From the spring of 1941, things began to improve and, while the battle was far from being won, winning became a realistic possibility. This improvement was due to a number of factors, all improving and interacting one with another. If credit is to be given, it should be to the early commanders-in-chief, Dunbar-Nasmith and Noble, who initiated the successful programmes.
There were more escorts; gone were the days of only one or two per convoy. They were better trained too, both individually and as groups. The massive Canadian effort in shipbuilding and training began to pay off, despite shortages of equipment and experienced crews. Operating bases in Iceland and Newfoundland helped in making good use of the ships. It became possible to run more fast convoys, enabling the minimum speed of independents to be raised, with a marked drop in losses. The task was eased by the fitting of early radar sets, soon followed by much-superior centimetric sets of the 271 series.
It was not all good news; there were far more U-boats in commission than before, though initially most were still in home waters working up and the actual number at sea was still low. Dönitz was able to use wolf-pack tactics with increasing success, although it was noticeable that a high proportion of successes were due to a small number of aces. As Allied measures made life more difficult for U-boats, Dönitz moved his submarines westward. This exposed the limited endurance of RN destroyers, only partly offset by primitive refuelling-at-sea techniques.
Both sides depended to an increasing extent on secure radio communications and in both cases these were much less secure than was hoped. The code breakers of B-Dienst and Bletchley Park played a more and more important part in the battle. The topics introduced here will be explored in more detail in following sections.
Numbers of Escorts
It is difficult to give the numbers of the various types of escort involved in the Battle of the Atlantic. First there is the problem of geographical limits; Gibraltar convoys are included but not those in the Arctic or South Atlantic. However, diversion of considerable numbers of U-boats to the Mediterranean in this period had an important effect on the Atlantic battle. This was not only an issue for the Germans: on the Allied side, destroyers too were frequently diverted from the main battle to other areas, particularly the Arctic. The figures below and those in later chapters are only approximate. The numbers were further reduced by the need for maintenance.
At the beginning of 1941 (figures for January 1942 in brackets) there were about 36 (73) RN and 13 (67) Canadian Flowers, together with a few sloops. In addition, there were about eighty destroyers, mainly RN but a few RCN, a number which did not change greatly over the year. These destroyers were mainly World War I veterans at the beginning of the year, but later, vessels of the A–I classes joined as they were replaced by more modern ships for fleet duties.
The increased number of escorts available in the North Atlantic made improved protection for convoys possible. A typical convoy escort might now consist of two to three destroyers and four to five corvettes. Such numbers meant that there would be a complete asdic screen ahead of the convoy, making submerged attack very difficult. Night surface attack by wolf packs was already Dönitz’s preferred method, but the long summer nights reduced the opportunity for such attacks. With better protection to convoys it was possible in June 1941 to raise the minimum speed for independent sailing from thirteen to fifteen knots, with a dramatic drop in sinkings. Up to June, losses of independents were running at some 200,000 tons per month but this fell to about 50,000 tons with the change.
With more escorts available there was much more chance of keeping the ships of an escort group together and even to find time for group training, discussed later. Western Atlantic command had begun to realise the futility of detaching escorts to pursue an old and distant contact.
The numbers game was far from one-sided. By the beginning of May the total number of U-boats available had risen to 139, further increasing to 184 by August and 250 by December. The effect of this vast increase was not immediately apparent, as the new boats required extensive trials and their crews needed lengthy training. In consequence, the number of boats at sea did not rise rapidly. For the nine months from April to December 1941, the average number of U-boats at sea was only about twenty-seven and not all these were in the operational area.
The Focus Shifts Westward
At the beginning of the year, the UK-based ships escorted convoys to and from 20º west. The presence of stronger escorts made attack in the area increasingly dangerous and by February the U-boats were moving further west. This movement was limited by the US-patrolled neutrality zone. These westerly longitudes were outside the range of Condor reconnaissance aircraft and the broader spread of the ocean facilitated evasive routing, which was becoming more effective with the breaking of the Enigma code.
Even while officially neutral, the USN played an increasing role in the Atlantic war and, although this book does not cover USN operations or ships, some mention is necessary. President Roosevelt announced a Neutrality Patrol on 5 September 1939 to keep war out of the western hemisphere. This was formalised in October at a pan-American conference, with a neutral zone extending for 300 miles into the Atlantic from the coasts of North and South America. Goods purchased in the USA had to be paid for in cash and carried in the purchaser’s ships. This was a great help to the Allies, as no German cargo ships were at sea. Initially, the sale of munitions was prohibited but such items were put on the same basis as any other in November 1939. British dollar funds were quickly exhausted and in March 1941 the requirement for ‘cash’ was removed under the Lend-Lease Act.
After the fall of France, US policy moved further towards support for the UK. The exchange of fifty old destroyers for bases is discussed in chapter 3. One of these bases, Argentia, Placentia Bay, Newfoundland, was commissioned as a USN base on 15 July 1941 and was to give invaluable support to the Atlantic battle. A staff conference in January 1941 led to a promise of US support ‘short of war’. The government of Iceland agreed to the stationing of a US division there from July, and USN ships escorted convoys as far as Iceland.
Support for Germany within the USA diminished during the spring of 1941 as a result of the loss of US lives in sinkings. Then, in September 1941, the destroyer USS Greer was trailing a U-boat, without attacking, when U-652 fired a torpedo at her, which missed, as did Greer’s retaliatory depth charges. Escorting convoy SC48 on 17 October, the USS Kearney was torpedoed but not sunk by U-568, thanks to her unit system of machinery.1 The first loss came on 31 October with the sinking of the USS Reuben James.
The actual support from the USN was quite small in the period before open war but the effect on morale in the UK of feeling that she was no longer alone was very great – yet another ‘gleam of light’.
A key factor in the breaking of Enigma came as a result of the westerly deployment. Four U-boats attacked outward bound convoy OB318 on 7 May 1941 and for the next four days and nights, sinking seven merchant ships (39,000 tons) and damaging two more. However, the depth charges of the corvette Aubretia and the destroyers Broadway and Bulldog brought U-110 to the surface. Bulldog came in to ram, but her CO decided to try to capture the submarine. A party went across in a whaler and U-110 was taken in tow. She sank, but not before her Enigma machine and code books had been removed. With justice, this has been called the most important incident of the whole battle.2
Map Phase 3. The battleground was now the whole of the Western Atlantic from the tip of Greenland to Freetown. There was still little effective air cover.
In August, U-570 surrendered to a Hudson aircraft. She was brought back to the UK, where much was learnt about the capability of the Type VIIC, its equipment and torpedoes. She was commissioned into the RN as HMS Graph.
The difficulty of interception with the westward spread meant that U-boats had to be arranged in long scouting lines, which, in turn, made concentration more difficult if contact was made with a convoy. In November 1940 twenty-six Italian submarines were based on Bordeaux. Dönitz did not expect much from his allies but hoped they would help in locating convoys. In fact, in five months’ operations, only fourteen of the twenty-six scored any kills. Their large superstructure made them easily visible, even at night, and they were untrained in surface attacks. Their signals discipline was poor and sighting reports, even if made, often failed to get through.
Designed in response to criticisms of the Flower class, and initially referred to as ‘twin-screw corvettes’, the River class were a big improvement, but did not enter service until 1942. The River class differed little one from another, but Nadder, shown here, is unusual in mounting Oerlikons on the forecastle.
After ten weeks, the attack shifted back to the east. It was hoped that the greater number of Condor aircraft would permit three flights per day and enable convoys to be located. However, it seems that there were only two examples of such location. On the other hand, the more numerous escorts enabled ships on an outer screen to locate a shadower in good time and at least force it to remain submerged and so lose contact. From mid-September to November only five convoys were attacked, losing forty-six ships totalling 198,000 tons.
Then, in November–December, twenty U-boats were detached to the Mediterranean, losing three in transiting the Straits of Gibraltar. Though they scored a few successes against naval targets (Barham and Ark Royal), their effect on the North African battle was slight, while they were sadly missed in the Atlantic. Another six were detached on Hitler’s orders to Norway, to operate against Russian convoys. Their sinkings were few.
The Royal Navy had done well in 1941 and it is true that there was a gleam of light, but the overall picture was still dark. UK and Commonwealth shipyards had built one million tons of merchant shipping in 1941 but losses exceeded four and a quarter million tons, half due to U-boats. It was estimated that the UK needed thirty-six million tons of dry cargo per year, but imports fell nine million tons short of this. Worse, oil stocks, already low, fell by a further 318,000 tons. Sinkings of merchant ships were still only at 38 per cent of the 1917 figures and were certain to increase as the new U-boats became operational.3 Between April and December 1941, thirty U-boats were sunk, an increasing number by corvettes, though destroyers were still the biggest scorers (see table 4.1).
Table 4.1: Causes of U-boat sinkings, April–December 1941
|
Surface vessels |
21 |
|
Aircraft |
4 |
|
Submarine |
2 |
|
Mines |
0 |
|
Accident and unknown |
3 |
Technical and Other Developments
The year 1941 saw the introduction of a considerable number of developments, both of equipment and of its operational use. Their full effect would not be seen until a later phase of the battle and was delayed by the U-boat assault on US coastal shipping.
Centimetric Radar4
Better things were on the way; a new lightweight, rotating aerial based on that used in the 281 was tried in Legion in February 1941. It could detect a merchant ship at nine miles, a destroyer at five and a surfaced submarine at one and a half miles. By September 1941, thirty-seven ships (thirty convoy escorts) had this new aerial 286P, compared with 210 ships with earlier, fixed aerials.
Early in 1940, there were three developments, including the cavity magnetron, which made possible the use of much shorter wavelengths in the 10cm range. By November, a working set was tried in Swanage Bay. It detected the submarine Usk, stern-on, at seven and a half miles, and with the conning tower only showing at four to four and a half miles.
Type 271 radar. The original lantern had thick wooden framing which caused side echoes (a). A lighter version was introduced with more perspex (b). The lantern was usually covered in port (c) so that the dimensions of the aerial (d) could not be measured since this would disclose the operating frequency. (John Lambert)
A very simple set was devised and the first twelve units were built at Eastney. In February 1941 the brave decision was made to order 150 Type 271 sets in advance of trials and increase the Eastney build to twenty-four. Though waveguides5 were known, there was no experience with them, and the aerial feed was a coaxial cable. This had to be short, so that the aerial was on the roof of the radar office. This unit was quite big and heavy and there were difficulties in fitting in some classes. However, the Flowers had sufficient space and stability for the new set. The aerial was contained within a lantern with teak pillars framing flat, perspex panels, sand-blasted so that the aerial itself could not be seen (which would allow its operating frequency to be deduced).
The first Eastney set was fitted in Orchis during March 1941, while she was under repair at Scotts (Greenock). Trials began on 25 March in calm weather, with destroyers and the Norwegian submarine B1. As usual on trials, there is always calm when rough weather is wanted. However, rough-weather trials began on 1 April, with Orchis rolling and yawing violently. The detection range against the small submarine was 2.1 miles in rough seas, 2.5 in calm; destroyers were picked up at about 6 miles, buoys at 1 mile and high land ‘off scale’ at over 14 miles.
The twenty-four sets from Eastney and another 100 from the Allen West company were designated Mk II and a further 1,000 from Metropolitan Vickers, re-engineered for production, became the Mk III. These late sets had almost identical performance to the Orchis prototype but required much less maintenance. By May 1942, a year after the Orchis trials, there were 236 RN ships at sea with 271 or its derivatives. Most ships with 286 had been converted to 286PQ with a rotating aerial and improved performance. A redesigned 1.5-metre set, 291, was tried in Ambuscade in February 1942. Also early in 1942 an ingenious scheme was put in force to aid ship-fitting of 271. All components and the huts were sent to a factory,6 where the ship’s radar office was completely fitted out. Installation on board took place in two successive boiler cleans, each lasting about a week. During the first, the steel supporting structure was built, and three months later the hut with superimposed aerial was dropped on and connected.
There were remarkably few problems. Early installations suffered from much larger side lobes than expected, a problem traced to the teak framing of the lantern. The solution was a six-foot moulded cylinder of perspex, tried in Hesperus in December 1942. In August 1942, a PPI display was tried in Watchman, and fitting followed by mid-1943. Type 272 had a waveguide feeding the aerial, so that the office could be a short distance from the aerial. This made ship-fitting less of a problem and many destroyers and sloops had the lantern on a short lattice mast aft. Type 273 had a different aerial and was used in big ships.
In August 1941, a team at Birmingham University developed the ‘strapped magnetron’, which enabled the transmitting power of the 271 to increase from 5kW to 70kW. On the first test, this extra power merely burnt out the cable, but once the cable was replaced with a waveguide, which doubled the aerial gain, all was well. The 271Q was designed to use the same office, panels and cabinets as the original set, so the update was simple. The first trial was on the Marigold in May 1942 (at the same time as Hedgehog trials), when it was found that a trimmed-down submarine could be detected at five miles, roughly double that of the early 271s. Production began in December 1942 (272 was not updated but replaced). Later and better types were to follow – but the battle was won by the 271.
High Frequency Direction Finding (HF/DF) – commonly ‘Huff Duff’
The German control system for wolf packs depended on two-way ship–shore communication using HF radio. They were well aware that such transmissions could be located using direction finding but thought that the apparatus needed was so large that it could not be fitted in a ship. The cross-bearing, or ‘cut’, obtained in mid-Atlantic between two shore-based sets would be too inexact for operational use. As a result, the Germans concluded that over 200 kilometres from shore was safe. HF signals travel in a straight line and hence direct transmissions can only be received by ships in line of sight, ie within the horizon, known as the ‘ground signal’. However, signals are also reflected in the ionosphere – the ‘sky wave’ – and this signal can be received at any distance.
The experienced operator could learn a lot from the signal itself. It was usually possible to distinguish between the ground wave and the sky wave – the former indicating that the source was within about fifteen to twenty miles. The strict German procedures made it possible to recognise some standard groups, even when the full message could not be read. Such groups might indicate a sighting report or the start of an attack. There would normally be a very short, high-priority sighting signal followed by an amplifying message. The HF/DF operator might well be able to recognise the signs of a wet aerial from a U-boat that had just surfaced. Sometimes it was even possible to identify individual U-boats from the characteristic key style of their operator.7
HF/DF aerial, used for both the FH3 and FH4 sets with insignificant differences.
The Allies were fully aware of the value of ship-borne HF/DF and went to great lengths to guard its secrecy. Intercepted messages formed an important input into the Ultra code-breaking effort and the system was given the same security level until well after the war. Unaware of HF/DF, the Germans ascribed unexplained contacts to the use of radar, so that even when their documents were examined after the war there was no indication that they appreciated the value of DF.8 In consequence there has been very little published on the way in which British and Polish scientists developed ship-borne HF/DF and the RN and other navies used it operationally. Indeed, Rohwer credits HF/DF as being the major factor in the great convoy battles of 1943.
In the autumn of 1940 an electronic surveillance vessel was fitted out. The ship chosen was the Beachy, already fitting out as a rescue ship. She was intended to listen for radar transmissions from U-boats and also for radio communications.9 Her first voyage was with convoy OB226 on 10 October 1940 and she appears to have confirmed the use of HF radio for communications. Beachy reverted to the rescue-ship role – the first in service – and was sunk by air attack on 11 January 1941. Work was already in hand on HF/DF and a prototype, FH1, went to sea in Hesperus in March 1940. It appears to have been unsuccessful as no more were produced and an FH2 went to sea in August 1941. This and other early sets could not distinguish between the true bearing and its reciprocal.
The biggest problem was in the aerial. It was soon apparent that it should be at the highest point of the ship, the foremast head. By the time that HF/DF went to sea, this site had been claimed by radar (286) and HF/DF had to accept a shorter mast aft. Accuracy was good within the ground-wave reception (fifteen to twenty miles). Later in the war, radar 271 moved down and HF/DF took the top site. Later still, Type 293 became the standard surface-warning radar and this was normally carried at the top of a lattice mast, with a pole topmast for the HF/DF. The height of the aerial left it exposed to icing and wind, which meant that its elements had to be strong but not too heavy, as escorts were already having stability problems due to extra topweight. Aerial design came under a Polish engineer, Struszynski, working at the ASE (Signal School) under the overall leadership of C Crampton. The earliest sets, the FH3, went to sea late in 1941, the prototypes in the fleet destroyers Gurkha and Lance in July 1941. Ship-fitting built up rapidly in 1942 (twenty-five escorts fitted by January 1942), reaching the planned total by the middle of the year. The FH3 set was superseded by the FH4 in 1943 but there was no obvious change in the aerial. The FH4 set had been developed by Plessey with ASE and featured a visual (cathode ray tube) presentation rather than the aural presentation of the FH3. The prototype FH4 went to sea in the ex-USCG Culver in October 1941 and was first in action in Leamington in March 1942, on convoy WS17. The French, too, had made progress; their team moved to the USA at the fall of France and formed the basis for the USN work.
When there was only a single HF/DF set with a convoy, there was a slight risk of plotting a reciprocal bearing, though the tactical plot would usually make it clear which bearing was relevant and the developed aerial made a reciprocal almost impossible.10 A single set gave no range, though the operator would usually recognise the ground wave as ‘close’ (fifteen to twenty miles). Two sets, or better three, would give a cross-bearing and hence an accurate position for the transmitter. Since there were numerous frequencies available for the U-boats, a greater number of ship sets increased the chance of an intercept. If this shadowing U-boat could be sunk or even kept down, there was an opportunity for a change of course to evade other boats.
HF/DF was particularly valuable in locating the tracker U-boat, which could be sunk or, at least, forced to dive, disrupting the wolf pack concentration. Priority in fitting seems to have been given to ships carrying the senior officer of the escort. Since his ship would usually be in the van of the convoy, it was logical to give second priority to the rescue ship, normally at the rear, to give as long a base for bearings as possible.
Howse describe a typical action due to HF/DF on 23 August 1942 with convoy ON 122. First contact was made on HF/DF by Viscount and the rescue ship Stockport. They directed the corvette Potentilla (Norwegian manned) to the transmitter and she made radar contact forty-five minutes later at two and a half miles. The submarine dived but was picked up by propeller noise at three-quarters of a mile and then by active asdic. Three depth charge attacks were carried out and U-256 was severely damaged. On the way home, she was bombed in the Bay of Biscay on 31 August and out of action for a year.11
By mid-1943, the Germans had abundant evidence of the use of ship-borne HF/DF, but they were obsessed by radar and the evidence was ignored. Decoded radio messages referred to HF/DF, and agents in Spain regularly photographed RN ships off Gibraltar showing the aerial. It was even painted out when some photos were reproduced for recognition books.12 Many U-boat COs were suspicious and noted that attacks followed quickly after a radio transmission. Very late in the war, the Germans developed the Kurier burst transmitter, which was almost impossible to DF.
Communications
Escort vessels needed rapid, secure communications with other escorts for tactical control, merchant ships (particularly for diversions) and with HQ on shore. The means available in 1939 included flags, which could be used only in daylight. Flashing lights could be used at any time but there were few merchantmen who could read reliably. Medium-frequency radio using Morse code was available but it could be intercepted and located at a considerable distance and hence was rarely used. Loud hailer and the older megaphone were used but messages were too liable to descend into mutual abuse. Flags and flashing lights were suitable only for short messages, for example a code word for a prearranged procedure such as a diversionary course change. This limitation emphasised the need for common training.
The River class Plym’s large quarterdeck is full of depth charge gear. Note the HF/DF aerial at the masthead. She came to an unusual end, being expended in an atomic bomb trial.
Ship-to-ship voice radio was introduced in 1941 using medium frequency but, like its Morse predecessor, it was easily intercepted and hence little used. The US VHF TBS was introduced to the RN in 1943 and proved very valuable and reliable. TBS greatly improved the escort commander’s control of his group.
Code Breaking
All German services relied on signal encoding via the multi-rotor Enigma machine, which they believed was unbreakable. Of their potential opponents, Poland came closest to breaking Enigma and early in 1939 they handed over their work to France and the UK. The German naval Enigma had one more rotor than those of the other services and proved very hard to break. During 1940 and early 1941 the British acquired a number of rotors, often as a result of captures of weather ships intended to obtain cipher material. As described earlier in this chapter, the big breakthrough came on 9 May 1941, when U-110 was boarded by a party from HMS Bulldog, which captured a complete machine with code settings and other material. Another weather ship was boarded on 28 June, giving settings for July.
For the rest of 1941, most naval signals could be read fairly quickly, enabling the Admiralty to divert convoys away from U-boat concentrations. U-boat Command had some suspicions of a leak of cipher materials and tried various tricks to make their messages more secure. However, they still believed that Enigma itself was unbreakable. The value to the convoy system was very great; Rohwer estimates a saving of 1.5–2 million tons in the second half of 1941. President Roosevelt had involved the USN to an increasing extent in the western Atlantic and decoded messages were used to reduce the number of clashes between US ships and U-boats.
It is difficult to compare the successes of Bletchley Park and B-Dienst, since there were periods in which one or the other had a partial ability to read enemy messages, but possibly taking too long for tactical use. In summary, the Germans could read British signals from September 1939 to August 1940, and from February 1943 to June 1943; the British could read German signals from June 1941 to February 1942 (the introduction of the four-rotor Enigma machine), and from December 1942 to the end of the war.
Training
The RCN had a work-up base at Pictou, Nova Scotia in 1941 and at Digby, also in Nova Scotia, in 1943. Both were susceptible to bad weather, even though training was transferred to Halifax in winter months. Eventually, in 1944, a satisfactory solution was found with HMS Somers Isle in Bermuda.
The need for group training was appreciated but there were a number of difficulties, notably that the shortage of escorts made it difficult to keep the ships of a group together. Before taking up his appointment as commander-in-chief, Sir Percy Noble went as a ‘passenger’ on the leader of one group, but before the voyage was complete every other ship in the group had been diverted on Admiralty orders.
Aces
Among both submarine captains and their hunters, a disproportionate number of sinkings were due to a small number of aces. Men such as Walker and Gretton and, on the other side, Kretschmer and Prien, and a very few others on both sides, distort all statistics. Much of their success was due to natural skills enhanced by dedication and hard work. However, once their reputation was established, men such as Walker got the best ships and the best captains and were selected to operate in areas where the action was keen.
There was one big difference between the fate of the British hunters and their German opponents. Some RN aces admitted to burn-out and were rested in shore jobs where they could pass on much of their expertise to a new generation. Walker died of a heart attack. Even when still serving at sea, they were encouraged to visit training centres, where they could pass on their ideas to others and even learn new ideas themselves.
U-boat aces usually went on till they were sunk. It is well said that ‘There are old U-boat captains and bold U-boat captains but there are few old, bold U-boat captains.’ Table 4.2 shows the top-scoring U-boat commanders in the spring of 1941.
Table 4.2: Top-scoring U-boat captains (spring 1941)
|
Commander |
Tonnage sunk |
U-boat |
|
Günther Prien |
245,000 |
U-47 |
|
Joachim Schepke |
230,000 |
U-100 |
|
Otto Kretschmer |
282,000 |
U-99 |
Within little more than a week, all three had gone. During the night of 7–8 March, U-47, with four other submarines (including U-99), was operating against convoy OB293. Prien attacked on the surface, just before dusk, under cover of a rain squall. Visibility suddenly cleared, showing U-47 in clear view of the destroyer Wolverine. Prien made a sharp turn and attempted to escape on the surface. The destroyer was slowed by heavy seas and Prien should have got away but, foolishly, he dived. Wolverine steamed in and released a pattern of depth charges set shallow. U-47 was fatally damaged; in particular, the propeller shafts were put out of alignment making them very noisy. Prien had not given up and surfaced to escape. The noisy shafts gave her away and Wolverine came in again. U-47 dived and another shallow pattern was followed by a massive underwater explosion. Wreckage came to the surface – one down.
Convoy HX112 was sighted on the evening of 15 March by U-110, which homed in four more U-boats, including U-99 and U-100, to the scene. The escort was the 5th Escort Group, led by Cdr Macintyre, with five elderly destroyers and two Flowers. U-100 sank a tanker that night, and the following night U-99 sank three tankers and two other steamers totalling 34,505 tons. U-100 then came in again and was picked up at 1,000 yards by Vanoc’s primitive 286 radar (the first such detection). Schepke tried to escape on the surface but the old destroyer was too fast for him and rammed U-100 amidships, crushing Schepke against the periscope standard. There were six survivors. Macintyre suggests that Schepke was deceived by Vanoc’s Western Approaches camouflage scheme.
Ness, a standard River.
Shortly afterwards, Walker made asdic contact with a submerged submarine. The depth charge crews had not completed reloading after a previous abortive attack but six charges were dropped in a snap attack. Immediately, a submarine was sighted surfaced and stopped astern. Every gun that would bear opened fire, but with no fire control there were few hits and little damage was done. Eventually the U-boat signalled that she was ‘sunking [sic]’. Forty men of the U-99, including Kretschmer, still wearing his cap, were picked up, to the delight of thirty-eight survivors of a merchant ship rescued by Walker a little earlier. Kretschmer had a special pair of Zeiss binoculars, presented by Dönitz, and these came to Macintyre as a prize of war.
Walker had made earlier contacts that did not lead to a kill, including one where the first contact was the visual sighting of a surfaced submarine wake.
The lessons were well known, even at that date, and it is surprising that three aces fell foul of them. Visual sighting was vital, though Vanoc’s radar contact gave a blurred view of the future. A submarine caught on the surface should stay there and run, heading into the sea. Even the faster destroyers had difficulty in catching a submarine at night and an unstabilised 4in gun without fire control was unlikely to hit, let alone cause fatal damage. Ramming was reliable, but costly in terms of the damage to the escort.
Twin-Screw Corvettes – River class Frigates
As soon as the Flower class corvettes entered service, there were complaints. Many of these were unjustified but it was clear that these corvettes, designed for coastal operation, were far from ideal for Atlantic work. A meeting was called on 27 November 1940 to consider the requirements for an A/S escort. Those attending included senior representatives of relevant staff divisions and A P Cole, RCNC, who was asked to make a rough estimate of the size and power of the desired vessel.13 It was noted that it might be necessary to ‘trim some of the requirements to bring them into the realm of practicability’. The ‘wish list’ included:
• good sea-keeping to keep company with a convoy in any weather;
• speed of at least 20 knots – 22 if possible – with good acceleration and astern power;
• endurance of 3,500–4,000 miles (UK–Freetown);
• small turning circle;
• simple machinery, easy to maintain;
• armament of one 4in HA/LA (two in bigger ship), two 2pdr, two Oerlikons, with simple fire control; if second 4in fitted, director with Type 285;
• four depth charge throwers each side and two traps (fourteen pattern) with 100 charges (it was noted that new methods of throwing depth charges were being tried, presumably ahead-throwing);
• asdics (unspecified), ASV (radar), W/T and R/T;
• SA gear (sweep for dealing with acoustic mines); wire minesweeping;
• stem stiffened for ramming but ‘rounded’ for asdic reasons;
• small crew.
DNC thought that this could be met by a ship of about 1,500 tons, 300–20 feet in length (soon revised to 350 feet). Diesel engines could not be relied on for continuous high speed. Steam reciprocating engines needed too much space and could not accelerate rapidly. Turbines of at least 12,000shp would be needed and would take about eighteen months to produce. The meeting then considered the conversion of the ex-US Town class to long-range escorts (qv). This ‘wish list’ conflicted with the need for the vessels to be cheap and easy to build using existing facilities – shipyards, engine works and armament manufacturers.
The design was directed by A W Watson (who had designed the P boats of World War 1; Goodall was to remark in his diary for 18 June 1940, ‘Watson is a really good man’), with Kimberley as the constructor in direct charge. A first study suggested that the requirements implied a ship of about 350 feet in length, with turbine machinery and a lightweight hull to warship standards. Production constraints dictated that such a ship could not be built in numbers; available slips limited the length to about 300 feet and there was no capacity for turbine machinery. The A/S armament would be the same as a corvette – asdic and depth charges – but far more of them. Design seems to have begun in about November 1940, when it was still referred to as a twin-screw corvette, and a design study approved by the end of the month. By December, Goodall was concerned that they were getting too complicated and Watson was told to simplify.
The River class Derg in 1943 seems to have minesweeping gear as well as fourteen-pattern depth charge gear.
Helmsdale was one of the few Rivers with turbine machinery. She remained in service longer than most, being used for trials of ASW equipment.
The hull structure was relatively light, though complying with Classification Societies Rules (Lloyd’s Register and British Corporation); some warship experience was incorporated and it was easy to build. The Societies’ surveyors supervised the building. The preparation of building drawings was divided among the shipbuilders. Some builders saved up to thirty tons by the use of welding.
The machinery consisted of two sets of Flower class triple-expansion engines, slightly modified with a drive off either end of the crankshaft to suit the right- and left-handed twin shaft arrangement. This gave a speed of just under twenty knots, at least faster than a surfaced U-boat. Six ships were powered by a twin shaft turbine plant of 6,500shp, giving twenty and a half knots.
Table 4.3: River class frigates (as designed)
Displacement (tons): 1,855 deep
Dimensions (feet): 301¼ oa x 36½ x 13 deep
Shp and speed (kts): 5,500 = 19.5
Fuel (tons), endurance miles @ (kts): 470, 5,000 @ 15 (clean bottom)
Complement: 114
The designed armament was two single 4in Mk XIX HA/LA guns (60º elevation, 1,300ft/sec muzzle velocity) and ten 20mm Oerlikons. Because of the shortage of guns, most carried only four Oerlikons. The single Oerlikons were later replaced by twin power mounts in most. Many had HF/DF at the masthead and 271 on the bridge. A few had the bridge set replaced by 277 later. Some had a 6pdr on the forecastle, either side of the Hedgehog.14
There were eight depth charge throwers (fourteen pattern) and 100 depth charges. Four throwers were later removed and the number of charges increased to 154, with asdic 144. There was a Hedgehog on the forecastle (Hedgehog had not been developed when the ship design was started). The first twenty ships were fitted for minesweeping but the gear was later removed, allowing them to carry an extra eighty-five tons of fuel. After the twenty ships of the first programme and four of the next programme, fuel capacity was increased to 650 tons, giving 7,500 miles endurance. All Canadian ships had the longer endurance.
The long forecastle gave good stability at large angles of heel, even when the inevitable weight growth took place. Similarly, the deep hull kept stresses low, reducing the risk of their breaking their back when torpedoed.
Rother was treated as ‘first of class’ (though beaten to sea by Exe) and carried out a full programme of trials in spring 1942.15 At a displacement of 1,900 tons, she achieved 19¾ knots with 5,535ihp at 184rpm. Cost was quoted as £240,000. Reports from sea were generally favourable, the long forecastle keeping much of the ship dry while the covered access fore and aft made life more pleasant. The low quarterdeck was wet, affecting the depth charge crews. The Hedgehog was exposed and wet. It seems that there was an intention to replace it with a split mounting either side of the forward gun, but only Monnow was so altered. The fact that few changes were made in service shows that the original concept was right.
Monnow was the only River to have a split Hedgehog on the forward gun deck. (IWM A22681)
A total of fifty-seven ships were built in the UK; Canada built ten for the RN (two were transferred to the USN) and thirty-nine for the RCN. Twenty-two more were built in Australia with slightly different armament, providing increased AA for the Pacific war. These took no part in the Atlantic battle and are not discussed.
The Rivers were very good ships when they appeared in 1942 but with so many later ships available they were soon paid off by the RN. Their only fault lay in their late arrival; what a pity that a prototype had not been built before the war. Some attractive and effective modernisations of Canadian ships, planned by Rowland Baker, showed what was possible.
HMCS Jonquiere shown post-war after a modernisation designed by (Sir) Rowland Baker, RCNC.