Torsional behavior of reinforced recycled aggregate flowing concrete hollow section beams

Fifteen reinforced concrete beams made with recycled aggregates and flowing concrete, which were tested under pure torsion to study the torsional behavior. The major parameters in this study were the percent of the recycled aggregates and the number of transverse reinforcements. The torsional response and crack behavior were investigated in this study. From the results, it can be noted that the increase of the recycled aggregate led to a decrease in the critical and ultimate torque of the beam, also, the increasing the number of transverse reinforcements led to the enhancement of the critical and ultimate torque. Numerical analysis by finite element method was conducted and gave a good indicator for agreement in the results of critical and ultimate load between the numerical and experimental study, as well as the angle of twisting of the beam. The second approach was Hsu ̓s softened truss model, the model proves it can predict the critical and ultimate torque for the beam and showed its ability to describe the behavior of the beams before and after cracks.


Introduction
The building industry generates a large amount of rubble that may need to be recycled and utilized as recycled aggregates (RAs) to replace natural aggregates in part or whole. Recycling helps to reduce waste and energy consumption, resulting in a more sustainable building sector. Where natural aggregate contains smoothtextured, angular, and elongated particles, recycled aggregate has rough-textured, angular, and elongated particles. The compressive strength, flexural strength, and stiffness of hardened concrete formed using recycled aggregate are slightly lower than those of concrete made with conventional aggregates [1][2][3][4][5][6][7][8]. Flowing concrete was employed due to the lack of workability caused by the use of recycled aggregates. In recent decades, the use of flowing concrete has increased dramatically. The concrete must have an adequate level of passing ability, filling ability, and stability in order to fill the entire space and flow through the formwork without any external effort [9,10]. Flowing concrete is defined by ACI committee 116 [11] as concrete with a slump larger than 7.5 in. (190mm) while remaining cohesive. Concrete that flows or has a good workability Concrete with low cement content and high cement replacement materials like fly ash (FA) and ground granulated blast-furnace slag (GGBS) would be a cost-effective solution to make long-lasting concrete. [10] [Nan Su]. However, flowing concrete is susceptible to segregation, and strong flow combined with high form pressure and low form leakage increases sensitivity to plastic shrinkage and early cracking [Peterson] [12]. The structural behavior of beams cast using self-compacting concrete including recycled concrete aggregate as coarse aggregate at various replacement ratios is investigated in this study. This research includes an experimental component that entails the casting and testing of fifteen beams. The percentage of recycled aggregate and the elongation reinforcement ratio are the two most important variables. The study's second goal was to see if Hsu's softened truss model [13] was valid in the scenario of recycled aggregate being reincorporated into the concrete mix. In addition, this research developed the "Softened Membrane Model for Torsion (SMMT)" theory [14] to predict the torque-twist curve, including the pre-cracking, post-cracking, and post-cracking responses for a concrete beam under torsion with flowing concrete containing recycled concrete as coarse aggregate, by adding a factor that causes the tensile strength and strain gradient of the c to change. Finally, using the ANSYS program 18.2, experimental beams were finite element evaluated.

Test program
The experimental program included the testing of fifteen beams under pure torsion. Table 1 shows the details of the beams that were tested. All of the beams have the same 300x300mm2 cross-section and are 1300 mm long. The longitudinal reinforcement was the same for all beams, Ø12mm at each corner. The transverse reinforcement and the percentage of recycled aggregates replacement were the primary variable. Fig. 1 shows the details of tested beams.  The first number of beam distinguishes being the amount of steel reinforcement and the second number representing the proportion of recycled aggregates. The reinforcing steel's average yield strength and modulus of elasticity were 420 N/mm2 and 1.98x105 N/mm2 correspondingly.
The recycled aggregates used in this investigation came from the demolition of concrete cubes that had been delivered to the lab for testing (see Fig. 2) Its grading complied with B.S. 882-1983 [23], and the aggregates' maximum size was 20 mm. To obtain the specified level of workability, the same amount of additive was employed in all mixes. It was made of cement, sand, and other materials. At 28 days, the average cubic compressive strength was above 32 N/mm2, with water-cement ratios of 1.0, 1.1, 2.1, and 0.42, respectively. Two types of concrete mixes were created: one with natural gravel as coarse aggregate and the other with recycled concrete aggregate with varying coarse aggregate replacement ratios. Table 2 shows the computed characteristics of concrete mixtures.  The Universal Testing Machine Model (855 IMFI System) was used to test all of the beams, which has a maximum load capability of 300 tons. The beams had a clear span of 1.3 meters, and the specimens were placed in a test machine with roller support at each end. To measure torsional rotations, a set of linear variable differential transformers (LVDTs) was installed under four arms that were projected normal to the beam axis. Three electrical resistance strain gages, each 10 mm long, were mounted at a 50 mm interval on one of the longitudinal bars' corners to record longitudinal strain.
Three electrical resistance strain gauges were installed on the beam surface in the middle with a 45-degree angle to the beam axis to measure the diagonal concrete strain. A hand-held microscope with an accuracy of 0.001 mm was used to record the data and measure their widths. Fig. 3 depicts the test setup. Table 3 shows the experimental results of the tested beams.  Table 3 shows that the beams with 100 percent recycled aggregates, group E, with various percents of transverse reinforcement (0, 10 @ 60, and 120 mm) showed increases in critical and ultimate cracks as the percent of transverse reinforcement increased; this had the same effect for all the groups (A, B, C, and D), but with lesser percent increases. The effect of recycled aggregate may be seen in beams T0R0, T0R20, T0R40, T0R60, and T0R80 that do not have a transverse reinforcement and contain recycled aggregates of 0, 20, 40, 60, and 80 percent. It can be shown that the critical and ultimate torques decrease as the recycled aggregate content increases, which is also true for other beams with transverse reinforcement of 10 @60 and 120mm, as shown in Figure 4. The curve for specimens without transverse stirrup was more regressive than the other curves, indicating that the influence of recycled aggregate increased with beams without transverse reinforcement.

Finite element analysis
The model of the tested beams was represented in this study using the ANSYS 18.2 tool. Concrete and reinforcement steel were represented using the 8-node SOLID65 brake element and the 2-node LINK180 discrete element, respectively. The qualities of the materials were the same as those used in the experimental program. Fig. 6 explains the finite element model and the approach to the applied load, as well as the boundary conditions. Loads were applied at the ends of the beam in opposing directions as torsional loads.    The ultimate and twist angles for tested beams have been determined to evaluate the validity of Hsu's softened truss model. Table 5 summarizes the findings of the tested and theoretically derived beams. The projected ultimate torque and angle of twisting at ultimate load were found to be in good agreement. Fig. 8 illustrates the experimental torque twist for the T6R80 beam, as well as two ANSYS and Hsu's softened truss modelpredicted curves. The truss model's capacity to predict the torque twist curve's initial stage before crack and post crack stages can be seen. Furthermore, the finite element method demonstrated its capacity to predict the stages preceding and following cracks.

Conclusion
The experimental program includes the testing of fifteen beams under pure tension to investigate the behavior of reinforced concrete with recycled aggregates for flowing concrete and various transverse reinforcing options. The results show that as the percentage of recycled aggregates in the concrete mix is increased, the critical and ultimate torque decreases. The cracking and torsional strength of beams with normal aggregates and maximum transverse steel reinforcement will be the highest.
The analytical prediction of the ultimate and angle of twisting by finite element and by Hsu's softened truss model is the second part of the investigation, and it was discovered that the results by finite element and experimental corresponding to it are in good agreement. The capacity of Hsu's softened truss model to forecast the ultimate torque and angle of twisting of the beam, as well as show the behavior of the beam at the first and second stages after cracking the beam, was also discovered.