Visual Analysis of Defects in Glass Fiber Reinforced Polymers for 4DCT Interrupted In situ Tests. (4th July 2016)

• Record Type: Journal Article
• Authors: Amirkhanov, Alexander; Amirkhanov, Artem; Salaberger, Dietmar; Kastner, Johann; Gröller, M. Eduard; Heinzl, Christoph
• Is Part Of: Computer graphics forum. Volume 35:Number 3(2016)
• Page Start: 201
• Page End: 210
• Publication Date: 2016-07-04
• Subjects: Categories and Subject Descriptors (according to ACM CCS) -- Image processing and computer vision [I.4.7]: Feature Measurement—Life Cycle -- Image processing and computer vision [I.4.10]: Image Representation— -- Computer graphics [I.3.0]: General—; Computer graphics -- Periodicals; 006.605
• DOI: 10.1111/cgf.12896
• Languages: English
• ISSNs: 0167-7055

Abstract

Abstract: Material engineers use interrupted in situ tensile testing to investigate the damage mechanisms in composite materials. For each subsequent scan, the load is incrementally increased until the specimen is completely fractured. During the interrupted in situ testing of glass fiber reinforced polymers (GFRPs) defects of four types are expected to appear: matrix fracture, fiber/matrix debonding, fiber pull‐out, and fiber fracture. There is a growing demand for the detection and analysis of these defects among the material engineers. In this paper, we present a novel workflow for the detection, classification, and visual analysis of defects in GFRPs using interrupted in situ tensile tests in combination with X‐ray Computed Tomography. The workflow is based on the automatic extraction of defects and fibers. We introduce the automatic Defect Classifier assigning the most suitable type to each defect based on its geometrical features. We present a visual analysis system that integrates four visualization methods: 1) the Defect Viewer highlights defects with visually encoded type in the context of the original CT image, 2) the Defect Density Maps provide an overview of the defect distributions according to type in 2D and 3D, 3) the Final Fracture Surface estimates the material fracture's location and displays it as a 3D surface, 4) the 3D Magic Lens enables interactive exploration by combining detailed visualizations in the region of interest with overview visualizations as context. In collaboration with material engineers, we evaluate our solution and demonstrate its practical applicability.