Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Identifizierung von Fehlereinflüssen auf die SAFT-Rekonstruktion händisch aufgenommener Ultraschallmessdaten

: Kodera, Sayako
: Krieg, Fabian; Kirchhof, Jan

Ilmenau, 2019, 61 pp.
Ilmenau, TU, Bachelor Thesis, 2019
Bachelor Thesis
Fraunhofer IZFP ()
SAFT; Ultraschall

Material inspection is the core tool to ensure the absence of critical flaws in fabrication and enable maintenance of large infrastructure. One of its methods is ultrasonic testing (UT). As the image quality of raw UT data is generally insufficient to visualize the inner structure of objects, its image quality can be improved by post-processing. Techniques used for post-processing are developed to process automatically acquired data, although UT measurements are mostly performed manually. Due to the lack of positional information, post-processing of manual data has been challenging. However, by using the measurement system capable of recording scan positions, post-processing of manual data can be performed, enabling to enhance the visual feedback to the operator in assistance systems. One of the prevalent post-processing methods is the Synthetic Aperture Focusing Technique (SAFT). So far, its application potentials to manually acquired data are little studied. Since manual measurements differ from its automatic counterpart, such as irregular sampling or inaccurate position recognition, its SAFT reconstruction is significantly degraded, when such differences are not considered. In order to determine the effect of the error sources on the SAFT image quality, this thesis compares automatic and manual measurement setups and derives five factors: coarser sampling density, unequal distribution of the scan positions, lower coverage of the inspection area, positional inaccuracy and varying contact pressure. As to the former two factors, there are already relevant studies from which we can assess their impact. Moreover, the former three are related to scan path decisions of the operator, which we cannot influence. Hence, this thesis focuses on the latter three factors by conducting two simulation studies. The first investigates the impact of the positional inaccuracy and the spatial coverage, while the second examines the effect of varying contact pressure. For the evaluation, we compare the obtained results with a SAFT reconstruction of a simulated automatically acquired data set serving as reference. By showing that the effect of those factors are tolerable or even negligible under suitable conditions, both studies demonstrate the feasibility of applying SAFT to manual measurement data. In the future, the obtained results can be used as indicators for developing measurement assistance systems.