Detect machine vibrations with order tracking analysis

Machine vibration can be a critical symptom of a pending problem that, left unchecked, can damage or expedite deterioration of industrial equipment. Causes for machine vibration include an imbalance, misalignment, binding, looseness or wear and tear. Unaddressed machine vibration can have serious business and health ramifications, spanning from expensive, unplanned downtime of operations to life-threatening safety issues.  

That is why machine vibration analysis is such a critical tool in predictive maintenance programs. When regularly performed, maintenance technicians can detect machine faults in the early stages and take appropriate action to avoid machine failure or replace parts. 

Motor vibration analysis is a broad field with numerous methods for detecting machinery faults. To identify a more advanced technique to determine the cause of motor vibration, engineers experimented with the little-used approach of order tracking, which is used to detect rotating machinery faults. Order tracking analysis is effective for vibrational analysis of motors, engines, generators, turbines, pumps and compressors. 

Vibration analysis to optimize asset health

These kinds of rotating machines generate vibration that can be analyzed to assess their general condition and the remaining life of their parts. Order tracking analysis can help differentiate the rotational source compared to random vibration in the overall output. 

The vibration signature analysis can detect faults such as bearing faults, unbalance, misalignment and gear faults. 

As a proof-of-concept, the engineers developed a piezo-accelerometer vibration sensor prototype to assess the health of bearings, which are often the source of most faults, unbalances and misalignments. Piezo-accelerometer sensors can be used to perform vibration analysis of electric motors to detect vibration generated by their bearings. 

Bearings have four key elements that can deteriorate with their usage: the inner and outer race, the balls, and the cage. Each of these elements generate a characteristic vibration frequency when faulty, which makes it possible to identify the fault just by analyzing the vibration signature. The exact frequencies depend on the physical properties of the bearings, such as the dimensions and the number of balls or rollers. The following describes how the piezo-accelerometer sensor prototype was set up using off-the-shelf components. 

Hardware setup for a piezo-accelerometer vibration sensor

The solution has five hardware components: A piezo-accelerometer vibration sensor, a brushless DC (BLDC) motor, an electronic speed control (ESC), a power supply (PSU), a data acquisition device (DAQ) and a computer. 

The CPU controls the motor speed by setting the PSU voltage (from 6 to 16 V) and by controlling the electronic speed controller (ESC) through a throttled pulse position modulation (PPM) signal. The DAQ measures the vibration signal from the sensor and the voltage signal from a motor phase (which is used to compute the motor rotational frequency). 

The DAQ supports a maximum sampling frequency of 200 ksps @16 bit in the range of ±10 V, so engineers had to divide the motor phase signal. They reprogrammed the ESC with the BLHeli firmware and disabled its proportional-integral-derivative (PID) control loop and set the min and max PPM throttle values to 1000 and 2000 ms, respectively, making it easier to fine control its speed (see figure 1). 

Key elements of the sensor were the signal processing algorithms; specifically, the rotation tracker and the order tracker. 

Figure 1: Reprogramming the electronic speed control (ESC) and disabling proportional-integral-derivative (PID) control loop made it easier to fine control its speed. Courtesy: Newark

Determining how to measure, filter results

Engineers tracked the motor rotations by measuring the divided voltage signal of a single motor phase. The brushless motor used in the prototype rotates once every seven electrical cycles. The processing of the signal was not trivial as the “form” the cycles depend on the frequency. 

Due to the irregularity of the signal, simple approaches to track rotations – such as identifying blocks without voltage “spikes” – did not work. Instead, engineers implemented the rotation tracking algorithm as a sequence of the following operations:   

• A high time resolution, low frequency resolution spectrogram is computed from the original signal. 
• The main frequency at each segment of the spectrogram is computed and then interpolated at every time unit of the original signal. 
• A filter bank of 5 bands ([100; 312.5], [312.5; 625], [625; 1250], [1250; 2500] and [2500; 5000] Hz) is created (the selected range includes up to the third harmonic of the signal). 
• The filter bank is applied to the original signal to generate five filtered signals. 
• These filtered signals are linearly combined at each time unit based on the computed main frequency at each time unit. 
• The linearly combined signal is used to detect all zero crossings with a negative derivative. 
• Zero crossings where the signal amplitude is below a certain threshold are discarded.   

Order tracking separates synchronous, asynchronous vibration

Order tracking is a technique that can separate the synchronous vibration from the asynchronous vibration (the synchronous vibration frequency is proportional to the shaft rotational frequency, while the asynchronous vibration frequency is not). An algorithm was developed to convert a time series into a rotational series, which is a series of equally spaced rotational increments of a reference shaft. The algorithm was implemented in the following ways: 

1. The vibration time series is expanded through FFT resampling. 
2. Rotational tracking is used to compute the negative derivative zero crossings. 
3. The computed zero crossings time locations are resampled to small rotational increments. 
4. The expanded vibration time series is sampled at the resampled small rotational increments (step 2).   

Signal analysis

Engineers focused on identifying the synchronous and asynchronous vibration components and implementing a real-time vibration analysis dashboard. To identify which vibration components are asynchronous and which are synchronous, they performed a linear ~180 s sweep from ~20 Hz to ~200 Hz. The sweep highlighted which spectrum components were affected by the shaft rotational frequency, and which were not. The sweep also helps with identifying synchronous components that could get damped at certain frequencies.
Through rotational tracking, engineers were then able to measure the rotational frequency curve. The sweep generated a vibration frequency domain spectrogram, which displayed asynchronous components as horizontal lines, and synchronous components as diagonal lines that scale with the rotational frequency by a constant factor, or order (see Figure 2). By computing the spectrogram of the order tracked signal, the synchronous components order frequency were identified. 

Figure 2: The synchronous components order frequency were identified by computing the order tracked signal. Engineers built a real-time vibration analysis dashboard to display the results using NI LabVIEW. Courtesy: Newark

Real-time information dashboard, vibration analysis

Engineers built a real-time vibration analysis dashboard to display the results using analytical software. The dashboard contained four controls: a power button to turn on the PSU (accessed through VISA drivers), a voltage slide bar to set the PSU output voltage, a throttle slider that controls the ESC (through a PPM signal), and a history control to set the size (in seconds) of the FFT window. The gauge indicators indicate the PSU readback voltage, current and power, and the motor oscillating frequency (the highest frequency component of phase signal spectrum). Two time domain plots display the vibration and phase signal. And horizontal frequency domain plots show the order vibration spectrum, the frequency vibration spectrum, and the phase voltage spectrum. 

Vibration analysis benefits, machine health

Vibration analysis is a great tool to test the health of machines and perform predictive maintenance, and order tracking analysis proved to be an ideal approach for detecting faults in rotating machinery, such as motors and engines.
Considering the price of the sensor and digitizing systems in comparison to expensive motors and engines, it may be a good idea to attach vibration sensors to rotating machinery to track the health of the machine. As machines get smarter, they will begin to implement their own vibration analysis system to track their own health and evolution of faults. 

Randall Scasny is a senior community content specialist for Avnet’s element14 Community. Edited by Chris Vavra, web content manager, Control Engineering, CFE Media and Technology, cvavra@cfemedia.com.   

MORE ANSWERS 

Keywords: discrete sensors, vibration detection   

CONSIDER THIS 

What benefits could your facility gain from machine vibration analysis? 

Do you have experience and expertise with the topics mentioned in this content? You should consider contributing to our CFE Media editorial team and getting the recognition you and your company deserve. Click here to start this process.