Material properties related to fire spalling of concrete

Despite almost 100 years of research, the fundamentals of fire spalling of concrete are not yet fully understood. Many different theories exist, each trying to describe the coupling between different phenomenon leading to fire spalling. Some of these phenomenon have been investigated and are presented in this thesis. The thermal properties at high temperature have been investigated using the Transient Plane Source (TPS) technique. Measurements of temperature in the cross-section of selfcompacting concrete with the addition of polypropylene fibres (PP-fibres) during fire exposure are presented. A verification calculation, based on the measured values of temperature in the cross-section of the fire exposed concrete, was compared with results from a full-scale fire test with the same concrete. Good correlation was obtained. Using a measurement system consisting of oil filled steel pipes, the pressure inside the concrete during fire exposure, was measured. Results from pressure measurement tests on self compacting concrete show relatively low internal pressure before spalling. In fact, the pressure was highest in a specimen that did not spall during fire exposure. Conclusions from the tests were that pore pressure only played a secondary role in the spalling process of the concrete investigated. Thermal stresses were believed to be the primary reason for spalling. With a small addition of polypropylene fibres (PP) in concrete the fire spalling can be limited or totally avoided. Comparative studies of concrete with and without PP- fibres were performed, and the function of PP-fibres was illustrated in small and large scale fire tests. When determining the free thermal expansion as a function of temperature a strain plateau was found in specimens containing PP-fibres. The plateau corresponds fairly well with the temperature where accelerated drying behaviour was shown. This is an indication that an accelerated drying creep phenomenon might be present. Further experimental work is needed to verify this postulate.