Ceramic feedstocks simplify design, molding problems

Thanks to a chemical gel, one of the barriers to widespread ceramic injection molding recently became a thing of the past. While the basic molding technology has been in place for more than 30 years, feedstock materials haven't exactly been easy to use. For one, removing the polymer or wax binders can be time consuming. For another, this debinding step produces a brittle as-molded part.

Instead, imagine molding ceramic parts that gingerly drop into a bin as they are ejected from the mold, stiff but elastic enough to hold their own before firing. Then envision reducing costs by as much as 40 percent over traditional ceramic processing. When Allied-Signal Engineered Materials created a new binder, made from water and agar (a seaweed-derived polysaccharide not unlike Jell-O), these visions became reality.

How real? At the company's Autolite plant (Fostoria, OH), a 22-ton production molding machine pumps out 1 million parts annually using the water-based feedstock. During a visit, IMM spoke with Cliff Ballard, director of powder injection molding ventures for AlliedSignal, about the potential for this technology.

"Compared to the traditional shape forming ceramic processes, molding offers to reduce the number of steps required and, thus, take out cost," says Ballard. "But polymer and wax binders have held back growth. Parts have 40 to 50 percent binder content by volume. Removing it generates toxic effluent and may take days. The end result is a brittle part that must be carefully handled. In addition, part thicknesses are limited to less than .25 inch."

By comparison, Allied-Signal's water-based system requires no separate debinding step - the water simply evaporates after molding, leaving intergranular pores open. Of course, concerns over toxicity are also eliminated. As for the agar binder, it adds only 2 to 3 percent by weight to the feedstock. According to Ballard, there are no part thickness limitations, although thickness does determine drying time. And like polymer/ wax systems, tolerances can be controlled to ±.3 percent.

What really sets this new feedstock apart, from a designer's point of view, is the greater freedom it allows. "We're producing parts that were formerly impossible to mold," says Richard Schultze, manager of ceramic applications. At the plant, workers are able to mold spark plugs with a 7-mm wall thickness, vs. 12 to 14 mm for standard processing. Parts with combined thick and thin cross sections, undercuts, and complex geometries all benefit from the moldability of the feedstock. As an example, he cites a turbine blade produced via traditional methods at a cost of $1300. "Our customers are now molding prototypes of that part, and the price could drop to between $50 and $150."

Savings are a big driver for ceramic molding. Last year, a vendor told Autolite it could no longer supply an intricate component, formed traditionally in two half shells, because the cost was prohibitive. Jason Hogan and David Lyons, part of the CIM team at Autolite, had a prototype mold running parts within eight days after the announcement. This year, Lyons estimates the company has saved more than $250,000 on these parts. "In fact, because of the freedom this process allows, we were able to add features that the vendor could not produce ," says Lyons.

Differential shrinkage rates for the new molding compound are greater than those for most thermoplastics, ranging from 16 to 20 percent (depending on the material). However, notes Ballard, these are the same rates found in traditional ceramic forming. This has a lot to do with the composition of the new ceramic molding compound - by volume, 53 percent ceramic powder (alumina, zirconia, or silicon nitride), 47 percent polymer dissolved in water. It is this latter component that makes the mixture act like a thermoplastic.

Processing is another area where the Allied system saves money. Cycle times are similar to those for thermoplastic parts. But because injection pressures don't exceed a 500-psi maximum, and temperatures rarely go above 185F, prototype parts can be molded using aluminum- filled epoxy or even SLA masters. There is no mixing in the barrel; instead, material is moved along by a screw with no compression, then injected through the nozzle at a consistency similar to toothpaste. Parts bounce out of the mold, with strengths after drying about four times higher than those of conventional green ceramics. Molding machinery need not be modified for the process. At the production facility in Fostoria, technicians use three 22-ton Boy machines and a 27-ton Arburg.

Melt flow rates for the ceramic are difficult to measure using traditional thermoplastic equipment. However, spiral flow tests confirm that the material will fill complex geometries at low pressure. A turbine vane test part, 7 inches long with a feathered edge, had no problem filling. When it comes to wear, the compound acts like a 30 percent glass-filled nylon. Total wear for production molds, in stainless 440C, rarely exceeds .004 inch for 500,000 shots.

Material properties for the first commercial Allied compound--AS194 alumina--are equivalent to traditional 94 percent Al2O3 materials. With an initial raw-material cost of $5.50/lb, the gel-based material can be easily recycled in-plant, either by grinding parts and adding water, or by adding still-wet scrap directly to the hopper.

The gel technology developed for this first ceramic compound is also being applied to metal injection molding materials with similar benefits, Ballard adds. He shares a product development roadmap, which targets commercial release for additional ceramics as well as stainless steel products within the next 12 months. Stay tuned.