[Introduction]The arrival of Industry 4.0 is changing the rules of the game in the manufacturing industry. Insights based on big data will make the entire production chain more transparent, flexible and resilient, enabling customized production while still achieving profitability, which is unimaginable under the traditional manufacturing model. Moreover, the resulting change is comprehensive, and its influence penetrates into all aspects of the production and manufacturing process, including the maintenance of production equipment.
Traditional equipment maintenance in the manufacturing industry usually includes corrective maintenance and preventive maintenance. The former refers to passive intervention after equipment failures, while the latter generally refers to active maintenance according to predetermined cycles or conditions. These two methods are obviously not the optimal solution for equipment maintenance – the former often has to bear the huge loss of unplanned downtime (the cost of unplanned downtime will account for nearly a quarter of the total manufacturing cost), while the latter is unavoidable Unnecessary intervention in the case of equipment operating well, resulting in greater costs.
If there is a method based on big data analysis and insight, which can understand the real-time working status of equipment in real time, discover hidden faults in time, and carry out “just right” maintenance in a targeted manner, this is of course a more ideal solution. This analysis of the health of machinery and equipment is called Condition-Based Monitoring (CbM), and the CbM-based maintenance approach is “predictive maintenance.”
Obviously, within the framework of predictive maintenance, operator intervention is only required when machinery and equipment exhibit certain early warning symptoms. Compared with traditional maintenance methods, its benefits in reducing equipment downtime, reducing maintenance costs, extending equipment life, and increasing productivity are huge.
Figure 1: Cost analysis of predictive maintenance versus traditional maintenance methods
(Image source: ADI)
Predictive Maintenance for Electric Motors
Since machines and equipment with rotating mechanisms such as motors are the main force in the manufacturing industry, CbM or predictive maintenance for motors has naturally become a key issue. This is also a hot spot that many leading technology suppliers (such as Analog Devices, ADI) are chasing.
When judging whether the working state of the motor is “healthy”, it needs the support of a lot of data, such as pressure, vibration, noise, temperature, etc. Among the measurable physical quantities, vibration is a particularly critical parameter, because many faults Early symptoms—such as ball bearing failure, shaft misalignment, unbalance, excessive looseness, etc.—all appear as abnormal vibrations and exhibit distinct characteristics when measured on the “Diagnostics” to decide if and what type of maintenance is required.
For example, when the ball bearing fails, when the ball touches the crack of the bearing, or the defective position of the inner ring or the outer ring, an impact will occur, causing vibration and even a slight displacement of the rotating shaft. This impact sometimes occurs. An audible sound (ie, a shock wave) that typically appears in the frequency spectrum as low-energy spectral components greater than 5 kHz. As the problem worsens, the low-energy spectral components will continue to increase, and when we capture these vibration signals through the accelerometer, we can respond in the shortest possible time.
Figure 2: Different motor fault types correspond to different vibration spectrum characteristics (Source: ADI)
Therefore, it is not difficult to conclude that as long as we build a system from state perception, data acquisition, data transmission, analysis and processing around abnormal motor vibration, we can build a complete set of CbM and predictive maintenance solutions. Let it assume the role of the “whistleblower” of motor health problems and eliminate the fault in the bud.
Figure 3: Motor CbM and Predictive Maintenance Solution Architecture
(Image source: ADI)
Choosing the Right Accelerometer
From Figure 3, we can see that the design of a complete motor CbM and predictive maintenance scheme requires technical decisions in multiple technical links, and as the “starting point” of the entire scheme, it is necessary to select the one that can accurately capture the vibration signal. accelerometer. This kind of acceleration sensor for measuring vibration needs to be installed near the monitored object, and bandwidth, reliability, size, power consumption, cost, etc. are all factors that need to be comprehensively considered during model selection.
In the past, piezoelectric accelerometers were more commonly used for vibration measurements in CbM, because these sensors have a wide bandwidth (typically in the range of 3Hz to 30kHz, even up to hundreds of kHz), which means that they can “observe” more Signals of abnormal vibrations in a wide frequency spectrum. At the same time, piezoelectric accelerometers have good linearity, SNR, and high-temperature performance, which are very important characteristics in industrial applications. However, the performance of piezoelectric accelerometers in the DC range is not good, so it is not suitable for low-frequency applications such as wind turbines, and its cost is also high, which will affect the further expansion of the application range.
In contrast, MEMS capacitive accelerometers have shown stronger momentum in motor CbM and predictive maintenance in recent years. MEMS capacitive accelerometers have DC response characteristics, strong shock resistance, and are smaller in size, lighter in weight, more suitable for mass production, and thus more cost-effective.
More importantly, with the advancement of technology, the problem of “lower bandwidth” that plagued MEMS capacitive accelerometers in the past has also been greatly improved. It is conceivable that the development of MEMS capacitive accelerometers is lowering the “threshold” of CbM and predictive maintenance, allowing this technology to penetrate into more scenarios of motor applications.
Figure 4: Comparison of two accelerometer sensors for motor vibration measurement
(Image source: ADI)
ADI’s ADXL100x series is a very eye-catching product in the comprehensive performance of MEMS capacitive accelerometers. This family of single-axis accelerometers, with measurement bandwidths up to 50 kHz, has a frequency response that covers the major faults commonly found in rotating machinery, including sleeve bearing damage, misalignment, unbalance, friction, looseness, transmission failure, bearing wear and Cavitation, etc.
Figure 5: ADXL100x family of MEMS accelerometers
(Image source: ADI)
ADXL100x MEMS accelerometers enable high-resolution vibration measurements, are available in different ranges from ±50g to ±500g, and have ultra-low noise densities of 25μg/√Hz to 125μg/√Hz over a wide frequency range, Provides stable and repeatable sensitivity and can withstand external shocks up to 10,000g.
Figure 6: ADXL1001/ADXL1002 can support high frequency vibration response >5kHz (Image source: ADI)
It is worth mentioning that the ADXL100x series of MEMS accelerometers also provide a feature that is usually not available in piezoelectric accelerometers, that is, integrates many intelligent features, such as an overrange detection circuit – when the specified g value range is exceeded by about 2 times. In the event of a severe overrange event, the circuit will alarm; at the same time, the ADXL100x can protect the sensor element in the event of a continuous overrange event based on an internal clock intelligent disable mechanism, which can reduce the burden on the host processor and increase the sensor node. degree of intelligence.
In addition, the compact LFCSP package, wide operating temperature range of -40°C to +125°C, and low power consumption of the ADXL100x series of MEMS accelerometers help facilitate their integration into industrial IoT applications. , becoming an integral part of Industry 4.0.
Figure 7: ADI offers a broad portfolio of MEMS accelerometers for CbM
(Image source: ADI)
Build high-performance signal chains
Of course, the performance of the MEMS accelerometer is outstanding, and it is only the first step in the signal and data link processing of the entire motor CbM. To create a high-performance motor CbM and predictive maintenance solution, a complete system is needed for support. It mainly includes the following key parts:
Primarily a MEMS accelerometer-based vibration and shock detection unit that provides high-precision measurements. Other sensing functions such as temperature sensing are sometimes required to be integrated.
High-fidelity data acquisition can convert vibration, shock, temperature, acoustic, pressure, voltage, and current signals captured by sensors into digital signals, which can then be converted into valuable data that forms the basis for insights and decision-making.
Small size, high efficiency solutions ensure that miniaturized smart sensors can operate reliably in harsh industrial applications.
Ultra-low-power MCUs or other host devices can make local decisions about events happening at industrial edge nodes, and filter more important data to the cloud to build more comprehensive insights.
The communication module can ensure reliable transmission of relevant motor health data to PLCs and Manufacturing Execution Systems (MES) in harsh industrial environments, accelerating CbM deployment.
AI-Based Cloud Insights
Artificial intelligence (AI), built on more powerful computing resources in the cloud, can detect and interpret data such as sound, vibration, pressure, current or temperature for continuous condition monitoring and on-demand diagnosis, and by interacting with experts in the field of CbM Keep learning and upgrading.
Figure 8: Motor CbM and Predictive Maintenance Scheme System Architecture
(Image source: ADI)
In such a long signal and data chain, to ensure the reliability and accuracy of data, it is necessary to select appropriate high-performance devices for each functional module.
For example, in order to ensure that the data acquisition module can achieve high-performance signal conditioning, it is necessary to spend some thought on the selection of analog devices such as amplifiers, ADCs, and reference voltage sources (see Figure 8).
In the selection of operational amplifiers, ADI’s LT6015 Over-The-Top® precision operational amplifier is a recommended device. This is a single-channel rail-to-rail input op amp with an input offset voltage of less than 50µV. These amplifiers operate from single and split supplies (3V to 50V total) and draw only 315µA per amplifier. They feature reverse battery protection at reverse supply voltages up to 50V. It draws very little current.
The LT6015 can drive loads up to 25mA and maintain unity gain stability with 200pF capacitive loads. The amplifier’s Over-The-Top® input stage provides additional protection in harsh environments. Their input common-mode range extends from V- to V+ and above, specifically these amplifiers can operate from V- to 76V input, with internal resistors protecting the input from transients below 25V from the negative supply. damage.
Figure 9: LT6015 Over-The-Top Precision Op Amp
(Image source: ADI)
In the selection of high-precision, low-power reference voltage sources, the ADR45xx reference voltage source is a good choice. This family has a maximum initial error of ±0.02%, excellent temperature stability and low output noise. The ADR45xx references offer both excellent temperature stability and noise performance due to the use of a new core topology to improve accuracy. The reference features low thermally induced output voltage hysteresis and low long-term output voltage drift, thereby improving system accuracy over lifetime and temperature.
At the same time, the maximum operating current of the ADR45xx series is 950μA, which provides excellent low power consumption characteristics; and the wide temperature range of -40°C to +125°C makes it suitable for a wide range of industrial applications.
Figure 10: ADR45xx References
(Image source: ADI)
Of course, ADI’s products for realizing high-precision data acquisition are far more than the above two products, but include a complete set of CbM signal chain and data chain solutions, which saves everyone’s complicated development work such as material selection and debugging.
Summary of this article
With the advancement of Industry 4.0, it has become a general consensus to replace traditional motor maintenance methods with CbM-based predictive maintenance. To create such a system that can sense the health status of the motor in real time and perform maintenance work on demand, it is necessary to build an overall solution around the entire signal chain and data chain. This requires not only cost-effective sensors, but also high-precision signal conditioning devices, reliable power management devices, and efficient data transmission and processing solutions.
At this time, the value of cooperating with a technology manufacturer such as ADI, which has a rich product portfolio and can provide one-stop solutions, is highlighted. The convenience and scalability brought by this can make the concept of Industry 4.0 penetrate into every “capillary” of the future manufacturing industry, and make the “body” of the manufacturing industry more robust and powerful.