Determination of the effects of the rebar-cement bond parameters variation in UHPFRC using FA and ANN

Abstract

The experimental study included the design and production of ultra-high-performance steel fiber-reinforced concrete (UHPFRC). The physical and mechanical properties of UHPFRC were investigated in a laboratory setting. To investigate the properties of UHPFR concrete, three types of concrete and over 70 samples were used. Following that, samples were created to test the anchors' load-bearing performance. Six concrete slabs with a total of 108 pre-installed anchor samples and six concrete slabs with 108 post-installed chemical anchor samples were created. The analysis of the test findings comprised all individual results as well as the definition of the relationship between the anchor's tensile load capacity and other parameters. To accurately determine the individual influence of the investigated factors as well as their combined impact, a factorial experiment, and artificial neural networks were used in addition to normal statistical numerical studies. It was determined that both approaches offer advantages. The results obtained show matches in certain parts. Due to the way data is processed in different ways, there are also significant differences between them.