Abstract: In a method of flaw classification on and in welds of planar or arcuate plate-shaped workpieces of uniform thickness by means of ultrasonics, the expected flaw areas of the workpiece are pre-examined manually and are then examined systematically using the ultrasound echo image method in separate cross-sectional planes. A B-scan is thus generated, and changing the position of the transducer on the workpiece and/or the transmitting direction enables defects to be located by calculating the delay time of ultrasound waves in the go-and-return directions within the workpiece under consideration of possible reflections at the rear side (8) of the workpiece. For defining the test conditions, part-specific parameters are considered, such as weld type and weld shape (e.g., double-V-weld and bell seam), weld geometry (e.g. seam angle), and part characteristics (e.g.

Abstract: Welded joints of austenitic steels are tested by ultrasonic sound caused to be incident on the joint at an oblique angle thereto. Information regarding the size and nature of reflectors in the joint can be derived from the amplitude of the echoes generated at the reflectors in response to the sound beam. To ensure a definite relationship between the amplitude of the echoes and the nature and size of the reflector, the frequency of the sound use for testing is below and upper frequency limit at which the wavelength of sound in the base material equals the largest peak-to-valley distance of the base metal weld interface in any of the consecutive areas of the interface on which the sound beam is incident at any time.