Visualizing Air Voids and Synthetic Fibers from X-Ray Computed Tomographic Images of Concrete

Abstract

A challenge in quality control for synthetic fiber-reinforced concrete is determining the actual spatial distribution of fibers. This paper presents the first computer algorithm to identify synthetic macrofibers within hardened concrete that has been scanned in an industrial X-ray computed tomographic scanner. The algorithm can also be used to obtain the spatial distribution of other inclusions such as air voids or steel fibers as well. Visualization of synthetic fibers was the primary focus of this work. The heterogeneous nature of concrete results in a noisy image which makes identifying contrast edge segmentation difficult for to the image processing. In order to identify only fibers, the air voids touching the fibers must be identified separately because they are similar in grayscale as the synthetic fibers. These air voids are assumed to be spherical in shape, and once identified can be extracted from the remaining fiber-aggregate-cement system. In this study, it was determined that the algorithm works best for straight macrosynthetic fibers where the pixel resolution is similar or smaller than the diameter of the fibers and if the fibers remain straight lines in the 3D matrix.