Spalling of concrete: Implications for structural performance in fire

This preliminary paper is a progress report on an analytical investigation into the implications of explosive spalling on the fire performance of reinforced concrete structural elements and whole structures. This study does not attempt to predict whether spalling will occur. For accurate prediction of the occurrence of spalling a complete and fully coupled hygro-thermal-mechanical (HTM) analysis is required, as described by a comprehensive review of current research into the parameters and mechanisms that influence spalling, including a review of physical spalling criteria. This paper describes the structural performance of spalled concrete elements, using finite element analysis where spalling is modelled by removing layers of concrete when a set of spalling criteria are met. The method is presented using a case study of a simply supported reinforced concrete beam, where the analytical results indicate that spalling invariably triggers an early failure (well short of the required FRR rating) of a beam exposed to the standard fire. • The moisture content of the member is less than 3% or the member is designed for internal exposure • The tabulated data is used to prescribe generic fire ratings for concrete elements (except for axis distances > 70 mm) However, if the designer expects the moisture content to be greater than 3% for beams, slabs and tensile members the effect of spalling on the load bearing function of the element is checked by assuming local loss of cover to one reinforcing bar or bundle of bars and calculating the reduced load bearing capacity. This check is not deemed necessary where the number of bars is high enough, it is assumed that an acceptable level of redistribution of stress is possible without loss of stability. Examples given of where the number of bars is high enough to allow redistribution of stress are: solid slabs with evenly distributed bars and beams of widths greater than 400 mm and 8 bars in the tensile region. Concrete structures are generally designed for fire using the tabulated data thus it is apparent that there are very few occasions when a designer must consider the possible effects of spalling on structural stability. If checks must be made they are then limited to only considering localised spalling. The tabulated data is compiled from the results of standard fire resistance tests of isolated elements, generally without any spalling having occurred. Therefore continued reliance on this empirical data to account for the possible debilitating effects of spalling masks the mechanisms of how an element or a whole structure truly performs in the event of spalling. Potential alternative load paths or stability mechanisms such as compressive membrane action are ignored. 3 MECHANISMS OF SPALLING Spalling of concrete is generally categorised as pore pressure induced spalling, thermal stress induced spalling or a combination of the two. 3.1 Pore pressure induced spalling As concrete is heated the free water vaporises at 100 o C and expands; thereby resulting in increased pore pressures. Migration of some this vapour to the interior of the concrete member, where it cools and condenses, will result in an increasingly ‘wet’ zone (sometimes referred to as moisture clog). At some distance from the hot surface the vapour front reaches a critical point at which a maximum pore pressure is achieved (further movement will result in a reduction in pressure). The distance of this point from the heated surface will depend on the concrete’s permeability. Pore pressure spalling occurs if the maximum pore pressure is greater than the local tensile strength of the concrete. However, no pore pressures have yet been measured which would exceed the tensile strength of concrete which suggests that pore pressure in isolation does not lead to the occurrence of spalling (Khoury & Anderberg 2000, Jansson & Bostrom 2008). 3.2 Thermal stress induced spalling Strong thermal gradients develop in concrete as it is heated, due to its low thermal conductivity and high specific heat. These thermal gradients induce compressive stresses close to the surface due to restrained thermal expansion and tensile stresses in the cooler interior regions. The surface compression may also be augmented by applied loading or prestress. 3.3 Combined pore pressure and thermal stress induced spalling It is most likely that spalling occurs due to the combination of tensile stresses induced by thermal expansion and increased pore pressure. Much debate still surrounds the identification of the key mechanism (pore pressure or thermal stress) (Khoury & Anderberg 2000). However, it is noted that the key mechanism may change depending upon the section size, material and moisture content (Davie et al. 2008).