Words of Wisdom: Rapid manufacturing on the horizonWords of Wisdom: Rapid manufacturing on the horizon

Machines for rapid prototyping (RP) have been in use since the late 1980s. Stereolithography, laser sintering, fused deposition modeling, and other methods have become mainstream at many organizations around the world. Until recently, they have been used only for building models and prototype parts for form, fit, and function applications. With advances in materials, build speed, and dimensional accuracy, coupled with a dose of creative thinking, some resourceful companies are applying methods of RP to the production of finished, manufactured parts.

For years, I have felt that if these devices could be ?tuned? for the manufacture of production-quality parts, the RP industry would flourish beyond its impressive growth of the 1990s and the meaning of RP would transition to rapid production. Many, however, have discounted the idea as being unrealistic and even absurd. How could a device for models and prototypes compete with highly developed and proven approaches to manufacturing, such as injection molding? When considering how most RP systems build parts by layer, it seemed neither practical nor probable to these naysayers.

Proponents of the idea have underscored the fact that tooling is a necessary evil in manufacturing. If it were possible to eliminate or at least reduce the need for tooling, rapid manufacturing would provide a significant advancement. Designing parts without the constraints imposed by molds and dies would make the product development process easier and much less expensive. Also, it would open up the possibility of producing custom designs that earlier were prohibitively expensive and impractical. Some current methods of RP permit the mixing of two or more materials, so this opens up a host of additional possibilities in manufacturing.

With methods of RP as the devices of manufacturing, true just-in-time manufacturing becomes possible. Manufacturers could ship parts to customers soon after finalizing the CAD data. With this comes the possibility of eliminating finished-parts inventory.

The remaining ?inventory? would consist only of containers of material waiting to be formed. What?s more, corporations could decentralize manufacturing with ease by installing systems that would receive CAD data from anywhere in the world and build parts on demand.

Dream vs. Reality

To some, this may sound like a dream. However, a handful of pioneering companies are using methods of RP to manufacture parts in a way that is much quicker and less expensive than conventional approaches. Boeing Rocketdyne (Canoga Park, CA) led the way when it produced laser-sintered parts in glass-filled nylon for the entire fleet of space shuttles. Similar parts are also in use on the International Space Station. Most recently, Boeing has produced complex air ducts in the same material for the F-18 fighter jet. Savings in time and cost are staggering.

People who are hearing-impaired also benefit from rapid manufacturing. Companies including Siemens (Piscataway, NJ), Phonak (Warrenville, IL), and Widex (Long Island, NY) are using RP systems to manufacture custom-fit, in-the-ear hearing aids. Siemens and Phonak are using laser sintering, while Widex is using stereolithography. Siemens produced and sold more than 150,000 laser-sintered hearing aids by midyear 2003, with a production rate of 2000 products per week.

The Jordan F1 racecar team is applying RP technology to the manufacture of its cars. As many as 20 different parts, ranging from cooling ducts to electrical boxes, are being produced with laser sintering. A service provider named 3T RPD in the UK is supplying parts in an astounding 48 hours.

Laser sintering has become the most popular RP method for manufacturing parts, but a range of other technologies is also being used. The Dimension product from Stratasys (Eden Prairie, MN) was used to produce 40 camera mounts in ABS for the M1 army tank. The original intent was to use them as prototypes, but they functioned well enough to be used as the final parts. Meanwhile, the Electron Beam Melting machine from Arcam (Sweden) has been used to produce metal drill bits for oil drilling rigs.

For years, Align Technology (Santa Clara, CA) has used stereolithography for the production of its Invisalign plastic aligners for straightening teeth. The company has stopped publishing production numbers, but at one time, it was manufacturing 10,000 parts per day. Meanwhile, Interpore Cross (Irvine, CA) is operating 32 machines from Solidscape (Merrimack, NH) to manufacture titanium spinal implants. Both Align Technology and Interpore Cross are producing RP patterns that are in turn used to manufacture the end product, so one might argue that this is really rapid pattern-making and not rapid manufacturing. Regardless, these RP tools are instrumental in rapidly manufacturing parts at both companies.

Applying RP to the manufacture of parts works best when unit cost is high and part quantity is relatively low. Also, it is best when the parts being manufactured are small but complex. The bigger the part, especially in height, the slower it is to produce on an RP machine. These systems are especially good at producing highly intricate parts as easily as simple ones, so the benefit of using this technology is the greatest when dealing with tough designs. Surface finish is a consideration, due to the layer-by-layer buildup of parts, so rapid manufacturing is best suited for products that do not have high demands in surface finish.

Many Markets

Other markets can benefit from rapid manufacturing (RM), such as prosthetics, dentistry, professional and collegiate sports equipment, antique restoration, museum artifacts, sculptures, custom aircraft, and military products. However, for RM to penetrate into these and other markets, a range of obstacles needs to be overcome. The first is the cost of material. Currently, RP materials range in cost from about $55 to $135/lb or more. As production volumes increase, materials prices will probably decline.

The removal of support material is a major issue, although the problem differs depending on the type of RP process and the shape and orientation of the part being produced. In the case of laser sintering, the support material is powder that must be cleaned from the part.

Photopolymer-based systems, such as stereolithography, have structures attached to them that require labor-intensive removal. The newer generation ink jet printing of photopolymer, such as that offered from Objet Geometries (Rehovot, Israel), permits the user to wash away the material. Likewise, Stratasys has developed a support removal system called WaterWorks that minimizes manual labor. Even so, removing the support material, regardless of the system, requires time, effort, and space that one must consider.

Process repeatability is another consideration. If three machine operators built the same part on three different machines, you?d likely end up with three similar parts, but not as similar as three injection molded parts. RP machines have been designed and used for building a wide range of prototypes, so I foresee a need to design machines from the ground up with manufacturing, not prototyping, in mind. It will make sense to produce machines for specific markets and applications. For example, hearing aids are very small parts and their manufacturers do not require the single-part build volumes needed by the aerospace industry.The changing of designer habits is another obstacle. For decades, designers have been trained to design for molding, casting, and stamping, as well as assembly. Adding draft to a part to prevent die lock conditions, for example, is not necessary when manufacturing on an RP machine. With RM, it will be possible to reduce the number of parts in a product by combining parts that would otherwise be assembled. Also, it is possible to produce working assemblies with today?s RP systems. CAD software products do not take into account the use of two or more build materials, nor do they allow for graded materials. Therefore, software developers will likely create new software products to accommodate these features.

The largest obstacle of all may be the culture and management at most companies. Management at some companies may not view this new method of manufacturing as a viable option. Others might discount it because it?s not conventional. But my hope is that change will be accepted. The results will be dramatic.