Superabrasives In The Groove

CBN abrasives have been around far too long to be regarded as new technology. But the term "emerging" may still fit because, while the potential of superabrasives is legend, the application has often proved tricky and in many cases downright disappointing. Shops that are willing to pay up to ten times more for a CBN (cubic boron nitride) wheel than its conventional counterpart rightly expect a big payback in terms of much faster material removal, or more parts per wheel, or better form consistency--and usually all three. Unfortunately, too many shops have found their experience with CBN, well, less than satisfying.

Wes Lee thinks the problem is not with the abrasives, but with the processes by which they are often applied, and in particular, the machines. Rather than slapping new high tech wheels on old machines, he contends engineers should begin with the inherent material removal capabilities of CBN, and then design a process with the speed and rigidity that lets the abrasive approach its true cutting potential. Properly done, material removal rates are possible that heretofore have been outside the realm of grinding--or perhaps better stated, grinding-like--processes. How much better? Mr. Lee says that in many cases superabrasive machining can achieve metal removal rates more than six times that of comparable creep-feed processes with conventional abrasive wheels.

All that said, it comes as little surprise that Mr. Lee's company, Edgetek Machine Corporation (Meriden, Connecticut), is in the business of manufacturing superabrasive machining systems. Indeed, Edgetek is one of the pioneers of this technology, which is only now just becoming to be widely commercialized. Edgetek was founded on the superabrasive machining concept in 1988 by Ed Elie. The company shipped its first two production machines in 1991 and currently has 45 machines installed at 30 sites. More than half of these machines have been put in place since the beginning of last year.

While it appears that superabrasive machining is finally on the way to finding its niche, Mr. Lee argues long and hard that they are just beginning to scratch the surface of the broader potential of this technology. The best applications right now, at least in the case of Edgetek's machines, mostly have to do with slotting operations. And the company has numerous examples of situations in which the move to a properly applied superabrasive process has produced dramatic cost reductions over conventional milling and grinding operations.

"We've seen at least a 40 to 50 percent increase in throughput on every job we've done," says Mr. Lee, in comparison to conventional abrasives. But he is also quick to point out that thinking only in those terms--CBN versus conventional grinding processes--will very much miss the point of what superabrasive machining is all about.

Officially, the process is referred to as high efficiency deep grinding (HEDG). Some people think of it as creep-feed grinding without the creep. But Mr. Lee will have none of the "G" word. This is a process, not just to apply a final size and finish, but to machine parts from solid to extremely high degrees of accuracy and surface finish. True to that notion, the machines themselves bear a greater resemblance to horizontal milling machines than to surface grinders. Of course, that's rather appropriate because they produce parts at near milling speeds.

Speed

More fundamentally, the reason Mr. Lee refuses to call his machines grinders is because that term brings along all sorts of baggage that doesn't necessarily apply. Like creep-feed grinding, an abrasive wheel is plunged at full depth through the workpiece, often creating the entire feature from solid stock rather than finishing a previously machined or cast form. Unlike creep-feed, the process happens at much higher speed, generally somewhere between 10,000 to 30,000 surface feet per minute (sfm), but potentially higher than 50,000 sfm. Moreover, considerably lower pressure is exerted on the workpiece so that there is virtually no mechanical deformation of the remaining material.

The speed has much to do with making the most of both the remarkable cutting capabilities of CBN crystals and the inherently higher speed capabilities of the wheels on which CBN is applied. The wheel is typically made of a pre-formed steel core to which a thin coat of CBN crystals is electroplated with a nickel binder.

It also can be spun at much higher speeds than conventional wheels, which Mr. Lee asserts is one of the essential process components frequently missing in failed CBN applications. It all stems back to the fact that most grinding machines are designed for conventional wheels that typically can't be run safely much above 10,000 sfm due to the centrifugal forces that want to pull the wheel apart. That's too slow to achieve the true volume cutting potential of CBN. Edgetek's standard 35-hp spindle, for example, can run up to 14,000 rpm, which means a 6-inch wheel will reach about 22,000 sfm; an 8-inch wheel approximately 30,000 sfm; and a 10-inch wheel 37,000 sfm. Optional higher speed spindles can run surface speeds over 50,000 sfm.

Combining these speeds with the free cutting action of CBN--the next hardest known material to diamond--results in remarkably high metal removal rates, even in extremely hard-to-machine materials. On one part, for example, a slot 6 inches long by 3/4 inch wide by 0.220 inch deep had to be cut in hardened D2 tool steel, approximately 62 Rc. With an 8-inch, 80-grit CBN wheel running at 29,000 sfm, the slot was cut with two passes and produced complete in just 36 seconds.

In another case, hex flats were superabrasive-machined on both ends of a shaft measuring 1/2-inch diameter by 2 inches long, made of M42 steel and heat treated to 60 Rc. Each flat was 0.475 by 0.225 inch, and the depth of cut was 0.040 inch at the deepest point. With a 180-grit, 6-inch CBN wheel running at 21,000 sfm, all 12 flats were machined in 21 seconds.

The Platform

The most obvious reason why such production rates can't be achieved on conventional grinders is that the spindles simply don't go that fast, but there is far more to it than just that. The machine must be designed with the accuracy and stiffness that anticipates such elevated process speeds.

Interestingly, while achieving high grinding metal removal rates has historically required exerting heavy pressure on the workpiece, Edgetek contends that the cutting forces of superabrasive machining are lighter even than that of the milling process. Still, containing high frequency vibration is a real concern, and to that end the machine castings are made from a polymer composite material, which damps vibration better than cast iron. The design also includes other proprietary vibration damping features.

With the high metal removal rates, the process does require a significant amount of power, and thus the standard spindle motor is 35 hp, and optional motors go as high as 67 hp. That power plant is designed to drive 6-, 8- or 10-inch wheels around a work cube that measures 18.6 by 13.8 by 12 inches.

But once you get past the notion that those "tools" are abrasive wheels, the rest of the machine looks and largely acts like a horizontal milling machine. The spindle is carrier mounted, horizontally, with the Z- and Y-axis movements executed via a machining center-like column, with the X axis under the table. The cantilever design extends the wheel center either 7.5 or 11.5 inches from the column face. A rotary table is optional, as is a fifth axis, combining rotary and tilt movement. Five-axis rotary configurations are suitable for machining airfoil or compressor blades, while a five-axis trunion type tilt table is more appropriate for machining nozzle or vane segments, shroud segments or air seals. Linear scales are used on the X, Y and Z axes so that the machines can achieve linear positioning accuracy of ±0.0002 inch and a repeatability band of half that distance.

Workpiece programming is accomplished in a conventional manner for CNC, with routines defined with standard G-codes. In fact, part programs can be generated with commercially available CAM systems--treating the abrasive wheel as if it were simply a very large diameter end mill--even for five-axis applications.

Where It Fits

Who should consider superabrasive machining? Unfortunately, there's no pat answer to that question because so much depends on the specific application. Certainly, shops doing creep-feed grinding should take a look. But in more cases than not, comparisons to alternative processes will not be so simple.

Some potential users are put off by the initial costs, but they should take the total investment value into account. The basic machine tool will start at about $250,000. And then there are the wheels, which are quite expensive in comparison to conventional abrasives--two to ten times more. Realize, though, that besides the much higher production rates, CBN wheels will cut many more parts. In one case, for instance, a 0.469 by 0.820-inch slot was being finished to 15 RAA in heat-treated 4140 steel. The finish cut removed 0.050 inch of stock from all three surfaces of the slot. In the initial continuous dress creep-feed process, 12 wheels were consumed a day at $65 a piece. On a five-day work week, that added up to $202,800 a year (52 weeks) in wheel costs. A superabrasive CBN wheel cost more than twice as much at $140, but lasted for an entire day. Added up over a year's time, the CBN process has a total wheel cost of $36,400--producing an annual savings of $166,400. And that example does not take into account the non-productive time consumed in changing wheels.

An appropriate perspective, then, is not to focus so much on the costs of superabrasive machining, but on the opportunities for cost reduction. The largest such opportunities will be where multiple process steps can be replaced with a single operation. The point in such cases is not just that the grinding is fast, but that the entire manufacturing process is simplified and more controllable, particularly if the alternative is a multiple-step process such as milling, heat treating and then grinding. What is done in less than a minute might have otherwise taken weeks if parts were sent to outside departments or vendors for intermediate operations. Superabrasive machining provides the opportunity to begin with a hardened blank, and then produce the part complete--or at least the slotted features--in a single setup. Moreover, because plated CBN wheels are dressed simply to unload the grit rather than to expose a new layer of crystals, the size of the wheel is much more of a constant, so that the process is less dimensionally variable as well.

As for which parts belong on a superabrasive machine, examining workpiece features and materials is probably the best way to begin. Flats or straight side walls are the obvious choice because those forms are easily imparted with the bottom or side of a standard abrasive wheel. But formed slots are probably the best feature of all, at least for now, because they combine these features with a higher volume of metal to be removed, which is the strength of superabrasive machining.

Actual metal removal rates are difficult to predict, given all the variables associated with fine-finish machining. Moreover, the process is still new enough that there is not a large repository of process data from which to draw. Mr. Lee says that in the majority of cases, Edgetek works out the process for a workpiece on a turnkey basis, though a number of users have been able to improve those processes with experience, and have engineered their own processes as new workpieces have been added to the production mix.

Edgetek reports success with a range of both soft and hard materials, including tool steels and super alloys, as well as some advanced ceramics and composites. They continue to develop new applications as they come, and do research to further prove out the process. A recent test, for example, showed excellent results in Rc 60 chrome vanadium steel and Inconel 718, a difficult material.

But regardless of the part, a discussion of where superabrasive machining fits must take into account the entire production strategy. Think not about grinding faster, but milling to grinding tolerances, and in the hardest, most difficult materials around.