sensor is available at a very low cost and is extremely durable. However, it is important to note that the unrelated magnetic objects have to be kept away from the operating environment since the Hall current sensor is vulnerable to magnetic fields.
Vibration: Accelerometer
Vibration describes the state of an object moving repetitively back/forward, right/left, or up/down, and usually can be expressed by the physical quantity of acceleration, which is the changes in velocity divided by time. Machining vibration and noise, or the so‐called chatter, are usually generated from reciprocating motions of the cutting processes, mechanical malfunction, or component wear, which are undesirable since the irregular and random behavior may lead to severe fatigue of the machine structure and further deteriorate machine performance, such as the unbalanced forces of rotating parts. These failure incidents sometimes may not directly happen in motor power or current but in vibration. Therefore, the vibration is especially crucial to machine status monitoring. In a word, vibration data are useful for monitoring the rotation quality of a spindle and its critical components (such as bearings inside the spindle) that affect cutting and product quality the most.
A piezoelectric accelerometer is the most widely used electromechanical device converting a dynamically mechanical change (strain, force, vibration, acceleration, …) into an electric signal. This electric signal is proportional to the piezoelectric effect occurred through mechanical changes during machine operations. The piezoelectric accelerometer is designed in a small size and has rugged construction that mounts it on the surface of specific axes and positions close to the vibration source.
In addition, the accelerometer provides good data quality with low‐loss signal in forms of high frequency and transient response. Some critical characteristics of high‐frequency can be detected and outputted in linear waveforms within microseconds. However, this high‐sensitivity property is also prone to obtaining erroneous data that needs to be de‐noised. For example, the installation position and the use of cutting fluid may affect the accuracy of the vibration signal.
Figure 2.6 illustrates how the accelerometer is mounted on the metal shell that surrounds the machinery spindle to collect rotation vibration. Vibration is very useful to monitor the rotation quality for the spindle and its critical components (such as bearings inside the spindle) which also affect the cutting and product quality. Thus, the proper installation position should be as closer to the rotor as possible. Typically, the sampling rate of an accelerometer is ranged from 100 Hz to 100 kHz.
Figure 2.6 Installation of accelerometer by stud mounting.
Note that, there are various installation types for the accelerometer: probe tips, direct adhesive, adhesive pad, magnetic base, stud mounting, or insulating flange. The most recommended approach is stud mounting, which possesses the best relative sensitivity and highest frequency response among all these fixing types. Various types and sizes for studs and captive screws with mounting threads are all available. By mounting the stud or screw to fix the accelerometer on a specific location can improve repeatability of signal and reduce collection errors. Thus, the instructive nature is another critical issue to be taken into deep consideration once the desired vibration source is on the cutting tool.
The installation orientation of the accelerometer is as important as the attaching location and types. The accurate signal means that it reflects real situation in a straightforward and noncomplex form. The collected vibration should be clear and easy‐to‐understand strained conditions so as to notify where the force is from, since vibration independently occurs in the X, Y, or Z axes. Figure 2.7 demonstrates a vibration data collection of the Z‐axis from a machinery spindle from Figure 2.6.
Figure 2.7 Vibration data collection of Z‐axis.
Once the direction of the accelerometer’s receiving surface is not orthogonal with one of the specific axes, the collected signals may contain multi‐axis characteristics, and that will increase the difficulty of analyzation.
Temperature: Thermal Couple
The machining temperature changes not only affect equipment operation and machining performance but can also reflect on product quality and component status during processing.
For example, the measurement of temperature in the cutting zone has high correlation with machining quality. In general, cutting temperature gets higher along with the increase of cutting speed, feed rates, and depths due to the frictional heat generated on the cutting tool‐workpiece interface. This increased temperature can soften the workpiece so that material can be removed from the workpiece easily; however, higher temperature might also accelerate tool wears.
Thermal couple is the most commonly used sensor to measure the temperature of target objects by directly converting heat into electricity through thermoelectric effect, which creates the temperature‐dependent voltage when the temperature difference exists between two different semiconductors inside the sensor.
Once the equipment or specific components are determined to be the monitoring targets, the measurement can be completed by measuring temperatures around the installation place. As shown in Figure 2.8, one patch‐type thermal couple directly contacts with the surface of the metal shell that surrounds the machinery spindle to monitor the temperature changes. Temperature information here reflects the operation conditions.
Figure 2.8 Installation of thermal couple.
When the quality of oil or grease for the rolling‐element bearings deteriorates, insufficient lubricant quantity and viscosity may increase operation temperature and cause bearing or the spindle malfunctions. Temperature information can be a very wide‐range of temperature up to thousand degrees at a low sampling rate compared to other sensing techniques. Thus, more storage space and de‐noising methods for filtering the signals are not necessary.
However, sensing distance is one challenge that has to be taken into consideration. Figure 2.9 indicates that the linear distance between the thermal couple and the spindle can also lead to the inaccuracy of the obtained temperature.
Figure 2.9 Distance between a thermal couple and spindle.
Another challenge is that sensing the real temperature of semi‐product directly during machining is rather difficult than sensing it on the equipment or components since the installation on the real
