Skip to content

Pushing the Limits with Ceramics

Ed Sinkora
By Ed Sinkora Contributing Editor, SME Media
Tooling up a four-spindle Danobat Overbeck IRD to grind ceramic necking dies. (Image provided by Danobat)

Development of breakthrough carbide-grade opportunities in machining is practically tapped out, according to Bernie McConnell, executive vice president, commercial, of Greenleaf Corp., Saegertown, Pa.

“Outside of doing different things with coatings, geometry, and edge preps, there are only so many ways you can mix and blend the carbide materials,” he said. “Most of the exciting technology development is coming on the ceramic side of the business.” That’s for both cutting tools and parts. “Whether you are talking about going super fast, facing extreme heat, or abrasive wear characteristics, ceramics are pushing the envelope in diverse applications. Today’s ceramic capabilities are unbelievable, and continue to get better.”

Cutting Tool Applications

Ceramics remain an excellent solution for high-temp alloys and abrasive materials, with the ability to achieve eye-popping rates. In fact, milling speeds in high-temp alloys range from 2,800 to 4,400 surface feet per minute (SFM), according to Robert Navarrete, national product specialist for parting, grooving, and turning at Iscar USA.

“For milling, I’d typically start at around 3,400 SFM, gauge the wear on the insert and adjust the parameters accordingly,” he explained. “For turning, depending on machine capability, you’d be between 600 and 1,100 SFM in high-temp alloys.”

But this isn’t anything new. Navarrete noted that a number of shops are familiar with these capabilities based on experience with whisker-reinforced ceramics, which is a decades-old technology. Instead, he touted newer silicon, aluminum, oxygen, nitrogen (SiAlON) ceramics. “You can run the same whisker-like parameters at a lower cost, because SiAlONs are cheaper to produce than whiskers,” Navarrete said. “I’ve never gone into a shop, spec’d out and quoted a SiAlON and been more expensive than a whisker. It hasn’t happened.”

The price difference can easily be 25-30%, he added, citing the case of a whisker-reinforced ceramic having a seven-minute tool life, versus six minutes for the SiAlON. But, with a 25% lower price, that works out to a 7% savings in favor of the SiAlON. “Over a year of usage, that’s a huge savings,” Navarrete asserted. “Ceramics aren’t cheap.”

Another plus: SiAlONs are “much more forgiving than a whisker,” he continued, and as a result they can run at lower speeds, without the plasticization that is normally required for whisker ceramics to cut. “If I had a part with an interrupted cut, like a forging with a scale or any type of crust, I would run a SiAlON, as opposed to a whisker.”

Iscar offers “tough” SiAlONs and “hard” versions. The latter is advertised as “whisker-like,” because it mimics the hard, brittle nature of whisker-reinforced ceramics.

Meanwhile, Greenleaf’s premium performance whisker-reinforced ceramics continue to have value-added applications, owing to their higher hot hardness and wider feed-rate capabilities, according to Martin Dillaman, the company’s global manager for engineering and applications. For example, he said, “you typically would not finish with a SiAlON in heat-resistant alloys, because they run at a higher feed rate that exceeds the surface finish requirement for the part. So you have to use a whisker-reinforced ceramic if you want to keep the speed up and finish.”

Dillaman added that many aerospace companies perform sonic testing on their heat-resistant alloy parts, checking for minor imperfections in the surface. But in every case he knows of, “the surface finish left by a SiAlON will not pass sonic testing, whereas our WG300 and coated WG600 whisker-reinforced ceramics have been approved for this testing. So there are definitely some areas where whiskers are not going to be replaced by a SiAlON.”

Greg Bronson, Greenleaf’s sales director for the Americas, elaborated. “Because the edge of whiskered-grade ceramics holds up better, you don’t get the same level of heat generation and the smearing of material that could cause parts to fail additional surface testing.”

Greenleaf’s GF1 chip form also contributes to finishing success. The form is ground into the top of the insert, up to the cutting edge, and is offered on all Greenleaf’s whisker grades. “It helps reduce cutting pressure, which reduces the chance for any type of failure during the quality inspection,” Dillaman said.

In general, the higher hot hardness of whisker-reinforced ceramics, versus SiAlONs, enables WG300 to run in heat-resistant alloys at 10 to 20% higher speeds, but at a reduced feed rate, which is required for finishing, according to Bronson. “Our coated WG600 would go another 20% faster, so you’d be about 30-40% higher than a SiAlON. And our nano-coated WG700 is another 20% faster, so you’d be close to 50-60%. You can run in Inconel with WG700 as high as 1,500 SFM, and SiAlONs are going to melt long before you get to that point,” Bronson said, adding that WG700 also excels in interrupted cutting of Rene.

While chemically similar to a SiAlON, Greenleaf’s well-known XSYTIN-1 ceramic is a phase-toughened silicon nitride. “The way it’s pressed causes a crystalline structure to grow inside the material,” Bronson explained. As a result, he said, the company can “generally run 20-30% higher feed rates with XSYTIN-1 than most SiAlONs,” even though XSYTIN-1 has a similar speed limit, because the binders break down at a similar temperature.

3D-Printing is Driving Demand

The newest call for ceramic cutting tools comes from the need to clean up the build plates after metal 3D printing. Bronson listed cobalt chrome for medical products, Inconel 718 for aerospace, Haynes 282 for rocket engines, Rene 220 for power generation, and Rene N2 as difficult alloys the company is being asked to deal with. These are difficult to machine materials when forged, and Bronson explained that printing them adds another degree of difficulty because the laser-sintering process leaves scale between the layers.

“That beats up carbide significantly. You get a lot of chipping and excessive wear. So customers were hoping ceramics could get through that.”

Such printed parts are usually cut off the build plate with wire EDM, leaving 0.200-0.300" (5.08-7.62 mm) of support structure to remove, Bronson said. “Then they want to clean up the base plate so they can print again. So they’ll take 15 or 20 thou off the plate.”

Greenleaf’s first foray was removing 31-35 HRC Haynes 282 printed on a stainless steel plate. Owing to the part configuration, the print required a highly variable set of gates, risers and support structures, Bronson recalled. “Some of them were thick, some thin. Some were tall. They were all over the place. So, by nature, it was very interrupted. Plus, Haynes is very abrasive.”

Oddly, he added, the softer stainless material presented an even bigger challenge for carbide. Once a tool designed for the tougher material hit the stainless steel, it would fail. That’s probably because the gumminess of the stainless steel inhibited chip formation. Greenleaf’s solution was to use the XSYTIN-1 ceramic in a small face mill with RNG45 inserts. The strength of XSYTIN-1 allows the user to “get away with a sharper edge,” Bronson said. “That allows you to get through the hard material, yet it doesn’t smear in the softer material, thanks to the sharper geometry.”

What’s more, where the abrasive nature of the material had forced the customer to take very shallow depths of cut, amounting to only 0.015" to 0.025" (0.38-0.635 mm), Greenleaf increased the depth of cut to between 35 and 50 thou. “And instead of having to take maybe five, or six or seven passes with carbide, we can take two or three with ceramics and clean up the entire plate,” Bronson said. In fact, on six such projects on which he’s worked, edge damage usually required indexing the carbide tool after a single pass through the laser-sintered material. Whereas the ceramic “lasted through every pass we had to make,” Bronson observed.

Edge Prep is Critical

A 2021 Manufacturing Engineering article on carbide tools, “An Improbable but Powerful Solution,” included the curious fact that ceramic inserts can look like hell and still cut well. The key to this capability, Navarrete explained, is the condition of the edge. “There is so much heat thrown into these inserts, especially if you’re running them dry, they look burnt. But the edge prep is still there. You might take it out and observe that it’s dark or charred, but once you start to cut again, it burns off—you could almost say—like a barbecue grill. Once the machining temperature threshold is reached, material residue heats up and comes off the ceramic and the edge prep is engaged again.”

Rather than worry about the misleading appearance of the insert, or focusing your attention on the edge, Navarrete recommended simply setting a time limit. “Ceramics are very shock sensitive, but also predictable.” He suggested starting with a relatively short, programmed time in cut (TIC) and then inspecting the edge. “I typically start with three and half minutes as a starting gage point. Once we find a sweet spot where everybody is comfortable, we can adjust accordingly by either adding or decreasing the insert time in cut.

We want to find a TIC where we can then safely index each time and know it’s not going to fail. Milling can be anywhere from seven to 10 minutes, predictably.”

Of course, all this depends on having the proper edge prep to begin with. “For finishing cuts in high-temp alloys, we’d spec out a honed edge,” Navarrete said. It’s the most free cutting. For “intermediate cutting, like semi-finishing or semi-roughing, we’d go with a chamfer, which we call a T-land. It’s a little stronger. If it’s purely roughing, with a lot of material coming off, or we’re going to machine scale or crust, we go with what we call a TE edge prep. That’s a chamfer and a hone. It’s an even stronger edge, but it also means more tool pressure. It’s meant to take the beating.”

However, he pointed out, the fragility of ceramics is often exaggerated. “Ceramics can take a beating, you just have to program it and approach it properly.” On that note, he cautioned, any switch from carbide to ceramic tooling requires reprogramming the cut. “You should not plug a ceramic tool into a carbide program.”

Grinding Ceramic Tools

Neither Greenleaf nor Iscar shares the secrets of how their tools are made. But Scottsdale, Pa.-based Better Edge gives a glimpse into the challenge of grinding the final cutting geometry. While the company focuses on specialized carbide cutting tools, Better Edge recently got a project to grind a half-inch diameter, short-LOC, six-flute end mill for an aerospace customer. The tools came in worn, so Better Edge cut off the cutting portion, then ground a new geometry into the remaining blank.

The Danobat Overbeck IRD grinds 81 Rockwell ceramic necking dies with a form accuracy of 0.0002” and an ID surface finish of 2 µin. (Image provided by Danobat)

Despite having no previous experience with ceramics, the company worked through the required deviations from its carbide methods quickly. There were just two problems. The first, explained Brian Shaffer, vice president of operations and quality, was wheel wear.

“Initially, we were using a Toolgal wheel with a harder, high-performance hybrid bond, RM644, our ‘go-to’ bond when looking for longer wheel life for fluting. But in fluting ceramic tools, it was wearing faster than we would have liked.”

To help solve the problem, Better Edge reached out to David Ginzburg, who is the president of Elberton, Ga.-based Toolgal USA Corp. Ginzburg advised trying the newer RM769G hybrid-fluting bond. That might have seemed counterintuitive, because RM769G is a softer bond, but it actually worked better and delivered longer wheel life. According to Ginzburg, the wheel’s different diamond quality and the bond’s better diamond retention was more free cutting and suitable for fluting ceramic. Cycle times were just a bit slower than carbide, roughly a half-hour per tool, Shaffer added.

The second problem with grinding ceramic is that it produces a fine, milky swarf that clogs coolant filters faster than carbide. Better Edge addressed this by changing filters more frequently, but decided that if they were to continue grinding ceramic they’d have to dedicate a filter to that application.

Getting a Mirror Finish

As mentioned earlier, ceramics can also serve as wear parts, or other components that must survive tough conditions. Greenleaf operates a “technical ceramics” division for such applications, and trains its tooling staff to keep an eye open. McConnell said one engineer recently noticed an operator changing out the carbide guide wheels on the measuring device for a large roll-turning mill. This gauge must continuously ride along a huge steel surface, measuring roundness and size, so there is significant wear. As a solution, Greenleaf offered the customer ceramics that last 10 to 15 times longer.

As another example, McConnell pointed to guide bushings for deep hole drilling. “In the past, these would have been carbide or a hardened steel. Now we’re making them out of ceramics and they last 20 times longer. So the possibilities are sort of limitless.”

If you need a third example, you can’t do much better than a line that produces 2,000 aluminum beverage cans per minute. According to Jim Beavers, sales manager for Danobat Group in Rolling Meadows, Illinois, the necking die that forms the can’s walls to accept the lid was traditionally made of carbide, “but over the years, they’ve discovered that HIP (hot isostatically pressed) ceramic has a longer life, because they have better heat transfer and wear capability.” The version Danobat sees most often is yttria tetragonal zirconia polycrystal (YZTP), which Beavers reported as “coming in at 81 on the Rockwell scale, and 1,300 on the Vickers. So certainly a challenging material.”

That’s not all. The die has a fairly complex form, inside and out, plus a slot. “The profile tolerance is two tenths. And the other important aspect is the tangency on the radius versus the angle where they intersect,” explained Beavers. It gets worse. “Customers are requesting the inner bore come from the machine down to two micro inches. So we have to deliver a polished finish.”

Danobat conquers these challenges with its Overbeck IRD machine, equipped with a four-spindle wheel head. Grinding the form requires three-axis interpolation (X, Z, and B0), as well as all four spindles. For example, grinding the outside of the form requires a large OD wheel, while much smaller ID wheels tackle the inner form. Getting the required surface finish also necessitates using multiple abrasives for roughing as well as finishing, added Beavers.

“Many people don’t appreciate how tight the dimensional and form tolerances are on these parts,” observed Daniel Rey, president of Danobat distributor Rey Technologies, St. Charles, Illinois. “And one reason Overbeck has been successful in meeting them is by using solid natural granite as a machine base. Among other things, this manages any temperature fluctuations in a shop. Overbeck also uses linear motors, which is another edge over some of the competitors.”

In addition, Danobat’s Beavers said paying the “utmost attention” to the thermal stability of the machine means incorporating liquid cooling in the workheads. Coolant filtration also must be at “the highest level,” he said. So, in addition to a standard system, the company adds canister filters rated down to five microns, plus chillers to maintain a constant temperature.

So, there you have it. Ceramics offer outstanding solutions to difficult machining challenges, and the challenge of machining ceramics themselves can also be met with the right technology and a willingness to push the limits.

  • View All Articles
  • Connect With Us

Always Stay Informed

Receive the latest manufacturing news and technical information by subscribing to our monthly and quarterly magazines, weekly and monthly eNewsletters, and podcast channel.