Signal model parameter optimisation for primary wave signal responses from ultrasonic pulse transmission tests

Signal analysis concerning signal recordings from ultrasonic pulse transmission tests on different testing materials (cement paste at the early stage of hydration, ambient air, tap water, aluminium cylinder). Semi-automated fitting of a damped sinusoidal signal model to natural signals. This record contains a presentation containing a brief description of the analysis and all analysis results.

In ultrasound signal analysis, the time range estimation plays an important role. The time range is the time span elapsing between the pulse excitation and the time when the sound wave hits the ultrasonic transducer. In the coherent literature, signal model fitting is addressed as one of the most accurate but also most costly methods to determine the sound wave's onset point, which is the upper limit of the time range.

In this study, a signal model is fitted to the first ascending flank of the primary wave in order to detect the onset point and to parametrise the signal model. To show the broad applicability range, signals originating from ultrasonic pulse transmission tests (UPTM) on cement paste, ambient air, tap water and an aluminium cylinder are used.

A parameter variation based on that signal model allows for examining the signal model parameter's value ranges. Thereby, the fitting error is minimised using the least-squares method.

The results show that the signal model fits properly to a variety of primary wave signal responses, but not to all of them to the same degree.

To foster comprehensibility and reusability, all materials and results created in the course of this presentation are made available publicly under open licenses.