A Kalman-based deconvolution technique applied to ultrasonic imaging of fibrous composite hydrogen reservoirs: preliminary results

Composite reservoirs are common in high-pressure storage of gaseous hydrogen and the cryogenic storage of liquid hydrogen. These tanks are subjected to a lot of stresses that can result in the propagation of cracks than can alter their fiability and safety. It is then necessary to perform tests on these reservoirs to find and locate such defects. Non-destructive testing using ultrasounds is a valuable approach to perform this kind of task. Unfortunately, ultrasonic testing of fibrous composite structures is made difficult because of the noise caused by the diffraction of the emitted wave on the fibres. Due to an overlap between the emitted pulse and noise spectra, standard filters are not efficient to eliminate such structure noise; it is therefore difficult to image clearly the defects present in these structures. We propose a deconvolution technique based on the Kalman filter to detect the different echoes coming from the tested specimen in order to perform ultrasonic imaging of its structure. Preliminary results of C-scan type images obtained from the deconvolution are presented here. (author)