Grinding Is High-Tech
Grinding is necessary to machine certain new materials, and new grinding techniques will expand the applications for this technology
By Stuart C. Salmon
Advanced Manufacturing Science and Technology
A process employed by Neanderthals ages ago, grinding has stood the test of time, and is as important to our survival today as it was then. One might think that we would know all there is to know about grinding, but nothing could be farther from the truth.
Grinding has branched into two well-defined areas; the first is abrasive machining, and the second is precision grinding. Abrasive machining is the range of processes that allows the removal of significant stock. It's akin to large-chip machining processes such as milling, turning, and broaching, whereas precision grinding belongs to the more-traditional realm of creating extremely precise and accurate forms, sizes, and surface finishes, with minimal stock removal.
Grinding and abrasive machining processes are essential to today's industries, not only because of the need for higher levels of precision, but because of developments in materials technology. Next-generation materials are being developed that can only be machined using abrasive technologies; these are not just difficult-to-machine metals, but ceramics, cermets, and whisker-reinforced polymers. Indeed, grinding and abrasives processes will be around for a long time to come, and it will behoove manufacturing to recognize and understand the technology. Unfortunately, we still rest in our ruts of what we know best. The traditional approach to machining a component tends to lead us to try to do the same old things better. Content in their comfort zone, the enemies of the latest manufacturing technology are those who have been successful with the old methods, and are therefore reluctant to change, or even consider a different way.
The better method—economically, ecologically, and technically—may be an alternative technology. Laser, waterjet, wire EDM, and fine blanking can all produce similar part configurations in sheet materials. But depending on the material, the accuracy, the surface integrity, and the production rate, one process might prove superior to another in certain circumstances. Similarly, parts that are turned, milled, and broached can be ground, oftentimes with results that are superior to the traditional approach. Beginning at the grassroots of schools and colleges, there needs to be a change in how manufacturing processing is taught. The approach must be more on how to make "stuff" rather than: "if the part is round, it had better go on a lathe" or, "if it needs a high degree of surface finish, then it had better be ground." There is a need to move away from the "process-oriented" way of teaching to the "manufacturing-methods" way of teaching, where the total cost (economically, ecologically, and technically) is the overriding factor.
The change from large-chip to small-chip machining is beginning to show up in the multiaxis machining centers that once held mills and drills and reamers in their tool banks, and now store grinding wheels and have programmable diamond-roll-dressing cycles. Admittedly, one has to be aware that a milling machine is not a grinding machine. The large-chip industry is hailing high-speed machining as a "new world," whereas the grinder has been running at 6000 fpm (30 m/sec) as a "normal" speed forever. There are production-grinding installations running 24,000 fpm (120 m/sec) on grinding machines today, without great fanfare. But it is this area, where the rotational speed of the spindle for a milling cutter versus a grinding wheel is so different, that the user must beware.
The dominant frequency of vibration in a milling operation is typically linked to tooth engagement. A 4" (100- mm) diam milling cutter with eight inserts, running at, say, 475 fpm (2.4 m/sec) transmits a vibration of approximately 60 Hz through the system. The dominant frequency of vibration in a grinding machine typically originates at the spindle, at once every wheel revolution. A 4" (100-mm) diam plated grinding wheel running at 9000 fpm (45 m/sec) transmits a vibration of approximately 12 Hz. This means that the vibrational stability range of the multiprocess machine tool needs to be particularly broad to accommodate cutters as well as grinding wheels. Now, it's important for those who take an old vertical-spindle CNC milling machine and replace the machining spindle with a grinding spindle, and therefore believe that they have made themselves a jig grinder, to note that the grinding may not turn out quite as well as they initially thought.
Grinding wheels do not work well under conditions of excessive vibration. Not only do they tend to wear rapidly, but workpiece surface finish suffers. Also, when combining machining and grinding on the same machine tool, the fluid type and filtration system must be considered. With respect to tool life, milling is better performed using minimum quantity lubrication (MQL). On the other hand, the proper application of a large flow of fluid through a nozzle (1–1.5 gpm for every horsepower in the grind—or 0.075–0.125 L/sec for every kW in the grind—is a good rule of thumb), where the speed of the fluid matches that of the grinding wheel periphery, is critical for a grinding operation. Consider also the filtration of the fluid, which may contain a mix of grinding swarf and abrasive particles, as well as large chips from the milling operation and the contaminated MQL fluid. Both the fluid-application method and the filtration system need to be all-encompassing. The MQL fluid must be compatible with the flood fluid used in the grinding operation, so as not to cause a breakdown in the bulk fluid's stability.
To avoid wheel dressing, plated superabrasive wheels are often employed in multiprocess, multiaxis machining centers. These wheels have no porosity, and can "pump" far less grinding fluid through the arc of cut than an equivalent vitrified grinding wheel. Where a vitrified grinding wheel can pump 50–60 gpm (5–6 L/sec) through the arc of cut, a plated wheel, even one running at perhaps four times the speed of a vitrified wheel, can only pump 2 or 3 gpm (2 to 3 L/sec) through the arc of cut. This reduction in flow not only decreases the cooling effect of the fluid, but also the flushing action of the fluid as it leaves the wheel periphery after the arc of cut. Given that fact, there are many installations blasting large volumes of straight oil at a plated grinding wheel, in the hope of somehow pushing more fluid through the cutting zone because of the sheer force exerted by the pressure of the fluid. In this case, excessive foaming and—worst of all— misting occurs that can create both fires and health hazards in the workplace. Cryogenics are a good alternative in this area, not simply liquid nitrogen, but liquid nitrogen inoculated with lubricating oil droplets. The proper fluid application can often lead to reduced costs and a cleaner, safer working environment.
As grinding wheel speeds increase—and they will continue to increase—plated superabrasive grinding wheels will be and are becoming more popular. The higher the wheel speed, the lower the grinding forces, the longer the wheel life, and the better the surface finish—an all-positive direction. Plated wheels can and have run at 50,000 fpm (250 m/sec) in the laboratory without fear of the wheels bursting. As wheel speeds approach sonic speed (the speed of sound is approximately 67,000 fpm or 330 m/sec), wheel bursting is not the issue—"wheel-off" is the concern. Should a bearing fail or a shaft break, then the wheel will come loose and free, in one piece, with its "single layer" of abrasive intact and designed to cut efficiently. Wheel guarding has to be properly addressed, and such machines should be used in areas where only fully automated operations are being employed, so as to avoid human injury or loss of life should the worst occur.
The next generation of grinding machines will be more extensively automated with respect to wheel changing, wheel dressing (if necessary), part manipulation, and grinding-fluid control. The machines will have a far-smaller footprint than the mammoth cast-iron and V-flat guideway machines of the 1940s and 1950s, some of which are still around today. Hydrostatic bearings, magnetic bearings, linear motors, and epoxy-concrete and glass-fiber-reinforced plastic bases are proven technologies that are being incorporated into some of the next-generation machine-tool designs. The control systems will be more user-friendly and based upon the level of the operator, be it the production engineer designing or researching a process cycle, or an operator needing only to make minor corrections or adjustments once the process is running.
The Smart machine is thought to be the ultimate answer, of course. And indeed, we can sense much of what is going on in the grinding process. It's easy to control and maintain spindle speeds and slideway speeds and positions. We can sense process forces and monitor power during the grinding operation. There are in-process gaging and grinding-burn detection systems, autobalancing, and both wheel and dresser wear-compensating systems. The one fly in the ointment, however, is mechanical lag. Sensors and feedback systems can make computational calculations and operate at electronic speeds. Yet once a decision has been made as to what needs to be done, say increase wheel speed or decrease a table feed, the mechanical lag time caused by the inherent inertia in the system is too great to make the computer-generated change in time. It is possible to attempt to predict errors or run a grinding process in a modeled algorithm that sets safe levels below or above which the process must remain, but this is not always the absolute optimum. Based on the projected cost of a Smart machine this sub-optimal outcome may be just fine.
Nanoprecision is another world of grinding altogether. The nano prefix is bandied about today to gain recognition for perceived advancement. "Put nano in front of anything and funding will waterfall your way" has been the prevailing thought. Let's put nano into perspective: Those who mill find a tolerance of 0.010" (0.25 mm) commonplace and 0.001" (0.03 mm) difficult. The decimal point only shifted one place to the left to make things "difficult." Those who grind find a tolerance of 0.0005" (0.013 mm) commonplace and 0.00005" (0.0013 mm) difficult. The decimal point shifted one more place to the left to make things "difficult". A nanometer, however, is 0.00000004" (0.0000010 mm), and the decimal point has shifted three places to the left! The world of nanoprecision is truly another world. I well remember, as an apprentice, being told that a human body gives off about 100W, so the results from a super-precision inspection machine could be drastically influenced by simply standing nearby. On this scale, a nanomachine would be sensitive to breathing. Nanogrinding machines do exist, but their design and the environment in which they operate are a far cry from the shop-floor grinders I am writing about. Nanogrinders are not a natural follow-on from the grinding machine designs of today. So we will leave them out of this article, and perhaps make room for them some other time.
The future for the production grinder is to be able to deal with higher wheel speeds and a multitude of tools that are not all necessarily abrasive tools. Able to use multiple machining fluids, the machine will be fully enclosed. The machine will be able to grind both flat, round, and complex surface shapes. It will facilitate automated tool and part changing with a level of in-process inspection not only for size, but also surface integrity.
Grinding ceramic parts requires diamond grinding wheels—resin bond for machines that lack vibrational stability, but vitrified bond or metal bond for those that have good vibrational stability and stiffness. Grinding ceramics is a challenge, as the swarf consists of fine micron and submicron particles that require special filtration. Ceramics play havoc with emulsions, but grind well in full synthetics or straight oils. They require high stiffness in a grinder, but not high power. Ceramics creep-feed well for high stock-removal applications, and can be ground to very high levels of surface finish using both grinding and lapping. Applications for ceramics are increasing in the semiconductor, medical, and aerospace industries, as well as high and low-tech applications—for instance lightweight armor and the likes of washerless faucets. Cermets comprise lightweight and highstrength materials that tend to work-harden, so that conventional machining and grinding is very difficult. Titanium aluminide is one such material that is beginning to be used in gas turbines and high-performance racing engines. Whisker-reinforced materials are easier to grind than ceramics andcermets, whether in a metal or a polymer matrix. The directional properties of the fibers, however, often lead to part distortion post-grinding (as opposed to the grinding process being the direct cause).
Hard-lubricant coatings on plated wheels have been shown to extend their life by at least 30% in sticky and gummy materials like stainless and nickel and cobalt-based superalloys. The coating is designed more to prevent wheel loading than to decrease the wear of the abrasive particle.
The move to more superabrasive applications is increasing. Grinding machines are becoming smaller for higher power, and for better stiffness and vibrational stability. Spindle speeds are still increasing, and large and small-chip machining are being incorporated into one machine tool. CNC controls are being designed to accommodate the unskilled operator. With the lack of apprenticeships and fewer prospective journeymen learning the "hands-on" skills of abrasive machining and precision grinding, there is an urgent need to control the process via a computer using feedback from sensors. Small-batch quantities and one-off toolroom applications, performed manually, will be more and more expensive due to the cost and time required to program a CNC control. More user-friendly CNC controls means more-sophisticated CNC controls to accommodate the less-skilled man-machine interface. Linked to the Internet, future control systems will be able to download grinding cycles for given workpiece/wheel combinations, as well as interrogate experts, both human and synthetic, to refine or troubleshoot grinding operations around the world. With the demand for higher precision and the need to machine the unmachinable, grinding and abrasive processing will be around for a long time to come.
This article was first published in the August 2009 edition of Manufacturing Engineering magazine.