Experimental thermo-hydric study to simulate natural weathering conditions

Abstract

Listen

Translate article

• Development of an experimental setup realistically reproducing thermo-hydric strain. • Estimation and localization of the generated thermo-hydric stress. • Comparison of stones through the assessment of long-term stone weathering. Laboratory aging tests allow to indirectly study stones behaviors on a monument. Classic aging tests are designed to compare materials resistance to different weathering mechanisms: they are more intense than the on-site conditions and do not reflect the real weathering intensity and kinetics. To study the real weathering on a monument and the materials behavior, an experimental setup using realistic experimental conditions was designed. The stones of a monument in Reims (France) were studied through this experimental setup: the Courville limestone (original stone) and the Savonnières limestone (restoration stone). The environmental variations were taken into account through a previous climatic monitoring of the monument. The realistic experimental setup was designed to reproduce, in a stone sample, the thermal and hydric strains experienced by a stone on a vertical wall. Temperature and humidity variations were applied unidirectionally through one surface, defined as the exposed surface. The temperature and humidity variations were based on typical variations identified with an in-situ monitoring of the basilica: sunny day, cold and warm rainy days, and frost day. The mild conditions, similar to the on-site conditions, did not visibly damage the samples and their behavior was studied through temperature and strain variations. The Courville limestone appeared more sensitive to the thermo-hydric variations of the typical days, particularly by the presence of water in the porous network than the Savonnières. Water generated high stress in the wet layer due to hydric expansion and at the wet/dry interface due to strain difference. Furthermore, the ice formation in the wet layer could induce high stress. On the contrary, the Savonnières limestone did not appear sensitive and little to no stress was generated during the typical days. The realistic setup allowed a simulation of the temperature and water repartitions in the samples close to on-site conditions. It provided relevant information in this specific environment and could help anticipate stone weathering on the monument.