I think this paper [Practice in Communities: how engineers create solutions ‐ the Bloodhound Guided Missile and the Hawker Harrier “jump jet”. by Jonathan Aylen and Mike Pryce] sums up the ascendancy of the Harrier approach to vtol:
In 1957 the initial design team at Kingston, known as the Project Office, numbered just 25 staff, out of a total in the Design Office of around 400. The Project Office looked after all the aerodynamic work for the company, including both new and existing aircraft and analysis of flight and wind tunnel test results [Williams Interview 26/04/2005]. Indeed, the Project Office was the location of all technical design work outside mainstream structural and mechanical design. This was in addition to their activities in new project design which required the generation of new concepts, writing proposals and liaising with suppliers and customers for such project designs. Staff in the Hawker Project Office tended to be young, with a mix of graduates and former shop‐floor apprentices among their number. . The Chief Designer, Sir Sydney Camm, had a special interest in his ‘Young Gentlemen’ of the Project Office, being both their harshest critic and fiercest defender. The rate of ‘turnover’ in front line combat aircraft and related technology meant that it was clearly seen that the future of the company was directly linked to the work of the Project Office.
The central innovation of the Hawker P.1127 was to use the same single engine to both lift vertically and then drive forward in flight, a system known as ‘vectored thrust’. Other ‘jump jets’ designed in the UK relied on separate lift and propulsion engines for each flight regime, often installed by the half dozen or more. The origins of the vectored thrust engine lay in the work of a French designer, Michel Wibault, who approached NATO in Brussels with the outline idea in 1956. NATO officials saw the idea had merit, and passed the concept on to the Bristol Engine company in England, with whom NATO were already working.
At Bristol the idea of vectored thrust was simplified and made more practical, the work being led by a young engineer called Gordon Lewis. An essential part of this work was finding aircraft designers who could explore the issues raised by the new type of engine. Bristol were linked by cross‐shareholdings to the Shorts aircraft company. But Shorts were more interested in the alternative lift jet system as they had a contract from the UK Government to develop a research aircraft using that approach. Shorts therefore used design work on vectored thrust as an opportunistic way to gain a meeting with NATO, where they put forward their own favoured type of lift jet design ahead of vectored thrust ‐ hardly a way to gain popularity with Bristol. Design collaboration between the two firms ended immediately, a clear illustration that communities of practice which coalesce around a boundary object – such as vectored thrust – can divide off irrevocably from those with a different focus.
Since vertical take‐off and landing jet aircraft were a ‘hot topic’, and as Hawker were desperate to secure their future after cuts in the UK fighter programmes, Camm wrote to the head of Bristol engines, Sir Stanley Hooker, to enquire about their work on the subject. A brochure on vectored thrust was passed back to the Hawker Project Office, but it was not initially seen as an attractive system. The brochure was initially picked up by Ralph Hooper, a project engineer at Hawker, who admitted he did so out of boredom with his main task of designing a flight control system on another project. He produced a few sketches of types of aircraft that could use the new vectored thrust engine, but none looked too promising. After a period of going back to his main work, he returned to the idea of the vectored thrust design and it was then that “the blinding flash of the obvious happened”. The original Bristol brochure proposed that only half the engine’s power would be ‘vectored’, limiting the weight it could support. Hooper realized that all the power could be vectored, allowing a much more useful aircraft design to emerge.
Hooper visited Bristol to discuss the idea with Lewis. It turned out that Bristol had anticipated the development Hooper outlined and put it into their engine’s patent application, although the more detailed scheme Hooper created was also the subject of a Hawker patent. Despite this apparent conflict of interest, with Bristol fiercely defending their patent rights against other engine companies, they waived them with Hawker. This was a direct result of the design community that developed between Bristol and Hawker, with both feeding new ideas into the design of the engine and airframe over a period of several years. Many decades later it was impossible for participants to recall who had created what, and it was never a point of argument between them.. The design of what became the Harrier and its engine evolved over time through an informal process of joint working between Hawker and Bristol engineers (located 120 miles apart) with novel technical problems dealt with in a mutually supportive way – demarcation was neither technically nor organisationally desirable. This was a far cry from the often fractious relationship between other engine and aircraft companies, illustrated by a cartoon from the time:
An informal approach to working was supported by management – Camm and Hooker at director level having known and worked with each other for many years. In addition, within the two firms management trusted their project engineers to make decisions and act on them, and provided resources when needed. At a briefing for NATO it is reported that Lewis brought the wrong costing figures to win NATO funding for the engine. Too late to correct the mistake, and having obtained NATO approval at the lower costs presented, Arnold Hall, Bristol’s director, accepted Lewis’s explanation and approved the significant additional spending the company would incur. A similar relationship operated with government – the Ministry of Defence ultimately funded the Hawker P.1127, but contract cover was only received a few months before the aircraft flew in 1960, with the design and manufacturing costs having been met by the company up to that point, a highly unusual state of affairs in the aircraft industry. However, Camm’s reputation with the MoD meant that Hawker’s directors were willing to place their faith in the innovative project his design team created. Such support ‘greased the bearings’ of the design process, and formed a vital part of the community that developed the Harrier.
The role of the community in obtaining support and sustaining practices within the Harrier design team is highlighted by the difficulty those outside the community faced in understanding the engineering sense of the design. Many analytical studies were produced that ‘proved’ that the vectored thrust system produced the ‘optimum worst’ solution to vertical flight by jet aircraft. It was seen in these studies that the lift jet system provided a ’better’ solution, with a hybrid of vectored thrust and lift jets producing the ‘optimum’ best. On the basis of these studies a number of aircraft were built in France, Germany, the UK and elsewhere. The Germans built an aircraft that used the optimum ‘best’ system, which turned out to be, in Ralph Hooper’s words, “amongst the most expensive and useless of all time”. The French had a similar experience, one of their lift jet equipped designs having the dubious distinction of crashing twice and killing two test pilots.
Central to the Harrier’s success was that it was the simplest, most practical engineering solution to the problem. In large part, this was due to the community nature of the way the engine and aircraft designers worked together. Occam’s razor shaped their way of working and the new fields of design knowledge the community required. This informal approach was not used on the disastrous prototypes produced by others. Instead, rigid contracts, separate design teams (often spread across a continent) and patent disputes were commonplace and complicated engineering and project failure the result.