Inclusion in water - phased-array - Conclusion

The predictions of the models available in CIVA for the calculation of a steel inclusion and an infinite plane echoes in water were evaluated by comparing the experimental results obtained with a phased-array probe operating at 5 MHz.

The results of the study and the steps followed to determine the probe parameters entered CIVA (reference signal, attenuation ...)  were presented. Several active apertures were evaluated by changing the number of active elements, and by application or not of a delay law.

We recall below the main results according to the configuration.


Configuration without delay laws

For large apertures, a very good agreement is obtained between experiment and simulation with SOV-COMPLETE and SPECULAR models. The differences do not exceed 2 dB. The agreement between experiment and the SOV model is less good than for the 2 other models when the inclusion is close to the probe and when the aperture is large. The maximum differences reach  4dB for the 20 elements aperture.

The distance dmax at which the amplitude of the echo is maximum is well predicted by the SOV-COMPLETE and SPECULAR models for large apertures. The differences do not exceed 1mm. However, the SOV model overestimates "dmax" of more than 10mm; the most important differences can reach 20mm for the larger aperture.

For smaller apertures, a good prediction of the evolution of the amplitude as a function of the probe-inclusion distance is obtained with a systematic prediction of the two maxima of amplitude when the active aperture is asymmetrical. The differences in position of the distance dmax are nevertheless important, the underestimation may reach 13 mm for 4 elements aperture with the SPECULAR model.
However, unlike large aperture, the smallest differences in the amplitude and position of the maximum of amplitude are obtained with the SOV model. These lower differences compared to those obtained with the SOV-COMPLETE model are surprising. Indeed, SOV-COMPLETE is a less approached model than SOV. The differences could be explained by the fact that an approximation of the SOV model offsets the approximation common to all models.

Regarding the echodynamic curves "XY" if the  inclusions, a good agreement is obtained for large apertures. For small apertures (1 and 4 elements), the 3 models overestimate the amplitude as soon as one moves away from the probe center (-6 dB width overestimated of 7.5mm maximum). These differences may be associated with the approximation associated with the mode of vibration of the probe in piston mode (the hypothesis is less true when the active aperture is very small).

Finally, as for the amplitude-distance curves, a better agreement is observed on echodynamic curves along the X axis with the SOV model for small apertures. These differences are not explained.


Configuration with delay laws

A good agreement is obtained between experiment and simulation. The position of the maximum of the specular echo and its amplitude are predicted with the 3 models at 2dB. A better prediction was obtained with the SOV-COMPLETE and SPECULAR models and after the focal area despite the prediction of a small oscillation of amplitude that does not appear experimentally.

This shows the contribution of the SOV-COMPLETE and SPECULAR models on predictions of the amplitude of the echo of inclusion according to the water path. Indeed, outside the focak area, amplitudes simulated with the SOV model are systematically underestimated. Maximum observed differences between experiment and model predictions can reach 5 to 6 dB. The discrepancies between experiment and the predictions of SOV-COMPLETE and SPECULAR models are less than 2dB (experimental uncertainty).

Concerning "XY"  echodynamic curves from inclusions, the 3 models predict well the -6 dB beam width for all the studied apertures. When observing the decay in amplitude beyond this area, the comparison of the experimental results show a clear improvement in predictions of SOV-COMPLETE and SPECULAR models. The differences are not bigger than 2dB while the SOV model (with plane wave approximation) underestimates the amplitude of 12dB at the same positions. Note also that this underestimation is stronger when the aperure is small.


Back to Results of the configurations with delay laws