| dc.description.abstract | Composite steel and concrete structures are an interesting solution in tall buildings, just where structural fire safety design procedures are rigorous. In the case of composite steel deck and concrete slabs, the exposure of the steel decking to the environment and its sensitivity to high temperatures presupposes rapid collapse in fire conditions, which has been resolved with the addition of reinforcements. This research evaluated the influence of the use of reinforcements on the fire performance of the composite slabs. It has been estimated that the replacement of positive (AP) by negative (AN) rebars will optimize the steel consumption of these slabs, since the steel decking already works as AP and the AN is more thermally protected. The AN also mitigate the sagging bending moments and the mechanical responsibility of the steel decking in fire, and also increasing the structural redundancy. To solve the problem, a numerical and experimental analysis of composite steel decking and concrete slabs structurally continuous in fire conditions was solved. The influence of the use of positive (AP) and / or negative (AN) rebars on the fire performance was compared. The slab without reinforcement was also analyzed. Numerically, two analysis criteria were adopted: (C1) constant temperature distribution with increased load on the slab over time, and (C2) variable temperature distribution with constant load on the slab over time. The C1 allows to determine the load capacity of the slab for a given temperature and / or time of ISO 834, while C2 its fire resistance rate (FRR). The stresses and deflections in the slabs were discussed for both C1 and C2 criteria. The numerical simulation was done in Abaqus software. The experimental research was used in the parametric calibration of the numerical models. 17 full-scale prototypes of composite slabs were built: 9 with dimensions of 886x4600 mm and 8 of 3000x4600 mm, evaluated at normal (five point bending test) and high temperatures (with horizontal furnace according the ISO 834 curve) with 540 days of age. These prototypes were tested with rigid intermediate support, forming 2 spans of 2300 mm. These tests allowed the calibration of the numerical model with 31 reading points. As a result, the C1 criterion showed thermomechanical responses, while C2 thermophysical-mechanical. It was clearly noted that the design procedure of EN 1994-1.2 and NBR 14323 is based on C1. However, C1 does not allow to identify the increase in the stress state in the slab due to thermal effects, such as tensile membrane action, which is only possible in C2. The C2 also makes it possible to identify the concrete cracks and the detachment of the steel decking in fire. C2 showed more realistic results, but it is more 7 complex and time-consuming than C1, as it involves thermo-physical-mechanical parameters and several tests to define the ultimate loading of the slab at each time of ISO 834. From the C1 and C2 perspective, the replacement of AP by AN was not interesting. The slabs with only AP showed the best results. Removing the AP compromised the ultimate moments of resistance for sagging bending, that becoming the weak point of the slab in fire. In this case, the use of AN was unnecessary, becoming underutilized. This is due to the loss of composite behavior in fire conditions, caused by the detachment of the decking before the initial 30 min. C2 also showed that the slabs with AP preserved the tensile membrane action for longer. The structural continuity was even affected by the concrete cracks produced by the thermal bowing since the first 10 min. Some adjustments are needed in the simplified method of EN 1994-1.2 and NBR 14323, which is based on C1. This motivated the proposal of new formulations, in an attempt to simplify the C2 phenomenons in the fire-resistant moment equations of the standard. A new method has also been proposed for the definition of temperatures in concrete, positive and negative rebars and steel decking. A new table for the definition of the thermal insulation was presented. | en |
