Sample size influence on the damage caused by salt crystallization in limestone
Résumé
Size effect is a known problem of mechanics affecting the extrapolation of laboratory results into conclusions regarding full structures. In the context of salt weathering, a gap exists between the understanding of the crystallization of a single crystal in a controlled environment and the damage of porous material due to a zone of coalescence of such process. The recent advances in quantitative x-ray micro-tomography analysis has been a step forward in the comprehension of precipitation damage in porous media. However, given the scale of the process, a search for ever higher resolution motivates the use of smaller samples. While the use of samples measuring few milimetres or even micrometers is useful for the visualization of different phases of the damage cycle, it might not be representative of the process taking place in full size pieces.
We present data on the influence of sample dimension on the process of salt damage. Cylindrical samples of Lavoux limestone were prepared in four sizes (4x8 mm, 8x16 mm, 25x50 mm, 50x50 mm) and subjected to wet-and-dry cycles until the onset of damage. All samples were sealed in all faces except the upper side. During each cycle, a 3,4 weight % sodium chloride solution was dispensed on the free face in an amount equivalent to 60% of the volume of the respective porous space. The samples were kept for one hour at laboratory conditions (T=22ºC; RH=60%) allowing for the solution to penetrate the sample. Subsequently, a 23 hours drying step was carried out at 40 ºC and 40% relative humidity. At the end of the weathering, quantitative x-ray micro-tomographies were compared with prior scans of the samples in their unweathered state to assess the deterioration in all sizes of samples. The assessment of the relation between sample size and sample damage should highlight the limitations of the use of very small samples.