where expertise comes together - since 1996

Web's Largest Portal of Nondestructive Testing (NDT)
Open Access Database (Conference Proceedings, Articles, News), Exhibition, Forum, Network

NDT.net Issue - 2016-03 - NEWS

Condition monitoring system for wind turbine machinery using non-contact acoustic sensors

TWI Ltd (The Welding Institute)176, Great Abington, Cambridge [United Kingdom]
NDT.net Journal
Issue: 2016-03
Case Study

Operation and maintenance costs constitute an average of more than 30% of the total cost per kWh produced by a new wind turbine. Manufacturers are increasingly seeking to develop new maintenance approaches to bring down these costs.

TWI, in partnership with RTS International, worked to support this aim through the CM Project, a research programme funded by Innovate UK, within the competition ‘Emerging Energy Technologies’.

The project sought to establish the feasibility of using a non-contact acoustic pressure sensor to monitor the condition of machinery in an onshore wind turbine. This was achieved through experimental field trials in a 2MW turbine lasting over a month.

Figure 1. Wind farm location for experimental trials

Acoustic sound and vibration monitoring

Vibrations from all the rotating components within the nacelle were monitored through the airborne noise generated by the surface vibrations of the machinery.

Figure 2. Sensing instrumentation

The noise was detected by acoustic pressure sensors (microphones) mounted at the nacelle inner surface.

Unlike contact sensors, which cover only the small portion of their contact area, these microphone sensors are able to monitor up to 75% of the noise emitted from the machine surfaces. The developed non-contact sensor array consists of a group of five acoustic pressure sensors arranged in a manner which allow its easy installation and set-up.

Field trials

Figure 3. Sensor array installed in the wind turbine nacelle

To determine the feasibility of the non-contact technique, acoustic sound and vibration monitoring were carried out during the UK-based turbine’s normal operation. An array of five acoustic pressure sensors and two accelerometers was installed. In addition to acoustic sound and vibration data, parametric data, ie wind speed and power output, were obtained for correlation with the acquired data, as shown in Figure 4.

Figure 4. Root mean square of the acoustic signals versus power

Contact us

For further information about the project, please visit the project website: www.cmdrive.co.uk.

If you would like further details, please contact us.

TWI Ltd (The Welding Institute)

More of TWI Ltd (The Welding Institute) published in NDT.net
2018-12-01: Enhancement of ultrasonic guided wave signals using split-spectrum processing technique
2018-11-01: Piezoelectric transducer excitation for guided waves propagation on pipeline with flexural wave modes
2018-11-01: A random vibration based platform for statistical damage detection in a population of nominally identical composite aerostructures
2018-11-01: Comparison of coded excitations in the presence of transducer’s transfer function in ultrasonic guided waves testing of pipelines
2018-11-01: iPerm: A guided wave pipeline monitoring tool for Oil & Gas industry
2018-10-01: NDT In Service Inspection Training and Certification Aspects
2018-10-01: Industrial 3D X-ray microscopy allows extraordinary detail to be revealed in microstructures
2018-10-01: TWI has appointed two new training managers, Martyn Wright and Stephen Wisniewski
2018-09-01: TWI Joint Industry Project on early-stage detection of high temperature hydrogen attack in carbon steels opens to partners
2018-08-01: Ultrasonic Phased Array Non-Destructive Testing And In-Service Inspection System For High Integrity Polyethylene Pipe Welds With Automated Analysis Software
All 178 Articles & News of TWI Ltd (The Welding Institute) in NDT.net

© NDT.net - The Web's Largest Portal of Nondestructive Testing (NDT) ISSN 1435-4934

Open Access Database, |Conference Proceedings| |Articles| |News| |Exhibition| |Forum| |Professional Network|