Structure health monitoring in natural environments: Pre-failure event location and full-waveform characterization by nanoseismic monitoring

Nanoseismic monitoring (NM) is a section of experimental seismology that focuses on the detection, location and characterization of extremely low-energy (0.0 ≥ M_L ≥ -4.0) source processes recorded by portable sparse arrays (Wust-Bloch and Joswig, 2006; Joswig, 2008; Wust-Bloch, 2010; Walter et al., 2012). The structural health monitoring technique presented here down-scales NM to suit engineering applications in shallow geomaterials, whereby pre-failure microcracking is monitored within the 1-200 Hz band and fully characterized at unusually short slant distances (10²-10^-1 m). Here, NM techniques are validated in lab by monitoring incipient microcracking generated by unreinforced concrete beams and limestone plates undergoing four-point bending tests. These ground truth (GT) tests show that progressive loading triggers a wide range of impulsive signals whose frequency and rate patterns evolve until complete material failure occurs. Following this, NM techniques are applied to monitor microcracking generated by unstable archeological caverns that were excavated in natural chalk. Although signals can be detected in unfavorable signal-to-noise ratio conditions by a single array at slant-distances beyond 10² m, reliable locations can only be obtained when several mini-arrays are deployed in the vicinity of caverns suspected to be unstable. Epicentral locations of microcracking events tend to cluster near free boundaries and in zones of high tensile stress, as predicted by numerical models computed for these caverns. These results confirm the operative capacities of low-cost NM techniques in locating and analyzing pre-failure microcracking processes occurring within weak natural media. Complex source processes, which occur at distances that exceed the detection range of standard acoustic emission and ultrasonic monitoring, can be fully evaluated using portable equipment that can be deployed within minutes with no prior infrastructure.