Abstract

The phenomena of water absorption in porous building materials have been conventionally represented using a parabolic partial differential equation with hydraulic diffusivity as the transport parameter. The equation in its linear and non-linear forms has been widely adopted in the modeling of hydraulic diffusivity in terms of water absorption coefficient and other non-linear functions of moisture content based on pertinent experimental data. The objective of this paper is to assess the sensitivity of moisture intrusion characteristics, namely, (i) the moisture content and, (ii) the moisture penetration depth, to different hydraulic diffusivity models. The stated characteristics have been simulated with a duly verified CrankNicolson scheme modified to reduce computational cost related to the iterative solution. The scope of the study includes the cases of two distinct materials – fired clay brick and OPC-lime-sand mortar – for which experimentally observed moisture intrusion profiles are available from independently reported studies. The simulated moisture profiles are observed to remain concave upwards initially and become linear at later stages in cases where the governing model is linear. Although this pattern closely conforms to the trigonometrical series solution for the model, it does not match the experimentally observed distribution. It is observed that the non-linear model provides better estimates of moisture intrusion characteristics. The moisture dependent diffusivity value matches the magnitude of constant diffusivity only at a near saturation condition. The study further reveals that, that the non-linear model performs better in the case of fired clay brick than for OPC-lime-sand mortar.