TECHNICAL KNOWLEDGE IN THE WORKPLACE

What do you need to learn once you start work as an engineer? Almost everything.

Case studies of engineering work can help us understand more about what you will need to learn. The case studies are based on detailed publications that provide further details.

Case studies also help provide further background for maps that include all the aspects of engineering knowledge that we have identified so far.³²

I have included case studies from several different engineering disciplines: you can skip the less interesting ones, but make sure that you read the one on technicians, please.

Case study 1: Aircraft engine design

G. L. Wilde, a senior aircraft engine designer with Rolls Royce, argued for improving engineering education in order to help prepare engineering designers for the kinds of work that will be expected of them.³³In doing so, he provided us with a first-hand outline of some of the technical knowledge that aircraft engine designers need. He advocated creating drawings with freehand sketching, learning to read drawings, and making parts, which consists of ‘thinking with a pencil’ with economy and complete-ness of communication – we already mentioned visual perception skills in the tacit knowledge section.

He identified relevant engineering science knowledge across related disciplines:

• mechanical design (form, function, material properties, stress analysis, and fatigue),

• thermodynamics, fluid mechanics, and aerodynamics (fluid flow, pressure, tem-perature, turbulence, and energy transfer),

• tribology (lubrication, wear, and surface treatments), and

• vibration (noise, acoustics, and vibration suppression).

Wilde also hinted at the contextual knowledge of materials, manufacturing methods, and working with dimensioned drawings.

He described the ‘humiliation’ felt by graduate engineers starting work in a large design enterprise with little contextual knowledge to complement their theoretical

and analytical skills. He proposed a master-apprentice arrangement in which young design engineers would work side-by-side with the most experienced designers before being transferred to detailed product development and enhancement assignments. He reported subsequent discouragement and demotivation, associated with inadequate supervision and guidance, which led to the loss of young engineers and substantial company investment. Pressure on engineering staff to complete demanding technical work competed with the need to spend time with less experienced engineers to help them gain knowledge that, Wilde asserted, could have been more easily learnt earlier.

He complained that design ‘is a skill not sufficiently well understood by company specialists and mostly undervalued . . .’

This tension between the time pressure on experienced engineers and the need to spend time helping young graduates continues today. Part of this is because young engineers have little idea about what they have to learn: I expect that you will gain a much better idea of what awaits you by reading this chapter.

We could classify the different aspects of technical knowledge that Wilde men-tioned in this way:

• Abstract knowledge, both mathematics- and science-based, from formal education to construct an abstract model of a machine, organism, or physical system needed to predict performance.

• Documentation, ways to represent the internal operation of machines, organisms, and physical systems.

• Manufacturing and assembly methods to produce finished assemblies.

• Design for manufacture: ways to design products that result in economic manufacture and assembly in a given context.

• Product definition, including how it works and how each of the components contributes to product performance.

• Knowledge of components and materials.

• Knowledge of component and material properties, both individually and in combinations.

Wilde went on to describe a design office hierarchy: designers, design section leaders, senior project designers, assistant chief designers, and a chief designer, all of whom were supported by specialist groups in stress analysis and engine systems, such as lubrication, fuel handling, and aircraft installation.

He then outlined a selection of the main technical constraints facing the designer, apart from the need to deliver the predicted technical performance within an agreed overall cost and delivery timescale.

Existing designs: the entire design builds on existing practice that is thought to be well understood, as it has provided hundreds of millions of incident-free flying hours (design from precedents was discussed in Chapter 3).

• Turbine blade shrouding, gas leakage, blade vibration, blade cooling, turbine inlet temperature and pressure (Figures 5.7, 5.8).

• Minimising turbine blade tip clearance, location of shaft bearings close to turbine discs to limit shaft deflection, and the use of turbine inlet nozzle vanes as the bearing support structure within the turbine housing.

Figure 5.5 Knowledge map extended with fields of knowledge reported by Wilde – arrows convey the main associations linking different fields of knowledge.

• Adjustment of seals and clearances to regulate pressure and flow of cooling air to the internal engine components, such as bearings and turbine discs, handling of engine axial loads, and the effects of cooling air pressure on lubricant flow in the bearings.

Turbine blade-cooling air needs to be arranged to maximise convective cooling within the internal passages of the hollow blades and, by allowing air to pass to the outer surface of the blade through tiny holes, to provide a protective insulating film of cool air around the blade’s external surface, ensuring that there is no chance of engine pressure fluctuations that would be strong enough to allow hot combustion gas (up to 700^◦C above blade material melting temperature) to enter the cooling air passages.

While it is easy to describe the knowledge categories, much of the knowledge used in the design is tacit, unwritten knowledge.³⁴

Aircraft engines require extremely high-quality blade manufacturing methods to achieve the highly complex, internal cooling air passages that are desired with material properties to prevent high temperature creep (gradual elongation) under high centrifu-gal and bending loads. This necessitates experimental investigation of different blade designs to find the best cooling airflows and internal passage arrangements. Experi-mental studies are also needed to investigate transient thermal expansion effects during rapid engine speed changes that cause changes in seal clearances, cooling airflows, and internal air pressure differences.

Figure 5.6 Large bypass ratio jet engine, typical on a contemporary passenger jet aircraft. The turbine at the rear of the engine is at the right-hand end of the picture (Rolls Royce, 1973).

Modern aircraft engines are expected to be extremely reliable. Failures are so rare that they often become headline news around the world. Engines will usually run for a year or more of continuous flying with only occasional inspections before they are taken into workshops for maintenance.

Wilde’s account shows how the accumulation of all this technical knowledge is a dominant constraint on the capacity of designers. It takes young designers many years to accumulate this knowledge within the firm as they perform their design work:

it cannot be obtained from other sources. Younger designers will acquire detailed knowledge in different technical aspects, but the design cannot be completed without taking the interactions between all these aspects into account.

This is not easy when the knowledge of all these different aspects is shared between many different specialist designers and the design process is constrained by limited time.

Wilde’s account lies at one extreme of engineering practice: high technology design work where technical performance close to the ultimate limits of our knowledge is the desired outcome. Most engineers work further from ultimate performance lim-its, but under much tighter time and cost constraints. The accumulation of technical knowledge is still important, but there is much less time to learn, so access to shared knowledge becomes the principal constraint.

Adding to the previous list of technical knowledge aspects, we could classify the following:

• Previous and current designs of assemblies and products, and implicit knowledge of proportion.³⁵

• Abstract models of products and assemblies that describe certain aspects, such as heat transfer, vibration, gas leakage, and fluid flow.

Figure 5.7 Cross section of the top half of a 3-stage turbine in a jet engine showing air cooling around turbine discs. LP (pale arrows) refers to ‘low-pressure air’ and HP (darker arrows) refers to

‘high-pressure air’. The engine centre line is below the bottom of this diagram (Rolls Royce, 1973).

• Ways to control the production work environment to prevent disturbances and contamination, ways to organise work to reduce the chances of error, and how to maintain high-quality standards.

• Measurement, testing, and inspection methods for product and components.

• Diagnosis methods for product.

• Failure symptoms of product.

• Component and material failure modes.

• Failure modes of product.

• Production faults and defects in product and components, symptoms of ‘trouble’.

Case study 2: Aluminium extrusion

Ravaille and Vinck described some much more mundane and routine engineering design work – designing aluminium extrusion dies.³⁶The special aluminium sections used for window frames, for example, are all manufactured by extrusion from dies.

Figure 5.8 View of nozzle guide vane and turbine blades showing cooling air flow. Modern turbine blades have many more holes and cooling air passages and operate in gas temperatures well above the melting point of the material (Rolls Royce, 1973).

Extrusion works just like toothpaste, and aluminium is extruded through the small hole at the end of the toothpaste tube. The die defines the shape of the final alu-minium extrusion. Many alualu-minium extrusions have very complex shapes, and some are hollow.

Of course, there is a big difference between extruding aluminium and toothpaste.

The special aluminium alloy, chosen to have the best possible properties for extrusion, is heated to become soft and easier to be extruded. It is forced through the die passages at great pressure. The die passages have to be carefully crafted so that the aluminium progressively forms the ultimate shape that emerges from the face of the die on to a conveyor that keeps the soft aluminium perfectly straight until it cools and hardens.

Although the researchers were more interested in the design process, their account reveals interesting aspects of daily work for three distinct groups of people: die design-ers, die makdesign-ers, and extrusion machine operators. The designers and makers rely heavily on experience with previous designs. Each has a different knowledge domain with a small amount of overlap. The designers seem to have little appreciation for the wider context in which the machine operators live with all the constraints of a manufacturing operation. Different operators provide conflicting feedback on die per-formance, so designers tend not to take much notice of their comments. The operators

never mention manufacturing (application) details that they think would be obvious to anyone in their position, so the designers never get to hear about them. Some operators change their dies and send simple two-dimensional sketches back to the designer, even though they have made complex three-dimensional changes that the sketches cannot adequately represent. The die makers have to interpolate the sketches and instructions from the designers. The designers provide a precisely detailed final extrusion shape, but fewer details on other aspects of the die. The designers have just enough understand-ing of the die-makunderstand-ing process to be able to generate a viable design: the die makers don’t let them know much more because they guard their own expertise and don’t think the designers need to know the details anyway. Although the die makers craft a three-dimensional object with complex internal passageways, the authors reported that the designers almost exclusively thought in two dimensions.

Another interesting observation came from comparing two different companies specialising in die manufacturing. One company used more or less fixed design rules that were developed in a separate R&D department. They stored details of previous work in numerical sequence, making it almost impossible to refer back to previous designs, even if the designers wanted to. The other firm relied heavily on looking up and adapting previous successful designs. Designers used tacit rules to choose an earlier similarly shaped extrusion and would adapt the die for the new extrusion from the design of the earlier one.

We can see evidence for technical knowledge in several domains here. First, we should note that the die designers and die makers work for firms that sell dies to aluminium extrusion makers. The machine operators are clients that use the dies. This is common in engineering work: the client is also involved in engineering work. The die designers have some basic knowledge of how the dies are used, but face the problem that they have many clients with very different requirements and constraints. They choose to limit their knowledge of how dies are used in practice by taking little notice of comments and feedback from machine operators, on the grounds that the comments tend to conflict with each other. The machine operators, on the other hand, have much more detailed knowledge of their particular situation and how they actually use the dies purchased from the die-making firms. The die makers have their own knowledge of how to make dies. The designers share some of this, but only just enough to enable them to do the design work in a way that the dies can actually be manufactured.

Interestingly, according to the authors, there is no satisfactory theory that enables the die designers to predict die performance. They cannot make accurate predictions, so they rely on knowledge of previous designs. One firm uses actual designs that designers can copy and adapt, while the other uses design rules that emerge from an in-house R&D department. The authors say little about the actual way in which the R&D department operates because their main concern was the die designers.

In this account we can find evidence of the following areas of technical knowledge, but the knowledge is not shared equally. Different people possess different aspects of knowledge, which may not be complete or even consistent with each other.

• Definition of product (designer, die maker, operator)

• Customer needs (technical) (operator, some with designer)

• Applications of the product (mainly operator, some with designer)

• Operating the product (operator)

• Maintaining the product (operator)

• Repairing the product (operator)

• Failure symptoms, signs of ‘trouble’ with the product (operator, some with designer)

• Failure modes of product (operator, designer)

• Diagnosis methods for product (operator, designer)

• Documentation techniques and standards, representing the product, ways to represent the internal operation of machines and physical systems (designer)

• Properties, models of product assembly for predicting performance and behaviour (note: in this case, this was not feasible because there was no satisfactory theory)

• Manufacturing methods, assembly of product, construction methods, time/cost, and resources needed (mainly die maker, some with designer)

• Knowledge of components, materials (designer, die maker, and some with operator)

• Locating required technical information in large amounts of mostly irrelevant written documentation (some designers)

• Design for manufacture: ways to design products that result in economic manufacture and assembly in a given context (designer)

• Previous designs for similar products and components, neat, well-structured, economic designs (designer, die maker)

Case study 3: Laboratory and photocopier technicians’ knowledge

Technicians develop special kinds of knowledge that most engineers seldom think about.

In their 1997 collection of studies, Barley & Orr raised the issue of technical knowledge in attempting to define technical work.³⁷ They suggested that such work

‘requires understanding and utilisation of abstract knowledge’, presumably mathe-matics or science acquired through formal education. They acknowledge that many technical workers, particularly technicians and trade workers, acquire their knowl-edge and skills through experience and working with more experienced workers (e.g. apprenticeships), rather than through formal education. They also suggest that,

‘technical acumen is the result of contextual and tacit knowledge’. In another chapter, Barley and Bechky identified many aspects of technical working knowledge required by biotechnology laboratory technicians. Doron also discussed similar lab technicians:³⁸

‘Technicians are usually charged with ensuring that machines, organisms, and other physical systems remain in good working order. Caretaking often requires technicians to employ theories, diagnoses, documentation, and other representa-tions drawn from the symbolic realm they support.’ (p. 89)

Yet this knowledge is much wider than the obvious specialisation of a technical worker:

‘[Technicians] focused more on the instruments than on the cells they analysed, knowledge of optics, lasers, for instance, and computers was considered more crucial than knowledge of cell biology.’ (p. 92)

‘Highly developed tactile sensory-motor skills also seemed to be important.’

(p. 99)

Avoiding sources of disturbance that could lead to mistakes also seemed to be an important issue for lab technicians. Such sources could be visitors, lack of concentration, being preoccupied with something else, being interrupted, doing new things, dirt, or contamination. Methods for preventing trouble included being well organised, almost to the point of obsession, knowing the proper place for every-thing (tools, instruments, and supplies), and careful scheduling of the work to avoid unnecessary time pressures.

‘Troubleshooting was, therefore, both a discipline and a way of life among techni-cians and research support specialists. A panoply of undocumented practices for increasing the odds of a technique’s success enveloped every procedure performed in the two labs and constituted most of the labs’ routines. Although mentions of documentation, habitual cleanliness, rules of thumb, and strategies for recover-ing from mistakes or confrontrecover-ing enigmas were conspicuously absent in published discussions of scientific methods, the fortunes of the labs largely rose and fell by their exercise. Such a state of affairs challenges not only standard conceptions of the technicians’ role, but prevalent images of the distribution of knowledge in sci-ence’s division of labour . . . To achieve this task, lab personnel constructed and employed a body of knowledge that most scientists lacked – a contextual under-standing of materials, instruments, and techniques that was grounded in hands on experience.’ (pp. 115–116)

We could classify the different aspects of technical knowledge that emerge from these accounts like this:

• Abstract knowledge: mathematics- and science-based, from formal education, to construct an abstract model of a machine, organism, or physical system needed to diagnose performance problems.

• Documentation: ways to represent the internal operation of machines, organisms, and physical systems.

• Knowledge of tools and equipment required for a technician to perform work, how to keep the tools in good working order, and how to use the tools.

• Failure symptoms and signs of ‘trouble’.

• Ways to control the production work environment to prevent disturbances and contamination, as well as ways to organise work to reduce the chances of error.

• Production scheduling and the effective deployment of productive resources.

Much of the knowledge is contextual knowledge and is drawn from first-hand experience. Much of the knowledge is also tacit knowledge, as the individuals cannot necessarily articulate it. Much of the knowledge, even if articulated, is unwritten, par-ticularly working methods to maximise the chance of success without trouble. Some of the knowledge is readily applicable in another context. For example, computer knowledge and skills can be readily applied in other contexts, except, of course, for specialised software knowledge. Like the die makers, much of the technicians’ knowl-edge is not shared with the scientists, who are the professionals attributed with mastery of the subject.

In reviewing the work of technicians, Orr described the wide range of information that he observed technicians using to help them diagnose problems.³⁹ They liked to

talk to the actual users, not necessarily the customer representative. Users could

talk to the actual users, not necessarily the customer representative. Users could