Cutting tool coatings benefit macroscale machining, so you might reasonably theorize that they’d be beneficial for micromachining. As long as the coatings are properly applied and thin enough to avoid blunting the miniature cutting edge, some researchers would agree with you. Then again, those same researchers have not yet definitively identified the benefits or the best way to coat microtools.

To understand how to more effectively coat microtools, universities are conducting research. Presented here is information about university research for depositing diamond and other coatings, determining preferred coating methods and examining how different materials react to coated tools.

Growing diamonds

One of the challenges of coating with diamond is getting it to adhere to the tool surface. Rather than depositing an interlayer on a micro-endmill for enhancing diamond adhesion, a consortium of researchers at the University of Wisconsin-Madison, the University of Pennsylvania and Argonne National Laboratory etched 2-flute, 300μm endmills using a solution of hydrofluoric acid, nitric acid and deionized water. That was done partly to create a mechanical interlock between the carbide substrate and diamond, according to Frank E. Pfefferkorn, assistant professor at UW-Madison’s Department of Mechanical Engineering.

The cobalt binder, which increases a tool’s ductility, weakens the bond between the diamond film and substrate and suppresses diamond growth by limiting nucleation, stated a paper titled “Diamond coatings for micro-endmills: Enabling the dry machining of aluminum at the microscale.” Pfefferkorn and Robert W. Carpick, who worked on the diamond-coating research when he was at the UW-Madison and is associate professor at the University of Pennsylvania’s Department of Mechanical Engineering and Applied Mechanics, wrote the paper with their graduate students and collaborators. “The main reason to remove cobalt from the surface is because it poisons diamond growth,” Pfefferkorn said.

The process requires selectively etching an optimal amount of cobalt without substantially weakening the already-delicate microtool. To prevent removing too much cobalt and sacrificing tool integrity, a tool must not have more than 6 to 8 percent cobalt content by weight. “We etch away all cobalt from a thin surface layer to prevent it from interfering with the diamond growth process,” Pfefferkorn said. “We control the depth to which we etch to minimize the impact on tool integrity.”

After etching, the UW-Madison lead team performs a seeding step by depositing small diamond grains on the substrate using an ultrasonic treatment with diamond nanopowder in acetone. The seeded grains act as sites where the diamond starts to grow (i.e., nucleation occurs). Agglomeration of diamond nanograins leads to uneven seeding and nonuniform diamond growth, so technicians perform an ultrasonic rinsing step in an alcohol solution to ensure that large agglomerations are removed, leaving uniform seeding, the paper stated.

Then, using a hot-filament chemical vapor deposition system the team designed and built, nanocrystalline and fine-grain diamond is grown on the tools. (Carpick defined nanocrystalline diamond as having grains from 10nm to 100nm and fine-grain diamond as having grains larger than 100nm but smaller than 300nm.) The deposition system includes a chamber where gases, specifically methane diluted in hydrogen, flow over a tungsten filament maintained at a temperature of at least 1,800�C.

The resulting coating is from about 60nm to 200nm thick. Diamond coatings for conventional tools are 2μm thick and thicker, which would be too thick for microtools. That’s because an uncoated microtool’s cutting edge radius is often less than 1μm. “The thicknesses that are used for larger-scale tools would be completely unsuitable,” said Carpick. “You would blunt the tool and therefore dramatically reduce its performance.”

As is it, a microtool’s features aren’t as precise as desired. “Micro-endmills already have cutting edge radii larger than we would like for the chip loads we’re taking when cutting,” Pfefferkorn said.

In addition to collaborating with researchers having expertise in hot-filament CVD, Pfefferkorn said the team chose that deposition technique because other CVD methods, such as plasma-based, tend to vary the coating thickness, with more material being deposited at the corners. “You get sort of a bulb at the sharp edges, like a dog bone,” Pfefferkorn said. “I’m not implying hot-filament CVD is the only method that can do this, but one reason we picked it is because it doesn’t have this dog-boning effect.”

Not only does the coating have to be thin, it must adhere to the substrate and be continuous and smooth. The latter attribute, though, can be difficult to quantify. “At this point, we are pushing the limits of what we can properly model, so we don’t know definitively how smooth it needs to be,” Carpick said. “We also think that a little roughness can be helpful because it may assist in preventing adhesion of workpiece material to the tool.”

Because nanocrystalline-diamond coatings can be so thin, they conform to substrate surfaces, including striations from the tool grinding process and gaps from the acid etching process that microcrystalline coatings would cover. “Microtools are plenty rough,” Pfefferkorn said. “We don’t need them any rougher.”

Endmilling aluminum

The group focused on machining 6061-T6 aluminum with the coated micro-endmills because, according to Carpick, industry desires to increase the material’s use in various components, including engine blocks. Additionally, aluminum sticks to tungsten carbide but not diamond because of diamond’s low friction coefficient and low adhesion.

For the machining experiments, researchers mounted an NSK-HES500 high-speed spindle with electric drive and ceramic bearings to the spindle of a Haas TM-1 CNC milling machine. The high-speed spindle was run at 40,000 rpm, and a fixed-feed rate of 500 mm/min. was used for all the experiments. Performance Micro Tool, Janesville, Wis., provided the endmills.

Machining was performed dry, but a humidity-control system fitted with two nozzles blew humid air across the tool tip. “Friction and wear are a lot lower when you have some humidity in the environment,” Carpick said.

“The cutting forces exerted on the diamond-coated tool when it’s cutting dry are lower than when we used an uncoated tungsten-carbide tool with metalworking fluid misting,” Pfefferkorn said, adding that aluminum chips adhere to an uncoated tool whether or not misting is applied.

Analysis of the cutting and thrust force data showed a considerable improvement in the amount of force required to dry-mill 6061-T6 aluminum with an uncoated endmill, a 0.5μm- to 1μm-thick fine-grain-diamond-coated endmill and a 200nm-thick nanocrystalline-diamond-coated endmill, according to the paper “Analyzing the performance of diamond-coated micro endmills.” The main cutting and thrust forces went from 2.14 newtons (�0.85 N) and 4.40 N (�0.44 N), respectively, for the uncoated tool to 0.49 N (� 0.09 N) and 0.34 N (�0.04 N) when adding a fine-grain-diamond coating. The nanocrystalline-diamond coatings further reduced the cutting and thrust forces to 0.18 N (�0.07 N) and 0.17 N (�0.02 N), respectively. This data shows the cutting and thrust forces for the coated tools were more balanced, while the uncoated tools exhibited a thrust force twice that of the cutting force. Reduction of the forces is attributed to the diamond coating’s low friction and adhesion.

Therefore, Pfefferkorn suspects that aluminum’s tendency to stick to tungsten carbide is a reason for an uncoated tool’s higher thrust forces. Continuous chips are unusual when endmilling—an interrupted cutting process—because a cutting edge only engages the workpiece for 180� of its rotation. However, the uncoated endmills occasionally produced continuous chips when cutting dry because a newly created chip adhered to the flute surface and was only moved after another newly created chip was pushed into it, essentially welding them together. “The chips must be pushed into each other with enough force to do that,” Pfefferkorn said.

The micromachining experiments also found that the channels created by a diamond-coated endmill show a highly patterned, uniform bottom surface, while an uncoated tool generated a sporadic surface finish. The nonuniform finish suggests a significant amount of heat generated during cutting. “Heat generated when machining is a concern for micro-tools, especially for high-speed machining,” Carpick said.

The research paper noted that these performance gains are only experienced as long as the diamond coating lasts. About 80 percent of the fine-grain-diamond-coated tools and 40 percent of the nanocrystalline-diamond-coated tools experienced delamination, which typically occurs after several minutes of machining. After delamination, a previously coated tool either performed similarly to an uncoated tool or suddenly and catastrophically failed. Therefore, further work needs to be done to improve coating adherence.

The heat is off

Regarding heat generation when micromachining, other university researchers arrived at a different conclusion about the amount of heat created. According to research conducted at Purdue University, microscale cutting tools don’t produce a significant amount of heat. That’s because they need to rotate at a high speed and any heat that’s generated is immediately removed with the chips, which are tiny but have a large surface-area-to-volume ratio, according to Mark Jackson, associate professor at the university’s Department of Mechanical Engineering Technology.

“The coatings are of no use in taking heat away because the heat that’s generated is very small,” he said. Jackson noted that the spindle speeds ranged from 250,000 to 750,000 rpm, depending on the workpiece material and loading conditions, and the temperature at the tool tip was from 27� to 33� C.

Grant Robinson, a former Purdue University doctoral student who conducted research on coatings with Jackson, concurred. “The coatings designed to reduce and remove heat at the macro scale are totally unneeded at the micro scale because heat is not an issue that contributes to tool wear,” he said. “The main contributor is mechanical wear caused by mechanical forces rather than thermal forces.” Therefore, he noted, coatings for microtools are only needed to provide abrasion resistance.

To determine the temperature increase when micromachining, Purdue researchers conducted finite element calculations and machined elements, such as sulfur, sodium and potassium, which melt at relatively low temperatures. “If we machine an element that melts at, say,

50� C, and we see signs of melting, such as small melted globules, then we can say the cutting temperature is around 50� C,” Robinson explained. “We didn’t see any sign of melting, so we concluded microscale cutting doesn’t produce a significant amount of heat.”

Because the cutting temperature is low, coolant isn’t needed, but Purdue researchers directed compressed air at the cutting zone to aid chip evacuation and oxidize the workpiece material. “If you don’t oxidize the metal quickly, the coefficient of friction increases—even with coated tools—and that contributes to an increase in temperature,” Jackson said. “This is because you have metal-to-metal bonding rather than metal-to-oxide bonding.”

The metals machined included 1020 mild steel, D-2 tool steels, copper, brass and various soft materials that strain-harden rapidly, such as niobium, tantalum and tungsten.

The researchers primarily tested commercially available tungsten-carbide endmills from 250μm to 750μm in diameter that were sent to a major tool coater for coating via physical vapor deposition. Those coatings included titanium nitride, titanium aluminum nitride, aluminum titanium nitride, titanium aluminum chromium yttrium nitride, chromium nitride and titanium aluminum chromium zirconium nitride. “We tested about 20 to 30 coatings,” Jackson said. Coating thickness ranged from about 300nm to 2μm to prevent blunting of the cutting edges. That compares to coating thicknesses from 3μm to 5μm-plus for conventional cutters.

Beyond coatings

In contrast to those who think coatings can boost productivity and extend tool life when micromachining, Robinson feels microscale cutting tools must be fundamentally redesigned to create chips effectively and micromachine properly. That’s because conventionally designed tools machining at the microscale need to overcome the problem of edge-radius-to-chip-thickness ratio. He explained that for a given edge radius, there is also a given minimum uncut chip thickness. “In other words, the material has got to be a certain depth for that edge radius in order for a chip to form,” Robinson said.

What happens, he continued, is that a microtool rotates so fast that it might only advance 1nm per revolution, which is an insufficient distance for the uncut chip thickness to have built up to the critical level needed for chip formation. As the tool rotates and advances, it burnishes and rubs the metal rather than cutting it until it reaches the critical value to form a chip. It may take several hundred rotations—or not. “I don’t think anyone has found out what the critical number is for the edge-radius-to-chip-thickness ratio,” Robinson said. Without that information, he added, coating a conventionally designed microtool further impedes it from cutting properly.

“In the area of micromachining, there are lots of unanswered questions about what the machining mechanism is,” said Jackson. “It’s probably going to keep me busy for the next 30 years.”

Coating options for microtools

Tool coatings can be particularly beneficial when micromachining difficult-to-machine materials, according to Jeff Davis, vice president of engineering, Harvey Tool Co. LLC, Rowley, Mass. “Those nasty materials with lots of nickel [and] lots of cobalt very often demand a coating,” he said, adding that the same can’t be said for others.

“A coating is definitely not necessary or mandatory when cutting aluminum or plastics. Aluminum cuts all day, every day with an uncoated tool,” Davis continued. He noted that the exception is a production shop that wants to minimize tool changes. In that case, a suitable coating would be zirconium nitride or titanium diboride, deposited via PVD.

Gary Lake, president of CemeCon Inc., Horseheads, N.Y., a provider of coating services and systems, concurred that titanium diboride is an appropriate coating when machining aluminum, but only when the workpiece material’s silicon content is less than 10 percent. When there’s more than 10 percent silicon in aluminum, TiB2 cannot effectively prevent the adhesive transference of material to the tool. Above 10 percent silicon, the workpiece’s abrasive nature would be reason to apply a CVD diamond-coated tool, he said.

For any application, the majority of coating companies use cathodic arc technology, according to Davis. That’s because up to 90 percent of the target material that’s vaporized and deposited onto tools is used, so less material is wasted, compared to other methods. “Also, the kinetic energy associated with this type of process leads to excellent adhesion,” Davis said.

The drawback to the cathodic arc process is that it deposits macroparticles along with the smooth coating. Davis described macroparticles as “molten globules,” which are titanium, a common coating element. “It’s almost like a little bit of splatter.” This splatter is not likely to impede chip control for a larger tool. But as tools get smaller, the negative impact of macroparticles becomes more pronounced, requiring a tool coater to adjust the process to minimize their size or avoid depositing them, according to Davis. He added that another option is to remove the texture after coating while maintaining the coating’s integrity.

“If the macroparticles stay the same basic size, then you’re getting texture on your tool that is not as slippery and likely to catch chips and cause chip packing,” Davis said.

“Macroparticles are totally unacceptable in a microgeometry,” Lake said. “With cathodic arc, you will end up with macroparticles of the metallic phase sticking to the surface. Because of microtools’ fragility, macroparticles cannot be polished off as is usually done on larger tools.”

Instead, Lake suggested that ionization sputtering technology, which CemeCon offers, is more suitable for coating microtools because it deposits a smooth, thin coating without macroparticles. “The coating thickness you would want to stay at is probably between 1μm and 2μm,” he said.

Lake added that carbide micro-endmills are primarily coated to enhance productivity, especially those used to machine difficult-to-machine materials. “I would expect a performance increase if the tool is coated properly, the same as it would be for any other carbide tool.”

Setting the technical challenges of coating microscale cutting tools aside, many commercial coating companies are hesitant to coat them because the delicate and brittle tools are prone to breakage during handling. According to Lake, there are at least three tool-handling steps during the coating process: individually removing tools from their packaging to place in a rack for cleaning, transferring them to the coating fixture and returning them to their packaging for shipment back to the manufacturer. “If I give you a good tool and you break it, then I expect you to buy it,” he said.

That wouldn’t be an issue if microtool manufacturers had their own coating equipment, but Lake noted that most don’t. They instead rely on outside sources that can damage the tools at an unacceptable rate, he said. As a result of the handling and technical challenges, Lake estimated that about 95 percent of microscale cutting tools remain uncoated.

Harvey Tool outsources coating of its microtools, and Davis concurred that it’s difficult to find a coating company that can properly handle tiny tools and is willing to coat them.

“The smaller a tool gets, the more likely the coater is to break the tool,” he said.