Metrology for Precision Motion Systems at PI USA

The straightness test is performed with a long plane mirror mounted to the stage’s moving table. The linear interferometer optic is placed next to the stage and is stationary. The mean path is orthogonal to the direction of motion.

This test is performed using a similar optics setup as the yaw measurement test described earlier. However, additional hardware is required, since the yaw optics can only measure up to about 10° of rotation before the beam path is broken. To overcome this, we use a special rotary calibration mechanism mounted to the stage table.

To learn more about this tool, see http://www.renishaw.com/en/xr20-w-rotary-axis-calibrator–15763

In the same way as the linear accuracy data is collected, the stage is moved in fixed increments (typically 5° – 30° per step) and a measurement is taken at each point. The stage is moved through its full range of travel, in both directions, for multiple passes. The data that is collected is plotted and the overall errors are calculated. The results are typically reported in arc-seconds or microradians.

Unlike all the previous tests in which a laser interferometer is used, the tests for radial, axial, and tilt error motion in a rotary motion axis require the use of different tools. Here we use a mechanical artifact called a master ball. The steel ball is made to be almost perfectly round and spherical. The ball is mounted to the rotating table of the rotary stage, and stationary capacitive probes are used to measure the motion axially or radially of the ball as it spins. The only difference between axial and radial measurements is the orientation of the probe relative to the ball.

The tilt error motion setup is the same as the radial, except two balls and two probes measure radial error motion at the same time. Knowing the distance between the two probes, the angular error motion can be calculated.

All of the tests described above can only measure one type of error motion at a time, one axis of motion at a time. Customers often ask us about “3D accuracy” or “overall precision.” Given the test methods we use, such specifications are difficult (if not impossible) to measure and therefore verify. We can make calculations to predict overall performance in average or worst-case scenarios, but we cannot verify those values with real work measurements.

Any metrology technique that can measure positioning errors in multiple degrees of freedom almost always depends on some sort of artifact, such a “golden wafer” or a “golden part.” Then a probe or a camera is used to measure the motion system’s ability to track or follow the master artifact. Such methods are comparative, depending on the known quality of the master artifact to verify overall performance. Since such artifacts are usually application specific, we don’t typically work with them at PI. Laser interferometer systems that measure multiple degrees of freedom at the same time do exist; however, these systems are usually used as control feedback system, not test metrology systems. And while they do measure multiple DOF at the same time, they are still only measuring individual errors, not true 2D or 3D spatial measurements.

Precision metrology is incredibly sensitive to environmental conditions. The measurements we make at PI are under tightly controlled conditions. We isolate the system as much as possible from thermal variations, air currents, vibrations, audible noise, and electrical noise. Replicating the results in the field may be challenging.