Explorative nonlinear pushover analyses for historical unreinforced masonry churches

Abstract

In the past decades, gas productions in the province of Groningen in The Netherlands have caused a significant amount of shallow human-induced earthquakes. Its building stock has shown to be highly vulnerable to these unexpected events. Among various building typologies, damage has been reported for approximately 12.9% of the churches in Groningen. From the perspective of conservation and prevention of loss of our historical and cultural heritage, the assessment of churches is of importance. As a majority of historical structures have undergone restoration works, that can be favourable for their seismic performance, it is worth noticing that for historical structures strengthening may not be necessary. Consequently, the question arises whether structural modifications during the lifespan of historical unreinforced masonry churches, positively influences their seismic performance. This thesis researches the seismic performance of the historical unreinforced masonry Dutch church Zandeweer, prior and post structural modifications, in the earthquake prone area Groningen. The church, built by the Moncks of Abdij van Aduard in the year 1230, underwent major renovation works in 1931. Its structure is composed of masonry walls, masonry ribbed cross vaults and a timber roof with timber cladding. During the restoration works in 1931 steel columns were integrated in the piers of one of the façade in the longitudinal direction, window openings in the curved façade were closed and concrete beams were added on top of the main walls. Its seismic performance was investigating in the positive longitudinal direction by studying its global nonlinear response and structural elemental behaviour and by indicating uncertainties and damage. A Nonlinear Pushover Analysis (NLPO) and the Simplified Lateral Mechanism Analysis method (SLaMA) were employed to capture the influence on the seismic performance of the church models. Although, the acquired results for the church models showed to a certain extend an insight in the influence of the structural modifications, an insight in the ductility behaviour of the structure was not acquired. Additionally, from the eigenvalue it was concluded that adopted single-mode analysis methods (SLaMA and NLPO) were not suitable, as the structural response of the church is governed by multi-modes. Moreover, with the highly unpredictable behaviour of the vaults in the structure the presented results were greatly influence by physical and numerical instabilities. For future studies is recommended to dedicate a detailed study to the numerical modelling and analysis of masonry ribbed cross vault structures. Furthermore, to conduct a detailed study on the church walls, for which subsequently various analysis methods can be employed and the influence of orthotropy of walls on the global response can be studied. Lastly, the analysis method can be changed by using methods that consider multi-modes simultaneously and that are known for less numerical instability problems than static analysis methods.