Seismic assessment of a detached masonry building using non-linear analyses

Masonry is one of the oldest and popular building materials used world over. The constituent materials are readily available and hence are cheap in most areas. So it serves as an economically viable building material. Though, its load bearing capacity under compression is high, it does not perform well under tensile forces, cyclic and lateral loads. In the Netherlands, specifically around the Groningen area, where natural gas is being extracted from the 1960s, masonry structures are currently prone to artificial seismic excitation due to excessive gas extraction and the resulting soil liquefaction. Such residential buildings having stood there for about a 100 years are being affected by these seismic vibrations and have developed cracks which can prove detrimental to both life and property if unchecked. To this end, Nederlandse Aardolie Matschappij (NAM) has spearheaded a research group along with TUDelft, ARUP and others to get a thorough insight into the strength and durability aspects of these existing buildings. Since there are different configurations and types of structures, it has been divided into several typologies. Further research, using shake table tests, has also been taken up at the TUDelft Stevin lab using these different typologies and/or scaled models. This project deals with the study of type T3a detached type villa, which is asymmetric in plan. This structure consists of clay brick masonry walls and timber floors and beams. Effectively, the current thesis project is a blind prediction on the strength and behaviour aspects of this particular structure. For the same, the structure is modelled and analysed using Finite Element application DIANA 9.6. Firstly, modelling of the structure is done according to macro-modelling technique with smeared cracking. Shell elements are used for walls and floors, and beam elements for timber beams. It is noted here that all the non-linearities have been focused only on the masonry part. Linear static checks and Eigenvalue checks are carried out. The time-history record was provided by the supervisor. Variational studies were performed on the structure to check their influence on the seismic resistance. These variations are - model without roof and with the roof part, the latter modelled under fair assumptions. Different type of beam end connections, seismic excitation along different directions of the structure were also included. Finally, as a check, monotonic mass-proportional pushover analyses were conducted. The results show that the given input signal at 0.16g scaling is not able to inflict serious damage, hence higher scaling is used. It is observed that the seismic resistance of the structure is higher along the X-direction of the structure than in Y-direction. Influence of beam ends on the seismic resistance is not significant. However, coupling between floors has been observed, by comparing their displacements at multiple time instants. The presence of roof part has a significant influence on the (box) behaviour of the structure, due to increase in overburden load. The pushover curves validate the time-history analyses by forming an envelope over the hysteresis curves from the latter. All the beams are provided along the X-direction of the structure, which provides significant stiffness to out-of-plane walls (X-direction excitation). It is also interesting to find that the same wall (IP1) undergoes maximum cracking for analyses along either X or Y directions. The pushover curves are in agreement with the hysteresis curves from time-history analyses which proves the validity of the results