MateriaLs' and Structures / Matdriaux et Constructions, Vol. 36, August-September 2003, pp 448-452 Drying of porous building materials: hydraulic diffusivity and front propagation D. A. Lockington l, J.-Y. Parlange 2, D. A. Barry 3 and C. A. Leech 1 (1) Environmental Engineering, The University of Queensland, Brisbane 4072, Australia (2) Department of Biological and Environmental Engineering, Cornell University, Ithaca NY 14853, USA (3) School of Engineering and Electronics, University of Edinburgh, Edinburgh EH9 3JN, UK ABSTRACT One-dimensional drying of a porous building material is modelled as a nonlinear diffusion process. The most difficult case of strong surface drying when an internal drying front is created is treated in particular. Simple analytical formulae for the drying front and moisture profiles during second stage drying are obtained when the hydraulic diffusivity is known. The analysis demonstrates the origin of the constant drying front speed observed elsewhere experimentally. Application of the formulae is illustrated for an exponential diffusivity and applied to the &ying of a fired clay brick. RESUME Le s&'hage d'un matOrkm poreux est d&~ritpar l'~;quation de dif~sion non lin~aire. Pour un co~fllcient de diffusion donnd, des' Jbrmules analytiques simples sont obtenues pour les pFvfils l~vdriques et pore" le J?o,t de s~;chage. Le cas, difficile ?l traiter, oh la surface du mat~riau est ~ventuellement sOche, est comid~rO en dOmil. L'analyse montre l'origine de la vitesse constante du fivnt de sOchage, qui a ~;t~ observOe dam des' Otudes r tnd@endantes. L 'application des" f&wules au s&hage dune brique d'argile est illustr&" pour un coefficient de diffusion qui dOpend eaT)onentiellement du contenu hydrique. 1. INTRODUCTION In a series of experiments using different building materials, Pel [1] and Pel et eL [2] observed the internal moisture profiles generated during drying. Magnetic Resonance Imaging was used to obtain the profiles. The experimental set-up simulated one-dimensional moisture migration to a drying surface as might occur in a wall or slab (see Fig. 1). Initially, the experimental cores were unifbnnly saturated. The observed moisture profiles were then used to determine the unsaturated hydraulic diffusivity using Matano's method [2, 3]. Typical diftiasivities are shown in Fig. 2. Note that the diffusivity declines rapidly with declining water content at higher degrees of saturation (as expected) before increasing sharply at very low saturations where vapour transport dominates. Of particular significance is their observation of a well-defined 'drying fi'ont' in the bricks and gypsum studied that receded into the material as drying progressed. The front marks tile transition from vapour-dominated transport to capillary flow (Fig. 3) and its observed speed of propagation was constant (see Fig. 4). In this paper we show fiom fundamemal principles the origin of this drying front behaviour and derive a simple analytical relationship between hydraulic properties, the drying front and the dynamics of the internal moisture profile. The results can also be applied to the simpler case when surface drying is not strong enough to produce a vapour zone near the surfhce. If the building material is initially quite wet, its surface will remain wet at least for a while, and this we term stage 1 drying. We limit ourselves to the drying process after the surl~ace becomes dry, hereafter referred to as stage 2 drying. If the material dries slowly, i.e., water transport within it is slow Fig. l - Schematic of drying problem. 1359-5997/03 9 RILEM 448