Chapter 3. After the Fall of France: June 1940–March 1941

The fall of France altered the whole geography of the battle. U-boats operating from Lorient and other French ports were many miles – days – nearer their operational station, so that their operational time was much increased. On the day after the French armistice, a train left for France carrying fuel, weapons, provisions, spares and maintenance personnel. On 3 July U-30 entered Lorient to embark fuel and new torpedoes. It is not easy to estimate the benefit of time on station; there was a concealed advantage, as much of the old route from German North Sea bases had to be traversed submerged for fear of attack. Hessler has estimated that the overall effect was about a week (25 per cent) extra on station per sortie. Kessler notes that before the fall of France, U-boat Command needed 2.35 boats in maintenance for each one on operations.¹ When the French ports became available, this ratio changed to 1.84 to 1, a 22 per cent improvement.

The use of the French bases eased the maintenance load and it was even said that French workers were better than the Germans at home. In the north, Norwegian ports brought the north-western approaches within range of the small Type II boats. There was a very small offset with the British occupation of Iceland, though it was early 1941 before the RAF had aircraft available to base there.

The advance of the German armies posed other problems for the Royal Navy. Many destroyers and escorts had been sunk or damaged during the Dunkirk operation and other evacuations, and repairs take time. Additionally, a considerable number of the best destroyers were held in south- and east-coast ports as part of the protection against a possible invasion. In July the Nore command alone had thirty-two modern destroyers and five corvettes, while more were held on the south coast. Only in September did the Defence Committee under the chairmanship of the prime minister agree that the threat of defeat in the Atlantic was greater than that of invasion and release the destroyers to the Western Approaches. The English Channel was closed to ocean shipping and all convoys were routed to the north of Ireland. The shortage of escorts was so severe that most convoys in June and July had only a single escort for much of the way.

In June 1940 U-boat Command had only fifty-one U-boats available, six fewer than at the outbreak of war, and many were undergoing refit or repair following the Norwegian campaign. Considerable progress had been made in rectifying the torpedo problems. By 9 June sixteen boats were assembled in the Western Approaches, and Dönitz planned wolf-pack attacks on convoys HX48 and WS3. Fortunately, his intelligence was incorrect and contact was not made. The German B-Dienst had broken the RN administrative code before the war, which gave general information on the sailing of convoys. Only in August did the Admiralty change to a more difficult code. Dönitz also hoped to rely on aircraft sightings. The first choice was the He 177 but the early examples were so prone to set themselves on fire that they were not used over the sea. The replacement was the FW 200 Condor, an adapted passenger airliner. It too, was unreliable, as the tail would fall off in violent manoeuvres, but it was still a useful, even formidable, opponent.

Phase 2 of the Atlantic battle. With the fall of France homeward bound traffic was routed through the north-west approaches, to the north of Ireland. Escorts were very weak both in ships and aircraft and U-boat attacks were concentrated in these approaches. A small number of attacks occurred off the West African coast.

Further attempts at wolf-pack attacks in August were largely ruined by bad weather but Dönitz interpreted the results as showing that he was working on the right lines. Then on 21/22 September five boats concentrated on convoy HX72, which had been sighted by U-47 following a signal interception by B-Dienst. In a series of attacks, eleven merchant ships of 73,000 tons were sunk. In mid-October two further attacks on convoys SC7 and HX79 confirmed Dönitz’s tactics. Some names that were all too well known already came to the fore: Kretschmer, Prien, Endrass and Schepke.

Monthly sinkings were well over 200,000 tons per month, up to 360,000 in June and October 1940. Viewed another way, the figures looked even worse; there were only about ten to twelve U-boats at sea at any one time and they were averaging 500–1,000 tons sunk per day each. The number of U-boats training or new boats working up had risen from twenty-four in June 1940 to eighty-two in March 1941. Clearly, worse was to come. There was very little to look forward to; as the USA was still neutral, there was no vast stream of new merchant ships and escorts coming from them as in 1918. There was no active assistance in anti-submarine work from the USN. Canada’s contribution to both merchant-ship and escort building was already becoming significant. It was perhaps the darkest hour of the battle, without a light at the end of the tunnel.

Available escorts were much as in the previous phase, though Flower class corvettes began to operate. Gladiolus was the first to complete and, appropriately, she was the first to score, sinking U-26 (with RAAF assistance) on 3 July 1940. Rhododendron scored on 21 November and Arbutus with Camellia accounted for U-47 on 7 March 1941. Altogether, fifteen U-boats were sunk from June 1940 to March 1941. Most of the surface-ship kills were with depth charge, though, unusually, Walker torpedoed U-99 and Vanoc rammed U-100. Even when there were several escorts to a convoy, they had rarely operated together and co-ordinated counterattacks were almost impossible with the primitive signalling devices available. The fifty ex-USN flush-deck destroyers began to operate towards the end of the period.

Almost all of the sinkings by surface ships were due to convoy escorts.

Technical and Other Developments

There were a number of developments in technology, training and other areas during this period for which Dunbar-Nasmith deserves credit. Few were of operational significance at the time but they contributed a great deal to the ‘gleam of light’ described in the next chapter.

Early Radar

Although the Admiralty was involved in the development of radar (known as RDF until June 1943) almost from its conception in 1935, resources were very limited and progress was slow.² What effort was made available was concentrated on air warning sets, and by 1938 Rodney and Sheffield had early development units of Type 79. This was a very successful set and remained in service well after the war. Fire-control radars came second.

Mansfield in 1942 is fairly typical of the ex-American Town class. She has one US 4in gun and a British 12pdr aft. Although less than ideal as ASW ships, the fifty old destroyers acquired under the Lend-lease agreement were a welcome reinforcement when they entered service at the beginning of 1941.

After the fall of France a requirement was formulated for a surface set to detect surfaced submarines or invasion craft. A test had been carried out some time earlier with a 1.5-metre ASV Mk I radar fitted to a Walrus aircraft standing on the slipway at Lee on Solent. It was capable of detecting ships at a distance of five miles. The decision to proceed was taken in early June and an aerial suitable for a small ship’s mast was made in little over a week. This came into naval service as the Type 286. The aerial was fixed pointing forward and the ship had to be turned to sweep a wider arc. The first sea trials took place with Verity on 19 June 1940. Only a few ships were fitted with the basic 286, which was superseded by the 286M, based around the better-engineered ASV Mk II. This still used the fixed aerial and also suffered from extensive side lobes, which were particularly confusing with the multiple echoes obtained from a convoy. It was not very effective at picking up submarines but made an enormous difference to station-keeping at night.

By March 1941, some ninety escorts had 286M. It covered an arc from the bow to just abaft the beam on either side. The simple aerial was quite heavy and several ships broke their masts when rolling heavily; Sardonyx broke hers in July 1941 and again two months later. The first success, and a very important one, came on 17 March 1941, when Vanoc detected and killed U-100, along with her commander, the ace, Schepke (see chapter 4). It was probably the only sinking directly attributable to a 286 with fixed aerial, though there were other contacts.

Canada developed a home-grown radar set, SW-1 (and the later SW-2), which was comparable with 286 but even less reliable. A more valuable Canadian contribution was in the supply of physics graduates who were trained as radar officers (sub-lieutenant) and lent to the RN. RDF ratings were not trained to maintain their sets in the early days. The Royal Australian Navy also contributed a number of officers and men.

Code Breaking³

By the mid-1930s, there was a general awareness in all navies of the part played by Signals Intelligence (SIGINT) in the submarine campaign of World War I and most took some steps to prepare for similar activities. There are three main types of SIGINT, although these may overlap. First comes full deciphering of enemy signals, preferably quickly enough for the intelligence to be used in current operations. Second is the location of a transmitter by direction finding. Finally, it may be possible to recognise a pattern in messages and their sequel, eg a message from a minesweeper followed by the passage of a high-value unit may be recognised as an ‘all clear’.

Simple precautions include minimising the use of radio and keeping messages very short. Codes were supposed to be changed frequently; a common error was to transmit the same text in both the new code and, for the benefit of ships which had not received the new code, a transmission in the old code. There was the eternal problem that operational use of intelligence will often disclose the source.

Before the war, the German B-Dienst had made considerable progress in breaking both the code and the super-enciphering of the British Naval Cipher No. 1 and the Naval Code. In the opening months the RN use of radio was limited and the Germans had little advantage from their achievement. During the confusion of the Norwegian campaign, however, radio was used to a greater extent and Germany gained from the ability to read messages.

After the fall of France, U-boat Command were able to read convoy-routing signals in slow time but the format of the signal indicated the pattern of movement. This would change in the next phase.

Rescue

Traditionally, sailors have always given high priority to rescue and, initially, it was hoped that this approach would be sufficient in the convoy war. The last ship in each column was supposed to stop and rescue survivors if any ship ahead was sunk. There were a number of problems; with eastbound convoys, all ships carried valuable cargoes, which would be put at risk, together with the ship itself and its crew, if it stopped in the vicinity of a submarine. There was a different problem with ships in westbound convoys, as most ships were in ballast, lightly loaded, with high freeboard, making them hard to steer, so that rescue was physically difficult as well as hazardous. Many rescues were carried out by escorts, which not only exposed them to risk but reduced the already inadequate number of vessels on the screen.

The quarter view of Mansfield illustrates some of the problems of the class. The very fine stern lines needed very large guards to protect the big propellers, and the quarterdeck was too small for an extensive depth charge fit.

The weakness of this approach was apparent as losses mounted after the fall of France. The loss of crews was as serious as that of ships. The training of both officers and men was lengthy and hence the pool of seamen was limited. It was also realised that morale would suffer unless it was known that survivors could be confident of a speedy rescue. At the end of September 1940, C-in-C Western Approaches drew the attention of the Admiralty to these problems and suggested special rescue ships. Action was rapid and the first such ship, Beachy, sailed on 9 October.

The ships chosen were mainly passenger/cargo vessels operating in the UK coastal trade. They were usually of about 1,500grt and 250 feet in length.⁴ They had merchant ship crews and flew the RFA blue ensign with the Admiralty anchor. It was thought that their small size would reduce the risk of attack, while their low freeboard would facilitate the actual task of rescue. The degree of alteration increased as time went by. Most were ballasted to their normal deep draught and many had empty drums stowed in their holds to maintain buoyancy and stability in the event of flooding. Fuel stowage was increased, often involving extra bulkheads, to enable a passage across the Atlantic without refuelling. There were also five Castle class corvettes, converted while building, and bearing ‘Empire’ names. They only entered service in late 1944 and played little part in operations.

Eventually, most could accommodate thirty officer survivors in two-berth cabins and 150 seamen in bunk spaces. Extra washplaces and toilets were installed and catering facilities were increased. A specially designed rescue boat was fitted on either beam. A properly fitted sick bay and operating theatre with an RN surgeon and sick berth attendant came a little later in the war, making them invaluable to treat minor injuries and illnesses within the 2,500–10,000 men of the convoy. They had a fairly full communications fit, with a team of RN operators, and carried HF/DF gear. Their station at the rear of the convoy made them invaluable in the sheepdog role. About thirty rescue ships entered service, of which five were lost to enemy action and one, probably, as a result of icing.

Personal survival gear was either lacking or deficient. An inflatable life jacket had been tested in 1939 and found to be dangerous, as it tipped the head of an unconscious man under water, but it remained in production throughout the war. There were no survival suits, so life in cold water was brief. The main vehicle for survivors was the Carley float, with a buoyant ring supporting a grating on which survivors sat, half immersed in the sea. Even this crude device doubled the chance of survival in the sea. It was easy to launch and in most installations would float off a sinking ship. No real thought had been given to the means of getting survivors from the sea or a raft into a rescue ship. After the war, the Talbot Committee concluded that the majority of those lost at sea in the RN had escaped from their sinking ship and died in the water. They found that no department of the Admiralty was responsible for survival aids. The merchant service was probably even worse off, relying on traditional kapok or cork jackets and totally exposed rafts.

Training

Training was a major factor in the Battle of the Atlantic. The best of tools is of little value if the operator is lacking in skill. The majority of officers and ratings were ‘Hostilities Only’, with little if any seagoing experience; indeed some, particularly Canadian, had never seen the sea. Training can be divided into three levels, though there was some overlap: basic equipment, ship and unit.⁵ It is notable that the most successful commanders insisted in spending considerable time when in harbour in practising the basic tasks, such as reloading depth charge throwers. The benefit of plenty of such practice loadings made it a little less difficult to load at night with the ship rolling and heaving while covered in spray if not green seas.

In early 1940 the Admiralty decided that a training base should be set up to train whole ships’ crews in the latest techniques and provide refresher courses. The site chosen was Lorient as a joint RN–FN base, but little had been done by the time of the fall of France – except for the appointment of Vice Adm (retired) Gilbert Stephenson in the rank of commodore.

A new site was chosen, at Tobermory, and it was operational by July 1940. The first headquarters/accommodation ship was a decrepit ex-yacht dating from 1891, renamed Western Isles, requisitioned in April 1940 and converted by Barclay Curle on the Clyde. She was renamed Eastern Isles in October 1941, the role of HQ ship being taken by a more up-to-date Western Isles (ex-Batavier IV) dating from 1902. The dates are significant, as they show the early recognition of the importance of training by the Admiralty. Western Isles only went to sea once – to qualify for duty-free drinks.

The syllabus varied from class to class, depending on the complexity of their equipment, but the course was usually two to three weeks. The role had changed significantly and concentrated on the work-up of newly commissioned ships rather than the original concept of a refresher course. The emphasis was on urgency and the unexpected; the commodore would often board a new arrival before mooring was complete and proclaim that the ship was on fire or that there was a man overboard. For many, this was their first ship and they were not long in uniform. Cdre Stephenson found that a short but tough period of close-order drill was useful.

Every weapon and sensor was exercised and there was a real, albeit ancient, submarine with which to exercise. Hours were long, though sleeping hours were usually undisturbed – except for teasing the sentries. The final reports exposed every weakness, with an ultimate sanction of repeating the course, rarely employed. Officers who did not meet the standards were removed. Twenty to twenty-five ships could work up at any one time. It was no wonder that Stephenson became known as the ‘Terror of Tobermory’. He worked up his thousandth ship in October 1944, drawing the inevitable comparison with Helen of Troy. Perhaps the greatest tribute to Cdre Stephenson is that the current work-up training for the RN is closely modelled on his plan at Tobermory. This RN course attracts ships from many other navies. Tobermory was overloaded and a similar establishment, HMS Mentor, was opened at Stornoway in December 1943, but closed when the rate of completion of new escorts dropped off. Another base was set up in Bermuda for the use of frigates built in the USA. Capacity was limited to twelve because of the need to provide sheltered berths during hurricanes.

The Towns

The Town class were sometimes referred to as ‘flush deckers’, which was a correct description of their profile, or alternatively ‘four stackers’, which was incorrect since two of the fifty ships transferred to the RN were built with three funnels and others were altered with two or three. As built at the end of World War I, they fell into four main groups, though there were variations between ships of a group.

Negotiations to obtain destroyers from the USN began with the fall of France, and an agreement was signed on 3 September 1940 to transfer fifty ships in exchange for the lease of bases in the West Indies. A considerable amount of preparatory work must have been completed, as the first eight ships arrived in Halifax, Nova Scotia on 7 September. It was originally intended to retain the US names, and this batch were commissioned as ‘HMS’, followed by the US name, on 9 September. The decision to give them new names based on places in the USA and UK that shared the same name was made on the 19th, but not implemented until 2 October. In the meantime a second batch of seven ships had commissioned with their original names.⁷

Leeds, one of a pair of three-funnelled Group D Towns. They did not take part in the Atlantic battle but were employed almost exclusively on east coast duties.

Churchill is similar in configuration to Mansfield.

On arrival in Halifax, the British or Canadian crew joined with the US crews for two days, learning how everything worked. An inspection led to a critical but helpful report. Hull and main machinery was generally in good condition, but auxiliaries, piping and wiring were in poor shape.

Two major problems soon became apparent. Corroded rivets in the shell plating allowed sea water to enter the fuel tanks causing, in the worst cases, complete loss of power. A temporary cure was found by carefully cleaning the tank and refilling with clean oil but the problem would recur for the rest of their short service lives. Another long-term problem was due to the operation of the rudder by long runs of wires and chains from the steering wheel. These would break or jam at critical moments, losing steering.⁸ In RN service it was found necessary to change all this gear every three months. Their turning circle of 770 yards at fifteen knots was nearly twice that of an RN destroyer.

Rockingham was refitted at Devonport from December 1940 to February 1941 and served in the Western Approaches until a refit at Southampton from May to August 1941. She became an air target ship in December 1943 and sank on a British mine in September 1944.

Brighton completed her refit in January 1941 and served in the Atlantic. She was badly damaged in collision with the cruiser Kenya in June 1941 and following another collision was laid up. She was lent to Russia in 1944 and scrapped on return.

The hull form had been successfully optimised for minimum resistance in still water, with very fine ends. At the after end, this led to a tiny quarterdeck, which made difficult the installation and operation of depth charge gear, while the big propellers projected far beyond the ship’s side and were easily damaged, particularly during frequent steering failures. At the other end, the freeboard was low because the sheer was insufficient to make up for the lack of a forecastle and hence the shipping of green seas was all too common. The bridge was well forward and easily damaged, while several ships also lost funnels. Forward of the gun, there was a large access hatch to the mess deck which was also liable to damage.

In the USN, the ship was controlled from an enclosed bridge with only a lookout platform above. The US 4in gun was criticised, mainly because of the flimsy nature of its fixed ammunition. The 3in AA gun had an elevation of 75º but could not be used at angles greater than 40º. As transferred, they mounted two triple 21in torpedo tubes on either beam but this was not a great success as the outboard ends of the tubes would dip into the water during a turn. The torpedoes were unreliable and ran some ten feet deeper than the setting.

Of the fifty ships, six commissioned with Canadian crews and, initially, the remainder with British. Australia and New Zealand were asked but did not have crews to spare. Later, several were operated by Allied navies, including a dozen by the Soviet Union.

The inspection report was taken very seriously and resulted in a set of Stage I modifications that were applied in varying degree to all ships, mainly at Devonport, while defects were made good. It was the spring of 1941 before most ships were operational.

The American guns were removed, with the exception of the forward one, and in some ships the beam guns remained for a short time. A British 12pdr was fitted on the after deckhouse. The after two sets of torpedo tubes were removed. To improve stability, the mainmast was removed and the foremast reduced in height. The three after funnels were lowered by several feet. Most were given fifty tons of ballast and the bilge keels were increased in depth. British depth charge throwers were installed. The asdic story was a little more complicated. Three ships had no outfit and these were given simple British sets (Type 133).⁹ The other forty-seven ships had USN Type QCJ/QCL, which was comparable in performance with contemporary RN sets but slightly more powerful. However, it lacked a streamlined dome, limiting use to speeds below fifteen knots. It was also more complicated and needed more skill to operate. There was no recorder; when a British unit was installed, the outfit was reclassified as British Type 141. Some were given a streamlined dome as Type 141A.

Stage II changes were applied piecemeal in many cases, depending on operational commitments – and the defect list. Changes comprised removal of the beam 4in guns, replaced by single Oerlikons, and of the remaining US torpedo tubes, to be replaced by one triple British mount on the centre line. Radar 286 was fitted at the masthead and Type 271 on the bridge in late 1941. The command was moved to an open deck above the old bridge; configuration varied. Later (late 1942), many received a Hedgehog forward. Some carried HF/DF on a mast aft.

The depth charge fit was complicated. Ships with the basic modification had a five-pattern fit with two throwers, two rails and fifty charges (Mk VII, light). The number of charges was reduced by six in ships whose funnels had not been lowered and by a further six if the starboard boat and equipment had not been removed. Ships modified in accordance with December 1941 instructions had a ten-pattern fit with two rails, four throwers (two ready-use racks per thrower) and sixty charges (fifteen heavy and fifteen light on the upper deck and ten heavy, twenty light in the magazine). There were twenty-four depth charge carriers. The three ships of Group D were used exclusively on the east coast and had one rail, two throwers and twenty charges.

Long-range escorts

At a meeting in November 1940,¹⁰ consideration was given to converting these ships to long-range escorts, noting that the hulls and machinery were believed to be in good condition. It was desired to obtain a speed of twenty-two knots, which would allow the two forward boilers and funnels to be removed, giving space for more fuel (eighty tons) – and feed-water, unless the evaporators could be improved. Endurance at fourteen knots would go up from 2,000 to 2,780 miles. Space would also be available for W/T and asdic offices. A stronger bridge of RN style would be built further aft. All torpedo tubes and US guns would be removed. A 4in HA/LA gun would be mounted on a bandstand with breakwater, further aft than the present A gun, and another fitted aft. Two single pom-poms and two Oerlikons would be fitted. Eight depth charge throwers and two rails (eight charges each) would be installed, giving a fourteen pattern. They would carry eighty-eight charges (thirty-four heavy and sixteen light on the upper deck and eleven heavy and twenty-seven light in the magazine). Asdic and radio were to be updated. It was thought that the work would take about six months, perhaps in the USA. Optimistically, it was noted that the USN still had another 120 similar ships! Later conversions would have had a fourteen-pattern fit with seventy-eight charges, soon changed to ten-pattern.

Chesterfield has a new bridge structure but is otherwise similar. It is likely that the bridge was fitted in Hull at the end of 1942.

Clare as a long-range escort. She has a different bridge a little further aft and has lost the two forward funnels and boilers. The full conversion took eight months and does not seem to have been thought worthwhile as only two more ships had similar conversions.

Three ships were converted to long-range escorts, a modification that predated the V&W conversions. Conversion took about eight months, although this figure means little as all the ships were in poor condition and a great deal of other defect work was carried out at the same time. It seems to have been concluded that conversions of the Towns were not worthwhile, as no other ships were taken in hand.

They had a short and fairly undistinguished career, participating in the sinking of eight U-boats. Most were paid off or transferred to non-operational service by late 1943. The value of these ancient ships will be discussed in chapter 10.

Ex-US Coast Guard Cutters

In the spring of 1941, ten cutters of the US Coast Guard were transferred to the RN under the Lend-Lease agreement. Strictly, they were of two different classes, but the effect on performance and appearance was negligible.¹¹

In several aspects they followed merchant ship practice. They were single-screw, turbo-electric, driven by an AC synchronous motor. Steam came from two boilers in a single boiler room. By warship standards their subdivision was poor, as none of the bulkheads ran continuously from top to bottom and there was very little subdivision between the weather deck and the deck below. The weather deck was of wooden planking with steel stringers and steel in the way of the guns. Their stability was suspect, though doubt has been cast on this story.¹² Ballast was fitted (thirty tons) and an oil fuel restriction of fifty tons was imposed. Their size and high freeboard made them comfortable sea boats.

The ex-US Coast Guard cutter Fishguard. Big, comfortable and with good endurance, they mainly operated on the UK-Freetown route.

Accommodation was exceptionally good, even by USN standards. The captain had a suite as well as a sea cabin and there were twenty cabins for officers and warrant officers. The wardroom even had a skylight. They had cafeteria messing with ice cream machines and drinking fountains. There were very large refrigerated spaces. Some of this luxury was lost to wartime equipment; for example, the laundry was converted to a radar office.

Displacement (tons): 2,116 deep

Dimensions (feet): 250 x 42 x 16 deep

Shp and speed (kts): 3,000 = 16

Fuel (tons), endurance miles @ (kts): 314, 8,000 @ 8

Complement: 160

Their original armament was one or two 5in guns, and changes were made as convenient. By 1945, most had one 4in HA/LA gun, one Bofors, two single pom-poms and seven Oerlikons. The ships were fitted with HF/DF and 271 radar and their anti-submarine outfit consisted of Hedgehog, four throwers with two rails (ten pattern) and 100 depth charges. Their endurance and spacious living standards made them very useful for the longer routes, such as UK–Freetown, but they were little involved in the main Battle of the Atlantic.

Lulworth, another ex-USCG cutter, in late 1943 shows few changes.