Gas and water permeability of cracked concrete

The goals of this research are to study the relationship between cracking and permeability of concrete and to compare the experimental results against theoretical models. Three concrete mixes are used; one of ordinary concrete, and two of high performance concrete, with similar composition save that one is reinforced with steel fibre. Two discs, 50 mm thick slices, are cut from the mid section of 110 x 220 mm cylindrical specimens and are dried prior to being subjected to permeability tests. The discs are diametrically loaded, as for a normal splitting test, until obtaining a predetermined lateral displacement, normal to the loading direction, also called the crack opening displacement (COD). A special removable device, equipped with two transducers, monitors the COD on both sides of the discs. When the predetermined COD level is reached, the discs are unloaded, and a gas permeability test is performed. After the last loading phase, gas and water permeability tests are both performed in order to compare the influence of the percolating fluid. Intrinsic gas permeability is measured, using a constant head permeameter. To take into account of the slip flow in undamaged specimens and the inertial flow inside the cracks, a specific procedure is applied. Experimental data show a coherent transition between Klinkenberg's and Dupuit-Forcheimer's laws. In the most cracked specimens, the water permeability and the gas permeability are equal if only the first percolating water flow produced in time is considered. Results suggest that permeability increases as COD cubed, as shown in theoretical models for cracked rocks. The correction factor, accounting for the tortuosity of the cracks induced in each concrete is compared and also discussed