Virtual vibrations

Adopting a condition-based predictive maintenance (CBPM) strategy is widely recognised as the best way of protecting expensive plant equipment - particularly that which is required to operate 24/7.

Vibration levels can offer an early indication of wear and tear, with the aid of monitoring systems that can raise alarms so that engineers can investigate the causes.

Stork Food & Dairy Systems is a keen advocate of CBPM as a means of better scheduling servicing and preventing unexpected failure.

The company specialises in the development and support of integrated processing and filling lines for the dairy and other industries.

The company has 40 Dairyfill rotary fillers in use with UK-based customers. These typically operate non-stop and if they do need to shutdown outside of scheduled maintenance it creates considerable and costly logistical problems.

“A filler is a complex machine, with several gearboxes and bearings, and most are operating at speeds of up to 300 2-litre bottles per minute,” says Luke Axel-Berg, Stork’s director of sales for UK and Northern Europe.

Stork had already been monitoring vibration levels to determine the health of its fillers.

However, in 2009, the company took CBPM to a new level. Working with the Centre for Intelligent Data Solutions, which operates out of the University of Portsmouth, Stork began developing an advanced condition monitoring system called ‘Virtual Engineer’ and for which it received funding under the UK Government’s Knowledge Transfer Partnership (KTP).

An early challenge was to determine how many vibration sensors should be used and where best to position them.

Here, Stork turned to Monitran, the UK-based OEM of sensors. Monitran made a site visit to see a Dairyfill machine in operation and recommended the use of 16 mechanically mounted sensors per machine.

What, where and why

The sensor mounting locations include the main bearing and hub, main drive motor, drivetrain, capper top and bottom bearing.

For data collection purposes Stork decided to initially monitor three fillers.

“Our team fitted the sensors to the selected filling machines and used HBM data acquisition units, which can cater for up to 16 analogue inputs, to convert the sensors’ analogue outputs to 24-bit resolution digital,” says Axel-Berg.

The vibration levels are sampled at a nominal 3kHz and data is then streamed over the internet to a data centre at the University of Portsmouth.

By sampling at such a high frequency it was possible to quickly build a detailed picture of ‘normal’ behaviour, on a machine-by-machine basis, for different modes of operation.

During the first year, anomaly detection algorithms looking for signs of deviation from normal behaviour identified six conditions that, if left unchecked, would have led to catastrophic failures.

These conditions included two instances of developing main bearing failures, water in a gearbox and a universal joint out of balance.

Moreover, a developing gearbox fault was detected on a machine that was also being protected by a third-party complex spectrum analysis warning system.

It was unaware of the worsening condition as the gear in question rotates at about 7rpm; a frequency sufficiently low to be masked by the ‘noise threshold’.

Since the initial trial period, a further eight conditions have been detected.

“These were typically more of the same,” says Axel-Berg. “However, on one machine, we encountered some previously unseen spurious vibrations. Upon investigation the source was found to be a fault developing in a height adjustment system, which we’d not set out to monitor.”

Axel-Berg says this reinforced Stork’s conviction that as time passed, a clear picture of normal behaviour and known faults would be established, with any new developing vibration patterns offering the tell-tale signs of a new fault.

“And good predictive maintenance is all about not being taken by surprise,” he says.

Based on the 14 failures that have been avoided, Stork estimates the savings are in excess of £2million.

“Virtual Engineer marks the first use of a remote, real-time, asset monitoring system within a CBPM strategy built primarily around vibration levels and signatures in a dairy application, and possibly within the food and drink industry as a whole,” says Axel-Berg.

Monitran sensors have since been fitted to a further eight fillers and Stork will be using Virtual Engineer to monitor their health.