An energy efficient wireless fatigue sensor system for long-term monitoring of assets

Metal fatigue is one of the main contributors to mechanical failure ina range of assets, from bridges and street lamps to pipelines and power turbines. During the winter of 2015 the Forth Road Bridge – a main route for around 70,000 vehicles crossing daily between Fife and Edinburgh – was forced to close for three weeks and had traffic restrictions imposed for several months due to a fatigue-induced crack found during a routine visual inspection. Estimates suggested the closure of the bridge cost the economy more than £40m.

Metal fatigue

Fatigue is caused by recurrent loading or stress to a structure causing the formation of micro cracks further weakened by repeated cycles of stress over an extended period of time. Metal fatigue can lead to sudden fracture and ultimately structural failure, even at stresses much lower than the structure would normally withstand.

Current approaches

A widely used approach for quantifying fatigue is by measuring strain and using a well-known ASTM-standard cycle counting algorithm to identify the number of stress cycles. This information is used by civil engineers to estimate the remaining fatigue life which is key to informing effective and efficient maintenance. Common sensing technologies include vibrating wire strain gauges and metal foil strain gauges; however, their use often requires a permanent power supply, which may not be readily or easily available to all assets, or high capacity batteries which are costly and difficult to maintain.

CSIC works to demonstrate better-performing and more energy efficient sensing

CSIC researchers have designed a wireless embedded sensor system that combines a novel low-power analogue electronics design and highly energy efficient software to deliver a monitoring technology that does not require mains power, allows for continuous fatigue monitoring and is durable over prolonged timeframes. Considering every component of the system under development and improving even one of these elements results in significant added value. This system only reports the strain-cycle information necessary to estimate the fatigue life of an asset. Increasing efficiency in this way means the system neither collects nor reports the vast amounts of data associated with current standard procedures. Low-power wide area network (LPWAN) technology enables the sensors to transmit data wirelessly over long distances for long periods. 

Experiments comparing the CSIC prototype with a conventional system with similar hardware characteristics to test the performance of the novel system showed a 9-fold increase in energy efficiency. Using a single 3.6V 19 Ah Lithium battery would power the conventional system for two to three months, whereas the CSIC solution will run for nearly two years. This solution is cheap to build and does not require additional cables and equipment associated with conventional systems.

The energy efficient wireless fatigue sensor system is ready to be deployed on a live site. CSIC is interested in hearing from organisations with suitable projects for potential collaboration.

Powering a sensor system to monitor dynamic performance of our infrastructure over the longterm is one of the greatest challenges to achieving the vision of smart infrastructure. This work shows that we can overcome this challenge by integrating state-of-the-art hardware and software tightly together. The data obtained from this system can lead to a new way to manage our infrastructure for significantly longer periods of time.
Professor Kenichi Soga, Chancellor’s Professor at the University of California, Berkeley and CSIC Investigator

Contact: Dr Xiaomin Xu
Team: Dr David Rodenas Herráiz and Dr Xiaomin Xu