Flow distance induced variation analysis of digitally segmented steel fibers in UHPFRC
Junil Pae, Sung-Hoon Kang, Namkon Lee, Sung-Wook Kim, Juhyuk Moon
Abstract
Ultra-high performance fiber-reinforced concrete (UHPFRC) is a construction material with outstanding mechanical properties and durability. The degrees of dispersion and orientation of included steel fibers are known to have a decisive effect on UHPFRC’s material properties. In this study, UHPFRC specimens with different volumes of fibers were prepared to investigate the variation of spatial information of steel fibers in UHPFRC. Each sample was scanned individually through micro-computed tomography (micro-CT) to analyze the three-dimensional fiber statistics. To enhance the accuracy of the analysis, bundled fibers were digitally segmented through reliability-based maximum likelihood estimation (MLE) and a watershed algorithm. It was quantitatively found that the number of contact areas of bundled fibers decreased and fibers tended to be aligned more horizontally as flow distance increased. However, at the casting position and the opposite region, there was irregular distribution, which has the potential to affect the designed material properties of UHPFRC.