- Understand the requirements "First and foremost, implementing a CM regime means that you have to know your plant," states Paul Poste, director and general manager at Pruftechnik. "A basic understanding of the way the machines behave and the way they should behave is essential. Information from ISO standards, machine suppliers and past plant operating experience will often provide the information needed." Ian Pledger, field service engineer at Schaeffler UK, concurs: "The key to good CM is understanding CM itself before looking at individual monitoring techniques. Equipment not fitted correctly and machines not used for their intended purpose are just two of the problems that can lead to false readings and a false understanding of plant condition."
- Match the system to the plant There is little point in trying to fit the plant to the CM equipment: the CM equipment must be aligned to the plant. "The one tool or technique that makes the difference is always the one that works best for the application or environment in question," says Pledger. "The first thing is to have a good relationship between the plant engineers and your CM partner. Secondly, it's important not to do too much too quickly – construct the portfolio slowly but effectively so that everyone is onboard at the outset. Thirdly, you need to decide if you're doing the CM yourself, whether you have a dedicated workforce, or if CM will be an ad-hoc activity. The latter is difficult because CM is not always a priority role for engineers who are pulled in many different directions and have limited resources." The key point here is to take the CM route that is most comfortable. Match the system to the plant and the skills available. Don't go for the most expensive option just because a salesman says it's the best – it may be, but it might not be appropriate to the plant. "If the plant has many process-critical machines, online monitoring may need to be considered, while if the process plant is small, start simple: routine visual inspection and straightforward vibration readings will suffice," says Poste.
- Vibration analysis Vibration analysis is arguably the most popular CM technology in the UK. According to Schaeffler, vibration monitoring is best suited to high-speed applications because it is easier to obtain more detailed information compared with, say, acoustic emission monitoring techniques. Vibration monitoring equipment ranges from handheld portable devices up to integrated systems with permanently mounted sensors. Here, Rockwell Automation insists that vibration monitoring is better applied in an integrated format that sidesteps the need to establish communications with the plant/machine controller. An integrated solution can monitor the vibration amplitude of multiple channels or possibly examine two measurement channels more accurately, as well as carry out FFT (fast Fourier transform) analysis for vibration characteristic studies.
- Thermography Where there are several monitoring positions in one location – for example, conveyor rolls in steel mills – thermal imaging is often the preferred technique because it is quick and easy to implement, and data can be obtained immediately. Thermal imaging cameras are simultaneously becoming cheaper and more powerful, while the parallel development of software is making the technology exceptionally easy to use, extending its scope of application. According to Eriks, the use of thermal imaging cameras makes it possible to target otherwise inaccessible components within machinery and identify parts that are functioning outside temperature limits. What's more, all of this can be achieved with minimal disruption, since thermal images can be gathered swiftly from a safe distance while machinery is still running.
- Acoustic emission monitoring "Traditional vibration analysis has provided a trusted approach to CM for the past 30 years, but it's a complex science and requires sophisticated knowledge and understanding. Acoustic emission technology, however, places the power of CM directly into the hands of every engineer," says Martin Lucas, managing director at Kittiwake. "There is a clear benefit in knowing what is going on at an exact point in time – not just when the engineer can get to a machine for a scheduled sample and analysis. Onsite diagnostic kits enable rapid testing and action, and online sensors remove sampling errors." Able to provide real-time information, the acoustic emission technique is based on the detection of the high frequency component of naturally occurring stress waves. It is suitable for continuously running plant as well as machinery operating intermittently or for short durations.
- Oil analysis Continuous oil condition monitoring of machinery and lubricant testing is fast becoming the established method of predicting and avoiding impending machinery breakdown, says Kittiwake, which insists that most experts in this area recognise that oil analysis will detect machine wear before vibration analysis. The true value of vibration analysis is its inherent ability to localise the problem (inner race, outer race, cage wear) rather than any ability to find a problem earlier in the failure cycle. In truth, the combination of oil analysis for early detection coupled with the advanced diagnostic capabilities of vibration analysis make the benefits of these two techniques far greater when treated as team-mates rather than competitors.
- Single device technology According to Geoff Walker, director at Artesis, CM has been difficult in the past because each stage in the process was complicated, requiring expertise to make the right judgements. For example, for continuous online vibration monitoring, sensor installation knowledge, software competency and data interpretation skills were required. This situation was compounded by the fact that past CM technologies only covered a limited set of parameters – vibration sensors could detect mechanical problems, electrical meters detected electrical problems, etc. "This situation has changed with the advent of model-based current and voltage analysis systems that use the electric motor as a sensor to convey not only what is going on in the motor itself, but also the driven equipment," says Walker. "These solutions monitor and detect incipient faults in electrical, mechanical and operational areas through a single device." Detection is achieved by measuring the voltage and current drawn by the motor, and by building mathematical models of the relationship between voltage and current.
- The OEE factor "Increasingly, plants are looking to boost OEE through CM," says Seamus Rice, ICM team leader at Rockwell Automation. "In the past, output rate in parts per minute was the key benchmark figure for machines. Nowadays, product quality and machine availability are also included because it is these three parameters that comprise OEE. "Manufacturers and machine users regard machine availability as very important," he continues. "Consequently greater emphasis is being put on preventative maintenance. Whereas in the past a cyclical replacement of components was normally undertaken irrespective of their functional capability, today the trend is towards condition-orientated exchange or repair. The question of which parts to replace is answered using a condition analysis based on CM data."
- Training One of the keys to running an effective CM regime is the investment that management is prepared to make to ensure operators are skilled in using the systems they employ. Thermal imaging, for example, is quite a science, or so says Jon Willis, training manager with Flir, where factors such as material thermal conductivity, emissivity, ambient temperatures and external influences such as HVAC, are significant. "Infrared cameras allow thermal irregularities to be spotted in an instant but in order for the operator to interpret the images correctly, he or she needs to know how different materials and circumstances can influence temperature readings," he says. "The thermal problem also needs to be put into context as not everything showing hot is a fault. Taking a course in basic thermography is undoubtedly the best action plan for any maintenance engineer seeking to adopt this technology."
- Data trends According to Pruftechnik, CM essentially means that machines get a regular 'health check'. This is usually taken in the form of periodic measurements that are compared to a standard or known operating condition. Depending on the extent of any change, machine condition can be further investigated or monitored more frequently. The key tool in this CM concept is to assess trends of collected data and respond to changes before the machine fails catastrophically. Taking this further, a PC program can be deployed to automatically establish and trend machine condition data. Further developments allow analysis of machine condition via algorithmic calculations (FFT analysis) to determine specific machine condition defects, while online monitoring systems can provide 24-hour measurement and alarming of machine condition.
Conditions for success
7 mins read
It takes the right mixture of tools, techniques and training to achieve best practice condition monitoring. But which ones really make the difference? Steed Webzell provides some pointers
No one ever said that good condition monitoring (CM) was easy to achieve, but certainly the latest technology is a huge abettor – gone (thankfully) are the days of listening to a screwdriver resting on a bearing housing wondering whether the 'pitch' sounds different to yesterday.
What follows are 10 recommendations for successful CM compiled after consultation with some of the leading players in this important technology area.
