Radial temperature distributions within solid concrete cylinders under transient thermal states

Abstract

Synopsis Radial temperature gradients develop within typical concrete test cylinders during heating and can introduce structural effects which distort the material strain response of the concrete. Analytical solution of the Fourier equation, assuming infinite specimen length and invariable material properties, indicates that the maximum radial temperature difference (ΔT) obtained during heating at constant rates is directly proportional to the rate of heating and the square of the radius, and inversely related to the thermal diffusivity. A finite difference solution for a specimen heated on all sides confirms these relationships and indicates that the temperature at the centre is unaffected by end heating for a specimen length/diameter ratio greater than 1–1·5 during heating at 1°C/min. This is confirmed experimentally for gravel and lightweight concrete specimens. The effect of variations in thermal diffusively is to cause two peaks in the ΔT versus surface temperature relationship at about 160°C and 550°C corresponding to the processes of moisture loss and dissociation of Ca(OH)2 respectively. A weighted average temperature was calculated which suggests that the most suitable location to place a thermocouple to measure the average specimen temperature is at 58% of the radius. Radial temperature dlfferentials can be minimized by reducing the rate of heating and particularly the radius of the specimen. They can also be estimated by a simple relationship which, if used in conjunction with the specific test technique reported here, can provide an evaluation of the thermal diffusivity of materials.