Affordable technology is available that allows shops to dynamically model the accuracy characteristics of their machine tools. Here's how these inspection devices work, and what they can reveal about your production equipment.
Can shops inspect their own machine tools? Why not? Many of us used to do tune-ups on our cars. And while the expertise and tools required to do a proper tune-up on newer cars are now beyond the reach of most of us, machine inspection is moving in exactly the opposite direction.
It was once necessary to hire a specialist who had the training and the specialized equipment to inspect a machine. The capital investment required for some of this equipment and the frequency of how often the tests should be conducted made it practical to go outside to do the inspections. Although complete machine performance testing is divided into fourteen major areas even now, there is a new, simplified one-day check with just four tests that can be both informative and convenient.
This one-day program is not meant to replace the entire B5.54 Standard from ASME (American Society of Manufacturing Engineers) for machine inspection, but to offer a quick way to get an understandable diagnosis of a machine's health. It's easy to learn and doesn't cost a fortune in equipment. Thus, a very significant measure of the benefits of machine inspection is now within the reach of even small- to mid-sized shops.
With industrial pressures to produce higher quality products, all methods of production, inspection and documentation have become very important. Still, there are many firms that are just now contemplating taking this next step in determining the true capabilities of their production equipment. So we have focused the following discussion on helping to remove some of the apprehension one might have. It explains the basic application of three inspection devices manufactured by Heidenhain: a telescoping ball bar for the contouring accuracy check; a new two-dimensional grid encoder that can perform the same contouring accuracy check and has the ability to identify actual CNC control problems; and a linear standard that can be used as an alternative to lasers in some applications for the linear displacement accuracy test. We will provide an overview of the tests, of the devices that perform the tests, and finally of how the information can be analyzed.
The DBB 110 double ball bar is a precision measuring standard mounted between two telescoping precision balls. A magnetic socket is mounted into the spindle of the machine and a second magnetic socket is placed into a small tower that is bolted to the machine's table or work surface. The ball bar is then placed into the two sockets.
How does the ball bar work? Imagine the spindle socket as being the center of a circle, and the ball bar as the radius. A program is written to move the machine table in a circle around the spindle, both in a clockwise and a counterclockwise direction. Once the table starts to move, the ball bar can begin taking measurements, which are output to a personal computer running evaluation software that captures and processes the test data. If there are any deviations that would cause the table to move away from the programmed radius, the telescoping portion of the ball bar detects this movement via the measuring standard inside.
The ideal path or circle would have no deviations from the programmed radius. However, if there are any deviations at all, these measurements are recorded by the evaluation software. The software then prepares the recorded data so that it can be printed as a representative circle.
It is here that things start to become interesting. The resulting shapes represent the errors in a way that often can be easily diagnosed. This data was taken on a typical vertical machining center. In this example, notice the small abrupt changes in the path at each of the axis reversal points (90, 180 and 270 degrees). This "footprint" is the result of backlash in the machine that has not been fully compensated for in the control. Once detected, it is a relatively simple matter to reprogram the backlash parameters in the control and get the machine back to peak performance.
There are many "footprints" that will identify such problems as:
The real advantage of the circular contour check is that it offers a quick snapshot of the machine's condition without having to take that machine out of production for long periods of time. Other machine inspection tests can take several hours to a couple of days. The initial ball bar test only takes about an hour total to set up and program the machine, and then to run the test. And once this has been done and some experience gained, all subsequent tests can be performed in ten to fifteen minutes.
The indirect benefits of this technology in many cases are more helpful than the actual diagnostic data. A direct benefit, for example, would be the collection of data that would determine the specific level of backlash or amount of radial deviation. Indirect benefits are those discoveries that save money or help justify the costs of purchasing the equipment. Here are two actual examples of these benefits as related from one user.
Ron Boudrie of Quality Die & Mold in Grand Rapids, Michigan, had been searching for ways to keep tighter quality control on his production and on his machines. After purchasing a ball bar and getting used to it, he started to realize some of these additional benefits. A few months ago, Quality added a new vertical machining center to the shop. Not long after the purchase, the manufacturer sent over a technician to upgrade the machine's software. To see if the upgrade had the desired effect, Ron decided to run the ball bar test on the machine after the software was installed. What he found was that the new software had actually degraded the machine's performance. The technician was located before he returned to his office and came back to reinstall the old software. This saved time and money for both parties.
On another occasion, Quality needed service on a different machine. One of the axis encoders required replacement in the middle of a job. Unfortunately, an encoder with the wrong pulse count was installed. In effect, the wrong encoder count will make the axis error wrong on a proportional basis. Once again, before resuming work, the ball bar test was performed and the error corrected.
These two real life examples could have cost dearly, but additional machine downtime was completely avoided. Some other indirect benefits include finding the most accurate machine in a group and choosing this machine to do the work with the tightest tolerances.
Compliance to required standards and record keeping also provide an audit trail of proof that machines have been inspected. Many customers are now making the circular contour check a weekly routine as part of their regular preventive maintenance procedures. The ASME/ANSI B5.54 procedures suggest that the contour test be conducted at different radii for different size machines. It also suggests one test in each direction (clockwise and counterclockwise) be performed at 10 percent of maximum feed rate and at 80 percent of maximum feed rate. This parallels very closely with the ISO 230.1 "Machine Inspection Methods." The idea here is to evaluate the machine under two different conditions in order to ascertain a more complete picture of its dynamic performance characteristics.
VM 101 linear comparator, is actually a linear scale, though it's probably not like any linear scale to which you are accustomed. The physical construction is a "U" shape of steel with precision graduations placed in the bottom of the "U." Rather than using the typical transmission-based technology, these graduations are based on a special diffraction technology. The comparator is available up to one meter (40 inches) in length. With an accuracy of one micron or less it is equal to or perhaps better than a standard machine inspection laser over the same length. This device is not intended to measure larger machines (over 80 inches), but it is useful for smaller machines and especially machining centers. In every respect it can be used as a direct substitute for a laser on the linear displacement accuracy test. In simple terms, the linear displacement accuracy check is one of the four tests in the "One-Day Test" as defined by ASME, and users will find it accurate, convenient, fast and economical when compared to other measurement technologies.
To operate, the steel scale is placed on the work surface, and aligned to the machine axis, and the scanning unit is affixed to the spindle. A special carriage is configured so that the scanning head can be magnetically attached to a clamp on the spindle. Once this is done, the scanning unit is released so there is no contact between the scale and the scanning unit. Next, the measurement takes place and the software does the rest. The entire process takes less than ten to twenty minutes per axis. Besides the the linear displacement accuracy check, this technology also can be used for linear repeatability checks. For axes from 40 to 80 inches, a step-and-repeat process can be used.
The KGM 101 grid encoder is the newest inspection device, and perhaps the most unconventional approach to machine inspection. The device consists of a steel plate with a circular encoded area about 5.5 inches in diameter. The encoded area also uses diffraction-based technology, but slightly different than that of the linear comparator. The big difference is that the plate is encoded bi-directionallythat is, one measuring standard can detect any movement in two directions simultaneously. It can read around 2D corners, so to speak. The surface is somewhat similar to a chessboard, but with the red squares higher than the black squares. The squares are not actually red and black, of course, and in fact the structure is more like a lattice coated with a protective material. But the raised portions of the encoder, in combination with a spindle-mounted, non-contact scanning head, can provide accurate 2D measurements within ±2 microns (0.00008 inch). This allows for any 2D movement over the grid to be detected and recorded accurately.
Why this is important is the most interesting thing of all. Naturally, a circular program can be written so that the scanner moves over the surface of the plate, exactly like the circular contour check. And, indeed, the grid encoder can perform the circular contour check as easily as a ball bar. But it also can detect additional errors that are too subtle for the ball bar test.
Why? Consider this: On any machine running a programmed circle, any mechanical errors on an axis become larger as the radius is increased. It is a simple matter of geometry. For example, a 0.001-inch pitch-down error over 40 inches of travel is naturally less at the 20-inch point. It is even less at the 10-inch point, and so on. At some point along the radius, the pitch-down error becomes so small that it is hard to detect.
If we agree to this in principle then it would be reasonable to say that any errors caused by other sources could easily be buried within the larger errors caused by the machine geometry. That is exactly where the ability of the grid encoder starts to show. Because it is not physically limited to the mechanical radius of a ball bar, smaller circles can be programmed to trace a path over the grid plate. Mechanical errors that show clearly at large radii become difficult to "see" on these small circles and the errors that are now present can be traced to the CNC itself or to its associated electronics. For example, excessive lead or lag errors, gain adjustments and interpolation problems are control problems, not mechanical problems. But they do affect the movement of the machine and the optimizing of the control. The grid encoder can show this as physical evidence and not just as a voltage or amperage value on a meter. There simply is no other device that can portray these errors as a movement.
The grid plate behaves just like a piece of grid paper placed below a pencil, and any 2D shape can be traced. In the magazine, a circle, diamond and square program has been prepared and traced. The scale of each grid is portrayed in the upper right corner of the printout.
We have elected to zoom in on the bottom of the diamond. Here you will notice the path programmed and the actual path. In this example, you can observe that this traced path did not follow the programmed path as it was trying to make the corners at a high feed rate. This is direct evidence of control errors that require attention.
The grid encoder is a versatile tool. For example, it is possible to trace a straight diagonal line across the plate to check linear interpolation. It is possible to test 90-degree turns at high feed rates. With lasers that are line-of-sight devices or ball bars that have a fixed radius, these kinds of tests are impossible.
Besides the measuring tools themselves, the measuring data acquisition and analysis software is also extremely important. With machine inspection, there are many computations that must be performed to make the collected data as useful as possible. Fortunately, software is available that is easy to use, and runs on PCs and even in a Windows environment. Also, for the devices mentioned above, a "counter card" is required for the PC running the software, which functions as an interface for the incoming measuring signals. In the case of Heidenhain's ACCOM software, the values are in accordance with the ISO 230-4 standard.
The software also makes it easy to program such routines as circular interpolation tests or curved path tests as well as more simple movements. The program asks for all the necessary parameter information and generates the corresponding NC program which is downloaded to the CNC.
Will these machine inspection tests tell you everything you need to know about your manufacturing processes? Of course not. There are many other factors that must be accounted for to ensure that the parts coming off a machine tool are within acceptable boundsmaterial considerations, tooling, fixturing, feeds and speeds, and more. But getting a good footprint will tell you how inherently capable your machine is at achieving certain levels of accuracy from the outset, as well as point out many correctable error factors. And that should be an enormously valuable quality tool for shops of all sizes.blog comments powered by Disqus