Acoustic emissions announce machine stress Unexpected machine failure can cost small and medium-sized enterprises (SMEs) a fortune. Novel, inexpensive non-destructive testing technology developed with EU support will enhance monitoring to enable intervention before catastrophic failure. Industrial Technologies © Thinkstock From replacing broken parts to lost productivity due to downtime to the cost of additional labour not to mention potential issues with dissatisfied customers, SMEs need protection against machinery loss. The methods currently employed by large enterprises are cost-prohibitive for SMEs. The EU-funded project MOSYCOUSIS created a smart acoustic emission (AE) sensor incorporating on-board data acquisition, conditioning and processing systems for real-time monitoring and preventive maintenance. AE is the sudden release of strain energy in the form of transient elastic waves from a material under stress. Particularly in rotating machinery in industrial plants, AE can occur during impacting, cyclic fatigue cracks, friction, turbulence, material loss, cavitation or leakage. Researchers developed an intelligent monitoring system through self-powered wireless sensor networks using microprocessors and miniaturised radio transceivers. Considerable focus was placed on fault condition characterisation using data generated from simulations, laboratory tests and industrial machinery. The resulting database contains comprehensive information generated by testing mechanical components and metallographic samples under static and dynamic conditions. Sensor hardware includes an AE conditioning module that can manage up to three channels for analysis by the fault detection algorithm and the power supply module that incorporates three possible energy harvesting modes that utilise waste energy in addition to a conventional supply. Software development consists of specific algorithms exploiting signal processing and AE theory to correlate AEs to faults in the machinery, to perform machine prognosis and to calculate life expectancy. The cost-effective MOSYCOUSIS condition-based predictive system will improve production quality and workforce safety while reducing production time. This along with reductions in workforce and machine downtime from failures will considerably increase profitability through lower overheads. The need for this system is particularly felt during this period of economic recession. Successful outcomes will thus increase the competitiveness of participating SMEs as well as end users with important implications for the EU economy. Keywords Acoustic emissions, machine stress, machine failure, SMEs, sensor, real-time monitoring, preventive maintenance, strain energy, wireless sensor networks
