Deflections of high-content recycled aggregate concrete beams reinforced with GFRP bars and steel fibres

Current design guidelines for reinforced concrete (RC) beams reinforced with Fibre Reinforced Polymer (FRP) bars neglect shear and shear crack-induced deflections in the calculation of total deflections. However, shear crack-induced deflections can be up to 25 % of the total deflections of FRP RC beams and thus they need to be considered in the calculations. At the same time, the use of Recycled Aggregate Concrete (RAC) reinforced with steel fibres is becoming more common in construction, but limited research has examined deflections in FRP RAC beams with steel fibres. This article proposes a new approach (based on Eurocode 2) to calculate more accurately the deflections of FRP RAC beams reinforced with steel fibres. To achieve this, nine under-reinforced normal concrete (NC) beams and nine RAC beams were tested in bending in three Series: (I) beams with steel bars as flexural and shear reinforcement, (II) beams with Glass FRP (GFRP) bars as flexural reinforcement and steel bars as shear reinforcement, and (III) beams with GFRP bars as flexural reinforcement and 1 % of steel fibres (by volume). The nine RAC beams were cast with RAC made of 100 % recycled concrete aggregates, a structural option that has received limited attention. The results show that adding steel fibres to beam Series III increased the average load at first cracking by up to 15 % and 17 % compared to Series I and II, respectively. Beam Series II and III were then modelled in Abaqus® to provide further insight into their load-deflection behaviour. It is shown that current approaches can accurately predict crack widths at the serviceability limit state of FRP RAC beams but not at higher load levels. The new approach modifies the Eurocode 2 equation using a factor βr that considers the weaker bond of GFRP bars and the additional shear crack-induced deflections. Compared to existing models, the new proposed approach calculates deflections more accurately and with the lowest average Test/Prediction ratio (T/P = 1.17) and Standard Deviation (SD = 0.11) over all beams in Series II and III.