September 2021

37 www.drivesncontrols.com September 2021 INSIGHT n depends and whose failure may cause a catastrophic bottleneck in production. As with most things related to condition monitoring and predictive maintenance, it depends on the details of the application. Condition monitoring of machinery has traditionally relied on vibration analysis, because wear and imbalances on bearings, shafts, rotors, gears, and other moving parts, causes unusual variations in the vibration patterns in an operating machine. Monitoring machine vibrations and recording and analysing the patterns, can therefore help to identify defects and possible failures. Vibration analysis and other forms of dynamic monitoring are still common methods of condition monitoring, but they are far from being the only option. Another traditional type of condition monitoring is performance monitoring. This uses observation and performance trending, and derives an indication of a machine’s health from its output and existing manufacturing performance measurements. Similarly, an indication of health can be determined by monitoring an asset’s electrical performance using power signature analysis, surge testing, motor circuit analysis, and similar tools. External approaches to condition monitoring include thermography (thermal imaging), ultrasonic monitoring, acoustic analysis, radiography, oil analysis and tribology, laser interferometry, and electromagnetic measurement. Every condition-monitoring application is unique to the asset being monitored, and the needs of the organisation using it. Depending on the application, one or more of these types of analysis could be used. Smart drives When it comes to monitoring rotating assets, modern drives are incredibly intelligent devices. For motors controlled by drives, significant amounts of useful data are already available to the maintenance engineer, who often needs only a way to access it and monitor changes over time. Drive parameters such as output voltage, DC link voltage, power, torque and rotor speed, if made accessible and easy to monitor, can all give an indication of the health of the machine, and monitoring how they’re trending can indicate potential impending failures. For example, Siemens’ Sinamics Connect 300 is a device that can collect parameter data (including power, current, speed, torque, motor temperature, converter temperature, converter state, and fault codes) from up to eight Sinamics drives, and upload it directly to Siemens’ MindSphere Industrial Internet of Things (IIoT) platform. It is then presented as time series data, with the ability to set up text or email alerts for when a value goes out of pre-defined bounds. Further analysis can be performed on the data using various applications available on the MindSphere platform. This has the added bonus of being accessible from anywhere in the world as it’s an IIoT platform. But this is only one way of making use of the data already available in drives. There are various other techniques for performing electrical monitoring of drive systems using parameter values from the drives themselves. For a more in-depth analysis of the condition of a motor, vibration analysis may be appropriate. Another MindSphere technology is the Simotics Connect 400 – a wireless device that attaches to the housing of an induction motor. It uses onboard accelerometers and other sensors to monitor the health of the motor via its speed, vibration, and temperature, and communicates these directly to the MindSphere platform. While these and similar devices are easy to install and retrofit onto existing drives and motors, some condition-monitoring systems are a bit more complicated to set up. A more in-depth vibration analysis system, such as Siemens’ CMS range, requires piezoelectric vibration sensors to be installed on the motor (and gearbox, if applicable). This would probably involve drilling and tapping holes in key locations, and would need advice from a vibration analysis specialist as to which locations are most appropriate. Whether you need an expert to interpret condition-monitoring results depends on what is required in terms of interpretation. With vibration monitoring, for example, it is possible to determine upcoming failure modes for a particular bearing in a machine. This level of detail requires the involvement of a vibration monitoring expert – or, at least, someone extremely familiar with vibration analysis tools such as Siemens’ X-Tools software. However, many applications of condition monitoring do not require this level of detail, and only need warning of an impending failure in time to arrange for a maintenance engineer to inspect the asset and identify the problem. Once the condition-monitoring system is set up, and the standards of healthy operation recorded, all that is needed is for someone to observe the state of current operation and compare it to the state of healthy operation to identify any significant deviations or trends away from the healthy state. With the levels of automation available today – from the simple ability to set up alerts when a parameter goes outside of a pre- defined boundary, through to automated trend monitoring and maintenance scheduling – it’s now easier than ever to perform condition monitoring. n Monitoring vibration levels in motors can help to identify defects and avoid possible failures. Whether you need an expert to interpret condition- monitoring results depends on the application and the results you require