The importance of capillary pressure and adsorbed water in the behavior of heat and moisture transport in concrete exposed to high temperatures is explored by incorporating their behavior explicitly into a computational model. The inclusion of these two phenomena is realized with a formulation of a modified model, which represents an extension to the significant work of Tenchev et al. Comparative studies were carried out, using a benchmark problem, and it was determined that while the Tenchev formulation underestimated the capacity for fluid transport in the concrete, resulting in an overprediction of pore pressures (which may affect the prediction of mechanical damage and spalling), the inclusion of capillary pressure had little effect on the results. More important was the accurate representation of the free water flux, which has a significant effect on the prediction of vapor content and subsequently pore pressure. It was furthermore found that, while the adsorbed water flux may be minimal when concrete is exposed to high temperature, its presence has a significant effect on the fluid transport behavior and the prediction of pore pressures.

This paper presents a new method for testing the behaviour of soils placed under tensile load and demonstrates its suitability for testing a number of soil types under various conditions including saturation, compaction and stabilisation. Validation of the results obtained for the soils at relatively low saturation has been conducted using the established Brazilian (indirect) test for measuring the tensile strength of brittle materials. A fair comparison has been found and the results highlight the limited applicability of the Brazilian method to soils at very low water contents at which the tensile failure criterion has been assumed using this methodology. Optical characterisation of the performance of both testing methods has also been conducted using Digital Image Correlation. The consistent, accurate measurement of directly induced tensile strains using the proposed new method has been confirmed, verifying its capability to apply a direct tensile stress in the absence of shearing, a problem commonly associated with other tensile testing methods. The developed technique has then been used to investigate the water content–tensile strength relationship for compacted, unsaturated soils and offers significant advantages in the characterisation of clay soils subjected to variable climatic loading