Impact-echo is successfully applied for the non-destructive evaluation of concrete structures. Its principle is based on the use of longitudinal waves generated by the elastic impact of a steel sphere. However, this impact also creates other wave types such as Rayleigh and shear waves. Especially when IE is applied on structures with compact dimensions the effects of these waves become crucial. The separation of the different signal components and the definition of effective criteria to distinguish between them is therefore highly relevant. In this study the duration and attenuation of the different signal components is explained and used to distinguish between the actual measurement information and interfering effects caused by reflections at the boundaries. It is explained why the different components of an impact-echo signal have different characteristics regarding their duration over time. Furthermore a procedure is introduced, which makes use of the relation between the signal duration in the time domain and the bandwidth in the frequency domain. It therefore avoids the application of a rather complex time-frequency analysis. This is demonstrated on simple sinusoidal signals in a first step and then applied to data collected from impact-echo measurements. Due to the low computation effort and the fact that only few input parameters have to be chosen, this method is also applicable to whole scan lines comprised of a large number of measurement points and the imaging of the results.
Abstract