Modelling of RC beam-column sub-assemblages under sudden column loss scenario
conference contributionposted on 09.04.2020, 09:35 authored by J. Sio, B. A. Izzuddin
The design of structures against disproportionate collapse is commonly achieved through structural assessment under a sudden column loss scenario. In this regard, a novel framework has been previously developed at Imperial College London, aiming to improve the accuracy of robustness assessment whilst maintaining a certain level of practicality. To predict the dynamic behaviour under sudden column loss, the framework transforms the nonlinear static response of a structure into a pseudo-static response through principles of energy balance. Furthermore, the response of an entire structure can be obtained based on the assemblage of its individual components. As such, the structural assessment under sudden column loss can be performed, for example, by simply providing the nonlinear static response of a beam-column sub-assemblage (BCSA) extracted from the structure. Complement to the above framework, this paper aims to propose efficient numerical models to predict the nonlinear static response of RC BCSAs under column loss scenarios. The BCSAs are modelled in the nonlinear finite element analysis program ADAPTIC using one-dimensional fibre-beam elements and a combination of joint/link elements. To reproduce the fracture of reinforcement, a uniaxial steel constitutive model is developed, in which a tensile softening branch follows an exponential degrading function. Two approaches are employed to assess the relevance of considering bond-slip. The first approach, which uses conventional fibre-beam elements, is more simplistic, with the inherent assumptions of linear cross-sectional strain distribution and full-bond between reinforcement and concrete. The second approach is more sophisticated since it relies on two separate elements for the modelling of concrete and reinforcement/bond, respectively. Lastly, parametric studies are performed to assess the sensitivity of the predicted nonlinear static response to various model parameters, including the tensile softening rate of reinforcement and the bond strength.
- Mathematical Sciences