Investigation of Cracking Behavior in Reinforced Concrete Panels with Bond-slip Reinforcement
MetadataShow full item record
This thesis aims to investigate the crack pattern of two reinforced concrete panels subjected to a tensile load. Numerical models for two reinforced concrete panels were established in the finite element program DIANA10. Results from the nonlinear analyses were compared with results from experiments performed by Dyngeland . An important aspect of this thesis is the implementation of bond-slip reinforcement instead of regular embedded reinforcement. Bond-slip reinforcement models are valid when it is assumed that there no longer exist a perfect bond between the concrete and the reinforcement bars. This results in slip, or relative displacement, between the concrete and the reinforcement. This slip causes interface tractions along the reinforcement bars. Two bond-slip models were assessed, one from fib Model Code 2010 and one model proposed by Dörr, which is called Cubic Bond-slip. It was seen that the two different slip-models described the bond stress-slip relation quite differently, which evidently had an impact on the crack patterns. It was also desirable to establish a basis for which material properties and iterative procedures that gave the best results. Important aspects related to discussion and evaluation are the experimental results, crack widths and crack spacing from the nonlinear analyses and theoretical crack widths and crack spacing calculated according to Eurocode 2 and fib Model Code 2010. It was discovered that the cubic bond-slip model by Dörr resulted in an earlier crack initiation stage for both panels. This is caused by the initial stiffness of the bond stress-slip curve, which is higher for the cubic bond-slip model than the fib Model Code 2010 bond-slip model. The results showed that the shear stiffness modulus for the bond-slip models had a large impact on the behaviour of the two panels.Of the tested numerical iteration methods, the Newton Raphson methods would result in divergence for both panels. It was concluded that a more sophisticated iterative procedure had to be used for the analyses of the two panels. The Broyden Quasi-Newton method gave stable results and all over load step convergence.