Retrofit Solutions Wellington

Finite Element Time-History Analysis of URM Buildings

Principal Researcher: Associate Professor Nawawi Chouw
Students involved: Claudio Oyarzo Vera

Nonlinear analysis of URM structures

In the last 10 years the use of dynamic methods in structural analysis has increased significantly. Especially interesting is the development of time-history analysis techniques considering the nonlinear behaviour of the system. Some years ago, these methods were only applied in academic studies, but today they are widely employed by many consulting engineering firms for the seismic analysis, design and assessment of structures.


For that reason, many of the latest generation seismic codes around the world dedicate sections to this topic. Usually they include criteria for the selection and the scaling of ground motion records for analysis, and some considerations are suggested for modelling and analysis. However, there are several topics that currently are not fully addressed, and more comprehensive explanations or further recommendations are required.


That is especially critical for structures such as URM buildings, because it is necessary to consider other special variables related to the nonlinear material behaviour, the modelling techniques and the structural interaction between elements.


Masonry is usually described as a composite material formed by units and joint, with different directional properties due to the planes of weakness at the mortar joints. The very low tensile strength of the material renders the use of non-linear constitutive behaviour as the more obvious choice.


There are different paths to follow when URM structures are modelled. The model can consider explicitly the interaction between the masonry units and the mortar joints, the mechanical properties of both material separately and their degradation mechanism (micro modelling). Or the model can consider the average properties of the composite material, by the application of homogenization techniques. The composite material should represent the mechanical properties and the degradation mechanism of both materials at the same time (macro modelling). There are also different kinds of model based on finite elements, discrete elements, stability limit analysis, and others.


Typically, URM struts are compounded by different structural element and material. Masonry walls, steel or timber columns, timber beams and diaphragm, are common elements in URM heritage buildings, and they work all together a single system during, for example, an earthquake. So it is important and necessary to analyze the structure as a whole system and it is not enough to analyze each element as an isolated unit.

Record selection methods for Time-History analysis

Currently, the responsibility of record selection for time-history analysis belongs mainly to seismologists. However, overseas experience has shown that seismologists tend to be prudent when asked to select records and assume that all features (magnitude, source-site distance, fault mechanism, etc.) matter to structural response. Communication with several local seismologists and structural engineers has confirmed that the situation in New Zealand is similar to the experience reported elsewhere. Therefore, considering this situation, the question is:

Is it possible to integrate the seismologist's and engineer's specialist knowledge, and recommend a suite of records to be used in time-history analysis when assessing the seismic response of existing buildings in New Zealand?

For this porpoise, a selection criterion for New Zealand is proposed considering the seismological characteristic of the country and the requirements defined in the New Zealand Standard. Furthermore, suites of records for different zones in New Zealand are proposed. The results of this research are presented in the following articles:

Structural damage assessment of URM buildings using finite element models

Difficulties in conceiving and implementing models for the analysis of masonry structures arise especially due to the intrinsic complexity of formulating anisotropic inelastic behaviour. Only a reduced number of researchers have tried to develop specific models for masonry structures considering different criteria for tension and compression, and the anisotropic characteristics of the material.

The state of the art shows that the finite element model seems to be the most adequate tool for continuum models in which structural elements are represented in detail and local failure can be clearly captured. Discrete element methods are an adequate formulation for large displacements, including contact update, and an independent mesh for each block, in case of deformable blocks; but the main disadvantages are the need of a large number of contact points required for accurate representation of interface stresses and a rather time consuming analysis, especially for 3D [1].


In this field some techniques of finite elements model are studied, specially focused of the identification, assessment and localization of damage due dynamic loading as a tool to assist the later retrofitting of the structure.


The relation between damage and the degradation of dynamic properties of the structure (natural frequencies, modal shape, damping) is explored, and also the instantly and cumulative effect over URM buildings of actual time-history earthquake records with different characteristic.

References:

1. Lourenco, P.B. (2008). Structural masonry analysis: Recent developments and prospects. 14th International Brick and Block Masonry Conference (14IBMAC). 17-20 February 2008. Sydney, Australia.