Proto Choice

Sizing up multiple micro prototyping options

Ready to prototype that new design? Today, there’s no shortage of options for modeling microscale parts and features. These include new and tried-and-true technologies that address key prototyping concerns, such as speed, accuracy, cost and quality.

On the other hand, most options also have significant downsides, forcing users of prototyping equipment and services to make tradeoffs based on what factors they think are most and least important in a particular situation. Fortunately, some suppliers have developed ways of improving their prototyping offerings that limit the sacrifices designers must make when selecting them.

Injection molding option

Sacrifices are certainly being made when designers choose traditional injection molding as the way to produce microscale prototype parts. While this process will yield a production-caliber part, the hard-steel tooling requirements make it relatively slow and expensive for making prototype parts.

In some cases, molders can cut time and cost out of the process, according to Aaron Johnson, marketing manager for Accumold, a molder based in Ankeny, Iowa.

To expedite toolmaking, for example, molders can opt for “pick-a-part” molds, which lack some of the automated action capabilities that speed up normal production molding. When the mold opens, the operator reaches in and retrieves the part—thus the name of the mold. In addition, prototype manufacturers can leave out some mold features that aren’t critical for the early stages of product development, Johnson noted.

Recently, Accumold implemented what it calls a Quick Mold Program that aims to cut normal moldmaking lead times in half. Employing QMP procedures, lead times are 2 weeks or less for simple prototype parts that can be produced by basic mold actions, according to Johnson. In the next phase of QMP, which will allow additional mold actions, lead times will be 3 weeks or less, he added.

Similar efforts to speed up prototype molding are being made by Empire Precision Plastics, a Rochester, N.Y., molding company. Like Accumold, Empire Precision can leave out some detail that would be included in a full-featured production tool but is not required for prototyping.

For example, some features can be made “steel safe,” which refers to the practice of leaving extra steel in some areas of the mold that can be shaped later to create required production features and dimensions. “When you’re manufacturing the mold, you don’t have to spend as much time on those features for the initial prototypes,” explained Markus Lettau, director of application engineering for Empire Precision.

Prototype jaw for a surgical instrument made via microresolution stereolithography process. Photo courtesy FineLine Prototyping.

In addition, production process add-ons, such as cooling-water lines and ejector systems, can be left out to speed up prototyping. How much time can be saved? On average, a production tool would take 8 to 12 weeks to complete, but Empire Precision can produce prototype tooling in 1½ to 3 weeks, according to Lettau.

No tooling needed

While even this relatively fast molding process requires time-consuming toolmaking, no tooling at all is needed to machine microprototypes using traditional processes such as turning, grinding and drilling.

For example, CNC machining of prototypes can produce stronger, more accurate prototypes than other methods; can be performed on a wide range of materials (including the material the part will eventually be made from); and can be used to create true-to-life models. However, the cost of machining prototype parts can be highly variable, growing with the complexity of the part. Machining is also generally slower, requires more human intervention and generates more waste than rapid-prototyping processes.

MiniMachine Inc. in Bend, Ore., has found innovative ways of prototyping parts with CNC machining. Prototypes are usually attached to a mandrel that comes from the same piece of material. This gives operators a way to hold and locate the parts during machining and inspection, said Mike Rosenboom, the company’s vice president.

Rosenboom advises those companies machining micro prototypes to do the job as completely as possible while the part is attached to the mandrel. “If you cut it off and then decide to do something else to the part, you may not be able to do it,” he said. “It’s monumentally difficult to fixture a tiny part to go back and do secondary operations.”

This is especially true for irregularly shaped parts, which require the construction of an oddly shaped fixture. In addition, there might not be enough surface area for an adhesive to hold a micropart in place when it’s subjected to cutting forces.

Another big problem encountered by Rosenboom is that machine spindles don’t turn fast enough to reach the correct speeds for the small cutter sizes that must be used when machining micro prototypes. He notes that newer machines have spindles that reach 15,000 rpm, but even that’s not nearly fast enough to work efficiently with very small drills. As a result, MiniMachine has adapted many of its CNC machines to use special air motors capable of 35,000 rpm.

Speedier prototyping

When maximizing speed and minimizing cost are your top prototyping priorities, the best choice will probably be a rapid-prototyping technique. RP is “dramatically” faster and less expensive than production toolmaking, potentially saving weeks of time and thousands of dollars in costs, according to Terry Wohlers, president of Wohlers Associates Inc., a product-development consulting firm in Fort Collins, Colo.

With no tool manufacturing or changes required, “you could turn out many versions of a small-part design within a day,” Wohlers noted. This lets users check their designs and make changes before committing to tooling, while the changes are still relatively inexpensive.

Generally speaking, RP systems capable of very fine detail produce parts made of photopolymers, according to Wohlers. These materials tend not to be suitable for actual products because of their inferior mechanical properties.

The oldest rapid-prototyping technology is stereolithography, in which a laser focused to a small point traverses the surface of a liquid thermoset resin, turning the liquid into a solid. Thin cross sections are created in this manner, one on top of another, until a 3-D part is formed.

Stereolithography is a good rapid-prototyping choice when surface finish is important. In addition, “it can give you the tiniest details, due to the finely focused laser beam,” said Rob Connelly, president of FineLine Prototyping Inc., Raleigh, N.C., which makes prototypes primarily via stereolithography.

On the downside, the process only works with special photopolymers. According to Connelly, these range from very flexible to very stiff, and some can handle temperatures up to 200º C before they lose their mechanical properties. “But there’s always a compromise in terms of temperature capability, moisture resistance and UV resistance, so you can’t get final production-part material properties with stereolithography,” he said.

Flawed but useful

Compared to molded products, many parts made by stereolithography are “very delicate” because they don’t have the same structural properties as thermoplastic parts, according to Lettau from Empire Precision. But they can be useful, especially in the early phases of product development.

“A customer might come to us with a stereolithography model that’s the first step toward making a production-type part,” he said. “The next step would be to go to a production-intent prototype mold that will produce real-world production parts that can be tested. Then you can make modifications and move toward a full production tool.”

Whether or not all these steps are necessary depends on the product. “If it’s a brand new product, many times you’ll take that initial prototyping step to get a good feel for what the part is going to be,” Lettau said. “If the part is a variation of an existing product, there may be enough data on the part to allow you to go straight to a production-intent prototype to see how the new features will function.” If a customer requests a stereolithography prototype, Empire Precision will outsource its production.

Though he maintains that machining produces stronger, more accurate prototype parts, MiniMachine’s Rosenboom also believes stereolithography is a useful tool that can be used to rapidly determine whether the shape of a design is correct. Like Empire Precision, when a stereolithography prototype is required to check engineering details before cutting begins, MiniMachine uses an outside supplier.

From an economic standpoint, Connelly believes stereolithography makes the most sense for complex micro prototypes. For CNC machining and molding, he noted, the price per part is highly variable, going from rather low to extremely high. For stereolithography, on the other hand, the price is somewhere in the middle but fairly constant, no matter how complex the part.

“If your part is a simple piece that’s easy to machine, then machining is a good option,” he said. “If it’s something so complex that machining or molding get to be difficult, then stereolithography can really shine.”

These prototype parts were produced using standard micromold hard tooling. Photo courtesy Accumold.

When prototype feature sizes get down to just a handful of microns, however, stereolithography becomes unusable due to limitations involving the laser beam and materials. Then designers must turn to micromolding and micromachining, which “take you down as low as you can go in this field,” Connelly said.

However, he added, his company’s new microresolution stereolithography system can handle smaller features than traditional systems, producing details down to 25µm in the drawing plane. The keys are equipment modifications that shrink the focus of the laser beam, as well as resin changes that allow the materials to work with a smaller beam.

Stereolithography has also changed for the better thanks to 3D Systems Corp. The Rock Hill, S.C., company—the inventor of the stereolithography process—has released the ProJet 6000 system, which is designed to make the stereolithography process easier for users. Instead of creating and loading huge and complex prototype-build files into the system, designers can submit build information from their laptops, then operate the system via a touch-screen interface.

“With this technology, you don’t have to have a Ph.D. in stereolithography to run the machine,” said Lee Dockstader, vice president of business development for 3D Systems.

Printer-based prototyping

In addition to stereolithography machines, Dockstader’s company makes 3-D printing systems that can be used for micro prototyping. Though the meaning of the term is sometimes broadened to encompass other technologies, 3-D printing mainly applies to systems that make 3-D parts using office-type machines that people can operate without extensive training.

Many 3-D printing systems use inkjet technology to deposit material in very thin, cross-sectional layers that eventually build up to form a complete 3-D product. Like stereolithography, 3-D printing is a relatively fast prototyping option, but it limits users to a small number of materials that don’t always exhibit the most desirable properties.

Why use 3-D printing instead of stereolithography? According to Dockstader, virtually any part shape can be produced using stereolithography, but supporting the features can be difficult. “When you get down to micro features, all the down-facing little bits of geometry need a support to keep them in place during the build,” he explained.

By contrast, easy support generation is a hallmark of multijet modeling, a 3-D printing technology that uses an inkjet print head instead of a laser print head. To produce a prototype, the jet head travels back and forth, laying down layers of the model material (an ultraviolet-curable resin) and wax support material at the same time. When the printing process is complete, the structure is placed in an oven and the wax is melted away, leaving the desired part shape.

With pixel dimensions of roughly 40µm and layers down to 29µm, multijet modeling is suitable for making micro prototypes. Applications include micro connectors, jewelry and any parts with small features that aren’t easily supported during the stereolithography process.

In addition to its support advantage, multijet modeling systems cost “quite a bit less” than stereolithography equipment, according to Dockstader. On the other hand, he added, stereolithography produces smoother parts than multijet modeling. In addition, he noted, small stereolithography parts have sharper, more clearly defined edges and corners than similar parts made via multijet modeling.

“With the inkjet printer, the drops are quite small, but it is a pixel-based system,” he explained. “So when you get down to very small features, they’re not quite as crisp as the ones created with stereolithography, which draws with a fine laser beam.”

Plastics no, metals maybe

Lasers can also be used to produce prototypes in a process called selective laser sintering. An SLS machine incorporates a high-temperature laser that draws on a hotbed of thermoplastic powder, sintering the powder into a solid. After each layer is produced, a roller deposits a fresh layer of powder on the bed, and the process repeats until the part is completed.

Because SLS involves engineering thermoplastics, the parts it turns out are stronger than stereolithography-produced parts. Nevertheless, Connelly points out that the process isn’t used much for microscale work due to the coarseness of the thermoplastic powder, which negatively impacts surface finish.

Prototype of a microfluidic device for lab-on-a-chip diagnostics made via high-resolution stereolithography. Photo courtesy FineLine Prototyping.

Medical parts printed on the ProJet HD 3000 production system from 3D Systems. Photo courtesy 3D Systems.

While the prospects for laser sintering aren’t bright in terms of plastic micro prototyping, it may soon be a contender in the arena of metal prototyping, thanks to a new technology called MicroLaserSintering, or MLS. Introduced last year by EOS GmbH, based in Munich, Germany, MLS sinters metal powder consisting of particles measuring less than 5µm, according to Joachim Goebner, MLS project manager for EOS.

The MLS system operates in a clean argon atmosphere, which is required because of the highly reactive metal powder. The process runs in a sealed process chamber that protects the operator from both powder and argon.

Parts printed on the ProJet HD 3000 show melt-away wax supports. Photo courtesy 3D Systems.

Featuring a focus diameter of less than 30µm, the MLS system can deposit layers less than 5µm thick. So far, MLS has only been used to process molybdenum and 316L stainless steel, but Goebner believes the system should be able to handle virtually all metals. He reports it can reliably produce walls 60µm thick and has sintered walls 32µm thick.

MLS is best suited for complex metal parts with thin walls and other small features. In making parts like these, Goebner said, MLS has an advantage over milling and other cutting technologies, which present problems for users who must fixture small parts and subject them to potentially damaging cutting forces.

At present, the main challenge facing Goebner and his colleagues is finding the process parameters for different materials that can be used in MLS. “This is very time-consuming and will take too long if we do it alone,” Goebner said, adding that EOS is currently looking for academic partners to accelerate R&D.

EOS is working with customers on several application-development projects involving medical devices and micromolds. The medical market is the main target for MLS, with others being automotive, aerospace and chemical/pharmaceutical.

Prototype gear made via EOS’ microlaser sintering process. Photo courtesy EOS.

If all goes well, EOS hopes to bring the first full-fledged MLS systems to market within the next 2 years. Due to the high cost of these systems, EOS executives believe it might be best to partner with selected manufacturing firms to create production centers that will provide MLS parts to those who want them. “We think that a limited number of companies will be willing and able to pay for this technology,” Goebner said. “But it will be much cheaper to pay for the parts.”

Whether companies buy MLS systems or the parts they produce, they’ll have to deal with a big challenge: part handling. “We can make parts that are so small they look like dust,” Goebner said. “Once, I lost some very small gears in my office, and I never found them.” µ

About the author: Bill Leventon is a New Jersey-based freelance writer. He has a M.S. in Engineering from the University of Pennsylvania and a B.S. in Engineering from Temple University. Telephone: (609) 926-6447. E-mail: [email]wleventon@verizon.net[/email].

Contributors

Empire Precision Plastics (585-454-4995) www.empireprecision.com

Accumold (515) 964-5741 www.accu-mold.com

Wohlers Associates Inc. (970) 225-0086 wohlersassociates.com

3D Systems Corp. (803) 326-3900 www.3dsystems.com

FineLine Prototyping Inc. (919) 781-7702 www.finelineprototyping.com

MiniMachine Inc. (541) 330-8641 minimachine.com

EOS of North America Inc. (248) 306-0143 www.eos.info