Punching shear of flat reinforced-concrete s labs under fire conditions

Abstract

This thesis examines punching shear response of reinforced-concrete flat slabs under fire conditions. The shear behaviour of concrete in fire is relatively poorly understood compared to its flexural response. Failures such as the Gretzenbach car park failure in Switzerland (2004) have prompted concerns over the punching shear capacity of flat slabs in fire. The shear behaviour of reinforced-concrete in fire depends on degradation of the individual material properties with temperature, their interaction, and more recently recognised, the effects of restrained thermal expansion. Through experimental testing this thesis aims to build a foundation understanding of the punching shear behaviour of flat reinforced-concrete slabs in fire conditions. A series of shear blocks, tested after exposure to elevated temperature (realistic fire temperature), were used to develop an understanding of the effects of elevated temperature on the shear transfer performance of reinforced-concrete. These tests allowed the complex interplay of shear-carrying mechanisms at ambient temperature to be extended to the case of post-elevated temperature. Fifteen slab-column punching shear specimens were tested under both applied load and extreme heating. In particular, the effects of restrained thermal expansion were experimentally investigated by altering the support conditions of the slab-column specimens. A purpose-built restraint frame allowed the boundary support conditions to be either fully restrained or unrestrained. This experimental series is the only series to have tested restrained specimens at elevated temperatures, though previous researchers have simulated the thermal restraint effects and reported the importance of restrained thermal expansion and curvature on the behaviour of punching shear. Parameters of slab thickness and reinforcement ratio were also varied to investigate their respective impacts on punching shear behaviour at elevated temperature. The thicker 100 mm reinforced slabs failed in punching shear, whereas the 50 mm and 75 mm thick slabs failed in flexure-shear mechanisms and the unreinforced slabs failed in flexure. Clear behavioural differences were observed between specimens with different support conditions. Unrestrained 100 mm thick slabs under sustained load failed soon after heating began, whereas none of the corresponding restrained specimens failed during heating. One restrained, heavily reinforced specimen failed during cooling, whilst under sustained load. This is the first recorded punching shear failure during the cooling phase of an elevated temperature test and may also be the first recorded test specimen ever to have failed during the cooling phase of an elevated temperature test. This failure highlights the unknown and potentially unsafe behaviour of structures during the cooling phase. Further structural investigation of the cooling behaviour of concrete flat slabs after exposure to fire, needs to be undertaken. Most of the specimens’ central deflection was away from the heat source (in the direction of loading) during the whole test, irrespective of support condition. The test setup was assessed to investigate the unusual slab-column deflection away from the heat source, however the complex behaviour observed during the tests cannot currently be explained. It is assumed that the degradation in concrete properties and non-linear material behaviour dominates over the thermal expansion of the slabs. Quantitative and qualitative comparisons are presented, though the quantitative data is impacted by size effect, non-repeatable heating application between tests and jack friction influences on specimens with low capacities. Eurocode 2 punching shear prescriptive elevated temperature design, extends the ambient temperature equation for elevated temperature use, by degrading the temperature-dependant parameters by factors. Support conditions are not considered, with the code specifically telling the designer not to consider in-plane thermal expansion effects, therefore consequently ignoring the premature punching shear failure that can occur. Furthermore, the ambient temperature equation is based on the regression of available experimental data at the time and does not consider the reinforcement as a shear transfer mechanism. The experimental capacities of the 100 mm thick, reinforced slabs that failed in pure punching shear mechanism were similar to the Eurocode 2 punching shear prescriptive design capacity, when directly compared. The unrestrained support condition was shown to be consistently, not conservatively predicted by Eurocode 2, whereas the restrained support condition capacities were conservatively predicted. It is comforting to know that the Eurocode 2 design predicts the restrained supported slabs conservatively, as real buildings are more likely to have supports closer to the restrained condition rather than the unrestrained support condition. A sensitivity analysis of the Eurocode 2 prescriptive design equation shows it is highly sensitive to the concrete strength degradation and not the variable, cp, which was used to make a support condition comparison in this thesis. This indicates how the Eurocode 2 equation for punching shear capacity lacks in its consideration of whole structural behaviour. The Critical Shear Crack Theory has been proposed as the background to a harmonised shear design approach, called Model Code 2010. The Critical Shear Crack Theory was safe in predicting the experimental punching shear capacities. There were large variances for the 100 mm thick slabs, however they are consistent with the original model comparison to test data. An expansion of the Critical Shear Crack Theory for elevated temperature requires further validation with experimental restrained thermal expansion tests, such as those presented in this thesis. Finally, a digital image correlation technique has been proven to be a reliable method to measure structural displacements of concrete at elevated temperatures. Digital image correlation allowed the crack locations and slab rotation angles to be visualized throughout testing. No other measurement techniques are able to provide similar versatility in fire testing such as that presented herein.