Experimental studies on continuous reinforced concrete slabs under single and multi-compartment fires with cooling phase

Highlights

• Fire tests were conducted on five continuous slabs under single and multi-compartment fires.

• The failure modes of these continuous RC slabs in fire were output and compared.

• The influence of number and position of heated compartments was investigated.

• A number of general test data was provided to validate the numerical and theoretical models.

Abstract

This paper experimentally investigates the effect of number and position of heated compartments on the fire behaviour of continuous reinforced concrete slabs. A total of five continuous three-span slabs are tested under single compartment fire, two-compartment fires and three-compartment fires. The measurements involve furnace temperatures, slabs’ temperature, displacements, restraint forces, cracks and failure modes. The results show that the cracking pattern, displacement and reaction force of continuous slabs greatly depend on the furnace temperature, the position and number of heated compartments. Transverse cracks and short cracks appear on the top and bottom surfaces of the heated compartments, respectively, and the number of cracks increases with increasing heated compartments. The heated middle compartment or the unheated edge compartments may experience upward or downward deflections. Explosive concrete spalling is observed in the tests, which affects the failure mode of slabs and water evaporation from the unexposed surface. The selection of the worst fire scenario depends on the furnace temperature, the layout of heated compartments and concrete spalling. The conventional failure criteria (temperature or deflection failure criteria) can result in conservative or unconservative predictions of fire resistance of continuous slabs, which should be improved to include the favourable effect of structural continuity.

Introduction

Recently, research on the structural performance of conventional reinforced concrete (RC) slabs in fire has received significant attentions. A review of references indicates that there have been a lot of experimental and numerical studies on the fire performance of isolated RC slabs [[1], [2], [3], [4], [5], [6], [7], [8]] and floors in buildings [[9], [10], [11], [12], [13]]. In addition, to understand the fire behaviour of RC floor slabs, several fire tests were conducted on the continuous slab panels (one middle panel, one corner panel, four panels and six panels) within a steel-framed building [[11], [12], [13]]. The experimental results indicated that deformation of the heated panels greatly depended on the boundary constraints around the heat panel within the floor. Due to the structural continuity and membrane action, the heated panels had good fire performance even under long-duration fire conditions or at large deflections. In other words, the RC slabs in steel framed buildings performed better in fire than the individual isolated members, and thus the prescriptive design approaches are conservative. Note that, although the slabs are usually restrained by beams in the floor system, the experimental tests were carried out to analyse the concrete continuous slabs, free from interactions with the surrounding structure.

In fact, a significant amount of research has also been conducted on the performance of continuous RC slabs. Both [14] conducted a total of 21 fire tests to investigate the effect of different factors (such as reinforcement ratio and load levels) on the thermal and mechanical behaviour of continuous two-span composite concrete slabs, and theoretical and numerical analyses were conducted based on the experimental results. The results indicated that increasing positive reinforcement and decreased degree of loading, increasing concrete depth often increased the fire resistance of the continuous slabs. Smith et al. [[15], [16], [17]] investigated the effect of boundary restraint conditions on the deflection and failure mode of the concrete slabs in fire. Note that the slabs were heated from the top surface, and the slab displaced continually away from the heat source (in the direction of loading) during the cooling stage. Guo and Bailey [18] conducted the fire tests on the composite slabs and found that the behaviour of composite slabs was dependent on the heating rate, maximum temperature and cooling rate.

No doubt, a two-span one-way continuous slab was simply supported at two ends which result in maximum flexural deflection representing the worst case scenario. However, these experimental results cannot be extended to the three-span prestressed concrete or conventional concrete slabs in fire. Therefore, Gao et al. [[19], [20], [21]] investigated the behaviour of one-way three-span simply-supported unbonded prestressed concrete continuous slabs (span-thickness ratio of 35) exposed to fires in the end span, middle span and end-middle spans. The results showed that the failure mode of the slabs was influenced significantly by the length of negative moment reinforcement, but was less affected by the level of prestress. Meanwhile, cracks along the short span appeared near the supports of the heated span due to larger negative moment. Hou et al. [22] investigated the effect of concrete cover thickness, load level, effective prestress and degree of prestressing on the fire behaviour of nine one-way two-span unbonded prestressed concrete (span-thickness ratio of 32) continuous slabs. The results indicated that the thickness of concrete cover, load level and fire induced secondary moments had significant influences on the fire resistance of unbonded PC continuous slabs. In addition, the degree of prestressing had a significant influence on the extent of concrete spalling of PC continuous slabs. Yuan et al. [23] conducted fire tests on four one-way three-span unbonded prestressed continuous RC slabs (span-thickness ratio of 35). They found that the sequence of spans under fire and the length of negative reinforcement had important effects on the deformation and failure mode. According to the above results, it was found that the concrete spalling and breaking of the tension reinforcements easily occurred within the prestressed concrete slabs. In addition, for two-span prestressed continuous slabs, the downward deflection in each span often increased during the fire. However, for three-span prestressed slabs, the deflection trend (downward or upward) and failure mode (position of damage hinge) of the middle span was mainly dependent on the sequence and position of the heated spans. Chen [24] conducted six fire tests on one-way three-span continuous RC slabs (span-thickness ratio of 35) where different spans (one edge span, one middle span and edge-middle span) were subjected to fire. For one heated span, the position and number of its failure hinge was dependent on the length of negative moment reinforcement and its position. Different from that of the prestressed continuous slabs, less longitudinal cracks appeared on the top surface of these slabs.

Based on the test results (one-way continuous slab), the following results can be concluded: (1) Due to the moment redistribution and restraint action, the two-span or three-span continuous slabs have better fire resistant performance than those of isolated one-way slabs. (2) For conventional or prestressed concrete continuous slabs, the length of the negative reinforcement has great influence on the position and number of the destructive hinge in each span. (3) For the edge or middle spans in one conventional or prestressed continuous slab, their deflections gradually increase with increasing temperature, but their deflection trends (downward or upward) may be different from each other. (4) Compared to those of the conventional concrete slabs, the concrete spalling and the fracture of the prestressed reinforcement easily occur in the prestressed concrete slabs.

Although the above tests have proposed these conclusions, several important problems also exist and must be addressed to achieve fuller and more confident insights into fire behaviour of the continuous slabs. On one hand, one problem is the effect of membrane action arising from two-way spanning, such as tensile or compressive membrane action, and the other problem is the effect of the lower span-thickness ratio. On the other hand, for the 3-span two-way continuous slabs, the effect of different fire scenarios on the deflection and failure mode of the edge and middle spans should be further investigated. This is due to the fact that the fire behaviour of 3-span one-way continuous slabs cannot be easily extended to analyse those of 3-span continuous two-way slabs, particularly the middle span. Meanwhile, the fire type (traveling fire or uniform fire) and number of floors in fire had a significant effect on the failure time and mode of collapse. One single worst-case fire scenario cannot be readily identified, especially considering the uncertainty in the number of fire-involved floors in a real fire, i.e. fires can occur in different parts and on different floors of a building. Even on the same floor, fires may occur in different rooms, which may locate at the edge, corner or middle [23,25]. Thus, Wang et al. [26] conducted the tests of four three-span continuous slabs (4700 mm × 2160 mm × 100 mm) under different compartment fires and investigated the effect of fire spreading direction, delay time (30 min and 60 min), total fire duration (Slab B1: 360 min, Slab B2: 400 min, Slab B3: 600 min and Slab B4: 600 min), the reinforcement ratio and arrangement on their thermal and structural responses. The results show that the fire spreading scenarios have important effect on the deflection trend of the middle compartment in the slabs. Increasing reinforcement ratio and using the continuous reinforcement can prevent or delay the failure of the continuous slabs.

In all, the above literature review shows that there is very limited experimental work addressing the behaviour of two-way continuous RC slabs in which the single or multi compartments were subjected to fire. Meanwhile, there is also limited experimental data on the behaviour of two-way continuous RC slabs during the heating and cooling stages. Hence, this paper presents the experimental results of five three-span two-way continuous slabs (4700 mm × 2160 mm × 80 mm) subjected to single-compartment, two-compartment and three-compartment fires, and the fire duration in each test was 180 min. Note that, apart from various configurations (slab's thickness, cover, bar diameter and concrete strength), the main difference between the present slabs and those in Ref. [26] was the fire scenario. The measurements of furnace temperatures, temperature distributions across the slab thickness, vertical and horizontal displacements, reaction forces are presented as well as the observations of cracking and failure modes. Meanwhile, the present results were mainly compared with the observation in Ref. [26]. Finally, the applicability of traditional failure criteria to determine the fire resistance of continuous RC slabs is examined.