1Dept. of Electrical and Computer Engineering, Michigan State University, East Lansing, MI, USA
2Dept. of Mechanical Engineering, Iowa State University, Ames, IA, USA
Abstract:
The development of digital twins (defined here as virtual and dynamic representations of real-world structures or systems that can change with time) can be improved with the use of non-destructive ultrasonic evaluation. Doing so gives research and developers more accurate models of structures and systems that can then be analyzed to improve efficiency, consistency, or reduce waste. Ultrasonic evaluation allows for the creation of digital twins that include surface and sub-surface data, something optical solutions cannot provide. This data becomes integral when inspecting additively manufactured structures since these are more likely to contain sub-surface defects that cannot be seen upon visual inspection. In this work, we collect ultrasonic data from multiple orientations of the additive sample and create a single unified digital twin. This improves the accuracy since single-pose ultrasonic scans can miss features that do not reflect energy. We utilize a data registration process by assembling the data from the multiple orientations and correcting for transducer position/spatial averaging, and the storing of the ultrasonic data into voxel (3-dimensional pixel) representations for graphical viewing.
Presenting Author: Zebadiah Miles Michigan State University
Ultrasonic Digital Twin of Additively Manufactured Samples
