Almost inevitably, their every attempt at meeting the requirements seemed to drive up size.

Initial CVF studies in 1999 indicated that the smallest practical design for a STOVL carrier able to support a permanent air group of 40 aircraft, and temporarily an extra 10 aircraft, would be about 38,000 tonnes.  Early published artist's impressions for the STOVL variant of the new carrier showed a ship of about 274 metres (900 ft) length.  However, both the catapult launched (CTOL) and the Short Take Off But Arrested Recovery (STOBAR) variants would require a larger conventional carrier design equipped with an angled flight deck and arrested wires for landing.  The smallest practical design for a STOBAR or CTOL carrier able to support the required size of airgroup was determined by the DPA to be about 43-46,000 tonnes.  BAE Systems stated that a CTOL or STOVL design would have to be about 10,000 tonnes larger than a STOVL design.  Both the DPA concepts published in 1999 and the early BAE concepts published in 2000 showed CTOL and STOBAR CVF designs significantly larger than the STOVL designs.  

From about this stage the CVF design began to very much reflect the strong influence of Admiral Sir Michael Boyce, then First Sea Lord and later Chief of the Defence Staff, who's maxim is "Air is free, and steel is cheap". 

It was recognised that size is not directly related to costs - the Royal Navy demonstrated (and not just to its own satisfaction!) that a large ship is not only cheaper to build in terms of cost per tonne but also has lower maintenance costs.

There were soon four principle size drivers influencing the CVF design. First, the sortie generation is clearly related to the number of aircraft and therefore the size of the hangar, emissions, and aviation support.  Second, there is the provisions of the ‘for but not with’ capability, for example allowing for the fitting catapults and arrester gears determines the length. Third, there is the overall complement required to be put in the ship, particularly given the more generous spaces for sleeping and recreation now required for all rates and ranks. Finally, there is the tankage volume necessary for unrefuelled range at the cruising speed.  There are others, but those are the most important points.

The large size also provides the greatest possible deck area, which not only facilitates aircraft handling but also allows the incorporation of large magazines and fuel stowage.  Magazines will in fact have a significant impact upon ship size and costs.   However the larger hull sizes will create some operational and support problems, for example a lack of suitable dry docks, which could be expensive to resolve. 

Certainly a large ship would ensure that (unlike the Invincible's) that there's plenty of room for mid-life update and expansion, and in an emergency the carrier could carry a much larger "surge" airgroup than normal. 

The STOBAR CVF design was eliminated in January 2001, at the end of Assessment Phase Stage 1, but the two competing carrier contractors (BAE Systems and Thales Naval) continued to develop designs based on operating STOVL or CTOL aircraft.

Numerous magazine and newspaper reports in 2001 and 2001 suggested that CVF would displace at least 50,000 tonnes (which was also mentioned by the Head of the DPA in May 2001), increasing to perhaps 60,000 tonnes for a CTOL design. 

French Navy and DGA officials after discussions and exchanges with their UK counterparts in early 2001 suggested that the planned weight for an empty CVF was now slightly over 55,000 tonnes, which would imply a full load displacement of about 70,000 tonnes!  While this was going too far, it perhaps indicates some of the extremely large designs under consideration at that point in time. 

In February 2002,  Jonathan Rich, then the CVF Programme Requirements Manager told a Naval Air Warfare conference that studies by Thales and BAE suggested that to deliver the desired fighting strength, CVF displacement needed to be between 52,000 and 58,000 tonnes when fully loaded with fuel and munitions. At the end of their intended 30-to-50-year lives, during which they would inevitably gain weight, the displacement would be up to 64,000 tonnes.  With lengths of 262 to 292 metres (860 to 958 feet), the ships would not much shorter than the 333-metre USN Nimitz Class carriers, and may be wider than the U.S. ships. 

BAE Systems identified in their assessment work that a key metric to the ships sortie rate (150 a day was initially the minimum requirement) was deck size, they found that a large clear flight deck reduces the constant reshuffling of aircraft and allows ancillary operations like AEW and ASW helo's to continue in their own deck areas without getting in the way of the JSF fighters.  This drove up the size (although not necessarily the displacement or cost in direct proportion) of the BAE CVF designs - they considered designs as long as 340 metres!  Perhaps unsurprisingly, BAE believed that they could comfortably exceed the current 110 (JCA only, about 130 total) sorties/day requirement. 

The BAE CTOL CV design dating to about May 2002 was 50,000 tonnes light, 64,000 tonnes full load, and just over 300m long.  The anticipated airgroup was 24 or 36 F-35's (CV variant), 4 E-2 Hawkeye’s, and 6 Merlin HM.1.  Aviation facilities included 2 steam cats with steam generators and  2 large deck edge elevators.  The alternative STOVL variant was 46,000 tonnes light and 58,000 tonnes full load.  No armament was carried except for ILMS/CIWS mounts of an undetermined model (e.g. Phalanx, Sea RAM, ...) plus a few manually operated light calibre guns.  In order to further maximise the available deck space, the island of the BAE CVF designs was regularly revised, the final STOVL variant incorporating a long low signature single island broken by an athwartship tunnel, (aka CVA-01 back in the 1960's), or perhaps the construction was more like two small islands connected by a flying bridge over a lift.

Thales announced in April 2002 that their CVF design was 290 metres long and would displace about 50,000 tonnes (presumably light load).  Emphasizing the importance of the ship/air interface, the design unusually had two "islands" for masts, bridge and control towers instead of the traditional single structure. 

Thales determined that the island must be large enough to house the bridge, flight-control centre and sensors such as radar.  "We found the best place to put the island was on the forward starboard quarter, similar to the French carrier [Charles de Gaulle]," says Simon Knight of BMT Defence Services. "So we put the bridge and the flight control there and we found immediately that we did not have enough room for the sensors.  ... We were also worried about where the forward starboard catapult would go and the space around that for manoeuvring aircraft preparing for take-off.  So then we looked at a second island."  Mr Knight may be being somewhat disingenuous here as the need for a second island (as an alternative to one very long one, aka the BAE Systems CVF design, and the existing Invincible's) is probably also driven by the trunking needs of the gas turbine prime movers. 

One of the many benefits of two islands (an idea first considered by the Royal Navy as far back as the 1950's) is that it improves resilience and damage control. 

In any warship design it is better not to put all the engines or sensors in one place, but widely separated.  But conversely it is also better to have the shortest possible length of ducting from the engine rooms directly up to stacks/funnels in the islands.  Widely spaced engine rooms and only a single island in the Thales design would have necessitated ducting the large uptakes along a tortuous route through the ship - wasting valuable internal volume and also slightly reducing power output.  For example the current small Invincible class carriers have an inconveniently narrowed width in the middle of the hanger where space has had to be made to accommodate the ducts. 

(Source: Thales CVF Team)

Adopting two islands overcame the problem of separating engine rooms and sensors.  It was claimed by Thales that the twin island design provides greatly enhanced operational survivability by allowing the physical separation of key ship (bridge - forward) and flying control (flyco - aft) functions in separate islands, while offering command and control redundancy for the ship as whole.  Additionally, air flow studies showed that the 2-island configuration improves the 'air wake' (wind-over-deck) environment for approaching aircraft.  The concept also separates some of the ships principal sensors, reducing interference affects.

Like BAE Systems, Thales decided that it was important to maximise the area of the flight deck, which they achieved by using exceptionally large side sponsons which provide a very wide flight deck - over 70 metres (230 ft).  Large sponsons relative to the hull size can cause ship stability and motion problems which restrict air operations in high sea states, thus to help reduce excessive ship roll motion damping devices such as large fin stabilisers are likely to be used.

The Thales design had two large deck lifts - one between the two large islands and one aft of the flyco position.  The large hanger deck is high enough to ensure that any aircraft likely to be embarked can be maintained anywhere, eliminating the need to move airframes around during maintenance (e.g. as had to be done on board the old HMS Ark Royal IV)

In 2000, BAE Systems spent private-venture funds during AP1 to develop a 'hybrid' CVF design.  Its concept envisaged a carrier configured with a waist catapult and arrestor gear for launch and recovery of CV-type aircraft, together with a ski-jump forward and a large recovery area for STOVL operations. Although not publicised, it is understood that Thales also briefed the MoD on a hybrid design concept.  While some in the MoD's equipment capability customer community welcomed the initiative, the DPA was markedly less enthusiastic about possible hybrid solutions. "The hybrid was not optimised for either STOVL or CV," said one industry source, "so we were told that we were answering a question that wasn't on the exam paper." 

Although the hybrid design was not formally being considered, its advantages were noted by the CVF IPT team, and it became apparent that the CTOL designs being proposed could be adapted to operate STOVL aircraft with minimum change (a fact that Thales were already emphasising by their use of a common hull) and could be modified if necessary with catapults and arrestor gear to operate a future generation of assisted launch and recovery aircraft.  It emerged during early 2002 that officials in the MoD's equipment capability community were seriously contemplating a so-called 'smart' carrier approach.  This would ensure that the ship design was sufficiently flexible to adapt to new aircraft, unmanned air vehicles and/or unmanned combat air vehicles entering service during the course of its 50-year service life.

Thus in early 2002 both BAE and Thales began edging towards flexible CTOL and STOVL designs with very similar characteristics - reflecting the MOD's increasing preference that the selected design could be converted from one to the other during build or a later refit - e.g. a STOVL design could be adopted that has an angled flight deck configuration and space and weight allowance to add arrestor wire engines and catapults at a later time.  Indeed Thales had been always stressed the commonality (circa 80%, compared with BAE Systems 65%) of their CTOL designs and STOVL hulls, and BAE also began  emphasising that their larger CTOL design could easily be adapted to a STOVL configuration and ordered as such, but the MOD would then have the option to convert it over to CTOL if the requirement appeared.

The final version of the BAE Systems CTOL design in 2002. Note the deck marking for HMS Eagle!
(Original source unknown)

At the time of the JCA selection decision in favour of the STOVL variant on 30 September 2002, a note to a MOD press release stated: "On current assumptions the two new carriers, known in MoD as the CVF project, are expected to displace between 60,000 and 66,000 tons. On this basis we expect these immensely powerful and impressive aircraft carriers will be the largest warships ever built in the UK, or indeed anywhere outside the USA and the Russian Federation. They are expected to cost no more than about £3 billion ($4.7 billion) to build. ... The two competing carrier contractors have each been developing alternative designs based on STOVL or CV aircraft. The CV designs could be adapted to operate STOVL aircraft with minimum change and could be modified if necessary with catapults and arrestor gear to operate a future generation of assisted launch and recovery aircraft. MoD has decided to take forward the CV design, modified to operate STOVL in the short to medium term, but with the flexibility to operate other types of aircraft after the F35 has left service."

The 'adaptable' CVF design will thus include provisions for the retrofit of catapults and arrestor gear at a later date.  Before the main landing deck is laid the necessary systems for steam-catapult launches and arrester-wire landings will be incorporated, so any later rebuild becomes easier and less costly.  The ski-jump will be of something of a "bolt on nature".  There will be voids where the cats and arrestor gear would be fitted, and although marked out for STOVL operations the flight deck form will effectively incorporate an angled deck - which could be easily modified in to a landing lane for arrested aircraft landings.  If the decision is eventually made to convert the carriers to CTOL type operations (including to support UAV's or UCAV's), the CVF's will return to a dockyard to have their 'top' taken off and remodelled during a major refit.  A new flight deck will be fitted with catapults, arrestor gear and angled flight deck all in place. The command centres will also need to be be altered, and appropriate provision for this will be incorporated into the initial build.

As a result of Secretary of State’s announcement on 30 September 2002, the baseline design used for planning and budgetary purposes changed from STOVL to a Carrier Variant (CV) based Adaptable Carrier design for the operation of STOVL JSF and rotary wing aircraft for MASC.

While the 'adaptable' decision will involve some increase in ship acquisition costs compared with selecting a purely STOVL CVF design, it will safeguard the investment in the two platforms (CVF and JCA).  It will also leave the UK well placed to adjust its programme in light of any changes to the USN's and USMC's JSF acquisition plans, or a disastrous technical problem or cost over-run associated with the F-35B.

The selected approach terminated the previous twin-track procurement strategy, whereby the CVF IPT has asked BAE Systems and Thales to develop CVF designs specially optimised for STOVL and CV air groups.  In effect, the decision to adopt a future proofed 'adaptable' design has de-coupled the carrier platform from the choice of JCA, resolving an obvious paradox: that a carrier optimised purely for fixed-wing STOVL and rotary-wing operations would clearly lack the flexibility "to be able to operate the largest possible range of aircraft in the widest range of roles". 

(Above) Thales CVF design, modelling and simulation activities in 2002/3

After the announcement Thales claimed that its design from the outset incorporated the flexibility to operate a mix of Short Take off and Vertical Landing (STOVL) aircraft along with helicopters and potentially Unmanned Aerial Vehicles.  In order to utilise the full operational potential of the CVF/STOVL JSF weapons system capability, Thales adopted a novel extended centre line runway option that will be used on occasions when it is necessary for JSF to carry increased payloads.  BAE also made changes to its CTOL design to produce an "Adaptable Carrier" variant.

According to MOD announcements in September 2002, basic CVF details included a length of about 290m long (951 ft), a maximum beam of about 75m (246 feet), a maximum draft of more than 31 feet, and displacement between 55,000-60,000 tonnes (presumably full load).  The ship would be able to accommodate up to 48 JCA, AgustaWestland Merlin ASW helicopters and maritime airborne surveillance and control aircraft.

Interestingly despite the official selection of the STOVL F-35 variant and an "adaptable" carrier platform, BAE Systems (and perhaps Thales) in October 2002 re-proposed a STOBAR configuration for the carrier, primarily for AEW reasons.   A STOBAR carrier would have a lower cost than a full CTOL configuration but would be able to operate a wider range of aircraft than pure STOVL.  Northrop Grumman were claiming that its E-2C/D Hawkeye 2000 could launch using a ski-jump launch, while the F-35C - which was increasingly seen as a likely choice for the manned element of the RAF's then planned Future Offensive Air System - could also operate from a STOBAR carrier. The MOD was not interested.  

By 20 November 2002 both consortia had submitted their final documents, concluding AP2. Both consortia agreed that the best solution was a mono-hull configuration and both were offering designs whose flight deck would be slightly shorter than that of the US Navy's Nimitz class (332m) but longer than that of the French Navy's Charles de Gaulle (261m).  Beyond this, however, the competitors' hull forms and approaches to construction are very different.  Thales (or rather its designers, BMT) had opted for the traditional format of a V-shaped hull of simple form, extended with box-shaped sponsons to provide extra hangar and flightdeck space, and with a distinctive two islands profile, the ship was 292m in length and displaced slightly more than 65,000 tons full load.  Meanwhile BAE Systems preferred a design (55,000 tonnes light, 64,000 tonnes deep) with a flared hull to provide extra internal space and to reduce the ship's radar cross section.  The flight deck of both designs were sized so as to permit simultaneous conventional take-off and recovery should catapults be installed at a later date.

By early December 2002 the CVF IPT had completed its evaluation of the proposals, marking the Thales design (which was in fact largely created by their design partner British Maritime Technology (BMT)) as being technically the more superior in several areas.  Some key differentiators appear to have been:

• Recognition of the operational and damage control advantages of Thales' twin island configuration.

• The two small islands also allowed the Thales design to have very widely separated engine rooms while wasting less internal volume on exhaust trunking to macks/funnels, and lose less valuable flight deck space than one very long island (despite BAE Systems efforts to recover some of the space with its lift and "flying bridge" idea).

• The potential design and build advantages of the full-electric all-pod propulsion system used by the Thales design compared with the less risky hybrid (shaft and pods) system of the BAE Systems concept.

• The two large high capacity (2 x F-35's) lifts of the Thales design were favoured over the three small (a single F-35) lifts of the BAE design, particularly as one the later emerged awkwardly in the midst of the island structure.

• The larger and higher hanger of the Thales design.

• Although the BAE design was more "stealthy" than the Thales design, this was not regarded as being of very great importance for an aircraft carrier of about 65,000 tonnes .

The fore and aft islands of CVF on model of CV displayed at Euronaval 2006 (Source: Thales)

The cost estimates of the two designs were very similar, and thus this did not become an over-riding deciding factor.  It was therefore not a total surprise to observers when on 30 January 2003 the MOD announced that the project would take forward and develop the "innovative" design put forward by Thales UK.  It was however more of surprise to learn that the project would be led by BAE Systems as the "Prime Contractor - responsible for project and shipbuilding management - while Thales UK would only be a "Key Supplier for the whole ship design".  [In April 2004 BAE Systems was downgraded by the MOD to a "major supplier" status alongside Thales, and KBR was appointed as Project Integrator in February 2005.]

By early February 2003 BAE Systems and Thales UK were locked in to a bitter dispute over how closely the final CVF design should be based on Thales' ideas, with BAE (used to dominating the design in other contracts for which it is the prime contractor such as the Type 45 and Astute projects) insisting that the Thales [CVF] design was merely a "concept" that it would use as the starting point for the ultimate design.  Nigel Stewart, BAE's Future Carrier managing director, said: "There isn't a French [sic] design.  All there is a concept.  We do not cut metal until 2006 so the next three years will be about producing that design.".  It does seem that Thales only did the minimum amount of work required to meet the requirements of AP1 and AP2 - perhaps being reluctant to spend its own money on an order it didn't really expect to win - while a more confident BAE Systems was by late 2002, and at its own expense, already moving on to the far more detailed design work, later claiming that utilising this work was the only way that the MOD's 2012 ISD for the first ship could be met.  But not surprisingly, that is not how BAE's "alliance" partner saw it. Alex Dorrian, the chief executive of Thales UK, told The Telegraph that his company's carrier design was " 50 per cent mature.  The MoD likes the Thales ship and the Navy likes the ship very much.  The MoD has said we want the alliance to take this ship forward. ... They want our ship. ...  It contains an enormous amount of detail which has been worked through over three years.  The MoD has said 'take the Thales ship to go forward' not 'pick some bits and use them as the basis of a new ship'."  For a while, the success for Thales in this dispute seemed to be potentially worth up to a billion pounds of the £3 billion contract CVF value, while failure meant that it would only get a few crumbs. 

However the new Prime Contractor (BAE Systems) imposed a number of design and systems changes on the Platform Design Team, this may have caused some growth in size and displacement as in July 2003 Navy News stated "The new aircraft carriers are the largest warships ever built for the Royal Navy ... weighing in at about 70,000 tonnes each. They will have a beam of 74 metres at the flight deck level and 37 metres at the waterline. Length will be 295 metres and draught up to11 metres".

In November 2003, a French official report described the adaptable CVF as displacing 65,000 tonnes, having an overall length of 288 metres, and a draft of 11-12 metres.  This report was rather out of date as by this stage cost concerns had intervened, and during the second half of 2003, the CVF platform design team investigated the possibility of reducing the size of the aircraft carriers and cost the carriers to enable them to built  within budget.

                                                                           Back to top                                       |<   <  Part 13  >   >|