Modelling of subsidence induced damage to masonry buildings

Abstract

Land subsidence is a well-known phenomenon that describes the progressive sinking of the ground surface relative to the sea level. In the Netherlands, the western and northern regions have experienced subsidence for centuries, due to the peat- and clay- rich subsurface, and a combination of natural processes and human activities. The evident outcomes of land subsidence include damage to structures and infrastructure, social unease and economic loss. In this context, individuals and public organizations are increasingly concerned with how subsidence impacts the built heritage.
The thesis focuses on two research areas: examining how soil heterogeneity affects settlements at the scale of structures, and developing statistical tools to estimate the probability of damage to unreinforced masonry structures affected by settlements. This includes integrating information from literature, measurements of existing structures, numerical modelling, and engineering judgement to achieve a better understanding of how subsidence processes impact structures.
For the soil heterogeneity aspect, an area in the Netherlands with available in-situ measurements was selected for analysis. The in-situ information served as input for numerical analyses designed to assess how variations in the thickness of soil layers might trigger or exacerbate spatially variable subsidence. The results of the analyses show that the spatial variability of soil layer thickness correlates with the spatial variability of the computed settlements from the numerical simulations. This suggests that the variability of the thickness of soil layers can cause differential settlement at the scale of structures, potentially leading to damage.
Regarding the probability of damage to masonry structures, the research incorporates both empirical and numerical data. First, damage surveys conducted on existing masonry structures in the Netherlands were collected. The collected information includes technical reports, photographs of the damage, and measurements of the buildings’ displacements caused by ground settlements. From these data, recurrent wall deformations were identified, including symmetric and asymmetric hogging and sagging settlement profiles. Analyses were carried out to retrieve probabilistic relationships between the intensity of the settlements and the probability of damage to structures.
The empirical insight was complemented by numerical analyses. Nonlinear finite element models were built to simulate the response of masonry structures to settlements. Initially, these simulations evaluated how various geotechnical and structural factors influence the vulnerability of buildings to settlements. The results confirm that building damage is significantly influenced by façade geometry in terms of length over height (L/H) ratio, masonry material properties and the shape of the settlement, and soil-structure interaction.
Additional numerical analyses were conducted to establish the probabilistic relationship between ground settlement intensity and structural damage probability, following a similar approach to the analyses based on empirical data. Overall, the developed fragility curves indicate that for a value of the angular distortion measured on the building equal to 1/500, the threshold recommended by international standards, one out of two buildings could exhibit cracks up to 5 millimeters.