It was probably Sir Ben Lockspeiser (formerly Director General of Scientific Research in the wartime Ministry of Aircraft Production) who played the most influential part in the decision to set up the Tube Investments Group Research Laboratory (TIRL) in the first place. Sir Ivan Stedeford was then Chairman of the Tube Investments Group (TI) of more than 50 manufacturing companies, which produced a highly diverse range of products including steel, steel tube, bicycles, machine tools, gas cookers, vehicle exhaust systems, rings for jet engines, and mechanical seals, among much else. Sir Ben, then a member of the TI board, persuaded him that basic research, which had contributed so much to winning World War II, would be good for industry and aid its postwar resurgence.
The Group was expanding rapidly and had set up, at Walsall Airport, the TI Technology Centre, which housed computer and operational research facilities. The core of the Group was the Steel Tube Division, which consisted of several companies, mainly in the Birmingham area, manufacturing steel tubes by a variety of processes. Over and above the essential production control facilities possessed by individual companies, the Division had set up the Department of Development and Research, which then became part of the Technology Centre. It addressed short-term customer complaints and production problems and conducted such research as might uncover the causes of perennial examples of both.
Set against this background, the commercial rationale for buying an old country house and transforming it into a basic research laboratory may not have been obvious at the time to some of the senior management. Nevertheless, Sir Ben Lockspeiser was not alone in urging the need for change. This was the era in which no blue-chip corporation was complete without its ‘ivory tower’: AEI (Associated Electrical Industries) at Aldermaston Court, Pilkingtons in St Helens, ICI (Imperial Chemical Industries) and BP (British Petroleum) all had, or were setting up, major laboratories for conducting basic research.
Close contact with a major university was deemed essential, as was remoteness from any of the operating companies that might be tempted to overload the Laboratory with production problems. Hinxton Hall, nine miles outside Cambridge, had been on the market for some time and fitted these criteria admirably. TI bought it in 1953 together with the plans of the Metallurgy Building at Harwell, which were used to erect a new laboratory building on land adjacent to the Old Hall (figure 1).
The task of setting up the Laboratory was given to Dr Phillip Bowden FRS, head of the Department for the study of Physics and Chemistry of Rubbing Solids in Cambridge—a title reflecting his classic studies of friction with Dr David Tabor FRS. Bowden's budget seemed to be unlimited, and he seemed to manage setting up a laboratory at Hinxton Hall almost as a part-time occupation. His main lieutenant, Jeof Courtney-Pratt (later head-hunted by Bell Telephone Laboratories), worked—also part-time—mainly at night and would leave scribbled instructions on the walls for the builders as to where benches or power points were to be installed. Reading the walls was their first task each morning. One of Courtney-Pratt's first acts was to roof in the stable yard and turn it into a workshop, which he filled with the best-quality machine tools that money could buy.
Bowden appointed one of his former research students, Dr T. P. Hughes, a chemist, as the Laboratory's first Director, and the first scientists were recruited in 1954.
The ‘ivory tower’ phase
Despite the fact that TI's turnover of several hundred million pounds was made almost entirely by manipulating metal, mostly steel, the Laboratory's initial preoccupations were irradiated plastics (Arthur Charlesby), the properties of metallic and ceramic whiskers (Jim Gordon) and the basic structure of materials (Jim Menter, later Sir James). Leading scientists in each field were recruited, given the best equipment with which to pursue their studies, and instructed simply to do good science. A Van de Graaf generator was installed for the work on irradiated plastics, and a Siemens Elmiskop 1 electron microscope was ordered for the materials studies. A newcomer to the Laboratory at this time would have been particularly struck by the almost total lack of formality, the sense of freedom and the prevailing infectious excitement. The microscope arrived late in 1954 and was the second to enter the country, the first having gone to the Cavendish. In the following year, Jim Menter, another of Bowden's research students and the first scientist to join the new laboratory, used it to achieve the first direct imaging of atomic planes in a crystal lattice. He resolved the (201̄) planes of platinum phthalocyanine—a distance of 12.0 Å apart—and succeeded in demonstrating the presence of dislocations (figure 2).1
This microstructural coup was effected in the butler's pantry in the Old Hall at Hinxton, where the extra thickness of the walls, originally intended to protect the family silver, was found to produce the lowest ambient magnetic field on the site and hence the best location for installing the electron microscope.
As early as 1957, it was being suggested that members of the Laboratory should at least start to pay visits to the operating companies to learn what they did and whether they had any problems doing it. During one such visit to a tubemaker, Howells of Sheffield, a sample of steel was produced that exhibited surface cracking that was alleged to have occurred at high temperature during the tube-making process. This specimen, examined under a microscope at Hinxton, revealed small regions at the surface that seemed to differ from the bulk of the sample. The links with Cambridge University now proved their usefulness: further examination of this sample by Peter Duncumb at the Cavendish Laboratory, using his newly developed X-ray scanning microanalyser,2 identified these regions as being enriched in copper and nickel, two of the elements that had long been suspected to be associated with such cracking. One day's work in a university had offered new insights into a recurring problem in a steelworks.
As a result of this experiment, it was agreed that Hinxton should build its own scanning microanalyser. Further collaboration between Duncumb and David Melford, a Hinxton metallurgist recently recruited from Cambridge, led to the successful design of the world's first such instrument for metallurgical use (figure 3b).3 This was also able to reveal the distribution of tin, the most troublesome element known to be involved in the cracking problem. A scientific understanding was thus developed on which to base what had hitherto been wholly empirical manufacturing practices.
The new instrument attracted major outside interest, and within four months of its commissioning in December 1958 it was licensed to the Cambridge Instrument Company for production under the name Microscan. Their first production instrument was demonstrated at the Institute of Physics Exhibition in January 1960. More than 80 instruments were sold worldwide, and its success encouraged them a few years later to license the scanning electron microscope developed in the Cambridge University Engineering Department and market it as the Stereoscan.
The further successes of Menter's group are well illustrated by this quotation from Walter J. Moore's review of the proceedings of the 1959 Lake George Conference, ‘The Structure and Properties of Thin Films’:
If anyone has not yet seen the exciting results of the Tube Investments Group (Menter, Bassett, Pashley) on the direct electron micrography of crystal structures and their imperfections, here is a good place to look. … almost 100 pages of Tube Investments research are reported. Maybe future scientific historians will ponder how and why at mid-twentieth century an English bicycle-maker was contributing more than the entire US steel industry to basic research on metals.
Peter Duncumb joined the Laboratory in late 1959 and immediately set about combining the technique of X-ray microanalysis with electron microscopy to enable the identification of submicrometre particles on extraction replicas. The outcome was the construction in the Hinxton workshops of EMMA, an electron microscope and microanalyser (figure 3a).4 This was, in turn, licensed for production to AEI.
EMMA attracted many visitors. One day an American arrived, bringing with him an air-pollution sample taken on top of one of the highest buildings in Chicago. It consisted of a filter paper on which a large number of aerosol particles had been collected. By now, the first production EMMA had been installed for testing in the butler's pantry in the Old Hall, replacing Jim Menter's Elmiskop 1 which had been moved into the new laboratory. As the electron beam was focused on each particle and its constituent elements were identified, the visitor annotated a photograph of the field of view, muttering ‘steel works’, ‘paint factory’, ‘cement works’, etc. as he did so. At the end of the session, evidently impressed by the power of the technique, he remarked that it seemed very strange to him to sit in the butler's pantry in an old country house in England and be able to make measurements that quite clearly revealed which way the wind had been blowing last Thursday in downtown Chicago. Eventually EMMA came to share a display case with the X-ray microanalyser in the Science Museum when the Laboratory had acquired more advanced versions of both (figure 3).
If your objective is to turn an empty country house into a first-class research laboratory complete with an international reputation in no more than five years, it must be admitted that Bowden had been amazingly successful. It was a priceless asset to be considered academically respectable by universities from which the Laboratory would wish to recruit in the future and to be taken very seriously by other major industrial laboratories.
These early years had firmly established the principle that the best way to solve many production problems and make advances in technology was to ensure that the underlying science was properly understood. This invaluable ethic was to stand the Laboratory in good stead on many occasions in the years to come and will emerge throughout this account of some of its activities. Industry's willingness to fund untargeted basic research, however, was now on the wane.
One day in July 1961, Hughes called the senior scientists together to announce that it had been decided the future interests of the Laboratory should lie in the fields of metal physics and physical metallurgy. The irradiation of plastics and consequent polymer chemistry and the studies of whisker properties would cease. He did not consider that he was the right person to direct the Laboratory in this future role and he was therefore leaving with immediate effect. Dr Menter would be the new Director.
A redefined mission
In effect, the ‘ivory tower’ phase was now over, and the Laboratory's new mission was to apply the facilities, skills and reputation it had accrued in ways that could benefit the Group in the longer term. But despite its newly acquired scientific reputation, the Laboratory had a huge credibility hill to climb with the companies whose efforts ultimately funded it. It is in the nature of any group research function that the scientists involved will find themselves effectively looking over the shoulder of those on the shop floor who operate a process day in, day out, and telling them how it could have been done faster or better—above all, differently. At best, this is inevitably a challenging human relationship that takes tact, patience and understanding to convert into a fruitful partnership involving mutual respect.
The recently commissioned microanalyser proved to be a considerable asset in breaking the ice in many such interactions. In the matter of steel quality, a frequently recurring issue in both raw material and product, it could identify non-metallic impurities and measure the segregation of alloying elements on the scale of a few micrometres in a manner that no other organization could match at the time. This enhanced the prestige of the companies in the eyes of their customers and also gave them ammunition with which to confront their suppliers if the raw material was at fault. It was thus an invaluable aid in dispelling the prevailing view that the Group Laboratory was largely an irrelevant expense as far as their day-to-day operations were concerned.
In effect, while responding as best it could to an immediate difficulty, the Laboratory had taken the opportunity to make a scientific advance that would yield continuing benefits in the longer term as well. Gradually a modus vivendi was being established through which science could increasingly be used to secure commercial advantage—as, for example, in the development of superplastic aluminium.
As long ago as 1934, C. E. Pearson5 had drawn attention to the fact that certain alloys could exhibit ductility, as measured in a simple tensile test, up to 20 times greater than that normally expected from metals. Pearson had obtained an elongation of 1950% from the bismuth–tin eutectic and then coiled the resulting specimen for an iconic illustration of the effect. No commercial use had been made of this phenomenon in the early 1960s. Indeed, TI had no interest in either bismuth or tin but nevertheless did make its money out of forming metal. In particular, it had an interest in aluminium as a half owner (with Reynolds Metals Corp.) of the British Aluminium Company. It was known from the scientific literature that a small number of ‘exotic’ aluminium-containing alloys could behave like the bismuth–tin eutectic. Could a commercial aluminium alloy be made to behave in this way?
Initial experiments with the Al–33 wt%Cu eutectic alloy allowed TIRL metallurgists to obtain some basic understanding of how to process the material to obtain superplasticity and, more importantly, what was happening at a microscopic level to give rise to such huge extensions. When these results were shown to the British Aluminium Company they listened politely and then pointed out that they were not in business to sell copper. If Hinxton scientists liked to return when they could demonstrate similar results in an alloy with less than 10% copper, they were assured that more interest would be shown. This was a challenge because superplastic behaviour had rarely been demonstrated in such dilute alloys and certainly not in dilute aluminium alloys.
Conscious of the fact that such behaviour requires a very fine grain size and that zirconium additions could promote this, Mike Stowell (later elected FRS), a physicist recruited from Bristol University, and Brian Watts responded to the challenge by adding 0.5 wt% Zr to an Al–6 wt%Cu alloy. Combining this formulation with the development of a radically new processing route, they demonstrated an elongation 10-fold that of the base alloy.6 The result became known as Supral, the world's first commercial superplastic aluminium alloy. True to their word, the then British Aluminium Research Laboratories at Chalfont Park now joined the project, contributing their expertise in ingot casting and rolling to produce sheet material on a commercial scale.7
Clearly, commercial use was likely to take the form of blow-forming sheet metal. Although blow-moulding was a familiar process for plastics, no experience and no suitable machinery existed for doing it with metals at temperatures of about 450 °C. In 1968 the Engineering Department had been set up to augment the Physics and Metallurgy Departments, and Frank Howard was recruited from the Production Engineering Research Association (PERA) to be its Director. The new department's Metal Forming Group, led by David Laycock, now set about developing both the forming technology for superplastic metals and the equipment with which to implement it.8 The superplastic forming press they built then (figure 4a) was still being used in production 30 years later.
The headrest for the Martin Baker ejector seat for jet aircraft was an early test product. A single moulding (figure 4b) replaced an assembly of a number of separate pressings that previously had to be fitted together and riveted. A significant saving in weight and assembly time was thereby achieved.
Convinced by Jim Menter (now a main board director) that there was a valuable niche market out there and a commercial opportunity, TI finally set up a new company, Superform Metals Limited, at Worcester to manufacture components by superplastic forming. The company survives to this day, as does its sister company in California. In 1981 this whole development attracted the Queen's Award for Technological Achievement.
The virtue of understanding the underlying science was never more apparent than in the Laboratory's contribution to the design of fluid seals. In the early 1960s, mechanical seals, which are used for the sealing of rotating shafts in process pumps and on ships' propellers, were a growing and important part of TI's business. However, the fluid environments in which the pumps operated were becoming ever more demanding, with increasing fluid pressures and temperatures. Controlling the distortion of the seal components under these influences was critical to performance, but design was only possible by trial and (costly) error. In 1964, Sam Wilkinson, the far-sighted technical director of the company concerned, and Alan Green at the Laboratory decided to initiate a programme of research; the Mechanical Seals Working Group was set up with representatives from both organizations. This technical working group went on to meet regularly for the next 24 years and transformed the understanding of how to design and manufacture an unbeatable, high-performance product with high added value. The dedicated facilities installed at Hinxton, although not exactly ring-fenced, can reasonably be regarded as an early example of the ‘embedded laboratory’, now a much more familiar concept.
Bill Graham was recruited from Heriot-Watt University via the National Coal Board's Mining Research Establishment, to lead this project in 1966 as one of the first of a stream of graduate engineers to join the Laboratory and contribute to mechanical seals and other, primarily product-oriented, research. At first, theoretical predictions relied on limited calculations performed by hand and, remotely, on a mainframe computer. But after the arrival in 1967 of the Laboratory's first computer, an IBM 1130, faster progress was possible. The research programme involved the design, construction and operation of highly instrumented test rigs and a considerable amount of computer modelling. The consequences of this were twofold, one predictable and one, surprisingly, not.
The aim was to develop a set of improved performance envelopes based on a better understanding of the hydrodynamics involved. What could not have been anticipated was that some of the new and critical design features would be so subtle that they were almost impossible to detect. As the reputation of the company's products grew, competitors appeared who offered what they thought were equivalent copies at lower prices. Unfortunately, these copies lacked the critical features of the originals, their performance was consequently poor and sales dried up. This ‘hidden technology’ was an unexpected bonus to emerge from the degree of basic understanding that had been achieved.
The computer programs developed for mechanical seal design demonstrated that the Laboratory could support TI's companies with the design of other products, and many quickly followed. Some examples from the 1970s are overhead conveyors and hoists, gas cylinders, bicycles and products for the construction industry such as structural hollow sections, purlins and open-web beams. The required special-purpose software programs were all written by Laboratory staff, and it was not until the early 1980s that the commercial software companies started to offer comprehensive, general-purpose programs. A further example of the wisdom of locating the Laboratory close to a major university now became evident when it purchased and started to use the Cambridge Interactive Systems (CIS) program Medusa in 1983 and introduced those TI companies manufacturing consumer products to the superior merits of three-dimensional design over two-dimensional computer-aided draughting. Hinxton Hall was an ideal test site for such new software.
With these tools the Laboratory was able to invite representatives from the company design, production and marketing functions to sit together in front of the computer screen. The designer's ideas could now be displayed and the product could be exploded or sectioned for the benefit of the production manager to show individual components, the tolerances required and the proposed assembly and fixing details; then shapes, colours and functionalities could be modified to please the marketing department, all in an iterative manner. Eventually, all three could approve the end result, even before the print-out of the first production drawing.
This multidisciplinary way of designing products in three dimensions ran counter to the traditional culture, in which these virtually separate company functions used only two-dimensional drawings to pass information between them. Ironically, it often seemed that those companies with most to gain from the new technology were the most likely to regard it as ‘rocket science’. However, the Laboratory was able to help them assimilate the new methods and to feed back information to the software designers on how to improve their product.
Sometimes, the impact of computer-aided design could, literally, save a loss-making company from failure. One such example was the replacement of an over-extensive range of cold-rolled steel purlins. The large range of purlin profiles and thicknesses resulted in very poor utilization of the company's rolling mills and serious financial losses. In an early example of this type of application, computer models of the market requirement enabled the design of a much reduced range of sections—optimized to capture key market points and reduce rolling mill downtime. The company has since become the UK's largest cold roll-forming company providing products for the UK construction industries.
Advances in ultrasonic testing
In the 1970s the Laboratory became increasingly involved in developing novel types of process control instrumentation. Since several TI companies were involved in ultrasonic testing of one sort or another, the Laboratory built a highly sensitive Schlieren system to explore the interaction of pulses of ultrasound with various types of material defect.
Glass is an acceptable model for steel if appropriate allowances are made for different velocities of transmission. To be able to detect and ‘freeze’ an ultrasonic pulse, travelling through glass at about 5500 m s−1, requires very brief, very intense flashes of illumination and a very sensitive optical system. The Hinxton system designed by David Marsh9 achieved this to a greater extent than had previously been reported, and pulses could readily be seen and resolved into individual wavefronts in most cases. In figure 5a, incoming pulses are slanting downwards from the right-hand edge of the glass block. Most of the energy in one such pulse has been reflected from the bottom surface in a bright pulse on its way to the top left-hand corner. In the bottom surface of the block, however, a small model defect 0.4 mm deep has been cut, and this has reflected part of the energy back in the direction of the incoming pulses and also produced a faster compression wave exiting from the block at the left.
Figure 5b shows a computer simulation of what was predicted should happen in the forward quadrant: the similarity is apparent. The computer program is based on a theory of how ultrasound interacts with defects, which was developed and constantly refined by comparison with such experiments.
If there is a defect in a raw material, it pays to discover this at the earliest possible opportunity before the material has much value added to it. For steel, this may mean while it is still hot from the rolling mill. But there is a snag. Ultrasonic transducers normally require some sort of contact medium to transmit the energy to the steel. It can be oil, water or jelly, but there has to be something. There were reports, however, that ultrasonic pulses could be induced in steel without contact if current pulses at ultrasonic frequencies were fired through the coil of an electromagnet placed 1–2 mm away from the surface. Such electromagnetic acoustic induction had been achieved in a laboratory, and at the beginning of the 1970s it was proposed to develop this technology at Hinxton Hall for application in practice in a steelworks.
As a consequence of the Treaty of Rome, there was European collaboration over coal, iron and steel before the UK actually joined the Common Market in 1971. If an industrial company and a research centre made common cause, they could apply to the European Commission for research funds. A joint application with the Round Oak steelworks at Dudley was successfully made for a grant to develop ultrasonic testing at high temperatures. Progress reports on the project were so well received by the Commission that the Laboratory was invited to apply for more funds so that suitable equipment could be designed, built and installed in the steelworks. This was achieved during 1974. Such external funding was welcome at the time because it not only allowed innovation to take place at reduced cost to the Group but also eased the Laboratory's funding situation, which had undergone a major change from January 1972.
People and resources
To underpin its redefined mission as one of increasing relevance to the Group's businesses, the Laboratory was no longer to be funded 100% from the centre; in future 50% was to come from the operating divisions. This was not such an abrupt change in the nature of the Laboratory's activities as might first seem. Only 20% was expected to consist of trouble-shooting support, wholly funded by the operating companies. The largest segment of the work, some 60%, was to be long-term, innovatory research, funded equally by the Group and the operating divisions. The final 20% would be wholly funded by the Group and consist of the exploratory ‘blue skies’ activity as before.
The thinking behind this change was not only to reduce central charges, the level of which seemed always to be regarded with suspicion, but also to increase the degree of engagement of a company with the progress and result of work being conducted on its behalf—in short, to enhance their feeling of ‘ownership’. At the same time, the provision for 20% of Group-funded, exploratory work would, it was hoped, allow some freedom for the continuation of basic research. In fact, the following three years demonstrated that this did not work out quite as intended. Much of the 20% Group-funded exploratory work went into demonstrating the virtues of proposals being put forward for the 60% of long-term innovatory research to be shared with the operating Divisions. The task of implementing the new charging system fell to Dr D. W. (Don) Pashley FRS, the third of the Laboratory's directors.
As already mentioned, Jim Menter had succeeded the first Director, Tom Hughes, in July 1961, and under his guidance much of the long-term ethos of the Laboratory became established. In 1966 he was elected a Fellow of the Royal Society for his distinguished scientific achievements, and in 1968 he relinquished his position as Director to concentrate on his responsibilities as a main board director of the TI Group. In 1972 he was elected Treasurer and Vice-President of the Royal Society and in the 1976 Birthday Honours he was knighted. Later that year he left TI to become the Principal of Queen Mary College of London University, a position he held until his retirement in 1986.
Sir James was succeeded as Director of the Laboratory in 1968 by Don Pashley. Don was elected a Fellow of the Royal Society that same year for his distinguished contributions to epitaxy and electron microscopy. After Jim Menter's departure he further increased the Laboratory's reputation in this field and became a world authority. He held the post of Director until 1979, when he was appointed to the new Chair of Materials at Imperial College (and as Head of the Metallurgy and Materials Science Department).
On Don Pashley's departure, Dr Peter Duncumb, who had been elected a Fellow of the Royal Society two years earlier for his contributions to the field of electron probe microanalysis, was appointed Director and General Manager of the Laboratory. With Dr David Melford as Director of Research and Deputy General Manager, the task of the management team was to press ahead with meeting the Group's needs for appropriate R&D on all timescales, a task undertaken with some success over the following eight years.
Peter Duncumb retired at the end of November 1987 at a time when the Laboratory was assured of a continuing role in TI's future. In fact, as will become evident later, events subsequently overtook it, and for the remaining 10 months of the Laboratory's existence—David Melford having also retired—no further appointment as Director of the Laboratory was made. Bill Graham as Technical Director and Bill Matthews as Commercial Director ran the Laboratory between them in this period.
As the Laboratory developed, it was recognized that changes in the organization were needed. Essentially, the challenge for any industrial laboratory is first of all to acquire an input, from universities and elsewhere, of the relevant range of skills and disciplines. These must then be nurtured within the laboratory, combined into appropriate portfolios, augmented by the critical addition of project leadership, and deployed to tackle problems of industrial importance. One discipline may thereby serve several companies, and one company may benefit from several disciplines. The ‘right’ degree of structuring had to be found to marshal resources behind projects that were well defined without stifling free thinking in the formative stages of others. Furthermore, the structure had to be credible to the main board of TI, which controlled the funding.
After several stages of evolution, the model arrived at was of two research departments, ‘Materials and Processes’ and ‘Engineering’, consisting of some 130 scientists and technical assistants supported by about 30 administrative and workshop staff. Six of the senior graduates, most of whom also led research teams, were appointed ‘business development’ executives. Their job was to seek out company needs from particular parts of TI at a strategic level, as a useful input to project planning and indeed to TI's corporate technical planning as a whole. This left the main project execution to 10 or so family-size ‘groups’ within the research departments, each of which would develop strong links with the companies they were serving, as well as with scientific institutions outside. The boundaries within the organization were not rigid, and there was much interchange between groups, helped along by occasional seminars. An excellent canteen for communal lunches, staffed mainly by ladies from the neighbouring Hinxton village, also provided space for social gatherings.
In these various ways the Laboratory had good contact with TI at all levels. Individuals developed specialized knowledge of considerable value to operational companies, and close personal relationships with senior company personnel. This created the opportunity for staff to transfer into the companies as part of their career development, thereby allowing fresh talent to be recruited into the Laboratory.
The Laboratory was never large. The total number of staff on site seldom exceeded 180, of whom 40–50 were graduates, including several women. The transfer of a graduate to a company was a natural, and in many ways desirable, step but it meant that his or her particular expertise was no longer available as a Group resource. In the years 1965–77 there were 32 such transfers. There were 11 more in the following two years. Maintaining a balance between career progression and the retention of a high level of relevant expertise and performance in a small laboratory was never easy.
The Laboratory was, however, particularly fortunate in the quality of its technical assistants. Some were already highly skilled when recruited from university departments, but a well-managed apprentice scheme produced many others. Some of these were later sponsored to go to university and, in one case, to take a PhD. There can be no doubt that much of what was accomplished owed a great deal to their formidable skills.
In the 1970s and early 1980s there were revolutionary changes in manufacturing technology, pioneered in particular by the Japanese. Among them were the emerging philosophies of ‘just-in-time’ manufacture—in which stocks of components and material were cut to the minimum—and also of zero defect tolerance. To complete the range of support that a central laboratory should be able to offer a manufacturing group such as TI, a small team of about six manufacturing engineers was therefore set up as part of the Engineering Department at Hinxton—a final step in the conversion from the ‘ivory tower’. These individuals played a key role by spending most of their time in the operating companies, helping them to assimilate the new trends and improve their efficiency. Dr Norman McBain, for example, played a leading part with an in-company team for 18 months in a domestic appliance factory. They successfully managed a seamless transition from traditional cast-iron construction to fabricated monocoque steel, without interrupting output.
The Laboratory also hosted a small group of corporate technical planners—never more than two or three in number—led by Alan Green. These were often MBAs and were picked for their business acumen and experience. Their role was to keep an eye on emerging technologies that the Laboratory might adapt and develop for the Group's benefit, and conversely to consider the best way to exploit innovations generated within the Laboratory itself. The ready availability of their commercial perspective was a valuable asset to the scientists and engineers. It was a role rich in opportunity at the time. One of them, for example, went on to teach innovation management and entrepreneurship at Cambridge University's Judge Institute. Another became the managing director of a company set up by TI in the USA to exploit the Tru-Form ring rolling technology described below.
Rings for jet engines were another of TI's most successful products. Most of the carcase of a jet engine, in fact, is constructed from one sort of annular component or another. These rings are usually made from expensive, tough Nimonic alloys and are not at all easy to fabricate. The rings, intended mainly for compressor or turbine casings, could be up to 18 inches wide, and the diameter and wall thickness could vary several times over this distance. It was a market in which it was particularly important that technical leadership should be maintained and, if possible, enhanced.
The traditional practice had been to take a thick piece of strip, curl it into a circle and flash butt weld the edges together. The width could only be 3–4 inches—as much as the welder could manage—and certainly not as much as 18 inches. The rings would then be welded together circumferentially and machined to the required dimensions, wasting a lot of expensive material.
Rolling the ring could introduce a certain amount of profiling but only at the expense of an increase in diameter, because the natural effect of rolling is to elongate. Faced with this problem, engineers at Hinxton Hall, led by Howard Jones, developed a rolling process that was able to thin the section by forcing metal to flow laterally—to increase the width of the ring rather than its diameter. This allowed rings of the required width and section changes to be rolled in one operation—very much nearer to net shape than previously—and wasted much less of the expensive material. This was named the Tru-Form process and the products were first exhibited to the public at the Paris Air Show in 1985—only four years after the development had begun. An observer of the TI company's stand at the Air Show would quickly have noticed that the world was divided into two kinds of people: those who had no idea how such rings were made and didn't care, and those who knew very well how they were made, cared very much, and had never seen anything like this before in their lives. A visitor wandered across from the stand of TI's main American competitor, stopped, stared, and ran back to fetch his colleagues. It was reported that their Technical Director arrived at work in the USA that morning and to his great surprise found himself on a plane to Europe before lunch.
In the days when steel was cast into ingots, there was one part of the process that always remained outside the control of the steelmaker. However carefully the steel composition was controlled in the furnace, when it was poured into an ingot mould the normal process of solidification caused segregation of the alloying elements, aggregation of non-metallic impurities and an unwelcome deterioration in quality.
With the advent of continuous casting it became possible for the first time to close this control gap by applying magnetic stirring to homogenize the melt during the solidification process. Round Oak steelworks had installed a machine that continuously cast a large bloom, 12 inches by 14 inches in cross-section, whose central soundness should be susceptible to improvement by magnetic stirring in the mould. The attraction of the stirrer design that finally emerged, after consultation between the head of the electronics group, Dick Whittington, and relevant university experts, was twofold. First, it employed mains frequency, and second, although the copper bars surrounding the mould had to carry a current of about 2000 amps, it was at a potential of only 1 volt. Insulation was unnecessary and no splashes of molten steel on the casting platform were likely to cause significant short circuits.
The laboratory test rig was engineered into a unit that could be dropped into place on the continuous casting machine at Round Oak as part of a normal mould-changing exercise, requiring only a single additional electrical connection. Video footage of the casting process confirmed that stirring was taking place, and sections of the resulting blooms demonstrated the hoped-for improvements in central soundness and homogeneity.
The whole exercise was unfortunately overtaken by events in that Round Oak was sold back to British Steel, which later closed it and what had once been the Earl of Dudley's Ironworks was transformed into the Merry Hill Shopping Mall. The Laboratory's expertise in stirrer design was, however, licensed to Concast AG of Zurich, which continued to consult members of the Laboratory even after it had closed in 1988.
Dual-phase vitreous enamels
In the last two years of its existence the Laboratory was still producing innovatory ideas even though they never came to fruition. Dual-phase vitreous enamel was one of several such developments destined also to be overtaken by events. It was tougher and hence more resistant to cracking and spalling than conventional vitreous enamel; it was therefore less susceptible to accidental impacts. Not only was this an attractive improvement to its performance in service but it also offered the possibility of a major reduction in ‘in-house’ rejects or rework and hence lower production costs. It had the further considerable advantage that it did not require to be laid down on enamelling-grade steel sheet but could be applied to cheaper material of normal quality. Indeed, the research shed a wholly new light on the mechanism of adhesion between vitreous enamels and steel substrates.10
The TI company for which it was destined—a major supplier of domestic appliances—was sold and left the Group in 1988 as part of a major reorganization of the business. Other companies that had traditionally drawn heavily on the resources of the Laboratory also left the Group as it was restructured into what were considered to be three strong and viable business areas. In contrast to many of the individual companies that had previously made up the Group, these new business areas were each judged strong enough to organize and sustain their own dedicated research programme.
It therefore appeared that the need for the Laboratory was over.
The Laboratory closed on Friday, 30 September 1988. In a very real sense it was a founder member of what later became known as the ‘Cambridge phenomenon’, the fruitful interaction between university and industry. This took several forms: the transfer and development of university research (as with the microanalyser), a ‘test bed’ for university ideas (such as computer-aided design software), the sponsorship of numerous PhD students and, of course, the recruitment of both graduate and postdoctoral staff.
The result was a fertile environment, balanced between university and industry, into which flowed scientific insights from one direction and an awareness of opportunities for innovation from the other.
Throughout the 34 years of its existence, the Laboratory produced a stream of technological innovations11 whose diversity reflected that of the engineering Group12 it served. It was also responsible for some significant advances in scientific understanding, and indeed four of its members were elected to Fellowship of the Royal Society and one to Fellowship of the Royal Academy of Engineering.
Perhaps its greatest achievement, however, is the fact that it seems to be remembered with affection by those who shared in these endeavours, as a very happy environment in which to have worked.
The authors gratefully acknowledge the help of Geoff Armstrong, Mike Burden, John Cooke, Norman MacBain, John Sawkill and George Ashton, who raided both archives and memories on their behalf.
↵1 J. W. Menter, ‘The direct study by electron microscopy of crystal lattices and their imperfections’, Proc. R. Soc. Lond. A 236, 119–135 (1956).
↵2 V. E. Cosslett and P. Duncumb, ‘Micro-analysis by a flying-spot X-Ray method’, Nature 177, 1172–1173 (1956).
↵3 P. Duncumb and D. A. Melford, ‘Design considerations of an X-ray microanalyser used mainly for metallurgical applications’, in Proc. 2nd Int. Symp. on X-Ray Microscopy and X-Ray Microanalysis, Stockholm (eds A. Engström, V. E. Cosslett and H. H. Pattee), pp. 358–364 (Elsevier, London, 1960).
↵4 P. Duncumb, ‘X-ray microanalysis of elements in the range Z = 4–92, combined with electron microscopy and electron diffraction’, in Proc. 3rd Int. Symp. on X-Ray Microscopy and X-Ray Microanalysis, Stanford (eds H. H. Pattee, V. E. Cosslett and A. Engström), pp. 431–439 (Academic, New York, 1963).
↵5 C. E. Pearson, ‘The viscous properties of extruded eutectic alloys of Pb–Sn and Bi–Sn’, J. Inst. Metals 54, 111–124 (1934).
↵6 B. M. Watts, M. J. Stowell, B. L. Baikie and D. G. E. Owen, ‘Superplasticity in Al–Cu–Zr alloys. Part 1: material preparation and properties’, Metal Sci. 10, 189–197 (1976).
↵7 R. Grimes, M. J. Stowell and B. M. Watts, ‘Superplastic aluminum-based alloys’, Metals Technol. 3, 154–160 (1976).
↵8 D. B. Laycock, ‘Superplastic forming of sheet metal’, in Superplastic forming of structural alloys (ed. N. E. Paton & C. H. Hamilton), pp. 257–271 (Metallurgical Society of AIME, Warrendale, PA, 1982).
↵9 D. M. Marsh, ‘Methods of visualising ultrasound’, in Research techniques in ultrasonic testing, vol. 2 (ed. R. S. Sharpe), pp. 317–367 (Academic, London, 1973).
↵10 R. F. Price, N. Fletcher and M. J. Stowell, ‘Dual phase vitreous enamels, part I. Preparation, application and properties’, Mater. Sci. Technol. 5, 1037–1043 (1959); C. Sherhod and M. J. Stowell, ‘Dual phase vitreous enamels, part II. Adhesion of vitreous enamels on steel’, Mater. Sci. Technol. 5, 1044–1051 (1959).
↵11 It is a feature of industrial laboratories that details of some of the most commercially sensitive research are not published.
↵12 In 2000, after 81 years, the Tube Investments Group itself ceased to exist, following a merger with Smiths Industries.
- © 2010 The Royal Society