Processing of ultrasonic array signals for characterizing defects. Part I: signal synthesis.

This work presents a novel procedure to characterize damage using an array of ultrasonic measurements in a generalized model-based inversion scheme, which integrates the complete information recorded from the measurements. In the past, we proposed some idealized nondestructive evaluation test methods with emphasis on the numerical results, but it is necessary to develop the techniques in greater detail in order to apply the techniques to real conditions. Our detection principle is based on the measurement and inversion of frequency-domain data combined with a reduced set of output parameters. The approach is developed and tested for the case of an aluminum specimen with a synthetic array of point contact ultrasonic transmitters and receivers. The first part of this two-part paper is focused on numerical synthesis of the experimental measurements using the boundary element method for a general ultrasonic propagation model. This part also deals with the deconvolution by comparing the data measured from the damaged and undamaged specimens. The deconvolution technique allows us to calibrate the data by taking into account the uncertainties due to mechanical properties, input signal, and other coherent noise. The second part of the paper presents the inversion of the measurements to obtain the parameters and ultimately to predict the position and size of the real defect.