Confirmation of Anomalous Diffusion in Non-saturated Porous Building Materials by a New Capillary Rise Absorption Test

A new one-dimensional capillary rise absorption test by direct weighing is presented in this note. The motivation is to obtain more accurate cumulative infiltration measurements in order to test the validity of a new theoretical description of the absorption process recently introduced in [1]. While in most of the technical specifications, weighing is generally carried out manually and requires to remove the specimen from water at intervals, the new apparatus allows a continuous automatic weighing of the specimen during water absorption. The typical performance of the experimental set-up is one measure per second with an accuracy of 0.01 g. Preliminary tests were conducted on four specimens of common red fired-clay bricks. The adjacent sides of the specimens were previously sealed with paraffin to prevent water infiltration from lateral faces of the material: this guarantees that the absorption process actually remains unidirectional. Whereas the amount of absorbed water is expected to increase as t^1/2 (with t the elapsed wetting time) according to the standard unsaturated flow theory, the cumulative infiltration (I) was found to systematically deviate from the simple t^1/2 relation. It is shown that absorption in brick scales as t^α with 0.57 ≤ α ≤ 0.59 (i.e. larger than 1/2), as previously predicted in [1]. This implies that the long-time predictions of both the amount and the penetration depth of absorbed water based on the classical t^1/2 relation are generally underestimated. Because this may also apply to the many deleterious chemical agents mediated by water, the consequences of water infiltration on the durability of building materials may also be dramatically underestimated. We suggest that the ASTM technical specifications for water absorption measurements should be reexamined at the light of these new results. The experimental procedure should be improved to increase the number and accuracy of cumulative infiltration data, especially at short times, to allow more reliable estimates of the absorption properties of porous building materials.