In the mining industry interest in vibration analysis is growing. Avoiding unexpected shutdowns is a simple way to increase productivity. Steen Christian Knudsen tells about FLSmidth's Smart Parts program.
In the old days, we used screwdrivers and fingernails – today, FLSmidth uses transducers and advanced calculation software. But the basic premise remains the same: as an audible fingerprint every machine in the world has its own unique vibration profile. Much like a physician listening to a patient's heart murmur, by listening to a machine’s vibrations, it is possible to predict and diagnose whether a bearing or a gear is in need of replacement or repair, and when.
With sensor cost reduced, increased power of computers, and the ability to connect to the internet everywhere – referred to by one of today’s popular buzz terms as the Internet of Things – the condition monitoring of machines is moving from offline to online.
Steen Christian Knudsen is Technical Manager overseeing FLSmidth's Smart Parts program: “We are creating some really amazing opportunities with the speed at which we can send vibration data and other data sets, and then correlate in many different ways with various operating parameters. It is now possible to know when to have spare parts delivered or have them printed in 3D, on site. Vibrations from rotating machines are especially useful in this respect, as they are so sensitive to changes in the condition of the machine.”
Investment in condition monitoring and root cause analysis is well worth it; vibrations from rotating machines such as SAG mills can indicate changes in the condition of the machine, as well as help sort irregularities between those that are urgent repairs and maintenance versus those that can be monitored or addressed during planned shut downs at a later date. With $100,000 in lost earnings for every hour that a gyratory crusher lies idle, it is crucial to know the correlation between reduced operation and prolonged maintenance.
Monitoring equipment no longer need to be decoded on site, but can be analysed by FLSmidth from the cloud, where it can be accessed by vibrations experts anywhere in the world.
Using different mathematical software tools, the vibration teams carry out signal processing on the vibration measurements, typically by analysing the frequency of the vibrations, the amplitudes and modulations in the vibrations and the frequencies of the frequencies – the so-called cepstrum. This allows the vibrations specialists to analyse when a machine component needs to be scheduled for replacement, and how to operate the machine to get the optimal production until then.
In May 2016, FLSmidth entered into collaboration with General Electric to integrate vibration measurements into their Predix platform, which will allow customers to compare their equipments' vibration data with other parameters, such as temperature and flow.
“Now that will enable us to discover some highly interesting connections. The software is able to correlate measurements with one another, recognise patterns and get the equipment to operate accordingly. By augmenting human intelligence the software makes experts more effective and it enables us to constantly optimise operations,” says Steen Christian Knudsen.
Contact: Steen Christian Knudsen, Technical Manager, Global R&D, FLSmidth: SCHK@FLSmidth.com