Damage of masonry structures relative to their properties: Development of ground movement fragility curves

International audienceThe objective of this paper is to study the effect of the properties of masonry buildings on the threshold values of the deflection ratio, which is used to assess the damage category of buildings. Based on the results obtained from a sensitivity analysis, fragility curves are drawn and used to develop a probabilistic approach to help assess the damage for different structure typologies due to ground settlement.The threshold values of the building deflection ratio are primarily empirical. The analytical method of Burland et al. (1977) was used to define analytical threshold values to assess the building damage in relation to the building properties: length L, height H, equivalent Young’s modulus E and equivalent shear modulus G. This paper starts with a comparison of various empirical and analytical threshold values of the deflection ratio. Next, a sensitivity analysis is performed using the analytical method of Burland et al. (1977) and modified by Saeidi et al. (2012) to take into account the anisotropic behaviour associated to the masonry mechanical properties and the existence of openings. The investigated parameters are the E/G and L/H ratios and the position of the neutral axis of the beam simulating the building. The results show a significant influence of the E/G and L/H ratios and suggest defining a range of the threshold deflection ratio associated to different damage categories for different building typologies. Fragility curves are then developed based on the definition of the building typologies. These fragility curves are a useful tool to estimate the probability of reaching each category of damage once the deflection ratio and the properties of the relative structure are determined. To improve their accuracy, several intervals of the properties of the buildings were used to develop these fragility curves. The choice of these intervals to draw the fragility curves was based on various bibliographic studies and numerical studies in the case of E/G and statistic studies in the case of L/H