The original role for rapid prototyping was simply to confirm the shape and feel of concept products. But as the technology changes, so has its role in design. Surgeons, for example, are using rapid-prototype (RP) models of bone and tissue to brainstorm strategies for surgeries. Complex organic models of skulls and tissue can guide the design of supports and bracing to repair fractures.

As more contract manufacturers and design firms use the equipment, the more uses they find for it. For example, a design team at B. Braun OEM/Industrial, Bethlehem, Pa., ( used RP parts to stimulate communication between far-flung company sites tasked with the design, manufacturing, and marketing of a complex pump.

The technology is changing fast with new materials that are more rugged and flexible than those from just several years ago, and equipment that works on desktops in offices. What's more, best practices for using rapid prototypes have evolved as more companies work the parts into their design departments. Here are a few of the best practices companies have discovered that wring more out of parts than simple go no-go decisions.

A few best practices

When to build a part of course depends on each project. But a good time is just before a group discussion, suggests Joel Bartholomew, lead designer at B Braun. “Passing a part around is a lot easier than sitting in front of a computer screen trying to explain how big it might be,” he says. In many cases team members cannot visualize a new design even from a 3D drawing. There are no set rules as to when to build RP parts, but when in doubt, err on the side of getting them. “Clients like to hold something in their hand and RP technology can produce a shape to hold in a matter of a hours,” he says.

Pass the part to as many on the team as possible. Engineers and designers like to get RP parts to sample their size and get an idea of a part's form, fit, and feel. Questions to ask include: Does the shape feel the way you wanted it to? Do joints allow easy attachments? Is there anything obstructing operators as they use the product?

“Send parts first to the marketing group and customers and ask if the design is headed in the right direction,” says Bartholomew. Then pass them to production groups and others such as R&D and quality to get their feedback.

“Make sure the manufacturing people examine final configurations,” he says. “They can use RP parts to design assemblies or welding fixtures. The parts are quick and inexpensive, so it's not a huge loss if they don't work,” he says.

Let patients and doctors handle parts as well. There may be focus groups that want to look at it for color and appearance, for photography, and for financing sources that need to “see” an idea before opening a checkbook.

And use the parts in packaging evaluations. “RP assemblies in final configurations are most useful here,” he adds.

Build the whole assembly when possible, not just a part. This allows testing assembly procedures. Bartholomew's pump, for example, let users evaluate the task of attaching several tubes into it.

Don't ignore other prototyping methods such as milling or vacuum forming. You might call these methods conventional prototyping, but they can still be fast. Engineers at photoetching companies, for example, can produce thin metal parts in as little as 24 hours. These can be complex parts that might snap into an RP housing as electrical contacts to test functional prototypes.

Vacuum forming, for instance, pulls a warm plastic sheet around a mold to create shapes in the material. It's another way to make many prototype parts and packaging.

Consider a step beyond the RP assembly and before final production. “Many RP materials fall short when used in field work or to demo the efficacy of the concept,” says Warren Haussler, president of design firm Keck-Craig Inc., Pasadena, Calif., ( Functional prototypes, those made with production quality materials, can be built for clinical testing in two to three weeks.

“After a design is nearly set, we go to prototype tooling of P-20 steel to prove out the design,” says B. Braun's Bartholomew. “These are good for about 100 parts. The goal is to do the proof of function and first launch with the P-20 molds. Then if something is too tight or pull forces too high, modifications can be made in tool-steel molds.”

Don't get hung up on moving as fast as possible. “Speed is not always useful,” says vice president of Keck-Craig Henry Keck. “A regular development cycle might be three or four months on the low side. But the designing mind may not work fast. People need time to think problems through and to make good decisions. The first solution is often not the best one. With sufficient time you can settle on better ideas because your subconscious has been working on the problem,” says Keck.

Think in terms of other materials as well. “Even models cut by hand from Styrofoam can show general shapes and sizes,” says Haussler. “Then progress to an RP case with circuits and other devices.”

Best practices in action

To show how the best practices apply to the real world, our experts describe how they have used the guidelines. “For example, we developed a peristaltic pump for filling bags with nutritional products. A disposable section of the pump was designed in the Allentown, Pa. facility and is built in an Italian factory. Additional hardware comes from Dallas and the marketing effort is out of Irvine, Calif. The project concluded with production of 13 molds so every part called for an SLA version. When in development, we mailed the latest version of RP parts to each facility to communicate better. It worked out well. It took four or five design iterations till we had something everyone was happy with,” says Bartholomew.

The disposable part required a lot of attention. “It is a key to the pump because it controls how nine tubes connect to it. To avoid user confusion, each line connects a different way, so it's goof proof,” he says. That feature is necessary because a mix-up in connectors would produce a costly cleanup and a safety risk.

In addition, production engineers benefited from the effort by making plans to weld two particular pieces. But in the design phase, there is no production line. “Nevertheless, we sent the manufacturing people RP parts and they planned the entire production sequence,” he says.

The ergonomics of assembly also benefited. “We gave the pump prototype to line workers and let them assemble the device to see if their hands could get inside. That prompted changes,” says Bartholomew.

A hand-held massaging device developed by Keck-Craig presents another perspective. “A client doctor invented a massager and wanted six working prototypes for evaluations,” says Haussler. “A cost comparison showed that six sturdy units produced by traditional methods would cost about twice that of RP methods. The client preferred the RP route so housings were ready in about 24 hours. Shortly thereafter we added the electronics and batteries. The understanding was these would be rather fragile. They were. Three came back after physical evaluations bandaged, patched, and barely working,” he says. Still, the broken housings pinpointed weak spots that needed strengthening.

A final example comes from Medtronic's Sofamor Danek prototype lab in Memphis, Tenn., ( The company develops spinal and cranial medical technologies. “Surgeons often come in with design problems,” says design engineer Richard Franks. “They'll explain their need to an engineer who models a solution, makes a rapid prototype, and has it in the doctors' hands the next day.” The company has in-house access to RP machines that make parts by fused deposition modeling (FDM) and photo-polymer jetting.

Company engineers recently designed a ratcheting counter-torque surgical instrument. It fastens set screws to a corrective implant on a patient's vertebrae and breaks off the screw heads at a preset torque. Existing methods required simultaneously using two separate tools. One drawback was a violent impulse from a set-screw head breaking off. Surgeons wanted to eliminate the jolt.

Franks' team produced an RP of a ratcheting device in FDM and the extension assembly in the UV-cured photo-polymer. The extension is two concentric tubes that slide one in the other. Franks used the UV-cured photo-polymer for the extension because of fine detail on the inner and outer diameters, and it produces the smoothest possible surface finish. The FDM machine built a working, durable polycarbonate ratchet that withstood testing on steel set screws and required only one hand to control.

After sending the counter-torque ratchet to three hospitals, Sofamor Danek learned that the tool design could be improved by rotating its handle 90°, information it might not have learned without working prototypes.

How prototypes cut time off surgeries

Generating a rapid-prototype model of a fractured skull or jaw lets surgeons plan more efficient repair operations. Associate Professor of Plastic Surgery at the University of New Mexico, Dr. Jon Wagner says that with the model of a jaw fracture, he can prebend fixation plates that will reinforce the fractured bone and more accurately select screws to hold the plate in place. “In addition, operations are shorter by a third to a half,” he says. And OR time runs about $2,000/hr.

To prepare for a typical operation, Wagner takes the CT of a patient, turns the point data into a solid model, and exports it as an STL file to a 3D printer in his office. Ready access to the equipment means models are available in five to six hours rather than two or three days a service bureau would need. “Then I can take the model apart at my desk, recreate the fracture, and put it back together with wax in a correct anatomical form,” he says. Standard titanium plates with holes are then bent to conform exactly to fracture areas on the model. “You can align these plates on the model and bend them precisely because you can clearly see from any orientation how they sit, making sure they touch the bone.”

“Ordinarily, a fracture would be put back together in the OR as best as possible and then a surgeon would spend an hour or so bending plates while the patient is under anesthesia. What's more, soft tissue covers the bone and it's easy to get disoriented,” he says. “So the benefits of a physical model in the OR make it easier to maintain anterior, posterior, and lateral orientation, and shorten the time and lower the cost of keeping a patient under anesthesia. Prebending the plates also ensures that they are formed as accurately as possible, and it probably eliminates a few redo surgeries.”