Influential Factors on the Hygrothermal Behavior of FireExposed Concrete

The mathematical model describing the hygrothermal behavior of fire-loaded concrete is inherently intricate, given the interacting physical and chemical processes within heated concrete. Striking a delicate balance between theoretical fidelity and the easiness of numerical implementation has prompted the development of various multi-physics models with differing complexities for hightemperature concrete. Nevertheless, the impact of various model assumptions on simulation accuracy remains unclear, hindering consensus in determining the most appropriate level of simplification for a high-temperature concrete model. Moreover, these multi-physics models typically require numerous input parameters, some of which are challenging to measure, impeding their practical application. This paper aims to evaluate the effect of several model assumptions and input parameters on the mass and heat transport characteristics of heated concrete. A fully coupled thermo-hydro model is first developed to investigate the distribution of hygrothermal-related variables within concrete under fire exposure, such as temperature, vapor pressure, and saturation degree. Subsequently, comparative analyses are conducted to examine variations in hygrothermal fields among different models, with and without considerations of dehydration, the Klinkenberg effect of gas flow, mass transport of absorbed water, and fluid gravity, successively. Furthermore, the significance of temperature- and moisture content-dependent material properties, including thermodynamic (e.g., thermal conductivity), hydraulic (e.g., intrinsic permeability, relative permeability of concrete to gas/water, diffusion coefficient), and physical parameters (e.g., porosity, density of the solid skeleton, density of vapor, moisture content), to the hygrothermal behavior is investigated. The effects of the fire loading curve and spalling model are also assessed. It is found that the mass transfer and heat conduction processes are strongly related to the dehydration model, initial moisture content, fire curves, and the involved spalling process, so careful determination of these input parameters is crucial. The absorbed water flux and fluid gravity have little impact on results, which can serve as evidence of reducing the model complexity ultimately for better technical applications.