Introduction

Economics and Political Commitments

The political and economic background to the Royal Navy over half a century is far too complicated to cover in part of a chapter, let alone when linked to the events of the Cold War.¹ However, many key events greatly affected the size and shape of the Navy and a brief note seems necessary. In 1945 the RN was the second biggest navy in the world with the navy in third place a very long way behind. On the other hand, most of its ships were of pre-war design, obsolescent and worn out from years of wartime steaming.

The first priority was to demobilise the wartime conscripts, whose departure revealed vast gaps in key technical areas, particularly those in the new field of electronic warfare. It proved increasingly difficult to recruit long-service volunteers, whilst conscripts took so long to train that it was doubtful if they were of overall value. Initially the Admiralty hoped for a large fleet, with a distribution very similar to pre-war days. The battleship was still seen as essential, and it was even hoped to complete at least two new ones. A large force of aircraft carriers was also required, but British naval aircraft were inferior and US planes could not be afforded.

The county’s economy was in dire straits and there was no money for a big navy. Steel was in short supply, so there was great pressure to scrap as many ships as possible, and certainly not to build more than a very few new ones, while ships already started were delayed again and again. The civilian labour force was needed to earn foreign exchange and could not be used on warship building.

Creators of the early postwar fleet. Directors of Naval Construction: Sir Stanley Goodall (second left) 1936–44; Sir Charles Lillicrap (second right) 1944–51; Sir Victor Shepheard (left) 1952–58; Director General Ships, Sir Alfred Sims (right) 1958–67.

(D K Brown collection)

The Admiralty appears to have thought that these were short-term problems and was slow to recognise economic reality. It was not fully appreciated that the equipment, attitudes and management skills of the UK’s industrial base were out of date. On the other hand, Britain was still responsible for the security of large areas of the world, the remains of empire. The potential threat from the Soviet Union became increasingly apparent, whilst minor wars and police actions stretched the Navy to the limit.

In consequence, there was a series of Defence Reviews aimed at reconciling commitments with available funding, which always ended in a smaller navy. Increasingly, the Defence budget was treated as a single entity, leading to bitter inter-service feuding. Sudden increases in the temperature of the Cold War, such as the Korean War, led to additional funding and new building programmes.²

Defence Reviews

In the context of this book it is possible only to give a brief outline of the way in which successive Defence Reviews affected the size and shape of the Navy and the design of specific classes. In 1945 the UK economy was in such a parlous state that it was almost impossible to think of new construction for the next five years. The Royal Navy was over-endowed with ships, though many, particularly the bigger ships, were of obsolescent design, and even the more modern escorts were of limited capability against the new, fast submarines. The new cruiser was ultimately abandoned, and the carriers of the late war programmes proceeded slowly, if at all.

The Government and the Admiralty Board gave some priority to research, since its demand from industry was small and the need for counters to the new threats was clear. The most conspicuous outcome was the Sea Slug surface-to-air missile. Less obvious was the work that went into the reduction of underwater noise from machinery, flow and propeller cavitation. This was paralleled by the development of new sonars – very large by the standards of the day – and ASW weapons. The design of prototype frigates to operate these systems proceeded slowly, and took into account the lessons from the Ship Target Trials programme.

The outbreak of war in Korea in 1950 provided the stimulus for rearmament with a modest frigate programme and a very large MCMV programme. However, by the mid-1950s there were further financial problems, which led to the ‘Radical Review’, and post-Suez problems, which ended the guided missile cruiser programme and caused several frigates to be cancelled. On the other hand, the destroyer programme and that of the Sea Slug came together in the ‘County’ class. Limited resources were made available to start the nuclear submarine programme.

This aerial view of Cleopatra in 1967 shows the lines of the Leander class frigate to advantage. The class was designed by Ken Purvis.

(MoD)

The next major review came in 1965/66 with the cancellation of the carrier CVA-01, all but one of the Type 82 destroyers, the fifth SSBN, and the Type 19 cheap frigate. The aircraft carrier had been progressed by the Admiralty, in the teeth of Treasury and RAF opposition, and with only lukewarm government support. Any such support died when the Labour Party came to power. Dennis Healey, the new Minister of Defence, rightly pointed out that he would not approve one carrier and that a case had to be made for the three in the long-term plan. This review led to a total rethink of the structure of the RN without the traditional fleet carrier. The ‘East of Suez’ role was to be abandoned, except for occasional cruises. On the design side a ‘Future Fleet Working Party’ was set up, which proposed a number of options, most of which led to actual ships.

The new fleet had three ‘Through Deck Cruisers’ (the Invincible class) as its core and, once RAF opposition had been overcome and Sea Harrier VSTOL aircraft were embarked, the fleet had some effective air defence. The Type 42 destroyer was smaller and cheaper than the Type 82 and could be built in some numbers. This was supported by the cheap Type 21 frigate until the more capable Type 22 AS frigate entered service. A considerable number of ‘Hunt’ class MCMVs were built. All these classes gave excellent service and were a credit to the Working Party. Further reviews came in 1974 and 1976 because of the effects of raging inflation on the economy. Building programmes were reduced, but the only new design to be cancelled was a commando carrier to replace Bulwark and Hermes. Little work had been done on the design and the cancellation did not cause the upset of the 1966 cuts.

Defence Minister John Nott’s review of 1981 cancelled the Type 44 destroyer with Mark II Sea Dart;³ it was also intended to scrap the amphibious assault ships Fearless and Intrepid and to sell the carrier Invincible to Australia.⁴ However, most of the cuts were cancelled on the outbreak of the Falklands War in 1982 and the Type 23 frigate was enhanced – it was allowed to have a main engine on both shafts! The Trident SSBN programme of four boats was not affected (see later discussion on ‘Weapon-Platform Ratio’).

The end of the Cold War in the 1990s brought a succession of reviews aimed at reducing defence spending – the so-called ‘Peace Dividend’. As a result there were fewer orders for new ships and older ships which still had many years of effective life were sold or scrapped. The most serious loss of capability was the abandonment of the dieselelectric submarine, with the four Upholder class leased to Canada and follow-on boats cancelled.

However, at the time of writing (2002) the future seems bright, with the commando carrier Ocean in service and the landing ships Albion and Bulwark nearing completion; other amphibious force ships are in hand. Two new aircraft carriers are being designed to operate the Joint Strike Fighter and six Type 45 destroyers have been ordered. Three Astute class attack submarines are in hand.

Other Factors

The atom bomb that brought an end to the Second World War, and the post-war Bikini trials, showed that fundamental changes were needed in the way that the sea war was fought and ships were designed. The later test of a hydrogen bomb brought a sudden end for schemes of shipbuilding after a war had broken out and to the large reserve fleet, as no-one believed that there would be long wars any more.

The NATO alliance recognised that the UK could not meet a major threat alone but needed allies. It introduced the need for inter-operability, and also led to a number of information exchange projects (IEP) in which ideas were traded on aspects of ship design (and other topics), and some joint research and production programmes were started.

Cost, Inflation and Budgets

During the period covered in this book the value of Sterling fell and fell, making comparisons very difficult both of the cost of individual ships and of the total value of the ‘Navy Estimates’.⁵ For nearly 20 years, 1956 to 1972, the cost of the Type 12 frigates (Whitby, Rothesay and Leander) formed a convenient reference point. The technical content did increase between classes, particularly with the introduction of the Leander class, but the cost, corrected to a common value of money (1984 £ in the table, right), was fairly constant.

Brilliant, one of the Batch I Type 22 AS frigates.

(Mike Lennon)

There was a slow cost reduction in later Leanders, probably due to the ‘Learning Curve’ (see Chapter 10) which may be identified in those ships built at a single shipyard. Yarrows was the only yard which built in sufficient numbers for the learning effect to be recognised, and its effect was reduced because ships were ordered in ones and twos rather than in bulk. The full benefit of the ‘Learning Curve’ is seen in the Vosper-built ‘Hunts’ (Chapter 10) where the man-hours required for the last ship were half that of the first.

Pugh has suggested that the real cost of military artefacts rises at about 7 per cent per annum above inflation.⁶ This effect is not obvious in the table above but then the technology changed little over the years concerned – confirming the view that the Type 12s were built over far too long a period. They needed large crews and their equipment was ageing, while the naval staff failed to recognise that a true successor would be bigger and more expensive. There was a marked reduction in the cost of the Type 23 frigates (or at least a reduction in the expected increase) which coincided with a much more aggressive approach to competition. Whilst competition played a part, it is probable that the introduction of building very large modules, outfitted under cover, was what made low tenders possible.

Weapon-Platform Ratio

The Nott Defence Review made great play with the so-called ‘Weapon-Platform Ratio’. In its crude form, as used by Nott, the cost of weapons was taken as bills paid by DG Weapons (Ensleigh) whilst that of the platform was the amount paid by DG Ships (Foxhill). The diagram, prepared by Peter Chamberlain, then head of forward design, shows the fallacy of this approach. The top bar shows the Nott breakdown (with items not included in either heading).

However, weapon and sensor crews are part of the weapon system and their accommodation, stores, etc are part of the weapon capability. The second bar shows this breakdown, painting a very different picture of a frigate’s capability. The third bar goes further and gives a truly functional breakdown of cost, showing that a frigate is a truly cost-effective mobile fighting machine.

The make-up of cost can be shown in various, equally valid, ways. This diagram shows how the money is spent with about 50 per cent of the production cost going to the shipbuilder for labour, materials, overheads and profit.

Navy Estimates

Again due to the changing value of the pound, Defence Estimates and warship building in particular showed massive numerical increases. The reality is that when formed NATO was promised 70 frigates and today the figure is ‘about 30’. True, today’s ships are far more capable, but numbers do matter. However, the number in active service has declined to a much smaller extent. This is only partly due to the abolition of the large reserve fleet; improved corrosion protection (see Chapter 13) has meant a great reduction in time spent replacing rusted plates, so a much smaller part of a ship’s life is spent in dockyard hands. The introduction of gas turbines, with ‘Repair by Replacement’, further reduced the time in refit. Personnel numbers shrank from 144,000 in 1949 to 70,000 in 1985, only partly reflecting the much smaller crews needed in automated ships with gas turbine propulsion. It is hard enough to recruit even to this lower level.

The so-called Weapon/Platform Ratio and functional breakdown of warship costs.

The make-up of cost showing that only about 50 per cent of total cost is attributed to the shipbuilder.

The conflict of quality versus quantity is an eternal one, but it has intensified in the years since the Second World War. No longer can designers seek the best, or even justify extra cost on the basis of greatly increased capability, but there is a fixed price and the problem is to get the best within that limit – all you want for £100 million!

Design Approvals Procedure

Warships are the most expensive single artefacts in the Defence Budget and it is essential that the Navy gets a capable ship at a price which the country – the taxpayer – can afford. Unlike almost all other defence equipment there is usually no prototype – 01 has to be operational after trials. Over the period covered in this book, approval procedures became steadily more elaborate, depending more on formal submissions and less on the subjective judgements of a few individuals, however well informed. These procedures changed frequently, perhaps too frequently, during the period and the titles of the committees involved changed even more frequently. However, the basic problems have not changed very much. The designer’s view is well represented by the game of Snakes and Ladders above.

A traditional representation of the reality of the design process.

(Redrawn by John Roberts)

Against a background of the Navy’s perceived role and an awareness of the likely extent of funding, ideas will be floated for new ships within the Staff. There will often be an input from technical departments on the introduction of new technology. Freelance designs were welcome;⁷though rarely adopted in toto, they quite frequently provoked a new line of thought – ‘Requirements pull, technology pushes’ – whilst the obsolescence of the existing fleet will be borne in mind. Politicians will wish to help industry and to provide employment.

Design Programme. This chart from the mid-1970s outlines the tasks in each phase and who does them.

Taking the ‘Tribal’ class frigates as an example of 1950s practice, formal design reviews were infrequent, and even those there were little more than rubber-stamp decisions already made informally. There were two formal reviews:⁸ the first at Sketch Design when the ship was basically defined, and the second when the design was ready to go out to tender. For both, the drawings and a statement of particulars would be laid out and approval would be given by the Board of Admiralty in a short minute. Since the Board included ministers, senior administrators (responsible to the Treasury for financial probity) as well as admirals, it was a more representative body than modern readers may imagine, although there were no engineers or scientists.

The informal contacts were frequent and invaluable. The most important was with the Staff (DTSD) – the ‘Tribals’ had a great advantage in being designed in Whitehall. Staff officers and constructors in Whitehall had offices very close together, and even quite substantial changes could be agreed in a few minutes of friendly conversation. Contact with the main design organisation in Bath was more formal and prone to misunderstanding.

A very detailed review of design procedures was given by Admiral Sir Lindsay Bryson, then Controller, in 1985, taking the Type 23 frigate as an example, and this will be taken as the end-point of this review, though they have changed again since.⁹ The requirement for a new design could arise from one or more of the following: a change in concept of operations, a need to replace old ships, a new threat, or the introduction of new technology, together with a window of opportunity for funding! Major weapon systems, and other systems such as communications that could be deployed in more than one class of ship, had a similar but not identical approval system. Note that the weapon system usually had several years’ lead over the ship programme.

The staff’s ideas resulting from these topics would be melded with the views of technical departments, leading to a series of design studies which might take from two weeks to a year, and might include unconventional craft such as hovercraft, SWATH, etc. These were considered by the Fleet Requirements Committee (FRC), who would decide which, if any, fitted best into the future role of the Navy and the resources available. This guidance permitted the Staff to start work on a formal Staff Target and Ship Department to develop their studies into a Concept Design.¹⁰ Note that weapons had a similar but separate approval system corresponding to the separation of Ship and Weapon Department on opposite sides of Bath.

There was an inherent problem in that the timescale for the development of a new weapon was very different from that of the ship that was to carry it. The diagram (left) shows that ship concept should start when the weapon is still in development, perhaps with a prototype on trial. It is useless to start the ship design earlier, as the weapon is sure to change, whilst if the ship design is left until later, it will go to sea with obsolescent weapons. This problem is exacerbated in collaborative projects and was a major factor in the demise of the NATO frigate programme.

Industry would be consulted at this stage on cost and export prospects, and a few study contracts might be placed on specific systems. The Staff Target and the Concept Design are as the chicken to the egg, and interact to a very considerable extent. During concept design risks and problem areas should be identified but not necessarily solved. Some quite major decisions are taken early, which may pre-empt other aspects. For example, it was decided at an early stage that the proposed Type 43 destroyer should have four SM-1A engines, which made a proposed increase in speed impossible.

The Concept Design and Staff Target then go together through a two-stage approval process. The Naval Projects Committee is a naval committee chaired by the Controller, and they will ensure the Navy is getting what it wants within the constraints of money, manpower, etc. When endorsed, the papers go to the Operational Requirements Committee, a tri-service organisation who will look at the proposals in terms of the national defence policy and resources.¹¹ At this stage very little effort or cost has been incurred.¹²Hence it is proper to consider some quite unlikely options, as it will be almost impossible to insert a new approach at a later stage. Many studies will be rejected at this point.¹³

The next design stage is Feasibility and this is concurrent with the development of the Staff Requirement. Feasibility takes the approved Concept Design and re-works it in greater detail; for example, weights and the positions of the centres of gravity are calculated in detail rather than scaled from previous designs, permitting a more accurate estimate of stability.¹⁴ Some structural design work may take place and layout will be developed. Overall, the object is reduction of risk; there should be no remaining technical problems and cost estimates should be accurate. Approval of the Staff Requirement and Feasibility design go through the NPC and ORC as before but, as the following design stages involve considerable spending and committal of in house resources, the Defence Equipment Policy Committee (DEPC) must also agree. The DEPC is concerned with the overall Defence Budget, its impact on national industry, collaboration and sales prospects. For simple ships, a contract may be placed at this stage. Feasibility design involves a considerable use of design effort and cancellation should be very rare at this stage.¹⁵ Approval to go ahead with the procurement of long-lead items may be given at this point.

The phasing of design: weapons and ship. The weapons programme must usually start some years ahead of the ship programme.

The procurement cycle around the middle of the 1970s, with approval stages.

Ship design proper can now begin. This may be divided into two phases, with Admiralty approval at the end of the first. When the design is complete, approval of the Minister and then of the Treasury is required before negotiations on a building contract can begin. The Treasury can take an active role supporting or opposing a particular proposal.¹⁶ The lead shipbuilder would probably be selected and they would participate fully in the later stages of design.

The approval procedures are long and complicated, but it is not easy to see how such important decisions that impact so much on the future Navy and on the national budget can be much simplified, particularly with some twenty staff and technical organisations involved. The time taken can be reduced by frequent contact at working level to ensure that the facts are clearly understood. Problems have arisen when junior staff have not ensured that their seniors know what they have agreed to. If the design is being carried out within the ministry – ‘in-house’ – work can continue during the approval process at each phase. Design by contract will usually come to a halt for many months while collaborative projects are delayed even more.¹⁷

The Procurement Cycle diagram on the previous page attempts to show how the technical task interacts with the staff, shipbuilder and minister, whilst the smaller diagram attempts to put a timescale on the process and re-emphasise the interaction with weapon programmes. Though the committee names (and initials) differ, the principles did not change greatly over most of this era – and ship projects were generally close to programme and budget.

Ship and Weapon Design Co-ordination Group

Ship and weapon design approvals took similar but separate routes. The Ship and Weapon Design Co-Ordination Group (SWDCG – pronounced ‘Shwepcog’), was a semi-formal grouping, usually chaired by the Director of Warship Design, to ensure that the programmes were consistent in both objective and in timetable. The Group operated through working parties, one for each ship project, usually chaired by the head of preliminary design (Chief Constructor). The preliminary design section was allowed to freelance for a short time, but if they wanted to develop any study, approval of the SWDCG was needed at the next meeting. Concept designs were approved by the SWDCG before being sent to central committees.

The main problem was that one man could represent the ‘ship (including machinery and electrics)’ but weapon work was so specialised that numerous representatives were needed from both AUWE and ASWE. Though the working parties did not work by voting, it was usual to add one or two extra men from Ship Department to avoid a single representative being overwhelmed by weight of numbers. A design working party is essential to bring together all departments involved. It needs a strong leader and a clear objective if it is not to break down into factionalism, as enshrined in the engineer’s adage:

Northumberland, one of the numerous ‘Duke’ class Type 23 frigates.

(MoD)

The Irregular Verb ‘To Design’

I create

You interfere

He gets in the way

We co-operate

You obstruct

They conspire

Design Methods

The problem is eternal but the methods used to reach a solution have changed dramatically in the last half-century.¹⁸ The problem is to design a ship capable of fulfilling operational requirements – interpreted as a specified weapon fit and performance – with economy and safety. The translation of an operational requirement into a technical specification is sometimes referred to as the ‘Interpretation’ phase, and it is essential to get it right.

The first stage, concept studies, involves a series of tradeoffs between cost and capability. It is a time for lateral thinking: novel solutions can only be introduced at this stage. Later stages are convergent, eliminating the less attractive options. It will take some 10 years to get the first ship into service so that, with a 20-year life, the designer will have to look ahead at least 30 years – more if the class is built over several years.¹⁹

The benchmark will almost always be a conventional, propeller-driven displacement ship, but it will sometimes be necessary to consider alternative vehicles such as hovercraft. Such comparisons are difficult: the ‘Advanced Naval Vehicle’ (ANV) will usually have exceptional performance in one role whilst being inferior in other aspects.²⁰ Cost, too, is difficult, as the pattern of spending may be very different – the hovercraft MCMV would have been much cheaper to buy than a ‘Hunt’, but much more expensive to run, and would have required new training facilities.

Comparison of different conventional vessels is straightforward. One foot on the beam hardly affects the cost, so stability need hardly be thought about in most cases, whilst a weight of structure scaled from an existing ship should be sufficient to satisfy strength criteria.²¹ Mutually consistent values for Weight, Space and Layout must be found, leading to cost estimates of the right order overall but accurately reflecting the differences between variants. Weight of the hull and its equipment can usually be scaled quite accurately from previous ships. However, a new ship will usually have novel weapons and sensors that may not even exist and are unlikely to be beyond the prototype stage. An intelligent guess is needed for their weight.²²

‘Space’ is more difficult as it may refer to volume as in tanks, to area as in mess decks, or length for the upper deck. Upper deck length impinges directly on layout governed by physical and electronic clearances needed. The superstructure and mast arrangement of the ‘Tribals’ was constrained by the trajectory of the AS mortar projectile. Upper deck length governed most post-war designs. Most design sections had equations (today they would be called ‘mathematical models’) based on experience with their class of ship. For frigates there was a complicated formula involving armament weight, crew number, speed, endurance, etc, all leading up to a figure for deep displacement.²³ Speed and power could be reconciled by intuition backed by R E Froude’s magical Iso-K books, which codified the hydrodynamic results of 80 years of model testing.²⁴

The designer’s desk: a mock-up created for the RCNC Centenary exhibition at Greenwich. At bottom left is an integraph (see text) with a pile of foolscap workbooks beside. Further back is a set of ship curves for drawing and a drum slide rule (Fullers) with a scale 84ft long. Centre is an integrator (a baby one – most were more than twice this size). Near the designer’s arm is a lead weight used to hold a batten to form a curve; it was said that if you needed more than six weights, the curve had no place in a ship. Behind, a photo shows students at work on the 8ft drawing benches typical of a real design office.

(D K Brown collection)

All this was done by hand using only a slide rule and a hand-cranked calculating machine and if the study was of a conventional type, two assistant constructors could produce a reasonable answer in half a day. All calculations had to be recorded in a foolscap workbook, dated and indexed.²⁵ Figures could not be altered but, if necessary, the old figure could be crossed out and replaced with a new one in red ink. The next stage, originally ‘Sketch Design’ later ‘Feasibility’, went over the same ground but in more detail using direct calculation rather than scaling where possible. A general arrangement drawing would be produced based on the form of a selected type ship that might be quite different in size and style. The form of the ‘Tribals’ was based on heavy cruiser ‘Y’ of 1944, itself derived from the Glorious of the First World War. This general arrangement formed the basis for many weight calculations. A small panel of deck structure would be designed, using the best method available at the time, and the weight per square foot calculated. Measuring the number of square feet on the drawing produced the answer.²⁶

The ‘elements of form’ which make up part of the Iso-K books include the midship section, load waterline and the curve of areas of the type ship in non-dimensional form. The curve of areas shows the area up to the water-line of transverse sections along the length. This still leaves the designer of the new ship plenty of scope to make changes to the form. In particular, the shape of bow sections to improve seakeeping was (and is) a matter of debate. There were real difficulties as the requirement for section shape below the waterplane would often conflict with that for the shape above water. Then began the ‘fairing’ process in which transverse and horizontal sections (waterplanes) had to be adjusted to obtain a consistent shape. (One or two diagonal sections would often assist this process.) It would take a draughtsman about a working week to get what was wanted.²⁷ Model testing would usually improve a form derived in this way, reducing power requirements by some 3–5 per cent.

Hydrostatic curves showing to a scale of draught the displacement, height of the centre of buoyancy, height of the metacentre, etc were calculated using lengthy arithmetic methods, converting breadth at each point to area and then to volume. It was usually more simple to measure half breadths from the centreline to one side and multiply the result by two. Forgetting to multiply was a common error and at least one design office had ‘x 2’ painted in large letters on the end wall. Curves of stability at large angles of heel (righting levers – GZ) were produced using a mechanical integrator, a large triangular device made of brass, running on a rail. It would take about 20 minutes to trace the pointer round each section in turn up to the heeled waterline for each point and some 20–25 points were needed.²⁸

Next came the strength calculation, in which the ship was assumed to be floating head on to waves of length, crest to crest, equal to the ship and with a height of 1/20 length; firstly with wave crests at bow and stern (Sagging) and then with the crest amidships (Hogging). Getting the draughts right in these conditions involved more lengthy and repetitive arithmetic. The distribution of weight along the length was estimated, the ship being divided into some twenty sections for the purpose. A curve of load (Weight minus Buoyancy) would be drawn and then another ingenious machine, the integraph, would be used to calculate shearing force and bending moment. The integraph was a beautiful device – poetry in motion – with chromium-plated levers moving fast in all directions and pens drawing graphs in coloured ink.

A later structural design approach involved calculating the probability of occurrence of different loads in random seas over the life of the ship. This was hard to apply and undervalued the effect of the largest seas, which imposed the greatest loads. In turn, this was replaced by a return to standard waves of length equal to the length of the ship, but the height used for all classes was 8m. The figure of 8m was justified from many years of wave height recordings, world-wide.

The bending moment, together with a decision on acceptable stresses, enabled structural design to begin. The midship section came first (the only section for sketch design) and usually two more would be drawn and calculated in detail, changing scantlings until stresses and buckling strength were acceptable. These sections enabled a more accurate estimate to be made of the weight of structure and the position of its centre of gravity, both height and longitudinally.

The sketch design (Feasibility) accurately expressed the design intention but the level of detail was insufficient to seek tenders for building. The whole process would begin again. There would be a much more detailed general arrangement with most individual spaces laid out. More accurate weight estimates of machinery and weapon systems should be available, while weighed weights from ships of earlier design but still under construction would be used to improve the estimates of weight for the so-called ‘judgement items’ such as electric cables and piping systems, which were always difficult. There would be a number of detailed calculations, including propeller design, shaft bracket strength, rudder and pintle strength, masts, etc.

The detailed design stage would end with the production of the building drawings, specification and book of calculations. The building drawings, 6–8ft long, comprised the sheer draught, inboard profile, plans of each deck, a ‘sketch of rig’ (outline profile showing masts, yards and aerials), and structural sections.²⁹ The specification was a thick book laying down the size of each item of structure, where it was to go and how it was to be painted. The equipment to go in each compartment was listed and so on. By about 1960 this had been replaced by the ‘General Hull Specification’ which, with a few exceptions, would apply to all classes. The book of calculations would summarise calculations of weights, stability, strength, etc made by two calculators working independently and checked and signed by the constructor and chief constructor.

All these drawings and documents would then be laid out for formal approval by other departments. They had to sign a special book that they agreed (subject to very minor changes at most). Finally the day came for the Director of Naval Construction to sign the drawings. This was a near-religious rite and was taken very seriously. At this point the Director accepted personal responsibility for the design.

Early Computers

Early computers were hard to program and the programs were difficult to use. In consequence, only very lengthy calculations, used often, went on the machine (work marked as lengthy arithmetic in earlier passages). The input of data was often difficult and as each program was ‘stand alone’, data input had to be repeated. These problems were exacerbated by staff moves, generally every three years.³⁰ It was not only ‘reactionaries’ who wondered over the value of computers at first.

There were exceptions: for the first time it was possible to study complicated flooding incidents causing both heel and trim.³¹ The so-called ‘finite element analysis’ was developed in other areas of work and applied to ship structures. The whole ship would be broken down into a mesh of tiny elements and the reactions one on the other studied. This technique was particularly valuable in identifying and correcting stress concentrations. The preparation and input of data was very lengthy.³² More recently, the use of finite element analysis has been extended to the flow round the hull.³³

Computer Aided Ship Design

About 1965, Ian Yuille, a naval architect working in a Science grade post, proposed an integrated suite of design programs – Computer Aided Ship Design (CASD). There was considerable opposition to this proposal, mostly on the basis that it was merely a device to interpolate between existing designs and would discourage innovation. There is some truth in this argument, but it was all too common in the past to stick firmly to old approaches while if one wished to innovate, the CASD system was an invaluable helpmate.

The author and his chief strongly supported the approach but thought that it was wrong to separate the development from Ship Department. The key to the system was that the hull form should be developed within the computer and then its definition could be used by any other sub-routine such as stability. This proved a very difficult problem and the first solution was still very difficult to use. A later method mimicked the action of the draughtsman (B-spline) and was fairly simple to use. It was 1988 before the system called ‘Goddess’ was ready to go but it worked from the start.³⁴ It has been updated and remains a world leader, proven by the number of requests to explain its workings at computer meetings. It was a little cumbersome to use in concept studies and a very simple system, CONDES, was devised to run on a desktop machine in some eight weeks’ intensive work led by Dennis Pattison and Simon Rusling.³⁵

Goddess was not just a quicker way of carrying out the design calculations. It enabled a large number of calculations to be performed that had previously been the subject of informed judgement, like flooding of large extent, buckling of large areas of stiffened plating, and many others. It could also produce detailed drawings, fully lettered.³⁶

A frigate on the computer screen of GODDESS, the modern design office.

The operator is Doug Pattison, now Director of Naval Architecture.

(D K Brown collection)

Forms to suit. There is no one hull form appropriate to all ships. Each of these is ‘right’, reading clockwise from right, for minesweeper, frigate, destroyer, fast patrol boat, and landing craft. Note that they are not all to the same scale.

(D K Brown collection)

Conclusions

A major warship is the most complex manmade artefact and the most expensive item in the Defence Estimates, supported on a turbulent sea rather than on solid foundations. There is normally no prototype – 01 has to operate at the conclusion of trials – and the life from concept may approach 50 years. No wonder that great care is needed to get both the requirement and the design right.

¹ See works by Eric Grove, Desmond Wettern, Norman Friedman and others listed in the Bibliography.

² Hurried programmes led to problems for the overloaded design staff and hence to errors in almost all classes. See, for example, the ‘Ham’ and ‘Ford’ classes in Chapter 10.

³ Most of the features of the Mark II Sea Dart were eventually retrofitted in the Mark I.

⁴ The main corridor of design headquarters was decorated with ships’ badges. The day after the Nott review was published, Invincible’s badge had a ‘SOLD’ label over it.

⁵ See Appendix 1.

⁶ P Pugh, Cost of Seapower (London 1986).

⁷ No longer true.

⁸ There was a less formal review when the design study agreed for development was selected. I think this involved Controller and ACNS.

⁹ Admiral Sir Lindsay Bryson, ‘The Procurement of a Warship’, Trans RINA (1985), p21. This was a very brave paper, read at the height of the ‘Short Fat Frigate’ controversy.

¹⁰ Note that UK terminology is used here. In the USA the meanings of Concept and Feasibility are reversed.

¹¹ Attempts would be made to schedule major programmes for each service so that peak spending did not occur at the same time. ‘Slippage’ made this difficult.

¹² At this stage there will only be a single option, perhaps with a few minor points left open. Unusually, the OPV had two options, which led to the ‘Castle’ class and to the hydrofoil Speedy.

¹³ Many good engineers hate to see their work discarded whilst others enjoy the exploration of novel schemes.

¹⁴ The author has doubts, see ‘Design Methods’ below.

¹⁵ The Common Hull Frigate, designed for mass production in wartime, was cancelled very shortly after the explosion of the first hydrogen bomb – a correct decsion, since the bomb rendered a long conventional war highly unlikely.

¹⁶ The Treasury tends to have the brightest administrators and submissions need to be written with great care.

¹⁷ I am sure this could be overcome but, so far, it has not been.

¹⁸ For a more detailed history of older design methods see D K Brown, ‘British Warship Design Methods 1860–1905’, Warship International 1/1995. Methods changed little between 1905 and the computer age.

¹⁹ It was always said that the principal design tool was the Admiralty Pattern Crystal Ball, Mk I, with lucky pin attachment.

²⁰ The OPV was remarkable in that several very different displacement ships were considered, together with hydrofoils (three types), hovercraft and airships.

²¹ Hull weight varies as LxBxD for most ships – L^1.3xBxD is slightly more accurate. L⁴B/D is often quoted but this applies only to fully stressed, longitudinal material, less than 25 per cent of the total.

²² My ‘guesstimate’ for the Type 965 radar aerial on the ‘Tribal’ was 2½ times the quoted weight, which proved right. Since it was at the top of a tall mast, this was very important.

²³ It was about 50 tons out in 2250 tons for the early studies of the ‘Tribals’.

²⁴ D K Brown, Warrior to Dreadnought (London 1997).

²⁵ Many have been preserved in the National Maritime Museum. They make it clear that obedience to the rules on index, date, corrections etc, were rarely observed.

²⁶ I was never very happy with this approach and believe scaling was actually more accurate than an approximate calculation, which, all too often, omitted something.

²⁷ All drawing work was supposed to be carried out by draughtsmen and they strongly resented graduates (assistant constructors) doing such work. On the other hand, the graduates had the hydrodynamic background needed to shape the form. There was usually a compromise in which the graduates drew for about one day producing a ‘guidance sketch’ after which draughtsmen did the fairing. The converse came when I was head of preliminary design and passed out a note giving the form parameters I wanted for the ‘Castle’ class. A few minutes later, the Senior Draughtsman came to see me – none of the draughtsmen had worked on a Sheer Draught since apprenticeship and would I tell them how to start.

²⁸ The work was eased (slightly) by using a special body plan with sections at a special, non-uniform spacing – Tchebycheff spacing – which eliminated one stage of arithmetic.

²⁹ John Roberts, British Warships of the Second World War (London 2000), includes reproductions of many such drawings.

³⁰ At least three times I completed a course on a particular machine only to be moved before I could apply my skill. Radiation-shielding calculations for nuclear submarines were the first use of computers in ship design (see Chapter 9).

³¹ I did such calculations for the ‘Tribal’ class, but that was a very simple case – even so, it took about three months – and time was available because of delays in the machinery programme.

³² I still feel that this technique is used too often in simple cases which can be solved by judgement – or even worse, used in place of judgement.

³³ Sometimes referred to as CFD – Computational Fluid Dynamics

³⁴ Government Defence Design System for Ships. The logo is the head of the Goddess of Wisdom, Minerva.

³⁵ This system illustrated the problem associated with the meaning of ‘Space’, which it read as Volume. In consequence, if there was a lack of space, it would increase depth to get more volume when it was deck length or area which was needed.

³⁶ One young man having shown that a set of deck plans and profile for a frigate could be produced in 20 minutes said: ‘And what’s the use? Whitehall will still take 20 years to make up its mind.’