Controlled excitation of ultrasonic transducers Arno VOLKER, Maurits VAN DER HEIDEN TNO Science Industry

摘要

The reproducibility of ultrasonic measurements complicates the interpretation of measurement results. The signal from ultrasonic transducers varies for to several reasons, like manufacturing tolerances in a batch, cable length, amplifier properties etc. A known problem that is caused by this effect is the quantitative interpretation of C-scan images from laminates, particularly the porosity present in a laminate. For this type of application, one requires a high level of signal fidelity, or in other words, system independent measurements. Ultrasonic transducers are commonly excited by a short electrical pulse, such that the temporal bandwidth of the transducers determines the shape of the signal. Instead, we propose to use an optimized sweep signal, such that after correlation the measured signal has a desired wavelet shape. The correlation is done with a specially designed correlation signal instead of using the transmitted sweep. During the design process of the sweep and correlation signal, the transfer function of the system, i.e., transducer and electronics is measured and transformed into a desired response independent of a specific transducer. This approach yields several advantages over conventional excitation, i.e., it provides higher temporal and spatial resolution, a better signal to noise ratio and allows system independent ultrasonic measurements. The controlled excitation makes it possible to widen the frequency spectrum of a transducer by boosting the signals at frequencies where the transducer has a lower sensitivity. A bandwidth up to 150 % is not difficult to achieve with this technique. The length of the sweep determines the signal to noise ratio, the correlation process has he same effect as averaging of signals, only it is much more efficient in time. Improvements up to 30 dB can be achieved compared to single pulse excitation.