[MIS19-06] Reproducing typical strains of a monument through aging tests to anticipate stone weathering.
Keywords:Cultural heritage, Limestone, Aging test, Thermal strain
Recurring temperature and humidity variations were identified on the Saint-Remi Basilica in Rheims (France), inscribed on UNESCO’s World Heritage List, with the two years long in situ monitoring. Typical events were categorized and modeled: a sunny day (24h cycle of temperature between 20 and 35°C), a rainy day (24h cycle of temperature between 20 and 35°C with a high humidity), a frost day (24h cycle of temperature between 10 and -5°C with a high humidity), and short events (20-minute cycle of temperature between 20 and 25°C).
The experimental setup used was designed to reproduce as close to reality as possible these thermal and hydric strains on stone samples. The largest side of the sample (10x5x4,5cm) was placed against a vertical heating plate surrounded by a thermal insulator, simulating a stone on a vertical façade exposed to the environment. The temperature of the plate was controlled with predefined temperature cycles, and rain events were simulated by removing the sample from the plate and placing it in 5mm water for 1h. Five cycles of each strain type (sunny day, rainy day, frost day, short events) were applied to the samples. Thermocouples and strain gauges measured heat transfer and differential micro-dilatation, allowing to monitor the sample’s behavior through the experiments.
Two typical stones of the Saint-Remi Basilica were studied: the Courville limestone, considered the original stone, and the Savonnières limestone, considered the restoration stone. They presented different properties and reacted differently to the thermal and thermo-hydric strains. The Courville limestone showed a more intense deformation than the Savonnières limestone under the same temperature conditions. The thermal strain during sunny day cycles and even the low-intensity temperature variations of the short events caused a significant response from the Courville limestone. As expected, the closer to the heating plate the gauges were, the higher the deformation was. On the other hand, the Savonnières limestone showed little to no deformation. With the introduction of the humidity factor, rain event simulated by imbibition, water transfer properties of the stones came into play. A supercooling phenomenon could be observed during the frost cycles, showing that freezing took place. For the Savonnières limestone, the water penetrated deep into the sample; the stone presented a constant and homogeneous deformation behavior during rainy and frost cycles. For the Courville limestone, the water stayed in a layer near the surface, inducing a dilatation difference between the water and the dry layers, causing a strain at the interface.
The number of cycles did not allow to observe damage of the stones, but it allowed to see the clear behavior difference between the two limestones. This difference can also be seen on the monument as the Savonnières limestone is still in a good state, while the Courville has undergone weathering (disaggregation, spalling/scaling). The deformation behaviors of the limestones during these typical events will be correlated to the frequency of those events in order to estimate the stones’ behavior throughout one year. The strains induced by environmental variations presented here have a relatively low intensity compared to the mechanical properties of the stones but can be damageable in the long term through fatigue effect.