Made to order

Professor David Wimpenny of De Montfort University’s Rapid Prototyping & Manufacturing Group explains how by stealth rapid manufacturing is becoming a mainstream process.

By now most engineers are aware of rapid prototyping as a method of producing models of new products by the precise sequential addition of layers of material. Since the introduction of rapid prototyping in the late 1980s the range of applications has evolved from design verification to the manufacture of tooling, metal castings and more recently rapid manufacturing.

Rapid manufacturing (RM), as distinct from rapid prototyping (RP), is the generation of production components (either directly on an RP machine or as part of the process chain), not prototypes.

The benefits of RM over many conventional production processes are easy to appreciate – it completely removes the tortuous and expensive process of producing tooling. In principle parts can be manufactured as soon as the design process has been completed and changed at the touch of a button.

RM is free from the design constraints associated with tooling or moulding. There is no cost penalty for part complexity and this opens up the potential for very sophisticated component designs, with substantial parts integration and even fully assembled products straight off the machines. Perhaps the most potent argument for the use of rapid manufacturing is that it enables very low production volumes to be economical.

Despite RM’s obvious benefits, however, there is still a long way to go before it is a mainstream production route. Existing rapid prototyping techniques are still too slow to support medium to high-volume applications, and the limited range and high cost of RP materials are still hampering the adoption of the process. The accuracy and surface finish of RP parts also fall below the standard required for some (but not all) applications.

The ideal applications for RM are small (less than 100x100x100mm), complex parts, produced in low volumes, where the surface finish is negotiable and the accuracy required is not too demanding. (There are many successful examples, though, that do not meet this precise specification).

Like many technological breakthroughs, RM has suffered its fair share of hype – comparisons with the Stargate Replicator perhaps grab headlines, but are misleading. Many of the initial examples from the US focused on exotic rather than mainstream applications, and this has reinforced the impression that RM is beyond the reach of most companies.

These exotic applications include the production of hundreds of parts for the International Space Station and space shuttles by Boeing’s Rocketdyne Division (although, as with many innovations derived from space programmes, this approach is now gaining more mainstream application in the manufacture of laser-sintered nylon parts for the F-18 fighter jet). The UK’s motorsport industry also depends heavily on the manufacture of parts for F1 racing cars by RM methods.

The use of rapid prototypes as wax patterns to investment cast items of jewellery is now relatively commonplace. In many respects the application is ideal – small, complex low-volume parts where the traditional route is particularly labour intensive. Moreover, many items of jewellery need to be produced in a wide range of sizes and, once designed in CAD, scaling of the design and production of RP patterns is relatively straightforward. Processes based on the printing of wax models and curing of liquid photopolymers are particularly favoured in this area – indeed the Indian jewellery industry, based in Mumbai, is now a major user of this approach.

Another important application for RM is the production of medical devices that are tailored to the patient. Previously patterns moulded to fit an individual’s anatomy would be used by highly skilled technicians to craft the final product manually. Now a patient can be scanned to ‘reverse engineer’ the required geometry. Using this information, accurate master patterns can be generated quickly and cost-effectively on RP machines and used to make the final product.

The dental industry has been quick to grasp the benefits of the approach, with companies such as Cynovad (Canada), CentraDent (Holland) and Align Technology (USA) producing investment cast dental crowns and vacuum-formed occlusal splints.

Materialise (Belgium) has developed the SurgiGuide which helps dentists to accurately drill holes for dental implants. Also, hearing aid manufacturers Phonak (Switzerland) and Siemens (USA) have manufactured over 250,000 hearing aid shells using a combination of reverse engineering and laser-sintering RM.

But use of RM isn’t just limited to exotic applications. Thousands of components are produced for more mainstream products every year. Why do we hear so little about these? Many companies regard RM as a trade secret which provides a competitive advantage that must be jealously guarded. Often firms get into it by accident – usually as the result of a problem.

A typical scenario is that an injection-moulding tool is late for the launch of a new product, so in desperation the company turns to RM to produce the missing parts until production parts become available. But once the firm realises it can have parts in hours not weeks it’s only a matter of time before it starts to use RM regularly.

Companies used to ‘exorbitant’ RP prices are often surprised to find how little it costs to manufacture several hundred small parts. When the cost of tooling and moulding is taken into consideration RM can prove to be more cost-effective even when producing 10,000 parts or more. In many cases it is not competing with injection moulding. For low-volume manufacturing the more conventional route is to machine or vacuum-cast parts. These routes can be expensive and often designers are forced simplify the design to reduce the cost of manufacturing. RM can provide a welcome alternative.

The technique has also been used to manufacture small components for specialised machines – for example, Renishaw often makes batches of its metrology products (up to 200 at a time) with parts produced using stereo-lithography.

As the demand for higher volumes of parts grows, RP service bureaus are changing the way they work and their pricing structure to reflect this new and important market. To support the growth designers will have to design for RM – this involves understanding the capabilities and limitations of the current RP processes. As more companies use RM they will demand better, quicker and cheaper products and this will provide a catalyst for the industry.

Work is on-going to provide a wider range of RM and, perhaps more importantly, supply reliable mechanical property data. Information on the long-term stability and durability of RM parts is also critically important – the working life of RP models is usually measured in days, but RM parts must provide reliable and safe operation over months and years. There is also significant research in the field of machinery development. These new machines will be specifically designed to meet the needs of production, not prototyping.

According to figures compiled by the RP guru Terry Wohlers (Wohlers Report, 2004), over the past 12 months 6.6 per cent of RP models (around 320,000 parts) were used in direct manufacturing applications – almost a two-fold increase on the previous year (3.9 per cent).

Moreover, this figure may be a significant underestimate as it doesn’t include models used to produce investment castings, many of which are utilised to support production not prototyping.

At the moment RM is growing by stealth but with growing confidence and experience, and further developments in machinery and materials, a point will be reached when the technology will be widely recognised as an alternative to conventional manufacturing methods. This will herald the start of a manufacturing revolution.