3D ultrasonic imaging and contour detection in sheet metal hydroforming

The paper describes an ultrasonic imaging system with a low number of transducers in a sparse 2D array for contour detection of specularly reflecting sheet metals in a hydroforming process. Image reconstruction is performed with a SAFT (synthetic aperture focussing technique) algorithm as a first step. Within the SAFT image, the contour of the metal sheet can be detected in a second step. Experiments and results are presented for a linear 1D aperture with 2D and a 2D aperture with 3D imaging capabilities, respectively. Contour detection results in contours with an axial reconstruction error of less than 1 mm for complex shaped 2D objects. Simple 3D objects have been imaged as well. Besides this, a modified approach (angle weighted SAFT, AW-SAFT) is presented to work around the lack of SNR under special conditions caused by specular reflections. Gaps in the contours can be "filled up" and poorly imaged parts of the contours can be amplified with this approach, whereas noise is reduced.