piloted by Wilfred Oulton, sank her with depth charges. There were no survivors.*

  Escorts of Halifax 237 drove Heinz Förster in U-359 off and down and he lost touch. But on the next day, May 10, another Rhein boat, the Arctic veteran U-454, commanded by Burkhard Hackländer, age twenty-eight, relocated the convoy, which was covered by Swordfish biplanes from Biter. One of the Swordfish forced Hackländer under, but his contact report brought up several other boats.

  Despite the air cover from Biter, four boats attacked Halifax 237. Two other Arctic veterans, Max-Martin Teichert in U-456 (of Drossel), age twenty-eight, and Heinz-Ehlert Clausen, age thirty-three, in U-403 (of Rhein), shared credit for sinking the 7,200-ton British Liberty ship Fort Concord. To get into shooting position, Clausen in U-403 fought it out with a Swordfish. The new Ritterkreuz holder Hans Trojer, age twenty-seven, in the newly sailed VII U-221 (of Drossel) sank the 9,400-ton Norwegian tanker Sandanger, which was straggling. Rudolf Baltz, merely twenty-two years old, the new skipper of the newly sailed veteran VII U-603 (also of Drossel), sank another straggler, the 4,800-ton Norwegian Brand. In retaliation, the escorts of Halifax 237 sank three veteran U-boats, all from the doomed group Drossel.

  • The VII U-89, commanded by Dietrich Lohmann, age thirty-three. She had been in the Battle of the Atlantic for a full year. She was sunk by British warships: the four-stack destroyer Broadway and the frigate Lagan, drawn to the site by a Swordfish from Biter.

  • The VII U-456, commanded by Max-Martin Teichert, which had been at the Arctic or Atlantic fronts for a year and a half. She was damaged by a B-24 of British Squadron 86, piloted by John Wright, who dropped a Fido homing torpedo. Teichert surfaced, radioed for assista
nce from any boats nearby, and held the B-24 off with his flak guns and survived the night. On the following morning, May 13, the British destroyer Opportune, commanded by J. Lee-Barber, bore down on U-456 and Teichert dived, never to be heard from again. The VII U-448 (of Elbe), commanded by Helmut Dauter, age twenty-three, came upon the bodies of two U-456 crewmen and gave them a proper burial at sea. Teichert was awarded a Ritterkreuz posthumously.*

  • The newly sailed U-753, commanded by Alfred Manhardt von Mannstein, age thirty-five. She had been in the Atlantic for a year and a half. A Sunderland of Canadian Squadron 423, piloted by John Musgrave, attacked U-753 with depth charges in the face of heavy 20mm flak. Two frigates, the Canadian Drumheller and the British Lagan, finished off U-753 with depth charges. Nothing further was ever heard from this boat.†

  Helmut Dauter in the Elbe VII U-448 joined a dozen other boats at the tanker U-459 on May 16. He refueled the next day, but he was unable to repair a disabled radio and was forced to abort. He reached France on May 26, completing a patrol of forty days. Like the overwhelming majority of boats that sailed in April 1943, this cruise was unproductive.‡

  Several Drossel boats continued to hold contact with Halifax 237. These included Paul Siegmann in U-230. In his book Iron Coffins, her first watch officer, Herbert Werner, claims that he, Werner, shot down a Swordfish. The biplane crashed very close to U-230, Werner wrote, and its armed depth charges exploded, atomizing the wrecked plane and the aircrew. The depth charges “kicked into our starboard side astern,” Werner continued, “but we left the horrible scene unharmed.”

  The loss of this vulnerable Swordfish and the near-loss of others to U-boat flak in the battle with Halifax 237 led the British to put restrictions on Swordfish tactics. Whenever possible they were to patrol in pairs. A single Swordfish was never to attack a U-boat, but rather shadow it until air and/or surface reinforcements arrived.