Identification of fibre parameters for spalling protection of concrete in fire

Since the first report of the high risk of spalling in high strength concrete at elevated temperatures in the early 1990s, there have been numerous studies of the mitigation of spalling in concrete. Polypropylene fibres have been found to be effective for spalling protection, but there are conflicting views in the literature about the effect of fibre geometry on this protection, so that the extrapolation of the data beyond this limit often results in confusion. Thus, a systematic investigation of the effect of fibre characteristics on spalling protection of concrete is required. Further, since spalling of concrete occurs due to a vapour-induced stress and thermal stresses, and the addition of polypropylene fibres can only mitigate the level of spalling against the vapour-induced stress, other mitigation methods against the thermally-induced stresses should be considered for full protection of concrete from spalling. However, the data reporting this issue in the literature are currently limited, and thus further research into developing the method for mitigation of spalling against the thermally-induced stresses is also required. In the present research, the first aim was to investigate the effect of fibre characteristics on spalling protection of concrete in fire. Various types of fibres (polypropylene, polyvinyl alcohol, cellulose and nylon) with different lengths, diameters and melting points were used for this investigation. It was found that the total number of fibres per unit volume (N) is the most important fibre parameter for determining the effectiveness of fibres for spalling protection. Other important fibre parameters include the length of fibres (Lf) and the melting point of fibres (Tm). These parameters are important because the main role of fibres in concrete is not only to create void spaces on melting, but more importantly to provide connections between pores forming a porous network that provides pathways for water vapour to evacuate. Of the three important fibre parameters, N affects the level of distribution of fibres, Lf affects the level of connection to pores further apart, and Tm affects the time for vapour discharge, in the concrete at elevated temperature. Hence, it is suggested that the calculation for fibre content for fire resistant concrete should be a function of N, Lf and Tm, rather than % by volume or kg per volume as commonly expressed. This is because the diameter of fibres was found to have little or no effect on the spalling protection of concrete, and the conventional measure for fibre content (% by volume or kg per volume) is a function of the fibre diameter. The proposed expression for fibre content provides a clear explanation for the conflicting views of fibre geometry in the literature. Furthermore, the design of fibres using this expression with other pertinent parameters (inter-aggregate spacing, mean size of coarse aggregate used, critical fibre number required for percolating concrete) found in the present study is useful to constantly predict the level of spalling protection and is possible to optimise fibre content for the required level of spalling protection. By optimising these three fibre parameters, it was found that the fibre content can be reduced up to four times to achieve the same level of spalling protection, compared to the fibre content designed by a conventional way. As it is unlikely that there is only one type of fibre satisfying the optimum level of fibre parameters for a given fibre content, i.e. high N, optimum Lf and low Tm, the present study investigates the effect of more than one type of fibre on spalling protection of concrete. Combinations of polypropylene, polyvinyl alcohol, cellulose and nylon fibres with various lengths were used for the investigation. It was found that for a given fibre content, a combination of two selected fibres with optimum lengths can achieve the optimum level of fibre parameters (N, Lf and Tm) for effective spalling protection. At this optimum level, the fibre content required is less than that of a single type of fibre for the same level of spalling protection. In another part of this research, the application of lateral confinement using metal meshes in concrete is investigated to develop a mitigation method that can resist spalling against thermally-induced stresses (thermal gradient and thermal incompatibility). It was found that by resisting the thermally-induced stresses, this metal mesh confinement improves the residual properties of the concrete at a structural level after fire exposure. The feasibility of metal mesh confinement along with fibre addition was examined by applying the technique in practice, and it was found that this technique works in practice. Since current fire codes worldwide have no specification for the design of fire resistant concrete for spalling protection, in the last part of this research, the effect of factors including water to binder ratio, air content, moisture content, aggregate size, aggregate type and mineral admixture type on the spalling of concrete at a material level is investigated, and then a design method for fire resistant concrete is proposed. In the proposed design, a detailed procedure is developed of how to design the optimum fibre content and of how to design the dimensions of the mesh confinement to achieve the desired residual strength of concrete after fire exposure.