Impact of air entrapment on capillary absorption in porous building materials

Abstract

When the water content of a porous material is high, air entrapped in the pore space is expected to affect water transfer through the pores. To understand the effects of air entrapment on water transfer in porous building materials in the high-water-saturation region, we examined the water transfer characteristics corresponding to significantly small air entrapment effects. First, we conducted two sets of water uptake experiments. In the first experiment, using three building materials, the time evolution of the amount of water absorption was measured at a low air pressure near vacuum (several kPa). In the second experiment, the water content profile during water uptake was measured using the gamma-ray attenuation method. The experimental results showed that low air pressure accelerated the water uptake by the brick and aerated concrete specimens, whereas water uptake by the calcium silicate board specimens was not significantly affected. These differences among materials were analyzed from a pore structure viewpoint. Moreover, gamma-ray attenuation measurements confirmed that the obtained water content profile was qualitatively similar at atmospheric and low air pressures, although the low air pressure increased both the water content of the material after capillary absorption and the wetting front propagation rate. Finally, simultaneous water and air transfer calculations based on the air and liquid water balance in a material reproduced the measured water absorption rates well, confirming that air entrapment and pressure development in the pores can significantly reduce the rate of water uptake and water content after capillary absorption. The calculation results also indicated that the air pressure in a material did not significantly increase at early water uptake stages where local water content was not high, which supported the general assumption that treating the liquid-water transfer in porous building materials as a one-component flow is valid in most cases.