Manufacturing engineers may find the following an all-too-common scenario: You've done your job well. You've agonized over the required three proposals and identified the best values in a CNC machine, toolholders and inserts. You've even made the machine tool distributor responsible for cycle time and Cpk. Best of all, your company has bought into your purchase recommendation. Yet, here you are just two weeks after start-up, and your quality control manager has already burst your bubble. She tells you the machine isn't producing parts to specification. Now you must identify and fix the problem -- right away. But how do you do that, considering all the variables that can affect a machining process? Unfortunately, process variables remain a fact of life in even the best-planned metalworking operations. Whether a company mills, drills, turns, bores or grinds parts with a modern CNC machine, many variables will affect the process, including work part material, tool characteristics, thermal instability and operator influence. Quality control technicians can easily spot machining variables' effects. For instance, if part size or quality is off, then a variable has affected the machining process. The next step is easy, too: The technician simply decides whether a given part is good or bad. Eliminating the quality problem caused by process variables inevitably proves more difficult. By definition, variables are factors that can't be predicted and planned around. However, they can be compensated for by combining some of the same gaging techniques used in the quality control lab with an automatic closed-loop process control system. These systems monitor work part attributes that can be controlled through tool offsets such as diameter, depth, length, distance and, in some cases, taper. When the gaged information falls outside preset limits, it automatically feeds back to the machine CNC, which activates compensating offsets. Automatic closed-loop process control offers the major benefit of making tool-compensation decisions the same way every time, with predictable results. For example, a turned diameter can be compensated for size, regardless of whether tool wear or machine thermal drift caused the size to change. The gage amplifier senses the trending size of the dimension being monitored. Then it outputs information used to automatically adjust the machine by the exact amount required to retarget the trend to the desired size. To accomplish this, a set of upper and lower compensation limits is preset at 50 percent to 70 percent of part-print tolerance. Whenever the process trend violates the preset limits, a compensation occurs. The gaging system does not distinguish between the particular variables that affect a process. Rather, it identifies quicker or slower change rates for that process's trend, continuously retargeting the trend to maintain in-tolerance parts production within consistent limits. This control mechanism not only keeps the CNC machine producing acceptable parts but also holds the machining process under tight control as well. At the same time, it provides quality assurance at an early stage of the manufacturing process, where it is most cost-effective. A flexible gaging solution Part measurements for closed-loop process control can use a variety of gaging systems. Modular multidimensional gaging fixtures for shaft or chucker-type parts work especially well in high-changeover manufacturing environments. Such systems combine high application versatility with error-proof measuring accuracy against a known standard. The gage user assembles modular fixtures for particular applications by selecting from a supplied set of off-the-shelf fixture components and combining them with supplied or existing metrology and gage amplifier units. Manual gaging performed with modular systems takes a matter of seconds, allowing many applications to provide for 100-percent post-process parts inspection. As with all closed-loop process control systems, the gage amplifier used with the modular gaging fixture senses the trending size of the part dimensions measured. Out-of-limit dimensional data then automatically feeds back to the machine CNC, which activates the tool offsets required to retarget the trend to desired size limits. Using a closed-loop process control system based on modular gaging eliminates scrap and maintains consistent parts production with dimensions close to nominal specs. Such a system also can produce SPC documentation when required by customers. When the need arises to change over from one part application to another, the gage user can quickly disassemble the fixture components, then speedily reconfigure them, generally using only a few hand tools. Machine-mounted touch probes The machine-mounted touch probe system represents another cost-effective meas-urement solution for closed-loop process control. Unfortunately, quality control managers often reject out of hand this highly practical measurement technology because it relies on the machine's coordinate feedback system to produce its measurement readings. Many quality control people can't accept the concept of meas-uring a part on the same machine that makes it. And they have a point -- if they plan to take a measurement, plot the data on a chart and do an SPC analysis. Because its measurements can be no more accurate than the machine tool's positional accuracy, touch probes are not recommended for on-machine applications where size data must be extremely precise. However, if the purpose is to control the machining process, a touch probe can meet the requirements at a far lower cost than alternative hard-gaging systems. With a locational repeatability of better than 40 millionths of an inch (0.00004"), it will successfully control any process where the dimensional tolerances to be maintained are no tighter than 0.001". What's more, when programmed correctly, the touch probe system offers as much flexibility as the CNC machine on which it is used. Generally, the system will provide process control for any parts that can be cut within the machine tool's work envelope. Closed-loop process control based on statistical monitoring and trend retargeting can be economically achieved on CNC lathes using a turret-mounted touch probe for post-process parts inspection. In these applications, the probe measures the part's critical dimensions after it is machined to determine where each dimension falls within programmed, statistically based process limits. The probing software then uses its statistical capabilities to analyze this information and make a compensation decision. Intermediate-process applications Touch probing applications for closed-loop control of part size work well on CNC machining centers that have capabilities for intermediate-process size control. Typically, intermediate-process size control, which uses an on-machine probe, involves finding the height or width of critical workpiece dimensions following roughing operations in order to determine the finish depths of cut needed. The probing allows maximum metal removal in the roughing operations, which helps optimize the finishing cuts for tool life, part dimensional size and surface finish. Of course, if the roughing operations remove too much material, the part measurements also will reveal this, identifying the part as scrap and eliminating useless finishing passes. Intermediate-process size control can be used effectively on CNC machining centers with any type of milling tools, including end mills, slotting cutters, face mills and shell mills. All these tools can be compensated by offsets based on part dimensional measurements. It should be noted that intermediate-process size control cannot be used with boring tools. Probing can establish that a bored hole has the right size and is in the right position. The boring tool itself cannot be offset, however, because doing so would cause it to bore a hole in the wrong place. These tools must be independently compensated for diameter size by available adjustment mechanisms that can expand or contract their effective cutting diameter. In-process control of grinding operations Still another form of nonstatistical proc-ess control in metalworking is in-process gaging and size control for grinding operations. This method of closed-loop process control, in use since the early 1950s, can accurately be defined as in-process because it monitors the size of the work part as it is being machined. The in-process control used in grinding operations does not require statistical tracking; the part measurements are made, and compensated for, on a real-time basis. This means that actual machine control depends on the variable conditions present for every part manufactured. In these applications, the measurement gage head itself must be very rugged in design to withstand long-term exposure to abrasive coolant and swarf. In addition, because parts with interruptions often are processed by grinding, the gaging system must be able to accurately measure true size, even when the interruptions fall into the measurement zone. Results obtained with in-process gaging fully justify its use because it is not uncommon to achieve part tolerances as close as 3 microns. Early intervention pays dividends All methods of closed-loop process control offer the common benefit of monitoring and correcting the metalworking process and eliminating quality problems early on. Rather than producing bad parts, then catching the problem in the inspection or even assembly stage, manufacturers can avoid producing bad parts to begin with. This can substantially reduce quality assurance costs, as illustrated in a concept known as the quality lever. The quality lever demonstrates that the earlier in the manufacturing process a quality correction or improvement is implemented, the greater the payoff will be -- both in fixing the process and reducing costs. To understand the point, consider the last stage of the manufacturing process, when a product gets shipped to the customer. Obviously, fixing a quality problem at this stage will cost plenty. Worse yet, it will have no effect on changing the process to prevent the same problem from recurring in the future. The story is different, however, at the first stage of the manufacturing process, i.e., product engineering. Theoretically, an investment in quality here will give a payback of 100 to 1 by the time the product is shipped. At the manufacturing engineering stage, where metalworking operations take place, investment in closed-loop process control based on modular gaging or probing will provide a payback about 10 times greater than improvements made at the later quality control, or inspection, stage. Waiting until the manufacturing assembly stage to make improvements will result in no payback at all. The quality lever clearly illustrates that the longer a manufacturer waits to make quality improvements, the smaller the benefits will be and the greater the cost. A closed-loop process control system builds quality into manufacturing processes upstream. Whether the system is post-process and statistically based, intermediate-process or in-process, it will help prevent quality defects from occurring downstream, where changes are very costly and do little or nothing to correct the manufacturing process. About the author Tom Stewart is sales and marketing manager in the standards products division of Marposs Corp., located in Auburn Hills, Michigan. He can be reached by fax at (248) 370-0621 or e-mail tstewart@qualitydigest.com. |
