Finite element modelling techniques for the assessment of tunnel-induced damage to masonry buildings

Prof. Harvey Burd, Prof. Chris Martin, Wing Nam Yiu

The construction of shallow tunnels (e.g. for new infrastructure developments) inevitably causes local soil deformations. These ground movements may, in turn, cause damage to nearby buildings.

A project is underway on the development of new procedures, based on finite element analysis, to assess the risk of tunnel-induced damage to masonry buildings. Risk assessments of this sort are typically conducted in design practice by imposing an assumed ground deformation profile on a flexible deep beam model of each façade of the building (e.g. Burland 2001). Although this assessment approach provides a useful indication of the likely risk of damage, it cannot be used to investigate detailed soil/structure interaction effects. In addition, the effect of structural details such as openings (for windows and doors) cannot be treated in a straightforward way.

Finite element procedures can, in principle, be used to address detailed structural and soil/structure interaction issues. However, unless efficient modelling procedures are used, the time and cost of the analyses may become excessive.

In the current project, finite element modelling procedures are being developed that, on the one hand, are straightforward to implement and, on the other, are capable of providing useful damage assessment information.

The following potential barriers to the routine use of finite element methods for tunnel-induced assessment procedures are being addressed.

(a) What level of detail is required in the modelling?
Which features of the problem need to be represented reasonably accurately and which features are less important.

(b) Availability of finite element software
To facilitate the practical application of the research, the project involves the development of modelling approaches that may be implemented using commercially available software. Current research is being conducted using Abaqus v6.11

(c) Tunnel-induced ground movements
A new – and relatively straightforward - approach to modelling the tunnel installation process is being developed.

(d) Constitutive models for the soil and masonry?
Parametric studies are being undertaken to identify the key features that need to be included.

(e) At what level of detail do the foundations need to be modelled?

(f) Is it possible to conduct useful calculations on a desktop computer?
Our initial results suggest that it is feasible to set up useful finite element models – involving non-linear behaviour for the soil and masonry – that will take a just a few hours to run on a desktop computer. Typical results are illustrated below.

Contours of maximum principal tensile strain