Development of advanced numerical models for segmental cast iron tunnel linings

Lead Research Organisation: Imperial College London
Department Name: Civil & Environmental Engineering

Abstract

The main aim of this research project is to enable the adequate reproduction in geotechnical numerical analysis of the behaviour of cast iron tunnel linings. This is an issue central to the design of most underground works in London, which are often carried out in areas where a number of tunnels built with this material are impacted by building redevelopment. During the recently completed EPSRC- and industry-funded Crossrail research project, high-quality data on the behaviour of cast iron tunnel segments, both through laboratory testing and field observations, were produced. The accrued knowledge in terms of the response of this material, which was cast according to procedures followed when the first deep tube lines were built, to solicitations similar to those verified around tunnels provided unique insight into a soil-structure interaction problem which often influences the design of new infrastructure in London. Indeed, when building in the vicinity of several deep tube lines, the designers are required to assess the safety of the tunnel lining, which is often carried out using simplified techniques due to the absence of available data. It is within this context that the proposed project will exploit the experimental data from the Crossrail project to produce new models for accurate incorporation of the behaviour of cast iron into the geotechnical analysis of tunnels and, crucially, produce new design guidelines for industrial use. It is therefore envisaged that this project will consist of two separate phases, with the first one focusing on developing new numerical techniques of increasing complexity for 2D and 3D analyses of cast iron tunnel linings. This will be followed by a detailed assessment stage using field data on the effects of the excavation of Crossrail tunnels on deep tube lines, where the crucial components of the developed models are identified and retained to formulate simpler, more accurate design procedures.

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/R512540/1 01/10/2017 31/03/2022
1966915 Studentship EP/R512540/1 30/09/2017 31/03/2021 Agustin Daniel Ruiz Lopez
 
Description The use of the underground space is essential to accommodate the increasing transport and services demand of large cities around the world for new infrastructure. As a consequence the underground space of these cities is highly congested and the potential impact of new construction on existing infrastructure is a major concern during the planning and execution of construction works. The present research focuses on the impact on existing tunnels, in particular, it aims to investigate the engineering response of historic grey cast iron (GCI) tunnel linings which are the most common typology of deep tunnels in London.
A number of methods are available to carry out the engineering assessment of existing tunnels both in terms of movements and internal forces to ensure that these remain within the prescribed limits. These methods can range from simple ones such as analytical and empirical solutions which allow quick estimates to be obtained to more sophisticated ones like numerical modelling. No matter which method is adopted, when the lining is segmental (that is when the structure is made of segments and rings connected together rather than a continuous tube) it is not clear how rigid the structure is and the rigidity is a key parameter for this type of engineering calculation. GCI tunnels are characterized by their segmental nature, with segments and rings bolted together across their respective joints, hence it is crucial to determine the influence of the joints on the rigidity of the GCI linings in order to obtain accurate solutions in the assessment of these tunnels.
This project is addressing the research with the development of a number of finite element models using the program ICFEP. The main body of research was carried out with a 3D model of a segmental GCI ring. In order to gain confidence in the modelling methodologies adopted for the ring model, the first part of the research involved a rigorous calibration and validation process of the numerical model by replicating a series of laboratory tests on a half-scale GCI ring conducted as part of a major project at Imperial College London in conjunction with the construction of Crossrail. The validation included analyses of laboratory tests at small and large distortions and the numerical results successfully achieved an excellent overall agreement with the experimental observations, providing the assurance needed to use the model and extend the investigation beyond the conditions considered in the laboratory.
Once the finite element model was validated, the research could move on to address the following objectives: 1) enhancing the use of analytical methods to provide design charts for the stiffness of the tunnel accounting for the effect of joints in various scenarios and 2) developing the formulation and implementation of a joint model that allows the use of simplified numerical models with more reasonable simulation times with respect to 3D models.
To sum up, at the current stage of the project, the following objectives have been fulfilled:
• A 3D finite element model of a segmental GCI ring was developed and validated with an extensive laboratory investigation.
• In order to provide guidelines for a better use of available analytical solutions, the 3D model was setup with appropriate boundary conditions to replicate the framework of the analytical solutions in question. Various design charts were obtained allowing the stiffness of the GCI tunnel at a range of in-situ conditions to be obtained. This design charts allow for the influence of the joints and the stiffness values to be taken into account and can readily be applied as input to the mentioned analytical solutions.
• The 3D finite element model was further utilised to investigate the rotational behaviour of the joints. A simple set-up allowed the response of the joint from the onset of rotation to the end of the capacity to be studied. The results guided the development of a new numerical joint model.
• Following the point above, a nonlinear joint model was formulated using a combined approach between analytical and empirical expressions that were calibrated with the 3D model results. The new model was implemented in the finite element code ICFEP and is currently being validated.
Future work
• The validation of the new joint model is ongoing, the ultimate goal of this process is to ensure that the results of the full 3D model can be replicated using beam elements.
• Once the model is validated, its influence in the soil-structure interaction problem will be evaluated by conducting 2D geotechnical analyses of underground construction in the vicinity of a GCI tunnel. The new approach will also be compared with others such as the modelling of the tunnel as a continuous ring or the modelling of the joints as hinges.
Exploitation Route The development of design charts providing the stiffness of the tunnel at varying scenarios actually translates into a new methodology for the assessment of GCI tunnels which can be adopted in industry/practice. Furthermore, the new tunnel joint model has been implemented into a finite element code that is used by a consultancy firm that has conducted numerous assessments on these tunnels in the past and the availability of this new joint model will be incorporated into their routine modelling procedures.
Sectors Construction,Transport

URL https://www.imperial.ac.uk/geotechnics/people/agustin-ruiz-lopez/