Discrete Damage Modelling for Computer Aided Acoustic Emissions.

This chapter is conceived as an essay on modern multiscale discrete damage modelling, providing a brief personal perspective about its foreseeable applications-implications for structural health monitoring purposes. In particular, it is argued that this sort of damage modelling could be potentially useful in damage detection by acoustic emissions (AE), which is a class of non-destructive techniques (NDT) used to capture damage evolution in a number of materials (e.g. from concrete systems such as bridges and beam elements to composites in aircraft components and pressure equipments) and from a number of external actions (e.g. sustained load, monotonic testing, fatigue, corrosion, etc.) (Biancolini & Brutti, 2006 ; Carpinteri & Lacidogna, 2008 ; Grosse & Ohtsu, 2008). With AE it is possible to “hear” the microcracking phenomenon and characterize the location and magnitude of a single microcrack (of size and “strength”1 beyond certain thresholds) acting as an acoustic source. Hence, it is routinely possible to plot the released energy of each crack as a time series or to map them over a 2D spatial domain by counting and locating individual acoustic events in time. Yet the analysis of this type of output is not straightforward and major difficulties exist, let alone sensitivity issues of equipment, material dependence, and other practical issues. The scope of this discussion covers two issues of general interest: 1. the randomness of the AE signal, 2. the need for structure-property relations as companion to AE monitoring.