Enhanced damage characterization using multiply scattered lamb waves

Lamb wave inspecting has being a potential solution for structural health monitoring of composite structures. In such applications, the Born (single-scattering) approximation is used, and the signals received after the wave scattered from the target are normally discarded. In reality, however, the structure often contains a variety of known singular features (e.g., edges, welds, reveted-lap joints and adhesive butt joints), and waves can scatter several times before returning to the detector. The information encoded in multiply scattered waves can give more information about the region of interest, because the extra reflections arise waves that interrogate the region from different directions. In this paper, we focus on multiply scattered signals, exploiting information that is usually discarded for damage characterization. Firstly, dispersion compensation is employed to compress the time duration of the multiply scattered waves, separating them in time domain. For each individual transducer, the edge of composite laminate could arise a virtual transducer distributed in a symmetrical position, which enlarges the element-number and the aperture of the sensor array. On this basis, a modified sparse reconstruction algorithm is employed for damage imaging. The experimental results show that the damage could be correctly detected and accurately localized even with a single transmitter-receiver pair through exploiting these multiply scattering waves.