Abstract

Owing to a complex solidification process, the microstructure of austenitic stainless steel welds consists of long columnar grains with varying preferential orientation. Ultrasonic inspection of such welds has always been a challenge with grain scattering and beam deviation originating from the preferential alignment as the main obstacles. Whilst the former may be to some extent circumvented by reducing the frequency of inspection, the latter requires some a priori information on the structure of the grains (essentially, the orientation of the stiffness tensor) within the weld. Grain stiffness map may be obtained from either a forward weld formation model, or measurements. The most accurate direct method - electron backscatter diffraction (EBSD) measurement - is a lengthy and costly process. At the same time, the level of detail available from an EBSD map is unnecessarily high from the viewpoint of ultrasound propagation. The time of flight of the wave, which is the most common feature of interest, e.g. in imaging, is affected by local variations of crystal orientation only in an average sense. Therefore, sufficient information on the distribution of the orientation sensor should be available from ultrasound measurement. Previous work confirms that such an inversion, based on ultrasonic measurements is possible. It is favourable to use a simple weld formation model, such as MINA or a geometrical description, to reduce the number of parameters to be identified during the inversion. In this contribution, we investigate how well a typical weld can be described by small-parameter models and the effect of the choice of the model on the inversion process.

How to Cite:

Kalkowski, M. K. & Lowe, M. J., (2019) “On the suitability of different wave description models for ultrasonic characterisation of an austenitic stainless steel weld”, Review of Progress in Quantitative Nondestructive Evaluation .