9

Nuclear Submarines

THE ADMIRALTY WAS QUICK to appreciate the potential of nuclear propulsion, particularly for submarines.¹ This interest was expressed in Daniel’s paper of 1948 to the Institution of Naval Architects which, though dealing mainly with nuclear weapons, also envisaged nuclear propulsion.²Early studies by the DNC Submarine Section under Newton and later Starks were based on a gas-cooled reactor somewhat similar to the bulky unit at Calder Hall, Britain’s first operational nuclear power station. The submarines were correspondingly large: the first of 1950 was a twin-screw design of about 2500 tons with an underwater speed of 25kts.³ Within a year, the reactor size had increased and with it that of the submarine. The next study was 3400 tons and speed had fallen to 22kts. A redesign for a manoeuvrable submarine and shock protection brought the displacement up to 4500 tons at 20kts, while the pressure hull diameter had increased from 25ft to 31ft. The obvious conclusion was that a gas-cooled, graphite-moderated reactor was not practical in a submarine.

Rowland Baker, later knighted, who led the early nuclear submarine (and Polaris boats) programme, completing them on schedule and within budget. He is seen here as a Constructor Commodore in Canada, where he designed the St Laurent class frigates. (D K Brown collection)

In the early 1950s submarine studies under Bob Newton were led by Sidney Dale, constructor, with Keith Foulger as his assistant.⁴ They were involved in the Porpoise class and later HTP designs as well. The team at the Atomic Energy Research Establishment at Harwell were also only part-time on SSN projects. This team included Dr J R Dunworth, later professor of Mechanical Engineering, and a young MEO, Lt Righton. As well as the traditional work on submarine design, the DNC team spent much time in an unsuccessful attempt to mount the complete reactor compartment on springs to protect it from shock. Eventually it was decided that the scheme was impractical and work stopped.

By 1953 further discussions within the naval staff had shown the importance of nuclear submarines, but the Government had decided that scarce resources for research and development in nuclear work should be directed to power stations. Studies at Harwell continued at low priority investigating both water and liquid metal cooling. At the same time the USN prototype water-cooled reactor went critical in 1953 and the submarine Nautilus commissioned in September 1954, thanks to the drive of Admiral Rickover.

In 1954 a small naval section was set up at Harwell under Captain (E) S A Harrison-Smith, liasing with the submarine section. This section was considerably enhanced in 1955 when enough uranium (U235) was allocated for R&D to proceed with the aim of getting a submarine plant running on shore by 1961.⁵ Attention was concentrated on a pressurised water reactor. A E Reeves was much involved in the design of the shielding, whose weight was an important aspect in the submarine design. This involved the first large-scale use of a computer in ship design. In essence, for each point outside the shielding the radiation dose from every part of the reactor system had to be calculated and added together. This meant a considerable amount of three-dimensional geometry as well as skill in programming the early, clumsy computers. The fact that measurements taken later during the radiation survey were close to the calculated figures reflects great credit on Reeves and the naval section. Newcomers were horrified by the cost of the many tons of lead required to absorb gamma radiation and even more when it was explained that the thick and flawless polythene required to mop up neutrons was much more expensive.

In 1956 Treasury approval was given to build a shore-based prototype at Dounreay, alongside, but not part of, the AEA site. It was hoped that this unit might be running by January 1960 with a submarine by mid-1962. This did not have high priority and progress was slow. The naval section soon concluded that the most suitable plant for a submarine was a pressurised water reactor using enriched uranium. Vickers Armstrong (Engineers) were to be the main contractor with Rolls Royce as the subcontractor responsible for the nuclear reactor. The core design had to be settled by late 1956 in order to have a zero power test⁶ by early 1957.

In that year Rickover, on a visit to the UK, invited a British team led by Starks to see something of USN work, which led to some changes in the British plans. Work on the guided-missile cruiser was stopped in 1957 and the design team led by Starks with Daniel as his constructor became the nuclear submarine team. These studies, based on earlier work by Brokensha,⁷ led to a nuclear plant very similar in configuration to the US S5W plant.⁸ In hull structure the British work was well in advance of the American thanks to the work of Bill Kendrick and others at NCRE (see Chapter 12), while L J Rydill in a paper of 1957 showed the problems of fatigue failure.

By 1957 there were three major projects in hand and all were making progress. The Dounreay prototype had been the lead project and a full-size wooden mock-up of the reactor and machinery spaces was nearing completion at Southampton in the old Supermarine works.⁹ The Harwell group had expanded to 160 professional staff and Neptune, a zero-energy reactor designed to provide information for the submarine programme, went critical on 7 November 1957 (it was moved to Derby in 1959). All this work was leading up to the all-British nuclear submarine which became Valiant. However, while the US offer to sell a complete nuclear submarine power plant (S5W, as in Skipjack) enabled the RN to get a submarine (Dreadnought) into service quickly, as discussed in the next section, it delayed the completion of the British work. There were those who were unconvinced that a shore-based prototype was needed now the US plant was available and there was some access to their design philosophy.

The biggest and most controversial decision to be made for the Dounreay plant was in the choice of the type of steel for the reactor pressure vessel and the primary circuit. The advantages and disadvantages of austenitic¹⁰ stainless steel and low alloy steel were argued with some heat, with corrosion of various types and weld reliability at the centre of the debate.¹¹ Eventually, the low alloy steel was chosen, which did rather better than expected as it formed a corrosion-resistant oxide film that helped to give an active life of over 17 years. Fabrication, however, was not as easy as had been hoped.

The replica submarine hull at Dounreay was built from mild steel which was not strong enough to contain the pressure from a major nuclear accident within the reactor compartment alone. It was planned that the over-pressure should be vented into a void space which had to occur very quickly indeed. This proved much more difficult than it sounds but was solved when toughened glass was suggested. The manufacturers (Pilkington) doubted if their material was consistent enough but some hilarious tests in which a number of panels were broken proved that it was reliable and consistent. The reactor compartment was surrounded by a big water tank which provided shielding and also kept things cool. When the structural work was complete the hull was water tested to 145psi internal pressure and after it was fitted out there was an air pressure test to 126psi. The energy contained in 10,000ft³ of air at this pressure was equivalent to some 10 tons of TNT, and all sorts of authorities had to be convinced that it was safe (including the author, who was standing on it!). The risk of a catastrophe was quite small but the chance of a valve flying off could not be neglected.

By 1959 most of the research work at Harwell was complete and the naval section was disbanded, with most of the staff moving to the Dreadnought project. Professor Edwards moved to the RN College, Greenwich, where he set up the long training courses in nuclear engineering which the submarines would need.¹²

Dounreay (DSMP) commissioned in 1963.¹³ There were initial problems with some of the nickel alloy smallbore pipework and this was replaced by chromium-molybdenum low alloy steel in 1964.¹⁴ Since then DSMP has given many years of trouble-free service and has been essential in the development of RN nuclear plants and in training their operators.¹⁵

Dreadnought

Towards the end of 1956 draft Staff Requirements for a British nuclear submarine (SSN) were agreed, and in February 1957 the Minister of Defence visited the USA when the US offered to release nuclear information. Later that year the uss Nautilus visited the UK and the Minister and First Sea Lord were given a demonstration cruise. In January 1958 the President and the Prime Minister signed an agreement for the UK to purchase a complete SSN propulsion plant, later defined as S5W. As part of the deal a small monitoring team from MoD led by Captain ‘Shorty’ Cotmiston was appointed to Westinghouse, Pittsburg in 1959. Another team with Constructor Commander Keith Foulger and engineer colleagues Roger Berry and Reggie Down worked within the Gorleston shipyard of the Electric Boat Company ensuring that all submarine design information relevant to Dreadnought was received in the UK. During this period there were frequent visits by Vickers, Rolls Royce, ship’s officers (designate) and by US to Barrow and Derby. Keith Foulger, later Chief Naval Architect, has said that ’seeing the first British nuclear submarine get off the ground’ was the most exciting time in his career.

In March 1957 the name Dreadnought had been chosen – Vulcan was a close second, followed by Thunder.¹⁶ In November the Dreadnought Project Team (DPT) was set up under Rowland Baker.¹⁷ DPT was responsible for production and programme but the team reported to DG Ships on design aspects.¹⁸ The after end of Dreadnought had to be identical to uss Skipjack¹⁹ in order to accept the US plant, while the fore end was derived from the earlier British studies and was dominated by the very large conformal array sonar. There was continuing concern that there might be a mismatch or misunderstanding between British and US technology.²⁰ Rolls Royce set up a new company – Rolls Royce and Associates (RRA)²¹ – to deal with Westinghouse and build all British submarine plants.

Dreadnought’s diving depth was set by the design of the US machinery systems but was still much greater than British submarines in service. It was realised that the speed, manoeuvrability and strength would lead to frequent dives to maximum depth and hence the fatigue life of the hull was a vital criterion. To ease the problem, the pressure hull ended in torispherical domes and similar measures were taken where the hull diameter was reduced in the machinery rooms. The steel was specially developed – QT35 – and it was decided that a detailed crack-detection survey should be carried out at regular intervals. There were early problems with cracking due to dirt inclusions in the steel but these were overcome with improved welding procedures. Dreadnought was decommissioned in 1983, 20 years after she entered service.²²

The bridge fin was further aft than in the US design, partly to permit an internal layout based on RN practice but also because model tests had shown that roll induced when manoeuvring at speed (‘snap roll’) would be less. After prolonged debate, it was decided to position the forward hydroplanes near the bow rather than on the fin as was USN practice as this gave better control at low speed, particularly at periscope depth, at the expense of more interference with sonar performance.²³

The launch of the Dreadnought at Barrow, Trafalgar Day (21 October) 1960. (D K Brown collection)

A considerable number of propeller designs were tested at model scale at Haslar. The interaction between the flow over the hull and into the propeller affected both efficiency and noise. Power is lost due to friction of the water in the flow over the hull and much of this loss can be recovered if the flow is channelled through the propeller. In general the bigger the propeller the more efficient and quieter it would be, but the bigger propellers needed to run at very low rpm – the biggest model propeller had a full-scale diameter well over 30ft and ran at impossibly low speed.²⁴ To our surprise we were past the optimum and a slightly smaller propeller would have been better.²⁵ It had long been realised that a single propeller was more efficient than two but the redundancy of two shaft lines was initially preferred by operators in all navies.²⁶ The success of uss Albacore made it clear that the advantages of the single shaft were overwhelming. A small retractable propulsor was fitted.²⁷

The fore end had to accommodate the very large sonar and six torpedo tubes using a novel torpedo discharge system which could work at great depths. It was thought that a very smooth surface finish would help to make the submarine quiet as well as increasing speed and Dread-noughtreceived a super-fine finish.

By about 1960 there was concern that design changes were delaying the completion date and Baker laid it down that every change would require his personal authorisation. The author had joined DPT in 1961 and initially was responsible for radiation shielding and nuclear safety for all projects (Dounreay and Valiant as well); after a year, when the Polaris project started, I found myself with a scratch staff looking after the completion of Dreadnought Getting Baker to sign a change order was a frightening task; one would be sent for a day or so after he had the file and more often than not the file would be thrown at you as you entered. Baker had been brought up on a working Thames barge and had an unusual command of the English language, used to the full when contemplating any change.²⁸

Everything possible was tested alongside at Barrow. The designers had promised that if there should be a nuclear accident, the leak rate should not exceed 1 per cent per day. This proved very difficult to achieve and the author spent many weekends at Barrow working on it under the eagle eye of a Health and Safety inspector called Norsworthy, but finally we succeeded. Then came the radiation survey: first a quick sweep when the reactor went critical for the first time to make sure that there was no gross problem; then a wait while the reactor was gradually worked up to full power. After various delays the full-power survey went off without problems. It is maddening to those involved when the media talk of neglect of safety in earlier nuclear plants. Everyone tried hard, supervised by men like Norsworthy, and got most things right. Speed trials followed off Arran. At that date submerged speed was measured by towing a float whose speed was tracked by a theodolite on shore. The only problem was that the landowner would not permit the use of the site during the rutting season of his deer!

Dreadnought, seen on early trials prior to commissioning (note the Red Ensign), still with a pendant number painted up. (D K Brown collection)

This brief account has only touched on the numerous problems associated with the design and commissioning of a nuclear submarine. For example, it has been said that the air purification gear was a more difficult design problem than the nuclear plant, and, on a personal note, even the instrumentation to check air purity kept the author busy. Oxygen was produced by electrolysis while carbon dioxide was removed by a scrubber – and tobacco smoke by CO and H₂ burner. Rubbish disposal was another long-term problem. Rubbish was bagged and blown out of an ejector like a small torpedo tube. Accommodation standards were vastly superior to those of earlier submarines.

Dreadnought completed in April 1963 on time and on cost, probably the only major defence project of the time to do so, due to Baker’s drive and determination and to the enthusiasm he engendered in his staff.²⁹ She would have completed 6 months earlier had it not been decided to re-braze all the joints in the sea water systems to UK standards³⁰ (it is believed that uss Thresher was lost due to a failure of a brazed joint).

There were a number of teething problems, mainly not very serious. The US-built turbines had to be changed to cure a vibration problem, which caused some difficulty as Rosyth had not begun to train welders to work to Grade A standards on QT35 steel.³¹

Particulars32
Dimensions:	265ft 9in × 32ft 3in × 26ft
Displacement:	3500/4000 tons
shp/speed:	15,000/28kts
Complement:	88
Armament:	Six 21in tubes33

Valiant

The basic concept was to graft the fore end of Dreadnought on to the British machinery under development in the Dounreay Submarine Prototype – Baker’s philosophy was to change one of the three main features (fore end, nuclear plant, main machinery) of an SSN in each class. Inevitably, other changes were introduced. The fore planes were moved a little further aft to reduce interference with the sonar; turbulence around the planes accounted for about 10 per cent of the total submerged resistance of Dreadnought. The diving depth was slightly increased to the limits of the British machinery, and she was given a retractable propulsor – known as the ‘egg-beater’ – driven off the battery as a ‘get you home’ unit in the event of a failure of the main machinery. Valiant could also use the main shaft in turbo-electric mode for quiet propulsion since the main gearbox was noisy even though both the main engines and gearbox were on a raft resiliently mounted to the hull.³⁴ Some tanks were resited and to obtain the right longitudinal balance – always a problem in SSN – the forward compartments were lengthened, resulting in a rather larger submarine.

A rare May 1966 photo of Valiant with a pendant number, which she carried for a very short time only. (MoD)

Churchill, of the later Valiant class. (D K Brown collection)

Conqueror, of the Valiant class, built by Cammell Laird. During the Falklands War she sank the General Belgrano, ensuring that major Argentine surface warships would remain in port for the rest of the conflict. (D K Brown collection)

Experience with the American plant in Dreadnought led to some changes from the DSMP nuclear plant. In particular, the primary circuit and its components were fabricated from stainless steel, and the reactor pressure vessel was low alloy but with a stainless steel liner deposited by welding. All systems were simplified and the number of valves was reduced.

Valiant was already well advanced when the uss Thresher was lost. Two British assistant constructors carried out a computer simulation of the sinking which pointed a finger at the likely cause – failure of a brazed joint in a 5in pipe. Following this loss there were major investigations on both sides of the Atlantic, not only into the loss itself but also into any other potential problems with SSN design or operation. These studies did not show any problems with the Valiant design. In the author’s opinion, Valiant was the finest British post-war design of its day, surface or submarine, and reflects great credit on Brokensha for the initial studies and on the Chief Constructor, Daniel, and his constructor, Foulger, who created the final design and saw it completed.³⁵ Manoeuvring limitation diagrams were produced for all classes setting out safe limits for speed and plane angle at different depths – eg full speed at maximum depth was not ‘safe’.

Particulars
Dimensions:	285ft × 33ft 3in × 27ft
Displacement:	4400/4900 tons
shp/speed:	15,000/28kts
Complement:	103
Armament:	Six 21in tubes

Diving depth has been quoted as 300m. Costs for the class varied from £24 million (Warspite) to £30 million (Conqueror).

Valiant was ordered in August 1960 and completed in July 1966, followed 9 months later by Warspite. There was then a gap in SSN orders due to the Polaris programme until Churchill (SSN-04) was ordered in October 1965, followed by two more repeat Valiants. (Conqueror and Courageous). Valiant was delayed by the priority given to the Polaris programme and following the problems at Dounreay all nickel alloy pipes and fittings were replaced.

Some cracks had been found in Dreadnought’s QT 35 plates and, though these were made good, American HY80 steel was used in some later Valiants until a suitable UK steel could be developed. All five gave good service, Conqueror becoming the first SSN to sink an enemy ship – the Argentine cruiser General Belgrano. They were paid off in 1990-2 when cracks were found in their primary circuits. Though these were repairable their remaining life under the defence run-down made this uneconomic.

The Polaris Programme

During 1962 the cancellation of the USAF Skybolt missile programme, also intended for the RAF V-bomber force, led to studies of alternative British nuclear deterrents. It was soon realised that the submarine-launched Polaris system was best suited to UK requirements and a technical team led by S J Palmer visited the USA to study the design of USN submarines of the SSBN-627 class. Armed with this information Prime Minister Macmillan and President Kennedy agreed at a meeting in Nassau late in 1962 that the Polaris weapon system and associated technology should be made available to the UK, and the formal agreement was finalised in April 1963.³⁶ The whole programme was to be run by Admiral Mackenzie, with Baker responsible for design and building the submarines and installing the weapon system, retaining the familiar initials DPT,³⁷ this time standing for Director Polaris Technical. At the same time he also retained authority over the SSN Programme.

Various ways of procuring the submarine force were considered, including a total copy of the SSBN-627 design, or the purchase of US-built boats. There was even a suggestion that Valiant herself should be cut in half, but this was dropped as likely to cause too much disruption. The eventual design concept was simple in principle: the missile compartments of a uss George Washington³⁸ were to be inserted between the bow and stern of a Valiant – but, in real life, nothing is simple. Accommodation was needed for a much larger crew to even higher standards making long patrols more tolerable. About half the increase in length over Valiant was due to the missile compartment and the remainder to associated requirements.

Within the missile compartment and weapon spaces the framing, decks, bulkhead, etc had to be identical to the US boats, which meant reducing the pressure hull diameter by 3in compared with 33ft 3in of Valiant.³⁹ External ballast tanks, fore and aft, were fitted as in Valiant, giving a surface reserve of buoyancy of 8 per cent. This was thought rather small and internal tanks, designed to take full diving pressure, were fitted to give a total reserve of 12 per cent.

The steam machinery, supplied by the initial reactor core, Type A, would deliver 14,250shp giving 21kts, submerged, clean. The later Type B core gave 19,250shp and 23kts as well as a longer life.⁴⁰ Surface speeds were 15.25kts and 16kts respectively. Battery drive would give 4.5kts, whilst a 700hp Ward Leonard turbo-electric mode would give a very quiet 6.75kts as the gearing noise was eliminated. The main machinery was mounted on a raft, as in Valiant, with a quiet speed of 15.75kts (6 months out of dock). The electrical load was greater than in Valiant and the steam turbo generators were modified to give 2000kW each. There were two diesel generators each giving 290kW on the surface or 230kW snorting. These provided power to start the reactor from cold, for battery charging, to provide DC power and for emergency propulsion. There was a retractable emergency propeller known as the ‘egg whisk’ which could drive the submarine at ‘more than 3kts’. A battery was installed to restart the reactor after a ‘scram’ (emergency shutdown).

Resolution, the first Polaris boat, off Portsmouth Dockyard. (D K Brown collection)

The design diving depth was 750ft and the main bulkheads were designed to take the same pressure. There were two escape towers, one in the fore ends and the other in the motor room. It was noted that the end compartments could not contain the whole crew (with margins) and that the shape was such that the submarine would not necessarily remain upright if resting on the bottom. Any nuclear accident was to be contained within the reactor compartment, which was tested to 240psi, a very severe test.

The submarine had to be kept within a narrow band of depths during missile firing, which required very clever hovering gear. It was decided to develop a British system which worked very well and was thought much superior to the USN system.

Four Polaris boats were ordered in 1963 and completed by 1969.⁴¹ In order not to disrupt the SSN programme entirely, Cammell Lairds was brought in as a second builder, with a constructor lent to the firm as project manager. Baker introduced the PERT management scheme with help from Electric Boat and the co-operation of Vickers, achieving the double of again completing the programme on time and on cost. Baker was awarded the KB on the personal recommendation of Earl Mountbatten (an unusual honour for a civil servant) in the New Year’s Honours List of 1968 and retired soon after. Great attention was paid to managing the refits so that the target of keeping two SSBNs on station was maintained throughout the life of the force.⁴²

Particulars
Dimensions:	425ft × 33ft × 30ft
Displacement:	7500/8500 tons
shp/speed:	15,000/25kts
Complement:	143
Armament:	16 Polaris A3 (warhead later replaced by the UK-designed ‘Chevaline’), six 2lin tubes

Swiftsure Class

The design of this class during the mid-1960s provided the first opportunity for an overall review of the features of a nuclear submarine in the light of experience. The principal performance attributes of a submarine are speed, diving depth and quietness and it was hoped to improve all three in the new class. Swiftsure built on the experience with Valiant but was in no sense an ‘Improved Valianf – it was a totally new design. The design team was led by Norman Hancock,⁴³ lacking in submarine experience but very experienced in surface ships and supported by a knowledgeable staff. Because of the Polaris programme it was not possible to assemble a full design team until 1964, though studies were in hand a year earlier under Hancock, assisted by W G (Bill) Sanders.⁴⁴Approval was given in 1961 for RRA to develop an improved Core B which would give more power and a longer life.

The pressure hull was made nearly cylindrical throughout its length, eliminating the contractions which reduced fatigue life in earlier boats. This led to a reduction in the capacity of ballast tanks and hence of reserve of buoyancy. A new British steel was introduced which was tougher and cleaner than QT35. Great attention was paid to the safety of systems which might be exposed to full diving pressure, and auxiliary machinery had a fresh water cooling system working at low pressure, itself cooled by the sea in a heat exchanger with very short lengths of piping exposed to full sea water pressure. Many of the components of the sea water systems were large castings of nickel-aluminiumbronze (NAB). It took a great deal of work to develop casting techniques which would guarantee the quality needed to ensure safety.

The external shape of the hull and appendages was improved. There was a very useful report from 1923 by the Royal Aircraft Establishment (RAE) on the shape of airships – also three-dimensional fluid travellers. There was an amusing twist in that two of the best forms tested by RAE had been suggested by AEW Haslar on the basis of the fast ‘R’ class submarines of the First World War. The aft end was made much more blunt, mainly to provide buoyancy aft but also because there was some small overall benefit in propulsive efficiency due to favourable interaction between propulsor and hull flows. If the ending was too blunt there would be a major loss of efficiency due to separation of turbulent flow. The hydrodynamics of the day was unable to provide a complete answer but it was known that separation would occur more easily on a model than on the full scale. Hence a form was selected which just began to separate on the model confident that there would be no problem on the submarines.

The bridge fin was reduced in size, even though this implied shorter periscopes. The forward hydroplanes were on the axis, just abaft the sonar where they were most effective in slow speed control. They were retractable to reduce noise and drag at medium speed though they were usually extended for safety at the highest speeds.

Some of the class had a pump jet in place of the single propeller.⁴⁵ A pump jet is similar to a water turbine consisting of a rotor and a stator, both with a large number of blades, surrounded by a carefully-shaped duct. It enabled the flow to be carefully controlled in velocity and pressure and could be either more efficient or quieter than a propeller or a bit of each. It was the culmination of a lengthy R&D programme at the Admiralty Research Laboratory, Teddington, by Alex Mitchell. Some years later this technology was passed to the USN. During the Cold War it was generally true that Soviet submarines were fast, deep-diving and noisy, while British boats were relatively slow and very quiet, with the USN in between.

Great attention was paid to machinery noise, with even more of the plant on a resiliently-mounted raft. At moderate speeds the cooling water circulated under the ram effect of scoops with the rather noisy circulating pumps switched off. There was a further re-examination of safety and many mostly minor improvements were made. Great efforts were made to simplify the machinery systems.⁴⁶ To improve on such a good design as Valiant was a real achievement.

The first of class, Swiftsure, was ordered in November 1967 and completed in March 1973, while the sixth and last completed in May 1981. Core B was installed at DSMP late in 1967 and went critical in August 1968. Core B was then run at high power for two years to confirm its reliability and check its life.

The Admiralty Development Establishment Barrow (ADEB) had been set up during the development of HTP propulsion and was now refurbished to assemble and test the Swiftsure secondary (non-nuclear) machinery⁴⁷ There were a number of problems; the first raft was insufficiently rigid and there were gearbox failures. This meant that the prototype unit had run for a few weeks only, instead of the intended 6 months, when the production unit for Swiftsure itself was ready. A risky but successful decision to go ahead was taken and the production unit began trials in April 1970.

Particulars
Dimensions:	272ft × 332ft 4in × 27ft
Displacement:	4400/4900 tons
shp/speed:	15,000/30kts
Complement:	116
Armament:	Five-21in tubes

Pump jet configuration. An accelerating duct will favour efficiency and a decelerating one will be quieter. The choice between pre- and post-swirl will usually be decided by ease of maintenance; the rotor and shaft is more easily withdrawn in a preswirl duct. (RINA)

There was rapid inflation while this class were building and hence the quoted cost rose from £37.1 million (Swiftsure) to £97 million (Splendid). Running costs are £3.8 million per year at 1976 prices.

The main sonar was dropped to the chin position, which necessitated moving the torpedo tubes aft and angling them, leaving room only for five, more than compensated for by very rapid reloading arrangements. Both noise performance and reliability exceeded expectations. Noise performance derived from continuous attention to detail and a vigorous de-bugging campaign of noisy items found during Swiftsure’s trials, whilst reliability derived from simplification and further attention to detail.

Trafalgar Class

There were a large number of design studies for the next class known as SSNOX, SSNOY and SSNOZ. These were splendid designs but far too expensive, and it was agreed that the Trafalgars should be ‘Improved Swiftsures’. SSN-13, Trafalgar, was launched in July 1981. Work had been started on Core Z in 1968 and completed in 1971. DSMP was due for a major overhaul and Core Z only began tests in 1974. Foulger was project manager from early design until near completion, when Arthur Cook took over.

They had the same machinery as the Swiftsure and much the same hull, lengthened slightly to get in extra equipment with a slight speed loss. The Swiftsure machinery had proved generally reliable in service and attention was concentrated on noise reduction. Reduction of machinery generated noise had reached a practical limit in most areas, as each further reduction of a decibel was costing more and more to achieve. Attention was therefore focused on attenuation and damping – the Trafalgars were the first to be designed for hull damping tiles. The only feature of the Swiftsures to give problems in service was the raft and this was redesigned.

ADEB was stretched to the limit and inconveniently situated, far from the building berths, so a new facility was created, the Submarine Machinery Installation Test Establishment (SMITE). It was finished in time for production tests for the last Swiftsure (SSN-12) and then used for later boats. Full-scale wooden mock-ups had been built for the machinery of earlier classes. These were time-consuming to build and even more to alter, and very expensive, so for Trafalgar it was decided to rely on a one-fifth scale model.⁴⁸ This could be scanned using a travelling telescope, passing information direct to the computer. This could then instruct bending machines making pipes and, as a result, Trafalgar had some 5000 shop-made pipes as opposed to under 500 for Swiftsure. The remainder were made the old fashioned and time consuming way by bending a wire to fit on the ship and taking it back to the shop as a template.

The ‘improvements’ relate mainly to noise reduction, seen as a continuing process in which quieter items were fitted in successive boats as they became available. They were the first boats to have noise reduction coatings (anechoic tiles) fitted on build, though most earlier boats have since been treated. It has been said that the Trafalgars are even quieter than an Oberon on electric drive. They were equipped to fire Sub Harpoon SSM.

The layout of the Swiftsure class. Other classes differed little. (PRO DEFE 24/238)

No mention has been made of sonar fits, electronic warfare sets or command systems, but it can be understood that all these were greatly improved from class to class.

Vanguard Class SSBN

In 1980 the government announced plans to purchase the US Trident missile system for installation in four new ballistic missile submarines. It was originally intended to buy the C4 missile but this had gone out of production and maintenance could not be guaranteed. In consequence, the D5 missile was purchased with eight independent reentry vehicles (MIRV) on each missile. The missile compartment is the same diameter as that of the USN Ohio, though shorter, since the British boats carry 16 missiles instead of the US 24. The missiles are maintained in the USA at King’s Bay, Georgia.

The bow and stern sections were developed from Trafalgar, though with many changes. The biggest change was the fitting of an entirely new reactor, PWR2, which had been on trial at Dounreay and represented a significant advance on the Trafalgar class.⁴⁹Vanguard had an early core which will need one change during the boat’s life; later cores will last the whole life of the ship. Forward, a new sonar suite⁵⁰ was developed to meet the essentially covert nature of the deterrent role. Advanced noise reduction measures were also embodied to ensure maximum stealth. A great deal of effort was also put in to easing the maintenance task and so maximising the operational availability of the deterrent force.

These are by far the largest submarines to serve in the RN and were built by Vickers at Barrow in an enormous assembly shop over the old Devonshire Dock, allowing a complete submarine, together with large sections of the follow-on vessels, to be assembled under cover. Conventional launching was replaced by a ‘roll-out’ procedure in which the complete submarine was transported on a system of synchronised rollers onto a ship lift platform (synchro-lift) which was then lowered into the water. A similar lift is installed at the Faslane base. They were ordered from 1986 to 1992. Vanguard completed in 1992 and entered service in 1994 after trials. (A note on the Astute class and later can be found in Chapter 14.)

A large replica section of a submarine was used to test the resistance of various equipments to shock from explosions. Those crossing the Forth bridges (in the background) were often surprised by the sight of these blasts. (D K Brown collection)

¹ D K Brown, A Century of Naval Construction (London 1983).

² R J Daniel, ‘The Royal Navy and Nuclear Power’, Trans INA Vol 90 (1948). Daniel had visited Hiroshima soon after the Japanese surrender and had attended the Bikini trials.

³ The word ‘about’ can be inserted in front of all performance figures in this chapter.

⁴ I am grateful to Keith Foulger for his contribution to this chapter.

⁵ Vice-Admiral Sir Ted Horlick, ‘Submarine Propulsion in the Royal Navy’, Trans I Mech E (1982) (54th Thomas Lowe Gray Lecture).

⁶ Reactor functioning but not delivering power.

⁷ Assisted by G H Fuller and D Henry.

⁸ The US team were suspicious of espionage (see A Century of Naval Construction) but this was not so.

⁹ The mock-up was started at Dounreay and transferred to Southampton by Vickers.

¹⁰ Austenitic refers to the crystalline structure of the steel. Steel alloys can have several atomic arrangements, with face-centred cubic form in austenitic. When alloyed with about 17 per cent chromium it gives a very corrosion-resistant material.

¹¹ See Horlick, Note 5 above, for a technical discussion.

¹² Greenwich had a zero-power training reactor, Jason, in the only reactor building designed by Sir Christopher Wren! At that time the local Council proclaimed Greenwich to be a ‘nuclear free zone’.

¹³ The author carried out the leak test to a much higher standard than Dreadnought.

¹⁴ Horlick, Note 5 above.

¹⁵ Horlick refers to some failures in the secondary (conventional steam) machinery.

¹⁶ My memory is that Upanatom was a front runner! ADM 1/26779 (PRO) gives this and many others.

¹⁷ See Warship 1995 for a biography of this remarkable naval constructor.

¹⁸ See A Century of Naval Construction for management aspects. Baker introduced many then novel procedures in management.

¹⁹ We did alter the transition pieces where the pressure reduced in diameter as Rydill’s work showed a serious risk of fatigue cracking.

²⁰ There was a sign on the bulkhead between the fore and aft sections which read ‘Checkpoint Charlie – You are now entering the American Zone’, mimicking a famous landmark of Cold War Berlin. Organisation of spares support from two hemispheres was a nightmare – I chaired the spares committee.

²¹ With Vickers Armstrongs and Foster Wheeler.

²² Rydill got it right on fatigue.

²³ It also permitted a smaller finall design is a compromise!

²⁴ Large diameter, slow-running propellers are not necessarily quiet but are a necessary step to silence.

²⁵ The author was working on propeller design at Haslar during this investigation. It was said that you became a true propeller designer when you ceased to think of a propeller driving a ship and started to think of the ship as a mere obstruction to the flow into the propeller (this took about 6 months).

²⁶ There was even a four-shaft study with a propeller in each quadrant!

²⁷ R P Largess & H S Horwitz, ‘Albacore – The Shape Of The Future’, Warship 1991.

²⁸ D K Brown, ‘Sir Rowland Baker’, Warship 1995.

²⁹ Dreadnought’s officers and crew were also outstanding, many of the wardroom making flag rank while several ratings took degrees.

³⁰ Even so, uss Nautilus took 6 months less.

³¹ At first I would be on the quayside to meet her after each voyage and sort out the problems. I gave this up when the only complaint was that the wardroom toaster was burning the paint!

³² Particulars in this chapter are from published sources.

³³ I was actually asked to design a mount for a 4in gun – for shore bombardment!

³⁴ The original name proposed for this submarine was Inflexible, which drew an immediate protest from the E-in-C that a fortune had been spent on flexible mountings and the name was quite unsuitable!

³⁵ Brokensha was the Principal Ship Overseer at Barrow for both Dreadnought and Valiant.

³⁶ For details of the complicated negotiations and organisation see P Nailor, The Nassau Connection (London 1988).

³⁷ Baker asked Flag Officer, Submarines if those involved in DPT could wear the RN Submariners tie. The reply was ‘No, you’re a lousy lot of b*******s and I’ve added the bend sinister!’ (in British heraldry the bend sinister denotes an illegitimate connection).

³⁸ The USN class name is Lafayette but UK papers always spoke of George Washington.

³⁹ This may not sound much but a nuclear submarine is so tightly packed that a 3in reduction is difficult.

⁴⁰ This and much of the section comes from the submission paper for the sketch design now in the PRO as DEFE 24/90.

⁴¹ Five were originally planned but one was cancelled by the new government. This meant that very elaborate maintenance, refit and stores support organisations were needed.

⁴² D K Brown, A Century of Naval Construction (London 1983).

⁴³ An empty space at the aft end of the machinery, needed for buoyancy, was often known as ‘Hancock’s hole’.

⁴⁴ A future head of the RCNC.

⁴⁵ P L Vosper & A J Brown, ‘Pumpjet Propulsion- A Splendid British Achievement’, lecture to RINA (Western Branch) 1996.

⁴⁶ Horlick: ‘what you don’t fit can’t give you trouble’.

⁴⁷ A ‘Battle’ class destroyer boiler provided steam.

⁴⁸ Technology advances very quickly and the use of a physical model was soon superseded by computer modelling.

⁴⁹ Much of this section has been contributed by Brian Wall, Project Director, Trident submarines.

⁵⁰ British sonar suite 2054.