Zhang, Jie; University of Bristol; United Kingdom
Zhang, J.; University of Bristol; United Kingdom
Velichko, A.; University of Bristol; United Kingdom
Wilcox, P.; University of Bristol; United Kingdom
Session: Applied PAUT 1
Time: 13:50 - 14:10
Ultrasonic arrays have been increasingly used in the non-destructive evaluation (NDE) field in recent years. Alongside this development, the Full Matrix Capture (FMC) technique and its related post-processing algorithms can provide improved array image quality and improved defect characterisation. This benefits from good understanding of array imaging algorithms and acoustic wave interactions between defects and their surrounding materials. There is therefore an increasing demand for efficient simulation tools to generate FMC array data sets to help to optimise array configurations for defect inspection and characterisation. Such simulation tools are achieved using either numerical method (for example, finite element or boundary element methods), semi-analytical or analytical solutions. Generally, the computation cost of using numerical methods is much higher than those of using the other two methods. Particularly, for simulating FMC array data sets in an immersion inspection, in a finite element model, the element size should be much smaller than the wavelength in water and this leads to a large model and high computation cost. One desirable solution for such simulations is using a hybrid wave scattering model which combines far-field scattering behaviour from a defect (termed the far-field scattering matrix) with a fast ray-tracing model. This has been successfully used to simulate array data sets when the far-field assumption is satisfied (i.e. the wave travelling distance in solid greater than the square of the defect size to wavelength ratio). However, the model breaks down when the far-field assumption is not satisfied. In this paper, the hybrid wave scattering model is extended to simulate FMC array data sets for large defects in an array immersion inspection configuration. For the defect, the extended model still only requires its far-field scattering matrix, but the information in the scattering matrix is used in a different way. Validations are performed through the comparison of the results from the developed model with finite element simulations and experimental measurements.