Measuring performance in real time

Real-time measurements during metalcutting can offer significant insight into machine tool performance. The parameters to be measured include position, velocity and acceleration of the axes; cutting force; sound; temperature; and acoustic emission, which is not the same as sound, but rather a high-frequency, structure-borne vibration.

These measurements are usually collected with a digital data acquisition system, including a sensor, digital-to-analog converter, computer, storage system and software for manipulating data. To prepare the measurement, the user must specify several parameters, such as the sampling rate, maximum frequency, observation time and frequency resolution. The choice of parameters is important because they are related.

All images courtesy S. Smith

Figure 1. With aliasing, a high-frequency signal (red) looks like a low-frequency signal (blue) because the sampling rate (green) is not fast enough.

Sampling rate: This is the rate of data collection, usually expressed in samples per second. Generally, users prefer slow sampling rates because sampling at a rate of several hundred times per second costs considerably less than a rate of several hundred thousand times per second. Fast sampling rates require more expensive data acquisition equipment and the transmission, storage and manipulation of larger data sets. However, the sampling rate must be fast enough to see the signal of interest. Mathematically, that means there must more than two samples per cycle of the highest frequency being measured.

Table 1. Summary of parameter interactions.