The Value Of Daily CMM Capability Checks

Coordinate measuring machines are not infallible. A variety of factors can adversely affect their measuring accuracy, including a crash, large temperature fluctuations, a malfunctioning probe head, use of an uncalibrated probe—even a person leaning on the machine gantry might be enough to take it out of square. If the CMM’s true capability is verified only once a year via laser or another calibration method, then it’s possible that the CMM might be inaccurately measuring parts for an extended period of time, unbeknownst to the operator. The question remains—did a workpiece fail inspection because of an imprecise manufacturing process or an inaccurate CMM?

At the General Motors R&D center in Warren, Michigan, Dr. John Agapiou, staff research engineer, and Mark Muczynski, CMM programmer and operator, are part of a team that evaluates equipment to enable the company’s manufacturing facilities to perform daily CMM checks on their own. Their goals are to identify a device that can perform its test quickly, provide reports that are simple to understand and yet is easy to set up and use.

Featured Content

Five Things You Get With Your Free Modern Machine Shop Subscription READ MORE

Registration Now Open for the Precision Machining Technology Show (PMTS) 2025 READ MORE

Zillow for Machine Tools? READ MORE

One device the team has been examining is Asanuma Giken’s Quality Master gage, which is available in the United States from Tecnara Tooling Systems, Inc. (Santa Fe Springs, California). This artifact-based CMM inspection method is said to enable 20-minute reliability checks and more comprehensive accuracy checks that can be performed whenever it is deemed necessary. The Quality Master is NIST-traceable, and it complies with ISO 9001:2000 and ISO/TS16949.

The system consists of a cylindrical artifact as well as data collection and reporting software. The artifact is made from material that is said to have one-fourth of cast iron’s coefficient of thermal expansion. This allows it to remain stable despite fluctuating environmental conditions. It has 12 precision-ground locating bushings—four on its top surface and four vertically oriented pairs located at 90-degree positions around its perimeter. Locating the bushings in such a way allows the system to collect X, Y and Z data in one setup. During a test, the CMM probes the outer surface and ID of the 12 bushings to determine their hole diameters and position relative to the others.

After collecting the measurement data, the system’s software compares the data to that of previous tests and creates a statistical report to identify trends that could show the existence of a problem. The repeatability test demonstration performed during this editor’s recent visit to GM’s R&D facility took 22 minutes to complete.

An accuracy test can be performed if it is felt that the CMM may have been damaged or is not taking accurate measurements. The accuracy test differs from the repeatability test in that the bushing probe routine is performed three times; the artifact is then rotated 180 degrees, and the test is carried out three additional times. The data collected from these multiple probing routines are averaged to ensure accuracy. The accuracy test, which takes approximately 2 hours, reveals X-Y-Z axial positioning, X-Y-Z straightness (pitch and yaw) and perpendicularity between the X-Y, X-Z and Y-Z axes. The tests reveal individual axis errors, rather than an overall volumetric error. This allows users to determine the positioning accuracy in discrete areas on the CMM’s table. The artifact is compact to allow this. Its diameter is 250 mm; its height is 245 mm; and it weighs 36 kg.

The accuracy reports that the software generates may also be helpful in reducing CMM service downtime. The report can offer clues that give the service person an idea of what the problem could be prior to the visit.

For CMMs without an articulating head, it is necessary to use a star probe to access the bushings on top and around the perimeter of the artifact. Regardless of the style of head, the same probe should be used for all tests, so as not to introduce an element of variability.