Reinforced concrete columns exposed to standard fire: Comparison among different constitutive models for concrete at high temperature

• The behaviour of columns under a standard fire can be adequately well described by beam finite elements.
• There are no major differences between any of the explicit and implicit models.
• The EC2 model allows the best agreement between experimental and numerical results.

Concrete behaviour at high temperature was investigated in depth since the 1970s, in order to highlight the main issues linked to its mechanical performance in hot conditions, such as chemical processes, kinematic behaviour (transient and creep strains) and evolution of the physico-mechanical properties. Thanks to these studies, a few constitutive models have been proposed in the literature for concrete at high temperature, with the aim of modelling reinforced concrete structural behaviour during heating. Within this context, a Beam Finite Element code for thermo-mechanical analyses has been developed by using a Fortran solver and GID as pre- and post-processor. A number of well-documented full-scale tests on reinforced concrete columns exposed to Standard Fire (without cooling) was simulated numerically, by implementing four different constitutive models proposed in the literature for concrete at high temperature. The main goals are: to highlight the role of some critical aspects regarding reinforced concrete members in hot conditions, in particular second-order effects, transient and creep strains (a), and to make a systematic comparison between numerical and experimental results in order to assess the reliability of both 1D numerical modelling (b) and the adopted constitutive models for concrete (c). The results confirm that 1D numerical modelling is generally consistent with the experimental evidence if transient and creep strains, as well as second-order effects are carefully taken into account. Moreover, the differences among the four investigated models for concrete behaviour in compression are quite limited.