TBM and Lining - Essential Interfaces

Nguyen Duc Toan Prof. Daniele Peila Dr. Harald Wagner

TBM and Lining Essential Interfaces

Student:

Nguyen Duc Toan

Dissertation submitted to the Politecnico di Torino, Consortium for the Research and Permanent Education (COREP), and D2 Consult Dr. Wagner Dr. Schulter GmbH & Co. KG in partial fulfillment of the requirements for the degree of

Master in

Tunnelling and Tunnel Boring Machines

Academic Tutor:

Company Tutor:

Prof. Daniele Peila

Dr. Harald Wagner

Turin, Italy October 2006

Abstract
Optimization of segmental lining design and construction, in close relation with proper selection and operation of the tunnel boring machine (TBM), are the two among major concerns for the owners, designers and contractors, in all tunnelling areas. The main task of this work is to deal with this subject, using both qualitative and quantitative approaches. It is challenging to achieve the attractive and effective mechanized tunnelling alternatives in saving both time and cost without a comprehensive and interdisciplinary consideration. The Parties involved should be aware of the proper approaches in adopting the mechanized tunnelling technology for a given tunnel project. Every TBM tunnel project needs to be feasible from both operational and engineering points of view, environmentally acceptable and value for money. A significant scrutiny on the critical cases of TBM excavation has been conducted to identify and rectify the obscure aspects that are often associated with TBM tunnelling, in terms of risk management and project management. Difficult or critical cases of excavation in various mechanized tunnelling techniques (with certain kinds of TBMs) are analysed in connection with face stability and ground reinforcement issues. The report identifies and describes both the technical aspects and the economic impact of the critical interaction between the TBM and the tunnel lining. The interaction between the soil and the TBM tunnelling process and a number of essential loading cases for the segmental concrete lining has been investigated to comprehend the lining behaviour, the risk of ground failure and the risk of surface subsidence. The parametric study was restrictively applied to the hydroshield tunnelling technique.

Acknowledgement
I would like to express my sincere appreciation to the following people who have helped make this master thesis materialized: Politecnico di Torino: I would like to sincerely thank my academic tutor, Prof. Daniele Peila for his nicely arranging a good placement for my internship, for his kindly keeping track of my Stage work in Linz, and for his support in writing of this thesis. Prof. Sebastiano Pelizza is always an inspiration to my striving and achievements. The two of them, as being the Director and Assistant Director of the master program, make the most contribution to its successfully realization and accomplishment. I would like to express my thankfulness to Prof. Pier Paolo Oreste and Prof. Claudio Oggeri for the technical materials, for their help in exploring the university library and their input in modelling. I would like to thank Prof. Marilena Cardu for the books on the blasting technology. I would also like to thank other professors of Politecnico di Torino for their useful lectures. International Tunnelling Association (ITA) and Sponsor Companies/Societies: I would like to gratefully acknowledge the ITA for its initiative and endeavour to activate and sponsor this unique study course. The lectures at the master course in Turin are a good source of reference for my work. I would like to convey my deep gratefulness to all my lecturers from a good deal of companies/societies/universities (as shown on the back cover of this report) and from different nationalities who have dedicated their time and efforts to come to Turin and teach us international students very high-quality lessons. Consortium for the Research and Permanent Education (Corep): Special thanks are due to Ms. Irene Miletto and Ms. Giusy Favasuli the Coreps Organizational Coordinators of the Master course in Turin. My sincere thanks are delivered to Mrs. Luisa Rosano the COREP Secretariat who always ensures the insurance coverage for my movement within and outside Italy, as well a good administration in general. Master class: I would like to thank all colleagues in the master tunnelling course in Turin, academic year 2005-2006, for maintaining a comfortable and pleasant atmosphere, and for their support in my studying in terms of discussions and material exchange, particularly, Mr. Daniele De Lazzari, Mr. Nick Chittenden, Mr. Kim Jin Ha, Mr. Bang Gyu Min, Mr. Nicola Donadoni, Mr. Marco Della Casa, Mr. Ciprian Eduard Partenie, Miss Katia Efpraxia Demirtzoglou, and Miss Lamprini Goli.

D2-Consult Team in Linz: I would like to thank Ms. Katrin Pesendorfer, Ms. Margarete Prendl and Ms. Michaela Zellner the D2-Consult secretariat, for their valuable assistance during the whole process of my internship in Linz, Austria from beginning of May to mid-July 2006. I would like to thank Mr. Ulrich Horny for allowing citations from his technical paper, and for his wholehearted and effective guidance on numerical modelling during my Stage. Thanks are due to Mr. Walter Pointner, without his explanation I could be hardly to interpret the technical drawings of the BEG railway tunnel project which are presented only in German language. I also highly appreciate his high sense of humour, which makes my stay in the Danube city worth remembering. Mr. Peter Ertl and Mr. Horst Wger, who helped me to find and explore the necessary contract documents and drawings, deal with computer problems and everyday life difficulties. Accompanying them to the BEG Project in Innsbruck - Southern Austria is a good memory of mine. Finally, heartfelt gratitude is conveyed to Mr. Andreas Beil, Dr. Harald Wagner and Dr. Alfred Schulter, Managing Directors of D2 Consult for their availability to any help I need. Their partly but valuably covering for the living cost is indispensable to my Stage period. A special point I would like to be grateful to them is that, they kindly allowed me to freely utilize all the company resources, such as a rich library, photocopy machine, scanner, printer, and limitless access to the Internet. The technical documents I collected in Linz are much helpful to my thesis finalization in Turin and will be greatly beneficial to my future career. The constant input through consultation with Dr. Wagner together with his writings is a never-ending source for my work. And I am very proud of being a "close friend of D2 Consult team" as allowed by Dr. Wagner and inspired from Mr. Martin Srb. Thanks are also due to the Brenner Eisenbahn GmbH (BEG) for the kind permission to use the company's respective information. Prof. Gunter Swoboda of the Innsbruck University (Austria) is appreciated for the nice talking at his Laboratory and for his helpful input on the analysis of the settlement induced by tunnelling, as well as other modelling aspects. Institute of Transport Science and Technology (ITST): My leaders in the Institute of Transport Science and Technology in Hanoi deserve my sincere thanks for their support in the first steps of enrolling in this master course. I am also thankful to my colleagues in the ITSTs Underground Structures Department for their consistently being kind and willing to help me. My family: I am deeply grateful to my parents, who have provided most of finance for my stay in Europe. I am greatly indebted to my wife Tran Thi Linh Chi for her support, understanding and patience. Much love and thank is due to my son Nguyen Ung Bach for his constantly missing me and passionately wanting me being back home. I would like to dedicate this thesis to my parents, my wife and my son.

Contents
Abstract Acknowledgement Table of Contents

1.

INTRODUCTION .................................................................................................................................1 1.1 1.2 1.3 BACKGROUND ................................................................................................................................1 OBJECTIVES ...................................................................................................................................3 CONTENTS OF THE THESIS ............................................................................................................4

2.

CONSTRAINTS OF A NEW RAILWAY LINE ................................................................................6 2.1 2.2 2.3 BEG COMPANY APPROACHING THE PROJECT ............................................................................6 THE PROJECT ALIGNMENT ...........................................................................................................6 TBM CONTRACT LOT H3-4 AND H-8 ...........................................................................................9

3.

ESSENTIAL INTERFACES OF EXCAVATION............................................................................12 3.1 OVERVIEW ON DIFFERENT TUNNELLING METHODS .................................................................12 3.1.1. General ...................................................................................................................................12 3.1.2. Classification of Mechanized Tunnelling Techniques ..........................................................13 3.2 GLOBAL VIEW OF TBM TUNNELLING .......................................................................................15 3.2.1 TBM Types..............................................................................................................................15 3.2.2 Operation of TBMs .................................................................................................................21 3.2.3 Conventional Tunnelling Versus TBM Tunnelling ..............................................................28 3.3 CRITICAL CASES OF TBM EXCAVATION ...................................................................................32 3.3.1. Risk Management for Tunnels...............................................................................................32 3.3.2. Critical Cases of TBM Tunnelling in Soil .............................................................................37 3.3.3. Critical Cases of TBM Tunnelling in Rock ...........................................................................41 3.3.4. TBM Tunnelling in Mixed Face Conditions .........................................................................51 3.4 GROUND REINFORCING ..............................................................................................................52 3.4.1. General ...................................................................................................................................52
3.4.1.1 3.4.1.2 3.4.1.3 3.4.1.4 Face Support ............................................................................................................................... 52 Failure Mechanism ..................................................................................................................... 52 Countermeasures to Ground Failure ........................................................................................ 60 Grouted Bodies ........................................................................................................................... 61 Subsoil Conditions ...................................................................................................................... 66 Shield Machines .......................................................................................................................... 67 Tunnel Lining and Excavation .................................................................................................. 69 Ground Improvement................................................................................................................. 70

3.4.2.

Case History: Metro of Turin .................................................................................................65

3.4.2.1 3.4.2.2 3.4.2.3 3.4.2.4

4.

INTERFACE BETWEEN TBM AND LINING ...............................................................................75 4.1 TYPES OF LININGS .......................................................................................................................75 4.1.1 General ...................................................................................................................................75 4.1.2 Reinforced Concrete Linings .................................................................................................80 4.1.3 Steel Fiber Reinforced Linings ..............................................................................................82 4.2 LINING DESIGN PROCEDURE ......................................................................................................82 4.2.1 Design Steps............................................................................................................................83 4.2.2 Loading Conditions ................................................................................................................85
4.2.2.1 4.2.2.2 4.2.2.3 4.2.2.4 4.2.2.5 4.2.2.6 Geostatical Loads ........................................................................................................................ 88 Thrust Jacking Loading ............................................................................................................. 89 Trailer Loading ........................................................................................................................... 90 Grouting Loads ........................................................................................................................... 91 Storage Loads.............................................................................................................................. 92 Erection Loads ............................................................................................................................ 93

4.2.2.7 4.2.2.8

Fire Loads.................................................................................................................................... 93 Other Loads ................................................................................................................................ 93

4.3 CONCEPT OF INTERFACE ............................................................................................................94 4.3.1 Contractual Interface .............................................................................................................95

4.3.1.1 4.3.1.2 General Aspects .......................................................................................................................... 95 Segmental Lining Optimization ................................................................................................. 97 General ...................................................................................................................................... 101 Machine Operation ................................................................................................................... 103 Guidance System....................................................................................................................... 105 Lining Ring Building ................................................................................................................ 106 Backfill Grouting ...................................................................................................................... 113 Back-up System......................................................................................................................... 115 Monitoring and Instrumentation............................................................................................. 116

4.3.2

Physical Interface .................................................................................................................101

4.3.2.1 4.3.2.2 4.3.2.3 4.3.2.4 4.3.2.5 4.3.2.6 4.3.2.7

5.

INFORMATION FOR SETTLEMENT STUDY ...........................................................................118 5.1 5.2 5.2.1 5.2.2 5.2.3 5.3 5.3.1 5.3.2 5.4 GROUND CONDITIONS ...............................................................................................................118 EXCAVATION AND SUPPORT......................................................................................................120 Shield Machine .....................................................................................................................120 Ring Configuration ..............................................................................................................121 Lining Material.....................................................................................................................123 NUMERICAL ANALYSIS TOOL ...................................................................................................124 Soil Models in Plaxis ............................................................................................................124 Hardening Soil Model ..........................................................................................................125 FLOWCHART OF CALCULATION ...............................................................................................127

6.

TUNNEL INDUCED GROUND DEFORMATION.......................................................................129 6.1 SETTLEMENT INDUCED BY TUNNELLING .................................................................................129 6.1.1 Volume Loss and Settlement ................................................................................................129 6.1.2 Settlement Calculation Approaches .....................................................................................132 6.1.3 Settlement Control Approach ...............................................................................................134 6.2 EMPIRICAL CALCULATION FOR SETTLEMENT .........................................................................135 6.2.1 Formulae ..............................................................................................................................135 6.2.2 Calculated Results ................................................................................................................144 6.3 FINITE ELEMENT MODELLING .................................................................................................149 6.3.1 Introduction ..........................................................................................................................149 6.3.2 FE Analysis by Plaxis 2D Professional ...............................................................................150
6.3.2.1 6.3.2.2 Geometry ................................................................................................................................... 151 Calculations ............................................................................................................................... 153 Geometry ................................................................................................................................... 161 Calculations ............................................................................................................................... 164

6.3.3 6.4 7.

Face Stability by Plaxis 3D Tunnel .....................................................................................160 SUMMARY ..................................................................................................................................171

6.3.3.1 6.3.3.2

CONCLUSIONS AND FUTURE WORK .......................................................................................173

List of Acronyms References Appendixes Curriculum Vitae

Chapter 1

1. Introduction

1.1

Background

Placement of the Internship The Master course in Tunnelling and Tunnel Boring Machines Edition V 2005/2006 is held by the Turin University of Technology (Politecnico di Torino) in partnership with the Consortium for the Research and Permanent Education (Corep) in Turin, Italy. The intense study period has brought rich and fruitful knowledge to all the international students, including the author. After that, the author has had a fruitful master trainee period (or internship/stage) in the Headquarter of the D2 Consult Dr. Wagner Dr. Schulter GmbH & Co. KG, located in Linz, Austria. The internship lasted more than two months. The author's host company tutor is Dr. Harald Wagner - Managing Director of the D2 Consult GmbH, and his academic tutor is Prof. Daniele Peila of the Politecnico di Torino.

The host company profile D2 Consult Dr. Wagner Dr. Schulter GmbH & Co. KG Hirschgasse 32 4020 Linz, Austria Managing Directors: Established: Harald Wagner, Ph.D., P.E. Alfred Schulter, Ph.D., P.E. 1985

Natural Duality is the founding concept of D2 Consult. D2 Consult Linz is the Head office of D2 Consult. Most projects have been being handled in Linz. Branch Offices: 1986 - Foundation of Branch Office "USA" 1996 - Foundation "D2 Consult Colombia" 1998 - Foundation "D2 Consult Prague"

2000 - Foundation "D2 Consult Berlin Activities:

Design and analysis of underground structures Tender documentation Technical assistance Construction supervision Project management Cost estimation

In the fields of Transportation, Energy, and Environment, with reference projects in all over the world. In connection with the purpose of the internship, two projects using tunnel boring machines (TBM) and reinforced concrete segments completed in Paris and Boston are introduced in the Appendix 1 and 2, respectively.

Jobsites From Linz the author also went to visit the BEG (Brenner Eisenbahn GmbH) Railway Project in Innsbruck, southern Austria. While gaining knowledge of the BEG project, under the guidance of the persons responsible for the checking of the project design, i.e. D2 Consult Linz Team, the author could have a thorough grasp of the upgrading work of the railway line on the Brenner Railway Axis. The upgrading focuses on the construction of the new high capacity line in the Lower Inn Valley in the Tyrol province of Austria. The author has also found background information about the projects history and milestones as well as the data on the BEG company. - Project Name: BEG (Brenner Eisenbahn GmbH) Railway Project - Location: Lower Inn Valley, Tyrol province, Austria (between Kufstein and Innsbruck) - TBM (Tunnel Boring Machine) Tunnel Sections: i) Contract Lot H3-4 from Mnster at Km 33.1 to Wiesing at Km 38.9, length L = 5.818 km. ii) Contract Lot H8 in Jenbach from Km 39.6 to Km 44.8, length L = 5.19 km. Two these TBM Lots are both at the beginning stages of construction.

Purposes of the Internship During the trainee period the author continued to gain greater knowledge of tunnelling technology. This enabled for a better preparation of the present thesis.

The sector of reference for the traineeship is the tunnel design, on the general subject Design and/or construction aspects of tunnel and underground works. The contents include looking over the parameters of tunnel construction; studying critical cases of TBM excavation; and studying interface between TBM and segmental concrete lining. The overall objectives are enrichment of professional experience; check and widening of the knowledge acquired during the lectures. The specific objective is project review and evaluation. The tasks are: Assistance in review and independent checking of structural tunnel design, including calculations of lining segments, temporary and permanent loadings, in coordination with actual construction and monitoring.

Investigation of technical aspects The following tasks which partly comprise the content of the present report have been accomplished during the internship period in D2 Consult Linz: Study of the BEG projects contract documents and drawings, with techniques in the field of conventional and mechanized tunnelling. This is a challenging task because all the Contract Documents are in German, and only a few ones are in English. This difficulty is partly released by the fact that, D2 Team, especially Mr. A. Beil, has allowed me to make quotations from the company's available English sources. Investigation of critical cases of TBM excavation and study on interface between TBM and tunnel lining, including: Review of details of structural lining aspects, during construction phase and in the service condition Review of ground movements and volume loss due to an advancing tunnel heading Performing some parametric studies on the tunnel lining calculation

After two months and a half from the beginning of May to mid-July 2006, an Internship report has been submitted to the Corep, Politecnico di Torino, and D2 Consult the host company. After that, the Internship report has been further developed into a full thesis as in the present form.

1.2

Objectives

This study is initiated in order to increase TBM applicability in both urban and suburban areas, as well as in other fields of underground works, by reviewing important engineering aspects of TBM tunnelling.
3

The report describes the critical cases of TBM excavation in general, and essential loading cases for the segmental concrete lining in particular. Both the technical aspects and the economic impact of the critical interaction/interface between the TBM and the tunnel lining will be analysed. From that interfaces, necessary lessons and/or reactions will be illustrated and envisaged, both from theoretical and practical point of view. This report is intended to integrate as many as possible the parameters/interdependent factors that come into play during lining design and subsequent construction of a quality structure. In order to illustrate part of that interfaces in the form of visible digits, numerical analyses for the problems of tunnel face stability and surface subsidence are carried out.

1.3

Contents of the Thesis

The thesis contains seven chapters followed by references, as described below: Chapter 2 introduces general information about the BEG company and the way approaching the Brenner axis upgrade project within Austria territory. The High-speed Railway Brenner in Austria is part of the European north-south railway axis, and the TBM Contract Lot H3-4 within the project is the subject of this study, among others. Chapter 3 describes the essential interfaces of TBM excavation, by first briefing on different tunnelling methods, then going more detailed into TBM tunnelling method, and addressing critical cases of TBM excavation. In briefing different tunnelling methods, classifications of mechanized tunnelling techniques are given. In running through the TBM tunnelling, available types of TBMs and their basic operation are discussed, together with a short comparison between conventional and TBM tunnelling. Finally, critical cases of TBM excavation are dealt with in several subtopics such as project management, risk management, and difficulties while driving tunnels in soil, rock and mixed ground. A separate part is reserved for dealing with the face stability and ground reinforcing problems in urban tunnelling, and presenting a case history of Turin Metro Line 1. Chapter 4 presents the interface between TBM and lining. Looking at certain types of tunnel lining will lead to the concept of contractual and structural interface. The investigation on these two interlinked interfaces also leads to a discussion on the possibility of segmental lining optimization. Next, critical

loading cases, among others, for the lining of tunnels driven by TBMs, are given. Chapter 5 contains the input data for parametric studies that will be performed in the Chapter 6. These include a flowchart of calculation, very short introduction to the numerical tool, ground conditions, lining configuration, and the loads to be considered. Chapter 6 contains computations of the ground volume loss and surface settlements induced by tunnelling. Both empirical approach and numerical modelling are carried out to compare one another and extract necessary conclusions. The numerical modelling can also gives member forces in the tunnel lining. Chapter 7 contains several conclusions obtained from the thesis. These conclusions have shown that the presented extensive analyses on project interfaces sufficiently address the TBM processes with many influential factors. These analyses are necessary to allow for economic and reliable technical solutions and other requirements from the Client, within the scope of the mechanized tunnelling techniques. This chapter also contains recommendations for the author's future studies.

List of Acronyms References

Chapter 2

2. Constraints of a New Railway Line

2.1

BEG Company Approaching the Project

The Brenner Eisenbahn GmbH (BEG) was founded as an infrastructure construction company owned by the Republic of Austria and since 2005 it is a subsidiary of the BB Infrastruktur Bau AG, which is a company of the BB group (Austrian Federal Railway). Since 1996 the BEG has been working on the implementation on Austrian territory of the European railway upgrade project for the Brenner axis. In the past years the BEG has organized the finance, completed the environmental impact assessments and obtained the necessary approvals for the first section of the project in the Lower Inn Valley. Construction work for the new line began in the summer 2003. The companys headquarters are in Innsbruck, Austria. For the southern leg of the new rail link, the BEG has been collaborating with the Italian National Railway (RFI) on the Brenner Base Tunnel project. Since 2005 the Brenner Base Tunnel SE has taken on the planning of the Tunnel. The BEG's consent for visiting the construction site, the BEG staff's considerate guide during the site visit, and its permission for use of the project information, are highly appreciated.

2.2 The Project Alignment

History In December 1994, the European Council pinpoints 14 priority infrastructure projects for the development of a common transeuropean transport network (TEN-projects). The upgrading of the railway line Berlin - Nuremberg - Munich - Kufstein - Innsbruck - Brenner - Verona is classified project number 1. In August and October 2003, the construction of the main lot between Vomp and Terfens (near Innsbruck) was started.

Finance The rail upgrade program in the Lower Inn Valley between Kufstein and Innsbruck is a European project. As part of the Berlin - Palermo TEN axis, it occupies a key position for future developments in international north - south traffic management. On the basis of the decisions taken to develop the Trans-European Transport Networks (TEN), the European Union (EU) is co-financing the Lower Inn Valley railway project, with 50 percent of planning costs and 10 percent of construction costs funded via the relevant EU budgets. At 2003 price levels, the first upgrade section of the Lower Inn Valley railway is expected to cost about 1.85 billion euros. In addition to EU funding, the necessary finance will be provided by the Austrian government.

Implementation For the Brenner axis upgrade, the European Unions transport-policy makers have decided on a step by step approach. First priority has been given to the section between Kundl and Baumkirchen in the Lower Inn Valley, a two-track line which currently handles more than 300 trains a day and where a sustainable increase in the volume of traffic is not an operative possibility with the existing infrastructure. To that extent the Lower Inn Valley can be described as the bottleneck of international north-south rail traffic over the Brenner and has to be upgraded to increase capacity. The new KundlBaumkirchen section is already under construction, and the second upgrade section from Kundl to Kiefersfelden is now in the planning stage. The BEG began main construction work on the new Lower Inn Valley railway in August 2003 with the award of the first main construction lot. Meanwhile six of ten main construction lots are under construction (Lot H2-1, H3-4, H4-3, H5, H6, H7); the ones to be awarded are on schedule (H2-2, H3-6, H8, H1). The tunnelling techniques used vary from conventional drill and blast to hydro shield machines for the crossing of the valley or special techniques like open cut with underwater concrete invert or tunnel excavation with jet grouting. Contract lots and state of construction works are shown in Figure 1.

Figure 1: Brenner axis upgrade project, its contract lots and state of construction works (BEG, 2006)
8

2.3 TBM Contract Lot H3-4 and H-8

Owner BEG Project as a Whole

BEG Brenner Eisenbahn GmbH The High-speed Railway Brenner in Austria is part of the European north-south railway axis. This section has a total length of 39 km and runs predominantly in the underground or in trough structures. The total length of the double track tunnels is about 28 km. 3 additional investigation tunnels have a total length of 9.8 km. During operation, they will be used as evacuation tunnels. The design in the 3 stages - preliminary, tender, and final design - comprises conventional methods within the principles of NATM, the methods TBM, Jet grouting, and cut and cover method under air pressure.

Details of Lot H3-4 Lot H3-4 Mnster - Wiesing has a total length of 5.8 km, with Mnster - Wiesing a minimum overburden of approx. 8.5 m. The TBM tunnel (Hydro-Shield-TBM) has an excavation diameter of approx. 12.90 m. The lining consists of concrete segments with 0.5 m thickness and fire protection inner lining with 0.2 m thickness. The tunnel crosses the river Inn with low overburden, as well as the motorway A12 and the existing tracks of the Austrian Railway. Services Provided by D2 Consult Review of preliminary-, tender- and final design Review of statical calculation Structural analysis for the fire loading case Consultancy services during construction Period of Work Details of Lot H8 Jenbach/Stans 01/2000 06/2009 Lot H8 Jenbach - Stans has a total length of 5.19 km (3.5 km with Hydro-Shield-TBM), with a minimum overburden of approx. 6.0 m. The TBM tunnel has an excavation diameter of approx. 12.90 m. The lining consists of concrete segments with 0.5 m thickness and fire protection inner lining with 0.2 m thickness. The tunnel crosses the motorway A12 and the existing tracks

of the Austrian Railway with low overburden. Services Provided by D2 Consult Review of preliminary-, tender- and final design Review of structural calculation Technical assistance Structural analysis for fire loading cases Consultancy services during construction Period of work 03/2000 12/2008

Some preliminary information on the tunnel cross sections, anticipated TBMs and lining segments are shown in Figure 2 to Figure 4. TBM for Lot H3-4 will be delivered to the site in beginning 2007, and TBM for Lot H8 to be delivered in autumn 2007.

Figure 2: Cross section with escape tunnel and escape shaft. Lot H8 Jenbach

10

Figure 3: Normal cross-section, two tracks upgrading with fire prevention lining. Lot H8 Jenbach

Figure 4: Anticipated TBM and segments (BEG, 2005)

11

Chapter 3

3. Essential Interfaces of Excavation

3.1 Overview on Different Tunnelling Methods

3.1.1. General The large number of demanding infrastructural measures realized over the last thirty years has brought numerous technical innovations to tunnelling. Tunnelling methods and technology vary depending on geology, tunnel location, length and geometry, local tradition etc. In tunnelling there are essentially three different methods of construction: Open-cut method of construction Cut-and-cover method of construction Closed-face method of construction Tunnelling using the open-cut method of construction initially works vertically from the surface of the ground to the floor of the excavation pit. This is followed by the actual tunnel structure, after which the excavation pit is filled in again. Only then does work continue in a horizontal direction. With the closed-face method of construction, also known as underground tunnelling, the tunnel is driven horizontally from a starting shaft (e.g. in an urban area) or a tunnel inset/adit (e.g. in the mountains). The cut-and-cover method is a hybrid method of construction that combines both open-cut and underground methods of construction. Tunnels with the overburden less than half a tunnel diameter are usually built by using cut & cover methods.

With the closed-face method of construction, various tunnelling methods are possible:

- Tunnelling with mechanical means, ranging from excavators equipped with ripper teeth, hydraulic rams, and roadheaders to TBMs of various designs. Excavation by tunnel boring machine TBM is always referred to as full-face mechanized tunnelling.
12

According to the definition of the French Association of Tunnels and Underground Space (AFTES, 2000), mechanized tunnelling techniques (as opposed to the socalled conventional techniques) are all the tunnelling techniques in which excavation is performed mechanically by means of teeth, picks, or discs. Within the mechanized tunnelling techniques, all (or nearly all) categories of tunnelling machines range from the simplest (backhoe digger) to the most complicated (confinement-type shield TBM).

- Sprayed concrete methods of construction, such as Sprayed Concrete Lining (SCL) or New Austrian Tunnelling Method (NATM), Norwegian Method of Tunnelling (NMT), and Analysis of Controlled Deformation in Rocks and Soils (ADECO-RS). The SCL/NATM and NMT usually involve drilling and blasting; these and ADECO-RS all belong to the conventional group.

- Special construction methods (pipe jacking, Microtunnelling, Horizontal directional drilling, Caissons). The special methods and above-said conventional methods are not within the scope of this report.

Tunnels are built today where the public requests them and not necessarily where the geological conditions would be more favourable. This makes construction technically more difficult, more exposed to risks, and more expensive. In most projects, financial factors and the related scheduling are the crucial elements for any decision which can compromise both the excavation technique and safety consideration. However, all the tunnelling methods should consistently aim at improving progress, cost, performance, and safety.

3.1.2. Classification of Mechanized Tunnelling Techniques Also according to AFTES (2000), it is vital to have an official classification of mechanized tunnelling techniques in order to harmonize the terminology applied to the most common methods. The following Table 1 presents this classification. The table breaks the classification down into groups of machines (e.g. boom-type unit) on the basis of a preliminary division into types of immediate support (none, peripheral, peripheral and frontal) provided by the tunnelling technique. To give more details on the different techniques, the groups are further broken down into categories and types.

13

Table 1: Classification of mechanized tunnelling techniques (AFTES, 2000)

14

From the Table 1, different mechanized tunnelling techniques can be re-listed, with a few relevant definitions provided in the next section: Machines not providing immediate support: o o o o o o o o o o o Boom-type tunnelling machine (Out of the scope of this report) Tunnel reaming machine (Out of the scope of this report) Hard rock TBM Open-face gripper shield TBM Open-face shield TBM Double shield Mechanical-support TBM Compressed-air TBM Slurry shield TBM Earth pressure balance machine (EPB) Mixed-face shield TBM

Machines providing immediate support peripherally:

Machines providing immediate peripheral and frontal support simultaneously:

According to Swoboda (1990), the future of tunnel construction will certainly be increasingly influenced by tunnel boring machines. A combination of excavation with tunnel boring machines and blasting appears to be the most economic solution for the future. In line with the topic of this report, only tunnelling operations with tunnel boring machines (TBM) will be considered.

3.2 Global View of TBM Tunnelling

3.2.1 TBM Types

A TBM is a complex set of equipment assembled to excavate a tunnel. The TBM includes the cutterhead, with cutting tools and muck buckets; systems to supply power, cutterhead rotation, and thrust; a bracing system for the TBM during mining; equipment for ground support installation; shielding to protect workers; and a steering system. Back-up equipment systems provide muck transport, personnel and material conveyance, ventilation, and utilities. List of main constitutive items:

15

Front face where the soil is excavated with special tools (shield or cutting wheel/cutterhead) Steering mechanism part with drive engines for forward movement. Control mechanism for deviation and inclination Removal installation for transporting excavated material through the machine to a separator or directly onto an independent transport system Installations behind the working chamber permitting either further soil improvements (i.e. with rock bolts, shotcrete or injections) or are used for preliminary investigations Support installations within the protection of the shield tail Eventually grouting the void at the shielded tail created between the lining and the subsoil.

In addition to the above-said technical classifications of the machines by AFTES (French Tunnelling and Underground Engineering Association), there exist other national classifications, such as those of DAUB (German Committee for Underground Construction) shown in Figure 5, JSCE (Japan Society of Civil Engineers) shown in Figure 6, and SIG (Italian Tunnelling Association), etc.

TM Tunnelling Machines TBM Tunnel Boring Machines SM Shielded Machines

SM-T Shielded Machines with Part Heading SM-V Shielded Machines Full-face

TBM TBM without Shield

TBM-S TBM with Shield

SM-T1 Face without support SM-T2 Face with partial support SM-T3 Face with compressed air application SM-T4 Face with fluid support

SM-V1 Face without support SM-V2 Face with mechanical support SM-V3 Face with compressed air application SM-V4 Face with fluid support SM-V5 Face with earth pressure balance support

Figure 5: Tunnelling Machines (according to DAUB)

16

Figure 6: Types of Shield (JSCE, 1996)

Some companies or even individuals also draw out specific classifications. The following are TBM types, according to Rehm (2006): Hydroshield/Mixshield EPB-shield Hard-rock TBM Single/double-shield Gripper shield

Shield with partial face excavation Micromachines

Range of diameter of TBM manufactured by Herrenknecht AG is shown in the Table 2 below. The worlds biggest TBM was used at the Groene Hart railway tunnel project in The Netherland 2000-2004, with a diameter of = 14.87 m. This record has been recently broken by a new EPB TBM with a diameter of 15.20 m, manufactured by the same company.

17

Table 2: Range of diameter of different TBMs (Rehm, 2006)

Another classification of TBM types and operational modes can be summarized as in Table 3 (Grandori, 2006).

18

Table 3: TBM types and operational modes (Grandori, 2006)

DSU = Double Shield Universal TBM

19

According to Pelizza (2006), mechanized excavation methods can be briefed in two groups: full face mechanized continuous excavation method, using TBM for the excavation of tunnels in rock. The main problem is to break the rock; full face mechanized continuous excavation method, using mechanized shields and with counterpressure against the face for the excavation of tunnels in soil above and below the water table. The main problem is the stability of the tunnel as well as the control of the groundwater.

Both the TBM (hard rock applications) and SM (Shield Machine, soft ground) fulfill the same purpose: ensuring systematic and automated subsoil excavation; providing an effective protection (the shield) for the labour force at the front; stabilizing the tunnel through quickly closing of the support ring; transportation of the excavated material

Shields with Special Shaped Cross-section Shields with special shaped cross-section are divided into two types: Compound circular shield (or multi-head circular shields, multi-circular face shield - MFS), and non-circular shield (Figure 7).

Twin circular face Compound circular face Shield with special shaped cross-section

Compound triple circular face shield Elliptical face shield

Non-circular face shield

Rectangular face Horseshoe face shield Half-circular face

Figure 7: Shields with special shaped cross-section (JSCE, 1996)

20

Selection of TBM The tunnelling strategy is based on using different types of TBMs. Each machine should be able to deal the best with the ground conditions expected. They must allow a pressure to be exerted by the machine against the ground in front of the tunnel to limiting ground movement and settlement. The size of the tunnel and the geological conditions of the rock determine the type and the configuration of TBM that is used. Relevant geological factors for the TBM selection are: grain size distribution, type of predominant mineral (quartz contents), soil strength (cohesion), overburden, heterogeneity (mix ground, weathering), and piezometric pressure (Kovari et al, 2004). The effect of unexpected geological conditions can be strongly amplified if the TBM has been wrongly selected.

3.2.2 Operation of TBMs

Generally, tunnel construction using tunnel-boring machines (TBM) involves three main processes: excavation, dirt/muck removal, and tunnel support. TBMs key specifications may include: shield diameter, machine weight, maximum torque (that is needed for the cutterhead rotation under the maximum thrust), machine installed power, cutterhead/cutting wheel speed RPM, gripper force, penetration, cutter diameter, number of cutters on the cutting wheel, thrust per cutter, etc. A distinction is basically made between open-type machines, hard-rock TBMs, slurryshield TBMs (hydroshields), and earth pressure balance (EPB) tunnel boring machines. Below is the briefing on the operation of several TBM types.

Hard-rock TBMs The machines for rock are built to advance through a hard material that is usually self supporting, and have tools made for breaking even the hardest rocks. The excavation is carried out at atmospheric pressure, and the extraction of the material is performed using trains, trucks or conveyor belts to minimize wear (Figure 8).

21

Figure 8: Unshielded gripper TBM schematic drawing (COE, 1997)

The application of the standard TBM types for long and large diameter tunnels in variable ground conditions would be risky, in particular: Open gripper type TBMs are too sensitive to poor rock conditions especially in large diameter range; Single shield TBMs cannot reach high performances in hard rock and are sensitive to squeezing ground and face instabilities; Double shield TBMs, while can achieve very good performance in good to fair rock, are still sensitive to squeezing ground and to face instabilities. The extremely difficult geological conditions was a good reason to develop a new type of TBM, that is the Double Shield Universal TBM (Figure 9) which, starting from a main general design concept, can be configured into different specialized versions to suit the particular project requirements and geology (Grandori, 2006).

22

Figure 9: Double shield universal type TBM (Concilia, 2006)

Compressed air (Air pressure) TBM A compressed-air TBM can have either a fullface cutterhead or excavating arms like those of the different boom-type units. Confinement is achieved by pressurizing the air in the cutting chamber. Muck is extracted continuously or intermittently by a pressure-relief discharge system that takes the material from the confinement pressure to the ambient pressure in the tunnel (Figure 10). It is possible for TBM to work by air pressure, when the soil itself is nearly impermeable against the air. This is only possible in rare cases. In addition, the use of compressed air introduces the risk of a blowout, that is, a sudden reduction of support pressure on account of rapid loss of air; in this case the air may escape to the surface by leakage through soils pores or by a heaving of the ground mass above the shield. In recent years, thanks to advances in technology together with increasing reluctance, mainly for medical reasons, to use compressed air working methods, slurry shield and EPB tunnelling machines have become widely used for tunnelling in unstable ground conditions.

23

a b c d e f

Excavation arm g Tailskin seal Shield h Airlock to cutting chamber Cutting chamber i Segment erector Airtight bulkhead j Screw conveyor (or Thrust ram conveyor and gate) Articulation (option) k Muck transfer conveyor

Figure 10: Compressed air TBM - Boom type (AFTES, 2000) Slurry or fluid support machines (Hydroshield) The Slurry Shield and the Earth Pressure Balanced shield (EPBS) have been developed in the recent decades for managing the instability of the excavation profile in unfavourable geotechnical and hydrogeological conditions, with challenge external constraints (see Figure 11 and 12). With a Slurry TBM, the unstable ground at the front is supported by liquid mixture (bentonite suspension) under increased pressure generating an even counterpressure. A filter between the existing ground and the support liquid (i.e. using bentonite suspension) prevents the liquid from penetrating and disappearing into the ground. Depending on the subsoil permeability, density and viscosity can be varied, pressure can be regulated by controlling the speed of the delivery and feed pumps. The excavation is done by a turning cutting wheel. The excavated ground material and suspension liquid is mixed by hydraulic conveyance via tubes with subsequent separation of the two materials. The most important deviating design feature of hydroshields from the slurry shields is the presence of a compressed air buffer through which the support pressure at the fluid supported working face is controlled by means of a compressed air regulating installation.

24

3 2 4

Figure 11: Hydroshield/ Mixshield with double-chamber system. c Cutterhead, d Bulkhead, e Air-cushion, f Submerged wall, g Slurry line, h Stone crusher, i Feeding line, j Erector (Rehm, Herrenknecht 2006)

Figure 12: Typical spoke type cutterhead of the Hydroshield/Mixshield (Hamburg 14.2 m, Berlin 8.9 m)

25

EPB - Earth Pressure Balance machines Instead of a hydraulic/bentonite suspension as in Slurry TBMs, the excavated ground in EPB is used as part of the supporting liquid and forms a ground slurry. This method requests ground which is homogeneous, soft and cohesive (see Figure 13 and 14). If the water content is too low or if small particles are absent in the grain size distribution, they must be added artificially (bentonite, polymers, foam). This is called soil conditioning. In this case, the environmental compatibility of the material for landfill purposes must be taken into consideration. EPB machine has the technical advantage compared to the Hydro-Shield that a separation plant is not required, hence - space and cost for these systems are unnecessary.

1 8

4 7 2 3 5
Figure 13: Shield machine - EPB technique. c Face, d Cutterhead, e Working chamber, f Bulkhead, g Thrust cylinder, h Screw conveyor, i Erector, j Segments (Rehm, 2006)

26

Figure 14: EPB TBM = 9755 mm used for Botlekspoort Tunnel project 1999-2000 (Rehm, Herrenknecht 2006)

The original single chamber design of the traditional Slurry shield was developed into a two-chamber system (Mixshield) in Germany by the companies Wayss& Freytag and Herrenknecht in the 1980s. The mechanical concept of the Mixshield is a very unique tunnel boring machine considering it size and type, which allows a conversion between the operation modes EPB shield and Hydro shield in a very short time. The geological range of application for slurry shield and EPBM is given in Figure 15.

Figure 15: Relevant grain size distribution for EPB and Slurry TBM drives
27

Mixed-face shield TBM / Universal TBM Modern technology has enabled us to design mixed (versatile) machines able to deal with different (even extremely heterogeneous) soils during a single project, which can operate both as EPB machines or Hard Rock TBMs with a few modifications to the cutting wheel and extraction system. Changeover from one work mode to another requires mechanical intervention to change the machine configuration. The universal machines are best used with the universal ring. The biggest challenge that the tunnelling industry is going to face in the new millennium is related to the design and the use of large diameters TBMs for the construction of long road and railway tunnels. One point deserves mention is that, because of the magnitude of the risk associated with rock mass conditions, a new double shield universal TBM design has been realized. However, DSU TBMs could be uneconomical where systematic treatment of the face is required for the whole length of the tunnel; in this case the use of an EPB machine equipped with advanced conditioned and fine compensation systems might be more convenient (Concilia, 2006).

3.2.3 Conventional Tunnelling Versus TBM Tunnelling

TBM tunnelling is characterized by a less adaptability to geology, better progress rates, possibly continuous tunnelling, and relating to longer tunnels. More and more tunnel projects are going to be mined by the safety TBM technique which in the past would have been excavated by the conventional method with all its uncertainties for the personnel. The advantages and disadvantages of using a TBM include the following:

Advantages of using a TBM Higher advance rates Continuous operations Less rock damage Less support requirements Uniform muck characteristics Greater worker safety Potential for remote, automated operation

Disadvantages of a TMB Fixed circular geometry Limited flexibility in response to extremes of geologic conditions Longer mobilization time Higher capital costs

28

Concilia (2006) stated that the mechanised method of excavation can be 4 times (small and medium diameter tunnels) faster than D&B, providing the correct machine has been selected.

Although any tunnel excavation will influence the immediate surroundings to some extent, the aim of mechanized tunnelling which includes excavation and support installation processes is to avoid and minimize stress relaxation. A comparison on the general aspects between conventional and mechanized tunnelling could be tabled as follows (Assis, 2006):

Conventional Tunnelling Geometry Flexibility Geology Flexibility Contractual Flexibility Political Flexibility Lower costs for short tunnels and cheaper labour

Mechanized Tunnelling Lower impact to ground Higher and more constant quality (industrial process) Lower load to workers Safer More accurate schedule costs and

Table 4 below gives another comparison between conventional shotcreting measures and mechanical drives in terms of constructional engineering and operational terms.

29

Table 4: Comparison of major criteria for shotcrete tunnelling methods and TBMs (after A. Haack, 1996)
Item no.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Operation al phase Construct ion phase

Phase

Assessment criteria
Supporting agent in face zone Lining thickness Safety of the tunneling crews Working and health protection Degree of mechanization Degree of standardization Danger of break Construction time - short tunnel Construction time - long tunnel Construction cost - short tunnel Construction cost - long tunnel Tunnel cross-section Cross-section form Degree of utilization of the driverelated tunnel cross-sections

Shotcrete tunnelling method

variable variable lower lower limited conditional higher shorter longer lower higher variable as desired generally higher

TBM
safer constant higher higher high high lower longer shorter higher lower constant generally circular generally lower

With respect to the question of comparative costs between conventional/hand mining and mechanized tunnelling, Sauer (2004) made a diagrammatic representation as shown in Figure 16. The diagram shows that conventional tunnelling is more cost effective than mechanized tunnelling for the cases of short tunnels (< 2.4 km), shafts and tunnels with changing geometry, and/or substantially changing geotechnical behaviour. There is an overlapped area where hand and mechanical mining may be equally considered and where a dual design is recommended. With tunnels longer than 3.2 km, Sauer showed a little difference of construction costs between hand mining and mechanized tunnelling.

30

Figure 16: Tunnel cost of Mechanized Tunnelling (MT) versus Hand Mining (HM) over tunnel length (Sauer, 2004)

In addition, according to Wagner (2006, oral consultation), the statements on the tunnel cost must be more related to ground conditions. It is impossible to establish a general rule on the cost, because it relates to geological conditions and to work progress (i.e. advance per day or per month). The tunnel costs may also need to be estimated using special TBM cost estimating software and cost database. Below Figure 17 shows a comparison of excavation and ring installation times between the TBM serial tunnelling and TBM continuous tunnelling. The first technique uses traditional ring type; the second uses unified rings, featured by simultaneous excavation and ring erection. The erection time of the first is twice longer than the second.

"SERIAL" TUNNELLING

CONTINUOUS TUNNELLING
TUNNEL LENGTH

RING 3 RING 2 RING 1

RING 3 RING 2 RING 1 ERECTION TIME EXCAVATION TIME PER RING ERECTION TIME INSTALLATION TIME PER RING

TE,C = 0,5 . TE,S

Figure 17: Modes of tunnelling with unified rings (Wagner, 2006)

31

Given the Hai Van Pass Tunnel in Vietnam (2000-2005) as an example; this is a double lane highway tunnel, the width is 11.9 m, excavation area is 95.2 m2, opening area is 73.3 m2, the length 6.2 km, excavated in accordance with the principles of NATM. If a combination of TBM and conventional excavations had been done, then the progress would have been faster 2 3 times. The scheme for this combination would be: both sides are excavated with NATM in lengths of 50 100 m, so that the max. conventional length is about 250 m, and the remaining main section is driven with TBM. It may be interesting if the same scheme is considered for the upcoming tunnel on the same highway No. 1A in Vietnam, namely Deo Ca Tunnel.

3.3 Critical Cases of TBM Excavation

Although the conventional tunnelling holds certain critical cases, they will not be considered in details in the present report. Other special methods of tunnelling such as open trenches, cut-and-cover structures (direct and inverse excavation), door-frame method, and immersed tunnels (caisson and door-frame methods), etc. are also beyond the scope of this project. However, it is possible to name a few of difficult conditions in conventional tunnelling such as: heavy groundwater leakage/high pressure inflow, very low overburden, soft ground, mixed face, swelling, squeezing, and rock burst, etc. It is recognized that the TBM performance is influenced by the rock mass quality, the selected machine type and the tunnel diameter. Advances in TBM technology and reliability have resulted in bored tunnels being successfully driven through ground conditions historically considered difficult. However, critical cases of TBM excavation from which risks emanate do not disappear. During excavation, the situation can become critical at any minute, meter, and under any circumstance.

3.3.1. Risk Management for Tunnels

It is well known, that tunnelling is not a risk-free technology. Tunnels are regarded as so called heavy risks, because each tunnel is a specific unique project on its own in a unique combination of ground / soil. The right construction method with the right experience parties involved are crucial for the success. The main most important factor however, the geology, is only known to a limited extent. Any accident during construction as well as in use provokes a substantial interruption and often a standstill till the problems are solved (Andreossi, 2001). The construction of tunnels and underground works are affected by potential risks not only for the different active Parties/main Actors (Owner/Client, Consulting Engineers, Contractor, Supplier), but also for the Public, especially in urban areas. Tunnelling projects must now consider numerous underlying political and environmental factors

32

which add to the overall complexity of a given project (IMIA, 2001). This is in conformity with the statement by Parker (2005), that realistically, not all risks associated with complex construction projects can be entirely avoided or mitigated, which calls for the need of risk management. According to its definition risk has two components: probability of occurrence w and amount of damage D. In a quantitative appraisal the product of these two factors defines the risk: R = w x D. In other words: Risk = Probability x Impact = Likelihood x Consequences = Probability of Happening x Cost of Event

Figure 18 shows the concept of reducing an initial risk by reducing its probability and impact. It is clear that residual risks are unavoidable and they should be shared among the Parties and systematically controlled by countermeasures.

initial risk consequences or effect/impact

NOT ACCEPTABLE residual risk

ACCEPTABLE

likelihood or probability

Figure 18: Concept of reducing initial risk by reducing its probability and impact

Several kinds of risks include (ITA, 2004): significant cost overrun risk work delay risk environmental risks risk of spectacular tunnel collapses and other disasters (potential for large scale accidents during tunnelling work) risk of damage to a range of third party persons and property in urban areas, (a particular concern with heritage designated buildings)
33

risk of public protests, caused by the problems of tunnelling projects

Risk management is the overall term which includes risk identification, risk assessment, risk analysis, risk elimination and risk mitigation and control. Risk management for tunnels is now routine for major projects worldwide. Tunnel construction imposes risk not only on all parties involved, but also on those not directly involved (note, that third parties - person or companies - have become much more claims conscious). Traditionally, risks have been managed indirectly through the engineering decisions taken during the project development. However, ITA Guidelines (2004) now recommend Systematic Risk Management Techniques instead. And through the use of a robust and transparent Risk Management Plan (RMP) adopted from the early design stages to the construction and operation phases, most risks can be effectively managed. Risk management tools include (ITA, 2004): Fault tree analysis Event tree analysis Decision tree analysis Multirisk Monte Carlo simulation

In a broader context, we could mention a Code of Practice for Risk Management of Tunnel Works, drafted by BTS (British Tunnelling Society) in association with Insurers, such as ABI (Association of British Insurers), IMIA (International Association of Engineering Insurers), and ITIG (International Tunnelling Insurance Group). This is a joint effort to face the more demanding challenge of future tunnelling projects. It is hoped that the risks will not be solely transferred to the insurer, but be fairly shared between the Parties involved. An international version of the Code is being prepared by BTS and likely to be completed soon. Its project stage basis is worth mentioning: 1. project development stage, 2. construction contract procurement stage, 3. design stages, and 4. construction stage. Of which, in the construction contract procurement stage, three highlights are: i) the use of FIDIC (International Federation of Consulting Engineers), ICE (Institution of Civil Engineers), National or Proven Form of Contract; ii) including of GBR (Geotechnical Baseline Report) in Contract Documents, and also in Subcontract Documents; and iii) including risk clauses in contract. The risk of unforeseen ground conditions (differing site conditions) and contractual claims cannot be overlooked; these can be administered fairly with the help of Disputes Review Board (DRB).

Below is given a brief introduction to a cost-risk estimating procedure (CEVP) and a decision making tool (DAT) in tunnelling.

34

The cost-risk estimating procedure CEVP (Cost Estimate Validation Process) recently developed by the Washington Department of Transportation (WSDOT). CEVP develops a probabilistic cost and schedule model to comprehensively and consistently define the probable ranges of cost and schedule required to complete each project, by incorporating uncertainty (uncertainty includes both risk and opportunity) (Reilly, 2005) (Figure 19).

Figure 19: Future costs are a range of probable costs (Reilly, 2005)

Because risks are explicitly defined, a risk management plan can be quantified earlier. This allows significant management and control of cost and schedule earlier in a project and allows a more explicit communication of cost and schedule (and changes thereto) with the public and key political decision makers. Obviously, it is desirable to use some decision making tools in tunnelling like DAT (Decision Aids in Tunnelling) to make more rational, informed, and effective decisions in tunnel design and construction. The software DAT is the product of a long research effort by: MIT (Massachusetts Institute of Technology, USA), EPFL (lEcole Politechnique Fdrale de Lausanne, CH), and GEODATA - Turin (I) for estimating the cost and duration of the construction of an underground project. The most important element of DAT is the possibility to consider diverse sources of geotechnical and construction uncertainties and variabilities (Grasso et. al. 2001, 2002). In Figure 20, the deterministic estimation and result of DAT simulations of a project duration and cost are visibly compared.

35

Figure 20: Scatter plots of a project duration and cost (after Grasso, 2001)

Both RMP and DAT have been successfully applied in recent years to a series of important deep and long tunnel projects like the California high-speed rail, the Pajares and Guadarrama high-speed railway tunnels in Spain, the new Lyon-Turin high-speed railway link (Grasso, 2006).

Experience with risk management for the Copenhagen Metro in Denmark (opening in turn 2002, 2003 and 2007) (ITA, 2004): The Contract defined the construction risk assessment work to be carried out by the Contractor. There were general requirements for all the construction risk assessments to be carried out for all construction sites and some further requirements to the construction risk assessment for the TBMs. The TBM construction risk assessment had to start immediately after signing of the Contract with an assessment of the conceptual design followed by an assessment of the detailed design with the purpose to contribute to the design of the TBMs. Furthermore, risk assessment of the TBM operation was carried out - providing input to the operation procedures. Risk identification and management (applied for the tunnel project xyz):

36

Installation risk resulting from TBM rotation TBM advance pressure Twisted thrust shoes Rigid erector hydraulics Transport conditions Sequence of installation Eccentricity of thrusters Sectional tension forces

Risks are related to both TBM tunnelling in soil (soft ground, e.g. urban area) and to TBM tunnelling in rock, therefore below is shown a number of critical cases that bring about risks.

3.3.2. Critical Cases of TBM Tunnelling in Soil

Basic soil classification Ground categories are broken down roughly according to the following table:

Table 5: Simple soil classification for construction purpose

Hard rock Dolomite Limestone Sandstone Granite Basalt Lava Gneiss Quartzite Soft rock Marl Hard clay Slate Dolomite Soft soil (without cohesion) Sand, fine gravel, gravel, coarse-grained, stones Flowing ground = soil with high water content Soft soil (with cohesion) Expansive or swelling ground such as clay-stones, anhydrite rock in contact with water, silt

Generally, in soft ground, majors concerns are opening stability and control of displacement field. Soft ground tunnelling is likely dominated by failure and admissible displacement criteria. Ground conditioning (improvement and reinforcement) might play an important role. In consolidated clay, the optimization of values and quantities of the slurry pressure and grouting pressure is required for TBM technology. In urban environment, major concerns are related to: shallow overburden, existence of nearby structures, foreign objects inside the ground, constraints for alignment,

37

restrictions for auxiliary works, and high visibility of damage. Ground conditions are normally challenging, characterized by recent weak geological formations near the ground surface, by frequently changing conditions due to the occurrence of lenses, layers, boulders, etc., and by presence of ground water above the tunnel or crossing the tunnel profile. Kovari et al (2004) listed the following specific features of metropolitan areas: Shallow overburden: Low overburden may be combined with a large tunnel diameter to create ground deformations (settlements) and collapse up to surface (Figure 21).

Figure 21: Full face mechanized excavation of shallow tunnel (Maidl et al. 1996) Existence of nearby structures: The next structures may consist of buildings, roads, railroads, bridges, underground networks, subways, etc. (Figure 22). The sensitivity of these structures to ground settlements as well as the potential damage to ground collapse may vary within extremely wide ranges. Surface monitoring in urban environment is fundamental to control the effects and the potential damages on pre-existing buildings, utilities and infrastructures. For example, in the first stage of settlement assessment, if the predicted settlement from bored tunnels is less than 10mm and the predicted ground slope is less than 1/500 (equivalent to damage risk category 1 as defined by Rankin, 1988), then buildings are not subject to further assessment.

38

Figure 22: Influence of a TBM drive on the neighbour buildings (Gruebl, 2006) Foreign objects inside the ground: The presence of frequently hidden subterranean obstacles is also one of the specific features of urban tunnelling using TBMs. These may include historical wells, ground anchors, sheet-piles, erratic blocks, archeological artifacts, abandoned utilities such as for gas and sewage, but also tree trunks, artificial fillings, etc. Constraints for alignment: Selection of both horizontal and vertical alignment generally meets with constraints. The tunnel is usually driven in public ground under main roads or streets. But it is not always possible to avoid under-passing buildings, roads and other structures, and this may cause various difficulties. For example, land acquisition costs would be very high and the foundations of the existing buildings would create complications during construction. Figure 23 shows the Brisbane North-South Bypass Tunnel alignment within the city central area.

39

Figure 23: Brisbane North-South Bypass Tunnel, Central Connection Option 7 (SKM Connell Wagner JV, 2005) Restrictions for auxiliary works: It is practically important to note the serious restrictions when selecting places of attack (launching shaft, access to TBM) and planning material transport from and to the construction site. Other restrictions include sinking drill holes for explorations, for ground water control or ground consolidation. High visibility of damage: In the urban environment damage to buildings and roads has a high visibility. The risk aversion is very pronounced, which may lead to a strong opposition to further underground projects in towns or even elsewhere. The loss of public confidence in the technology can be looked upon as a kind of damage.

Other difficult cases (in a general nature, after JSCE, 1996): Underground tunnel connection: Two shield tunnels can be connected head-on, or one tunnel is connected into the side of the other tunnel. For head-on connection, auxiliary measures such as chemical grouting, high-pressure jetting and mixing method or ground freezing are generally used. For side-by-side connection, reinforcement of the existing tunnel must be studied to prevent any damage to it. Underground space enlargement: Construction of an underground station between two tunnels, a space for shield machine assemble or underground space establishment, a branch-off tunnel or connection of a tunnel with some angle. Because the ground is already loosened by the preceding tunnel construction work, the ground shall be stabilized by auxiliary method as required. The tunnel shall then be carefully enlarged by excavating a partial face and by supporting the ground with steel supports or special segments.

40

Kovari et al (2004) then summarized the main features permitting safe and economic tunnelling in soft ground under urban conditions using TBMs with slurry or EPB type of face support as follows: Efficient TBM technology Reliable design procedure Improved methods of conditioning Advanced grouting technology Reliable risk management

Practically, Cross London Rail Links (Crossrail, U.K.) can be referred to as a large tunnel project in urban area today. The project has plans to ensure that the tunnel boring machines are performing as required, the TBM parameters together with the information from the ground movement monitoring need be relayed to a tunnel monitoring and settlement control room. This will be in addition to the contractors monitoring arrangements and will be in place and operated throughout the tunnelling activity. The control room will have displays of real time surface, subsurface and tunnel movement monitoring together with TBM tunnel progress and TBM parameters.

3.3.3. Critical Cases of TBM Tunnelling in Rock

Some of the (obvious) high-risk factors (multiple unexpected events) that TBM tunnels may suffer from include (Barton, 2006): significant fault zones adversely oriented planar clay-coated joints very weak rock, or very hard massive rock (high UCS rocks) very abrasive rock very low stress, very high stress exceptional stress anisotropy highly fractured or karstic zones with high volumes of stored water (severe water inflows) high permeability

In addition, Barla & Pelizza (2000) wrote, that important or difficult ground conditions for TBM tunnelling include boreability limits (when boring through rocks with very high strength), instability of excavation walls (a clear limit for open type TBMs), instability of excavation face, fault zones and squeezing.

41

It is hoped that the majority of the limit conditions will be coped with by adopting special methods and procedures of advance. Below are selected and explained to more details some of the above high-risk factors. 9

Significant fault zones

The cause(s) of a tunnel collapse or TBM cutter-head blockage (machine is trapped) in a tunnel are usually clear to the tunnelling engineer only after they have happened. Fault zone stoppages and difficulties/delays in making drill-and-blast by-passes for TBM/cutterhead release may eventually lead to the abandonment of TBM itself or abandonment of TBM option. So, before the event it would often be necessary to be exceptionally pessimistic to have foreseen the unthinkable. The unthinkable is often the combination of several adverse factors, which separately are expected though serious events, but when combined are, quite logically, unexpected events (Barton, 2006). Figure 24 shows a TBM stuck in a bulk of caving-in materials; this implies, that "risky" means to free up the trapped TBM cutter head or shield.

Figure 24: By-pass situation for the double-shield (11.7m) TBM at Pinglin tunnel, Taiwan (Shen et al. 1999)

Short sections of crushed shear zones with clay and gouge material may cause serious time delays for TBM excavation. Spiling rock bolting is very efficient under such circumstances, provided the fully grouted rebar bolts can be placed efficiently, which requires proper drilling equipment. Figure 25 exhibits a situation of TBM pull back and liner disassembly due to debris flow from a fault gauge zone.

42

Figure 25: TBM pull back and lining removal to overcome the debris flow (Oggeri, 2006)

Figure 26 shows the TBM utilization while boring through a fault zone. The utilization factor is an important parameter, and it varies according to the actual geological conditions.

Figure 26: Geological conditions and productivity (Concilia, 2006)

43

Pre-grouting has been proposed by Barton as a measure to reduce risks, because preinjection can increase Q index of the rock mass. It is well known, that the Q-values (the rock tunnelling quality) are estimated from the following expressions:
Q= RQD Jr Jw x x Jn Ja SRF RQD0 Jr Jw SIGMA 20 q x x x x x x Jn Ja SRF F10 / 209 CLI 20 5

QTBM =

where: RQD = Rock Quality Designation, is characterization of the degree of jointing; Jn = number of joint sets. RQD/Jn = quotient representing a crude measure of relative block size ; Jr = rating for the roughness of the least favourable of the joint sets or filled discontinuities; Ja = rating for the degree of alteration or clay filling of the least favourable joint set or filled discontinuity. Jr/Ja = quotient representing the roughness and frictional characteristics of the joint walls with or without filling materials. It is crude measure of inter-block shear strength; Jw = joint water reduction factor; SRF = Stress Reduction Factor; Jw/SRF = quotient representing a crude measure of active stresses; RQDo = RQD (%) measured in the tunnelling direction; SIGMA = rock mass strength estimate (MPa) found from a complicated equation including the Qo value measured in the tunnel direction (the same as the six first parameters); F = average cutter load (ton, ~10kN ) through the same zone, normalised by 20 tons; CLI = Cutter Life Index; q = quartz content in percentage terms, %; = induced biaxial stress on tunnel face (MPa) in the same zone, normalised to an approximate depth of 100 m.

44

Pre-injection/pre-grouting may cause moderate, individual effects to every parameter (6 parameters) contained on the first expression, thus increase Q index of the rock mass. The concept of multilayer pregrouting is represented in Figure 27. The first grouting operation is to create around the tunnel an outer reduced-permeability zone using "blocker" grout. The second grouting is to make an inner permanent strengthened, low permeability zone using stable ultra/microfine cementitious grout. The third is an extended, strengthened, low permeability zone ahead of the excavation face.

Figure 27: One of ELKEMs Multigrout concepts

The idea of reducing risks by pre-grouting also lies behind the fact that, the relative time for tunnelling and the relative cost of tunnelling normally decrease in accordance with the increase of rock classes quality. So, if Q could be detected before tunnelling, and if Q could be improved during tunnelling, then both time and cost could be reduced accordingly (Figure 28).

Figure 28: Relative time expenditure (left) and relative cost (right) of tunnelling in relation to Q value

45

Finally, Barton drew out the following comments on the risk to TBM tunnelling from faults: High risk factors are often combined in an unexpected combination when TBM get stuck; Risk can be reduced by appropriate use of standard techniques (geological logging and rock mass characterization, core logging, hydraulic testing, seismic profiles between holes); When tunnel depth is great, each of the above require extrapolation and risk increases, making probe drilling (even) more important; Barton stated that: Dont automatically assume that long tunnels need TBM this will also reduce risk! The assumption that TBM go faster than drillingand-blasting in long tunnels introduces several increased risks: a) adverse rock quality statistics (extreme-Q-value problem) b) need central rock qualities to improve TBM deceleration (negative gradient) of the penetration rate and advance rate with increased tunnel length or time of measurement c) less favourable problem solving conditions for the contractor in TBM tunnel Seismic velocity probing needs careful correction for stress/compaction effects as the longitudinal wave velocity VP in front of TBM may increase without rock quality improvements (e.g. the deeper rock does not always mean better quality, but just more highly stressed); A way to improve effective rock quality and control water, and therefore to reduce risk, is to (try to) perform pre-injection ahead of the face.

However, it should be noted, that Blindheim (2005) already recommended the QTBMsystem not be used. In addition, Palmstrom and Broch (2006) also think it is not likely that Q is suitable to express the effects of pre-grouting. They went further that QTBM is complex and partly misleading and is not recommended for use in its present form, i.e. to allow estimates of penetration and advance rate for TBM (Figure 29).

46

Figure 29: Suggested relation between penetration rate (PR), advance rate (AR) and QTBM values (Barton, 1999b) 9

Very high stress

Serious problems of high rock stress phenomena in TBM tunnels involve intense rock spalling in hard and massive rock, which causes significant safety hazards (e.g. during cutter change) and delays in work progress (e.g. loss of support for the TBM gripping pads). Rock stress may also lead to occurrences of mild buckling of schistose rock. It is important to be able to perform rock bolting efficiently immediately behind the cutter head in order to maintain work safety and firm gripping for continued boring. The need to handle rock stress problems in bored tunnels has been due to rapid development and application of tunnel boring. The experience on this issue should be applied from the planning stage of TBM projects where such problems may be encountered, and the well preparedness should be highlighted. In massive and strong rocks the spalling can occur in a concentrated and intense manner in bored tunnels, therefore utilizing TBMs with an open configuration and not shielded TBMs could be an advantage. 9

High volumes of stored water, under pressure

The only available tunnelling technique that can keep the ground water in-leakage near zero, is the Earth Pressure Balance Machine (EPBM), full face mechanical excavation using a pressurized shield and gasketed concrete segment installation. Such machines are for soil excavation and are limited to shallow depths (typically less than 15 meters) (Garshol, 2003).

47

In hard rock tunnelling this alternative is not available, even if a TBM and concrete segments are used for the excavation and support. Without pre-injection the leakage volume could locally become far too large, between the time of exposure and the time of segment erection and efficient annular space backfilling. With a serious local water inrush at hand, such segment handling and grouting would also be very difficult. Environmental restrictions are also a part of TBM tunnel excavation projects. Even such tunnelling may require a strict ground water control, because of the normal potential consequences on the ground surface. There are two ways of handling water inflow problems: 1. Pumping out of the water 2. Injection There are clear limits as to the quantities of water that can be pumped through pipes, or that can be handled by gravity drainage systems, provided that reasonable practical and economical frameworks are applied. The limitations are even more pronounced when considering the face area. Tunneling on a decline will experience problems already at inflow rates of only 1 to 2 m3/min. Water at a high static head may cause water jets spraying the whole face area, causing very difficult working conditions, especially at low water temperatures. If inflows have already occurred (through cracks and joints), post grouting is very difficult, costly and often unsuccessful, especially at high pressure. To take advantage of spiling rockbolts and to execute pre-injection, it is an obvious prerequisite to be able to drill the necessary boreholes in the right positions and at the correct angle. In drill and blast excavation this is simple, but in TBM projects it has repeatedly turned out to be difficult. The owners should accept only bids that contain the drilling method proposed by the contractor. The drilling method should be sorted out before start of the TBM operation (Figure 30 and 31).

Figure 30: Tailor made hydraulic drilling equipment mounted on hard rock TBM (Garshol, 2003)

48

Figure 31: Borehole length and net coverage per grouting stage, plan view (Garshol, 2003)

In short, a well planned use of probe drilling, pre-injection locally and pumping will normally be the optimum solution. The risk of major water inrushes can be virtually eliminated. 9

Squeezing rock conditions

Excessive rock pressure may cause the failure of the tunnel support resulting in large rock deformations, with the tendency to reduce the cross-section of the opening. This phenomenon is referred to as squeezing rock behaviour. Low strength and high deformability of the rock as well as the presence of porewater pressure facilitate squeezing. The following rock types are specially prone to developing large pressure and large deformations: altered gneiss (chemically altered/metamorphic igneous rocks), schist, phyllite, serpentine, shale, clay, mudstone, tuff, and certain kinds of flysch (Kovari, 1998). Mechanized excavation (use of Tunnel Boring Machines) in squeezing rock conditions is characterized by a certain degree of difficulty. It is generally agreed at the present stage that experience and technology have not progressed far enough to recommend without some reservations machine excavation in such conditions. The major difficulties can be listed as follows: Instability of the face; Relative inflexibility in the excavation diameter; Problems with the thrust due to reduced gripper action, for gripper type machines; Difficulty to control the direction of the machine, in soft or heterogeneous ground.

Instability of the face may be a problem only when severe squeezing conditions exist, which exert important face extrusion that could be difficult to be controlled. It is also difficult to anticipate precisely the type and magnitude of tunnel convergence (i.e. the reduction in size of the opening in course of time). With a slow advance rate, the danger of TBM blockage in a squeezing zone (i.e. a fault zone) is increased. Stoppage of the TBM can be brought out by a number of factors such as inflow of water,

49

advancing face, overbreak, or machine breakdowns. In the worst case as such, with a continuing displacement the machine may be squeezed, leading to difficulty or impossibility to restart the machine. The type of machine, i.e. shielded or not shielded TBM, will be selected based on the degree and extent of squeezing rock conditions along the tunnel length. Both shielded and open TBMs may be sensitive to rapid or large convergences plus wall instabilities, causing the problems of support installation and gripping. Therefore, increasing the diameter of the cutter head (overcutting) is often foreseen, so that the gap between the shield and the excavation contour from the usual value of 6-8 cm can be adjusted to 15-25 cm (Figure 32). Mediate squeezing conditions could be coped with by TBMs that are specifically developed to accommodate some radial deformation of the tunnel perimeter as the machine advances (Barla & Pelizza, 2000).

Figure 32: Solution for radial overcut by increasing the excavation diameter Controlled overcutting device

Squeezing ground behaviour is characterized by the occurrence of large rock pressure which may lead to the failure of the lining. Therefore, there have been attempts to limit the loads on the lining ring by means of installed strain elements. The plastically deformed joints cannot be sealed watertightly. Loads can also be limited through the use of substances of a defined firmness which collapse when a certain resistance limit is passed. These substances are either fitted on the outside of the segment or added to the grouting material.

In another development, for the squeezing type of deformation phenomenon the following are possible counter-measures for mechanised excavation (Grasso, 2006): plane head; overcutting; requalification of the rock mass ahead; bolts in shotcrete layer 1 (Open TBM);
50

short shield/absent/ with loopholes (Open TBM); conical shield (Double shield); bentonite injections (Double shield); deformable support (Open TBM); extended gripper surfaces (Open TBM)/ thrust on segments (Double shield); shotcrete in layer 2 (Open TBM).

3.3.4. TBM Tunnelling in Mixed Face Conditions

For a specific long tunnel project with changing geological conditions in different segments/sections, the tunnelling can be completed in phases using TBMs. Successful and economical completion of each section of tunnel requires the proper choice of TBM type and ground support, based on the ground conditions expected. Several different TBM types with various ground support methods may be needed to accommodate the various ground conditions. Development of TBMs during the past over 20 years has brought about so-called universal machines (polyshields, multipurpose shield, Figure 33), used for both soft ground and hard rock. The continual improvement of various extraction techniques have led to machine types capable of penetrating more heterogeneous subsoil, that is respectively a mixture of soft soils, unconsolidated ground and rock, thereby enhancing their rather limited flexibility.

Figure 33: Polyshield machine, EOLE Lot 35 B, Paris, France (Wagner, 2006)

51

To end up this section, it is fair to cite the following statement by Barla and Pelizza (2000): The unfavourable conditions can be produced by either a rock mass of very poor quality causing instability of the tunnel or a rock mass of very good quality (i.e. strong and massive rock mass) determining very low penetration rates. However, it is to be observed that when using the full face mechanized excavation method, the influence of the rock mass quality on the machine performance has not an absolute value: the influence is in fact to be referred to both the TBM type used and the tunnel diameter.

3.4 Ground Reinforcing

3.4.1. General This part is devoted to the problems of face stability and ground reinforcement for TBM drives in cohesive and cohesionless soils.

3.4.1.1

Face Support

Face stability is fundamental to avoid failure. In closed shield tunnelling, there are three main ways to support the face: by compressed air, by slurry, and by excavated soil (EPB) (Figure 34).

Figure 34: Face support in closed shield tunnelling (Kovri et al, 2004)

3.4.1.2

Failure Mechanism

Ground and groundwater pressures at a workface can be unbalanced: If the shield advancement rate and muck discharge rate are not synchronized in an EPB shield or a slurry shield, the pressure inside the chamber becomes different from the ground and groundwater pressure, at the face become unbalanced. If the pressure in the chamber is

52

smaller than the ground pressure, surface settlement occurs. In cases of contrary, ground heave occurs. Explaining and predicting face stability by statical calculations has already been addressed by a number of authors, of which some are mentioned here. Upper and lower bound 2D solutions for clays (limit equilibrium solutions) can be founded in the work of Davis et al (1980). Limit-state-design-based solutions have also been proposed by Leca and Dormieux (1990) for sands. A three dimensional failure scheme consisting of a soil wedge (lower part) and a soil silo (upper part) has been given by Jancsecz and Steiner (1994). Anagnostou et al (1994, 1996), based upon the silo theory according to the tridimensional model of sliding mechanism proposed by Horn (1961), provided a good understanding of the mechanics of face failure and determined face support pressures when using a bentonite slurry support as well as EPBM. In the following only two methods will be briefly introduced, i.e. Leca and Dormieux (1990) and Anagnostou and Kovari (1994, 1996).

a) Limit-state-design-based solutions by Leca and Dormieux (1990) Leca and Dormieux (1990) proposed upper and lower bound solutions for the face stability of shallow circular tunnels in frictional materials (sandy soils) (Figure 35). The question of determining the retaining fluid pressure to apply to the tunnel front is a three-dimensional problem, and was studied by using limit state design method. Such retaining/supporting pressure T can be achieved by using compressed air, bentonite slurry or earth pressure (EPB shield).
A S C P D T T = retaining pressure Section A-A S = surcharge = unit weight of soil P = unsupported span S C = unconfined compression strength

C = 2
T

c ' cos ' 1 sin '

c' = cohesion ' = friction angle

Figure 35: Simplified geometry for the front stability of a shallow tunnel (after Leca and Dormieux, 1990)

Both safety against face collapse and blow-out were considered based on the motion mechanisms of rigid conical blocks in front of the face, then a failure criteria was proposed for a cohesive and frictional soil.

53

Three failure mechanisms have been considered namely MI, MII and MIII, which are shown in Figure 36, with the assumption that the unsupported span behind the face P is zero. MI and MII are failure mechanisms due to the collapse of one conical block and two blocks, respectively, whereas MIII refers to blow-out failure in case of very shallow tunnels bored in weak soils with the pressure T becoming so great that soil is heaved in front of the shield.

f' H C D
T

, c', '

c' = cohesion ' = friction angle

(a) Failure mechanism MI

S 2

(')

C D T V1
1 S

V2

'

' 1

, c', '

2
12

1
2

() (b) Failure mechanism MII

C D
T '

, c', '

(c) Failure mechanism MIII

Figure 36: Conical failure mechanisms (a) MI, (b) MII and (c) MIII (after Leca and Dormieux, 1990)
54

Both MI and MII are characterized by only one parameter, the angle between the axis of the cone adjacent to the tunnel and the horizontal. MIII is also characterized by ; the geometry is similar to that of MI except that the cone is inverted and the discontinuity velocity V along the failure surface reversed. The problem is analyzed in terms of five dimensionless parameters: C/D, S/C, T/C, D/C and Rankine earth pressure coefficient for passive failure KP (or KA for active failure). An upper bound solution is then found, given that in order for the set of external loads {S/C, T/C, D/C} to be stable, the power e of the loads applied to the system and the power PV that can be dissipated inside the system during its movement must satisfy

e P V
or in the form

N S QS + N Q QT
for collapse mechanisms MI and MII and

N S QS + N Q QT
for blow-out mechanism MIII, where the parameters associated with mechanism MI are given below as an example. NS and N are weighting coefficients that depend on the angle , as in the following relations:
2 RD 1 NS = tg cos ( 2 ' ) cos ( 2 ) RE
3 1 RD N = RB 1 3 tg 3 RE

QS, QS and Q are three loading parameters as in the following relations:

QS = ( K P 1) QT = ( K P 1) Q = ( K P 1)

S +1 C T +1 C D C

RB, RD, and RE are parameters for the simplification of expressions (i.e. "convenient coefficients"):

55

RB =

cos ( ' ) cos ( + ' ) sin 2 '

RD = sin 2

2H sin 2 ' D

RE = sin 2 ( ' ) sin 2 ( + ' )

KA and KP are Rankine earth pressure coefficients for active failure and for passive failure, respectively:
KA =
KP =

1 sin ' 1 + sin '

1 + sin ' 1 sin '

The amount of material involved in the three above failure mechanisms is limited, but such geometries could be representative of initial ground movements that could lead to larger scale failures. These solutions can provide reasonable estimates of critical face pressures.

b) Analysis method of limit equilibrium by Kovri and Anagnostou (1996) The stability of the face involves the assumption of a simplified failure mechanism in the ground ahead of the face. The three-dimensional model of Horn (1961) shown in Figure 37 is close to reality and is simple to handle, which will be used for both slurry and EPB modes of operation. It consists of a wedge in front of the face and a prism extending up to the surface in the state of limit equilibrium. The support force S as a function of the inclination of slip surface is also shown on the right side of Figure 37. Geometrical parameters and notations of the computational model for face stability design are given in Figure 38.

56

S
support force

Smax

critical inclination of the slip surface

Figure 37: Failure mechanism consisting of a wedge and a prismatic body (after Horn, 1961, adopted by Kovri and Anagnostou, 1996)

Figure 38: Computational model for face stability design - Geometrical parameters

i) For slurry-shield-driven tunnels, Mohr-Coulomb failure condition and drained conditions is assumed. Complex interrelations between the various parameters (shear strength and ground permeability, suspension parameters, slurry pressure, geometric data of the tunnel, and safety factor) are given below.

57

Considering the failure mechanism, at each point on the slip surface, the mobilized shearing resistance is given by

c tan + F F

where and F denote the normal stress and the safety factor, respectively. The mean effective vertical stress v along the interface plane between the prismatic body and the wedge (i.e. the silo pressure at the tunnel crown elevation) is given by

'r c v = 1 e tan H ( tan

/r

r c )+ (e tan
t

tan H w / r

e tan H / r

r = 0.5 D tan / (1 + tan ) is the ratio of the volume to the where parameter circumference of the prism.

In view of the stabilizing effect of a bentonite slurry, to prevent a seepage flow towards the excavation face, the pressure pb in the slurry must exceed the pore water pressure pw in the soil. The effectiveness of slurry support depends on the infiltration distance of the suspension into the ground e. The suspension will come to a standstill after the penetration reaches a distance emax, owing to its yield strength.
e emax = 1 1 n b w v 1+ w f s 0 k
p p d10 (DIN 4126) = fs0 2 f

emax =

where, n = soil porosity; b = dynamic viscosity of slurry; w = dynamic viscosity of water; v = excavation rate; k = soil permeability with respect to water; p = pb - pw is the excess fluid pressure, a main feature of slurry shield tunnelling; it can be varied by adjusting the air-cushion pressure in a hydro-shield; fs0 is the stagnation pressure gradient which is an experimentally measurable constant; 2 f s 0 = f , with f is the yield strength of the slurry, and d10 is the effective grain size d10 (characteristic grain diameter) of the soil. The membrane-model can be assumed with the formation of an impervious membrane (a seal or a filter cake) on the face. In this case, face instabilities are hardly to occur when a slurry shield is used, even when the ground has extremely low shear strength.
58

But when the slurry penetrates into the ground, the validity extent of the membranemodel will be questioned. With slurry penetration into the ground, face stability assessment will be different. The support force S can be derived by following simple expression:
S e = 1 if e < Dtan S0 2 D tan S D tan = if e > Dtan S0 2e

where S0 denotes the support force of the membrane-model (i.e., at e = 0). The stabilizing effect of the slurry is attributed to the mass forces associated with the pressure gradient inside the suspension saturated ground. Relationship between the pressure gradient fs and the stagnation gradient fs0 is given by

f s = f s 0 + n w

b v w k

The lower the pressure gradient, the lower the safety factor, and face instability occurs when the pressure gradient is lower than a critical value fcr. But of interest is that a slurry shield machine can cause significant ground heave (upward movements), e.g. during the construction of tunnels through very soft silty clay and peat, if the pressure gradient is excessive.

ii) For EPBM, at limit equilibrium, the necessary effective support pressure s' (acting on the tunnel face) is a function of the tunnel diameter D, the overburden H, the piezometric head in the chamber hF, the elevation of water table h0, the shear strength parameters c and , the submerged unit weight of soil ' and dry unit weight t, included the effect of seepage flow:

s ' = f ( D, H , hF , h0 , c, , ' , t , wedge , prism )

with prism = ratio of horizontal stress h to vertical stress v within the prismatic body and wedge = ratio of horizontal to vertical stresses within the wedge The effective stabilizing pressure in the working chamber (s') is in the general form of the limit equilibrium condition:
s ' = F0 ' D F1c + F2 ' h F3c h D

D h0 D 2 s ' = F0 ' D F1c + F2 ' h0 F3c D 2

59

The theoretical minimum face support pressure for tunnelling in dry soil is given by the following equation:

s = F0 t D F1c
where F0, F1, F2, F3 = dimensionless factors from nomograms, as function of H/D and '; h0 is water level, hF is the piezometric head in the chamber, h = h0 - hF is head difference between chamber and ground, and it should be kept as low as possible. If the material in the working chamber is in a fluid state, s' = 0 and solving the equation for h the necessary water pressure for equilibrium is derived. The stability of the tunnel face is guaranteed through the joint effects of the effective stress s' and the pore water pressure p in the working chamber. Stabilization measures will depend on the interplay/compromise of the tunnel geometry, ground improvement, and support of face.

3.4.1.3

Countermeasures to Ground Failure

Ground settlement has the potential to damage overlying buildings and other installations/infrastructure including utilities. This can range from small internal cracks in plaster to effects on the structural integrity of the building (e.g. due to excessive differential settlement/angular distortion), although in most cases there is no discernible effect on the structure itself. The application of the appropriate measures to control and mitigate against the effects of settlement can reduce this impact to acceptable levels. Constructional measures to reduce the risk of ground failure include: Prevention of unbalanced pressure at the face. Ground improvement: Consists of grout injections; jet grouting; and freezing. Grouting operations aim at increasing the strength and stiffness and/or reducing the permeability of the ground. Prepared stations for TBM: These are predetermined stopping locations for EPB or slurry TBM for maintenance purposes in densely urbanized areas on a long stretch and under difficult geotechnical conditions. Grouting for block (boulder) stabilization: To treat loosened blocks between the cutterhead and the face, which may damage the tools on the head and cause overexcavation leading to local instabilities. Real underground structures: Such as forepoling, jet grouting arch, pipes umbrella (pipe jacking of a series of tubes filled up by concrete), or even heavier/more complicated structures to allow an adequate reduction of risk.

In the following only one ground reinforcement technique will be discussed, that is the execution of grouted bodies.

60

3.4.1.4

Grouted Bodies

Generally, TBM drives should avoid the involvement of extra work (ground improvement) as much as possible. But in specific cases, grouted body may form an integral part of EPB and slurry shield drives. Figure 39 shows six possible cases of executing grouted bodies. Case 1 has only a modest grouting in the roof area, since the ground has a sufficient average cohesion but locally mixed with materials possessing no cohesion. This unsystematic consolidated body needs not statical calculation. Cases 2 to 5 indicate the grouted bodies with a well defined shape and size and clearly defined shear strength parameters. The most important and most frequently applied body is presented by case 2 and 3. Case 2 is considered preferable by Kovri, however, Kochen (1992) deemed that geometry in Case 3 is better. Grouted bodies according to the cases 4 and 5 are extremely work intensive and costly. The case 6 shows the prepared stations for the preplanned maintenance work in the TBM's working chamber.

Figure 39: General layout of grouted bodies (Kovri, 2004)

Systematic grouting operations can be executed from inside or outside the TBM as shown in Figure 40.

61

a)

b)

c)

d)

Figure 40: Injection schemes: a) From inside TBM; b) from the surface; c) from an auxiliary adit; d) from a vertical shaft (Kovri, 2004)

Grouted bodies have to achieve the minimum required strength and a satisfactory homogeneity. Figure 41 shows a simplified sketch of virtual arches within the grouted slab (according to Case 2 in Figure 39) used for design considerations. The assumption here is that, the grouted material is stressed to its limit state. The uniaxial strength of the grouted body will govern the state of stress in such virtual arches under a uniformly distributed vertical load q. In failure state (state of limit equilibrium), by assuming that the bending moments at the supports and in the centre of the arch are zero, then there exists the following simple relation between the load, the uniaxial strength of the grouted body and the geometrical parameters:

q* = 8 c
where: q* c = load in limit state

hd b2

= uniaxial strength of grouted body

b, h, d = the width/span, height, and thickness of the virtual arch

62

Figure 41: Statical action of a grouted body above the tunnel, both in transverse and longitudinal directions (Kovri, 2004)

In designing a grouted body, a good relationship between its geometry (shape and size) and its uniaxial strength c (degree of homogeneity) should be achieved. For a given load q and a factor of safety SF it can be beneficial to increase the size of the body in favour of a reduction of the requested uniaxial strength c of the body. The planning of the grid of drill holes and the grouting intervals along the individual drill holes will depend on the shape and size of the grouted body as well as the type of grout and the amount per cubic meter of soil.

In practice, the face support pressure p is often considered together with the cohesion c of the untreated ground and the safety factor FS. The consolidated slab should be properly designed with its thickness h and span . Thanks to the grouted slab in the roof area of the tunnel, the support pressure needed to stabilize the face will consequently be smaller than the case having no grouted body. With the presence of the slab, the weight of the prismatic body in the failure mechanism has been isolated; therefore, support pressure is only required to sustain the wedge part (see Figure 42).

63

ground surface W1 q D = 10 m H = 19 m = 300

H annular grouting

W2

lining

shield

a) without grouted body

consolidated slab

h W2

lining

shield

a) with grouted body

Figure 42: Support pressure in cases of with and without grouted body

The practical meaning here is that, in the case of an EPB machine above the water table and a small value of ground cohesion and without important objects to be underpassed, an open mode of operation can be envisaged. In the case of collapse of the working face, the main purpose of grouted body is to bridge over the resulting void thus avoiding a failure reaching up to the ground surface. But a through-going grouted body also reduces to some extent the risk of failure of the working face. In summary on the discussion about the grouted bodies, following major benefits can be pointed out: reduction of support pressure high safety against collapse reduced ground settlements safety during work in the chamber

64

control of time schedule

It should be recalled that, grouting operations for ground improvement in TBM tunnelling are uncommon and always expensive; therefore this kind of work should be specified in the technical specifications.

3.4.2. Case History: Metro of Turin

GTT and Overview on the Project GTT (Turin Transportation Group) is the company that manages the Torino public transport network. GTT is the concessionary for design, construction and management of the Metro Line 1. The construction of Torino Metro Line 1 is one of the main infrastructures in the public transportation plan for the Torino area. The first section from Collegno (depot) to Porta Nuova includes 9.5 km tunnel and 15 stations (Figure 43).

Figure 43: Alignment of Torino Metro Line 1 (Crova R., 2006)

The civil works design was governed by the VAL (Automated Light Vehicle) system characteristics shown in Figure 44. The train is 2.08 m wide, 52 m long; and its maximum passenger capacity is 440 people (6 pass./m2). Because the train width is 2.08 m, then a single 6.8-metre diameter circular tunnel contains the double track line has been chosen, which was bored by TBM. The tunnel Metro construction works has been divided into Lot 3, Lot 4, and Lot 5. At the lowest point of the longitudinal profile, the tunnel runs at significant depths (approximately - 28 meters). Not to mention the advantages and/or constraints for these deep tunnels, their disadvantages include deeper and more expensive stations, and excavation work below the water table.

65

Figure 44: Rubber wheel trains of Turin Metro (Crova R., 2006)

3.4.2.1 Subsoil Conditions The first section of Line 1 has been completely excavated in the upper part of the fluvial-glacial and fluvial deposits. These deposits present horizontal and vertical discontinuous levels (lens) with different grain size distribution and varying degrees of cementation. Figure 45 shows the grain size distribution obtained from macro-samples (0,5m3) from stations/ventilation shafts excavation. Geo constraints in construction method choices were: - soil: gravel with sand and a high percentage of hard rock blocks and pebbles, from low to very low percentage of silt and clay; - presence of random layers (lens) of loose material (sand and/or gravel); - random degree of cementation; - groundwater table

66

Figure 45: Grain size distribution (Crova R., 2006)

The ground grain size distribution was abnormal and different from the design forecasts, due to the absence of fine materials (less than 2%), and the almost systematic presence of cobbles and boulders in very high percentages. This extreme ground grain distribution did not allow to efficiently operate the TBMs in EPB mode; significant soil conditioning problems were encountered with segregations of the muck in the excavating chamber and consequent operational and face stability problems (Grandori et al, 2003).

3.4.2.2 Shield Machines From those "geo" constraints, all the contractors chose EPB TBMs to carry out excavation beneath the water table under the railway links, for the best rate of construction, and for lower costs compared to traditional methods. However, disadvantage here is that, this is the first use of TBMs in the very coarse Torino soil. The TBMs used in Lots 3 and 4 are new models from LOVAT Inc., the third one used in Lot 5 is a second hand one produced by NFM.

67

The following table gives the main characteristics of the shield machines.

TBM model Tecnical data Excavation diameter Cutter head power Cutter head speed Propulsion thrust Torque Shield length Back-up length

LOVAT RME 306 Series 20600 (Lots 3 and 4) 7,802 mm 2,100 kW Variable: 1 - 2 rpm 76,000 kN 20,400 kNm 10.0 m 98 m

NFM TBM - EPB Mod. 13310506/001 (Lot 5) 8,030 mm 2,000 kW Variable: 0 - 2.4 rpm 91,350 kN 15,000 kNm 9.1m 100 m

Figure 46: TBM LOVAT RME 306 Series 20600 used for Torino Metro Line 1

68

3.4.2.3 Tunnel Lining and Excavation The tunnels were lined with pre-cast 30 cm thick segments in reinforced concrete, connected by bolts and EPDM gaskets to insure water tight conditions. Even with rather small curves and consequent assembly offsets of the segment ring, there is no water passage within the tunnel. Each 1.5 m long ring consists of 6 normal elements plus one key element that enables the closure of the ring, a universal lock that permits to adapt the ring to any kind of radius, from the minimum to the linear one, by a simple rotation of every ring compared to the previous one along the tunnel axis at a given angle (Figure 47). The injection of mortar behind the segments, performed immediately at the beginning of the excavation procedures, ensures the reduction of superficial collapse and the correct confinement/bedding of the lining.

Figure 47: Tunnel lining configuration of Torino Metro Line 1

Tunnel excavation with TBM started in October 2003 from Fermi station (Lot 3). It has been necessary to find a correct balance of consolidating agents within the excavated materials in the excavation chamber, to make the excavated material homogeneous. Because the subsoil of Turin is extremely varied in terms of grain size distributions and mechanical behaviour, soil conditioning was done to better the support pressure at the excavation face; lubricate the muck material in order to facilitate its passage through the screw conveyor; reduce the engine power at the excavation shield; avoid TBM engine overburden

69

A large quantity of conditioning agents (emulsions of water, air and foams with addiction of polymers) has been used. Lot 3 and Lot 4 run completely outside the water table and underneath the central aisle of Francia Street. A light consolidation treatment was carried out in the tunnel crown area, in order to avoid a collapse of the excavation face. The EPBM was operated with a partially filled working chamber, even under low or very low pressure (partial close mode). The TBM of Lot 4 was operated with an empty excavation chamber (open mode). There were no problems with face stability, partly because of the favourable soil condition, which is in average more cemented than the prevision. Lot 5 was executed in more difficult conditions: below the water table; below the old buildings; with very small radius curve; and simultaneous horizontal and vertical curves. Various consolidation techniques were planned for the protection of building foundations and reducing or cancelling the influence of the excavation on the preexisting structures, and to improve the soil mechanical characteristics where necessary. The TBM performance with a negligible pressure applied at the face, with provision of a small slab of grouted soil at the crown, is characterized by an effective average advancement of 10 meters per day. The first Metro section of Lot 4 had been completed in time for the Winter Olympics held in Torino in February 2006.

3.4.2.4 Ground Improvement Soil improvement solutions have been implemented where the assessments indicate potential risk of damage to the pre-existing structures. Such interventions include improving the properties of the ground and mitigating the deforming effects induced by tunnelling by means of low-pressure cement injection grouting. A consolidated slab is created above the tunnel section in order to avoid any localized instability from developing around it. Different grouting geometry have been defined, based on relative position between the tunnel and pre-existing structures, as well as site accessibility and surface site areas use. The project includes full-face cement grouting in the areas adjacent to the stations where the TBM will enter into or exit from the stations: the diaphragm walls in these particular areas will be partially demolished to let the TBMs in and out. In accordance with the environmental conditions, the drilling and grouting operations were done from the surface and/or from in service shafts and tunnels (Figure 48).

70

a)

b)

c) Figure 48: Some schemes of grouting works (Crova, 2006)

It was required that the EPB technique be used to its maximum capacity. In addition to the injection of bentonite, a new plant was also set up for the preparation and injection of heavier substances obtained by mixing calcareous filler mixtures with water, to integrate the ground conditioning with foams by way of optimized FIR (foam injection rate/ratio) and FER (foam expansion ratio). All the critical points and most of the obstacles were overcome successfully; the excavation was carried out always in close mode (pressure at the top of the tunnel around 2030 kPa). A systematic consolidation of a slab over the tunnel section was decided to be associated with EPB 02 operating mode (i.e., the pressure in the top of the excavation chamber was kept between 0.1 and 0.3 bar, the penetration rate up to 50 mm/rev, and the quantity of conditioning agents FIR 30%). Professor Kovari as expert of the Contractor then determined two type of slabs (Grandori et al, 2003) for Lot 3:
71

Light slab: non structural (too lightly consolidated and too small to be able to form a structure that can be taken into account in a statical calculation) having just the function of increasing the ground cohesion in the most critical crown area of the tunnel. The depth from the surface to the tunnel axis is approx. 16 m in average (Figure 49).

Figure 49: Light grouted slab on top of tunnel o Structural slab: having sufficient dimension and subject to heavier consolidation in order to assure a structural behaviour to be taken into account in the static verification of the tunnel stability (Figure 50).

Figure 50: Structural grouted slab on top of tunnel

72

The light slab associated to EPB 02 operating mode systematically all along the metro tunnel drive. The structural slab associated to the EPB 02 operating mode in the most critical area nearby important buildings and under important subsurface and surface services. The continuous control of the weight of the excavated muck furthermore allowed to detecting in real time few local small over-excavations that were immediately filled by additional grouting. For the purpose of completeness, we also like to mention the execution of special grouted soil structural bodies to protect the main buildings during the construction of Lot 5 of the Turin Metro Line 1 tunnel. Grouting design had to be extremely accurate since it had to take into consideration various aspects such as: o o o o the presence of underground lines not to be hit by drillings works; the presence of buildings, structures, underground constructions that were limiting the drilling possibility; to limit the disturbance caused by the drilling and grouting operations on surface; the presence and level of the underground water table.

Four different shapes of injected body were finally designed as in Figure 51. The shape (51.a) is the treatment scheme for the top part only for the passage of the TBM. The shape (51.b) is the treatment scheme on the crown and at both sides up to the springline for driving under street. The shape (51. c) is the grouting scheme on the crown and at both sides below the springline for building areas. The shape (51.d) is the treatment scheme completely around the tunnel for the maintenance of TBM. The first three schemes are the intervention of preventive consolidation by means of cement injection. The last scheme shows preventive consolidation and waterproofing by means of cement and silicate injection. All these shapes were designed to form tridimensional structural bodies to avoid that any over-excavation and/or instability occurred at tunnel level could be transmitted to the surface buildings.

73

Figure 51: Different grouting bodies schemes at Lot 5 of Turin Metro Line 1 (Concilia, 2005)

During grouting execution all injection parameters were recorded to increase the knowledge and behaviour of soil and to detect the local presence of fine lenses in case of low absorption of injected fluids. Quality of injected mixture was controlled during operations both in terms of density and of compressive strength, in order to compare with the designed values ( = 1.25 t/m3 and f = 0.10 - 0.12 MPa).

74

Chapter 4

4. Interface Between TBM and Lining

4.1 Types of Linings

4.1.1 General

Lining is a structure to secure a tunnel space by withstanding the earth and water pressures. Lining consists of primary and secondary one. The lining is in general a ring structure assembled with prefabricated segments, but is built in some cases with castin-place concrete. The secondary lining may be placed mainly with cast-in-place concrete if required. Segment is lining material for the shield tunneling. In general, it is made of reinforced concrete or of steel. Several segments are assembled to form a circle, multi-circle or other shape.

Tunnel lining should be designed and constructed to high standards across the project. This will cover issues including, how intermediate construction stages need to be taken into account, settlement mitigation measures and the impact of backfill grouting. It should satisfy the principles and requirements for safety, serviceability and durability. The following types of lining need to be distinguished: pipe lining (pipe jacking) in situ lining segmental lining

With pipe jacking, behind the shield and guided by means of a pilot head, prefabricated pipes, which are pushed forward from a launch shaft, constitute the final lining. With machine drives, excavation and installation of the pipe are not normally concurrently operations. The lining is constituted of individual pipes inserted in the

75

launch shaft, and the ever growing string of pipes is jacked forward by means of push cylinders. Without a tailskin and an erector, the shield is very compact (Figure 52).

Figure 52: Pipe jacking diagram. Typical job sites (Herrenknecht)

In situ linings do not require pre-fabrication. They are cast right in the shield area. The most modern and promising variant is extruded concrete, where the concrete is applied under pressure. Behind the shield the tunnel lining is concreted in rings behind a shuttering (the lining is slipcast behind a sliding form). To advance, the shield can push against the hardened concrete section, which must have reached the required strength (Figure 53).

Figure 53: Hydroshield with telescopic trailing shield for concrete extrusion, Mtro de Lyon, 1993 (Maidl et al, 1996)

76

TBM tunnels can also utilize sprayed concrete for initial (with steel fibres) and final lining (without reinforcement). Integrated concepts for TBMs and mechanized excavation systems has been shown by Melbye (2005), where design and build are based on modular concepts to meet most sprayed concrete requirements and specifications in tunnelling and shaft sinking. The example of a TBM tunnel with a full, circular sprayed concrete lining is given in Figure 54 & 55. In this case, the arch effect can be calculated and the bond strength is no longer a factor.

Figure 54: The example of a TBM tunnel with a full, circular sprayed concrete lining (Melbye, 2005)

Figure 55: Robot spraying manipulator is integrated into a large diameter TBM (Melbye, 2005)

With segmental lining, which is the focus of this report, rings made from a number of segments are installed within the protection of the shield tail. The lining segments are pre-cast and transported to the place where they will be positioned.

Following are the segment and ring nomenclature (Figure 56 & 57).

77

Figure 56: Ring nomenclature

Figure 57: Segment nomenclature

Segments and rings can have various types: segments can take one of the rectangular, trapezoidal, or honeycomb forms (see Figure 58); rings can take one of the straight, right/left, or universal forms. Segments are connected together by some types of joint design (conservative with tongue and groove or advanced with plain surface) and joint connector (bolt or dowel) (Figure 59).

78

Figure 58: Segment shapes

Figure 59: Circumferential and radial joint connectors. Plabutsch Double Lane Highway Tunnel in Austria (1998) (Wagner, 2006)

Activities that affect one single segment include pre-cast process, handling, storage, and assembling to constitute a ring. Activities that affect on the ring include TBM drive / tunnel excavation, tail void injections, and ground loads. Many kinds of materials are used to constitute the segmental lining: concrete, steel, steel fiber, connectors-bolts (plastic and/or steel made), gaskets for the waterproofing (elastomeric neoprene rubber and/or hydro swelling material). For compressive EPDM-sealing gaskets, it is better to use a wide and flat than a narrow and high gasket. In designing the segmental lining, one can use different Norms and Standards, such as European standards or local standards of the specific country (ACI - American Concrete Institute, DIN - German Institute for Standardization, BS - British Standards, SNiP - Russian Construction Norms and Regulations, etc.). Generally, just a right tune of the various coefficients is adequate to meet all the requirements in different locations, since the basic materials are well known (concrete and steel). However,

79

specific attention must be paid to the new generation of the well known materials, i.e. high performance concrete, and steel fiber reinforced concrete.

4.1.2 Reinforced Concrete Linings

Reinforced concrete segments are by far the most commonly used. The segmental reinforced concrete lining is used to satisfy many construction and/or environmental requirements, for instance: the need for an immediate support (mainly for an excavation in an instable ground); the need to control carefully the ground movements induced by the tunnel excavation; to avoid the drainage of the groundwater and therefore to build a waterproof tunnel; provide the counterbalance for the TBM advance; to avoid the installation of a secondary lining.

With respect to a segmental lining installed in the rear of a single or double shielded TBM with or without the face pressure control (Figure 60), the essential issues are to make choice of it, to pre-dimension it, and to perform the static verifications for it. The basic elements that must be known include: a) the geometry (general and detailed) of the chosen segmental lining; b) the actions undertaken by the single segment and by the full ring all along their life; c) the characteristics of the materials constituting the various parts; and d) the norms and standards to be applied.

Figure 60: Area near to the face where tunnelling operations with the use of a TBM occur

80

Geometry and relevant tolerances of a lining segment are given in the following Table 6 and Figure 61.

Table 6: The segment tolerances - limit deviation Description Width: Thickness: Length of an arc of circumferences: Radius of intrados: Diagonal of the segment: Position of the holes for the connections: Position of the hole for the erector: Depth of the groove for the gasket: Planarity of the faces which meet other elements: Tolerances 1.0mm -0/+3.0mm 1.0mm 2.0mm 1.0mm 0.5mm 2.0mm -0.5+0.1mm 1.0mm

Figure 61: Geometry and relevant tolerances of a lining segment

Nevertheless, the most important principle must be to keep the assembly tolerances as generous as possible so that they can be maintained on site, bringing about the ring erection personnel's work as easy as possible.

81

4.1.3 Steel Fiber Reinforced Linings

Using Steel Fiber Reinforced Concrete (SFRC) for precast tunnel segments has several advantages, because of the following (Plizzari & Tiberti, 2006): SFRC is a tough material Smaller crack width (durability) High resistance to impact loading (as it may occur during transportation or the segment placement) Better control of possible detachments of concrete cover Industrialization of the production process Smaller area for stockpiling conventional reinforcement (rebars or wire mesh) Steel fiber reinforcement is present in the cover, narrow corners or holes where conventional reinforcement can hardly be placed

However, attention needs still be given to the SFR lining subject to loading conditions. Loading conditions range from the transient phases to the serviceability phase. Transient phases include storage, transportation and placement of segments, and thrust phase. Under serviceability phase (i.e. during the expected life of the tunnel), the lining sustains ground and water action. During pushing of the hydraulic jacks, an optimized reinforcement should include rebars (for localized stresses under the jack) and fibers (for diffused stresses in the segment), so that bending cracks and concrete spalling will not occur. Because of the FRC toughness, the load increment on the segment could be achieved; in other words, the FRC toughness allows for an increase of the segment bearing capacity.

4.2 Lining Design Procedure

There exist various competent methods of designing shield tunnel linings, and this report does not give priority to any one method. Today, the design and dimensioning of a reinforced concrete segmental ring are still carried out under consideration of its ultimate state. Limit states analysis allows checking of both the structure's factor of safety with respect to failure and its satisfactory behaviour with respect to serviceability. The main characteristics of the two limit states are recalled in the following table.

82

Ultimate limit states

Serviceability limit states

Failure of a section due to crushing of concrete Excessive deformation of steel Instability of shape (buckling, bulging) Loss of static equilibrium at ring erection

Excessive opening of cracks (infiltration, corrosion) Excessive compression of concrete causing microcracking Excessive ring deformation

However, this subsection only introduce a few basic concepts of TBM lining design, so as to help understand better the TBM/lining interface which will be presented later.

4.2.1 Design Steps

Generally, design steps for TBM tunnels could be as follows (ITA, 2000): Step 1: Define geometric parameters Alignment, excavation diameter, lining diameter, lining thickness, width of ring, segment system, joint connections Step 2: Determine geotechnical data Shear strength of soil, deformation modulus, earth pressure coefficient Step 3: Select critical sections Influence of overburden, surcharge, groundwater, adjacent structures Step 4: Determine mechanical data of TBM Confinement pressure, overcut, shield tail conicality, TBM length, total thrust pressure, number of thrusts, number of pads, pad dimensions, grouting pressure, space for installation. All these structural parameters associated with TBM characteristics may have potential impact on ring stress analysis. Step 5: Define material properties Concrete: compressive strength, modulus of elasticity Reinforcement: type, tensile strength Gasket: type, dimensions, allowable gap, elastic capacity Step 6: Design loads Soil pressure, water pressure, construction loads etc. Step 7: Design models Empirical model, analytical model, numerical model Step 8: Computational results
83

Response: axial force, moment, shear Deformation: deflection Detailing: reinforcement, joints, groove

The design of a shield tunnel lining often follows the planning works, according to a sequence as shown in Figure 62.

Planning of Tunnel Project

Alignment Plan/Profile Cross Section

Survey/Geology

Function/Capacity to be given to Tunnel

Specification/Code/Standard to be used Inner Diameter Load Condition

Assumption of Lining Condition (Thickness, etc.)

Model to Compute Member Forces

Computation of Member Forces

Safety Checking for Lining

NO Safe and Economical

YES NO Approval YES Execution of Construction Works

Figure 62: Flow chart of shield tunnel lining design (ITA, 2000)

84

4.2.2 Loading Conditions The tunnel lining behind the TBM must be capable of withstanding all loads/actions and combined actions without deforming, especially during ring erection and advance. Single-shell reinforced concrete segmental rings behind the TBM, can be designed to fulfill those demands. Secondary lining can also be constructed with cast-in-place concrete as a structural member of the segmental lining. There are many loading cases for the segmental lining of tunnels driven by TBMs. This part provides information on some actions that should normally be considered in the design and construction of tunnel lining. The following loads shall normally be considered in designing the lining of the shield tunnel (JSCE, 1996): (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) Vertical and horizontal earth pressure Water pressure Dead weight Effects of surcharge Soil reaction Internal loads Construction loads Effects of earthquakes Effects of two or more shield tunnels construction Effects of working in the vicinity Effects of ground subsidence Others

Various combinations of the loads can be considered according to the purpose of the tunnel usage. Table 7 gives a classification of these loads from the design point of view. Table 7: Classification of the loads for shield tunnelling Primary loads
1. 2. 3. 4. 5. 6. 7. 8. 9.

Secondary loads

Vertical and horizontal loads Water pressure Dead weight Effects of surcharge Soil reaction Internal loads Construction loads Effects of earthquakes

Special loads

Effects of two or more shield tunnels construction 10. Effects of working in the vicinity 11. Effects of ground subsidence 12. Others

85

Figure 63 below shows the excavation process of TBM in order to analyze the necessary loading cases. Cross section 0 is far away from the face, the initial state of stress in ground is not affected. Cross-section 1 is right at the face, where TBM cutting wheel is in interaction with the ground. Cross section 2 shows the loading condition due to convergence of the ground on the TBM shield. In the cross section 3, lining is subjected to the grouting loading Pg. Cross section 4 is within the part of hardened grout, the lining is subjected to permanent equilibrium loading Peq.

Figure 63: Excavation process of TBM

Concerning the segment loading, the steps which involve a specific verification for the single segment are: handling in the TBM; assembling to built-up the ring; thrust for the advance; longitudinal injections (primary grouting); and radial injection (secondary grouting) - but normally no more used nowadays. Figure 64 gives the notations in the usual calculation method and the modified usual calculation method (Effective stress method).

86

Figure 64: Notations in the usual calculation method and the modified usual calculation method (JSCE, 1996)

The notations used for the structural calculation of lining are defined as follows: Bending moment (N), axial force (N) and shear force (Q) (for member forces, the directions indicated in Fig. 65 are assumed to be positive).

Figure 65: Notations of bending moment, axial force and shear force

The temporary loads during segment transport, ring assembly and TBM advance with grouting pressure in many cases are more important than the final loads from earth and water pressure. Therefore, the following different groups of loadings may have to be carefully considered.

87

4.2.2.1 Geostatical Loads This load case analyses the load effects on lining segments and ground. An example is shown in Figure 66.

Figure 66: Example load cases

Dead weight is a load in the vertical direction, distributed along the centroid of lining, and is included in this geostatical load case. Dead weight of the primary lining shall be calculated by the following equation.
g1 = W1 2 .Rc

where, g1 = Dead weight in unit length of the primary lining exerting along the centroid of lining (per unit length in longitudinal direction); W1 = Dead weight of the primary lining (per unit length in longitudinal direction); Rc = Radius of centroid of the primary lining.

88

The water load is set according to the associated water levels with consideration of a density of w = 10 kN/m3. If the problems of geotechnical engineering (stability and settlements) are combined with groundwater/pore pressure considerations, it would be rather difficult to deal with, as it increases the complexity considerably.

4.2.2.2 Thrust Jacking Loading The functions of the linings during tunnel construction are to sustain jack thrust for advancing a shield machine and to withstand the back-fill grouting pressure. The linings have also the function as a tunnel lining structure immediately after the shield is advanced. Thrust force of shield jacks (Figure 67) is a temporary load which acts on the segment as a reaction force against it while advancement the shield machine and is the most influential load to the segment among the construction loads. Several verifications must be done for the jacking load effects on the segment, such as contact pressure, bursting forces in the radial direction, and bursting forces in the circumferential direction.

Figure 67: Thruster pads distribution and spalling and cracking of concrete cover

Segments can be analyzed using FEM-method principles organized, for example, in the Structural Analysis Program SAP2000 (Computers and Structures, Inc. - CSI, 2004) to investigate the behaviour of the lining following linear elastic Hooke law. Modelling can be done with barrel shell elements. Boundary conditions of model are restrained on side of segment where it has contact with already installed ring. Loads from thrust jacking shoes are distributed over the length of thrust shoe (Figure 68).

89

Figure 68: Shell element generation (SAP2000 package)

With regard to the BEG tunnel, thrust jacking forces could be redistributed to the lining over thrust shoes contact area. Different extents of assumed forces are presented in Appendix 3. Two maximal pressures (10,345 tons and 9,052 tons) would be used as basis for loading analysis and further on for dimensioning of segments. Thrust jacking loads are the design criteria to define necessary reinforcement for the lining.

4.2.2.3 Trailer Loading Trailer chassis and other service loads can be applied on lining, including main bearing loads, divided by number of wheels (Figure 69). The loads induced by the trailer and by any fixations in the segments normally do not influence the reinforcement. During discussions with TBM manufacturer, it is necessary to state whether "Main Bearing Load" will be included in this type of analysis or not.

Figure 69: Trailer load distribution

Considering the BEG Lot H3-4, for instance, if after excavation of the tube the TBM will be dismantled and transported back through tunnel, then "Main Bearing Load"
90

could be assumed e.g. as approx. 165 tons. Also, additional assumptions may have to be defined as follows: a) "Main Bearing" forces are acting on same distance as trailer load wheels (3500 mm) b) 165 tons will be divided over 4 wheels (41.1 tons/wheel); additionally a horizontal load which is on the safe side is set of 15%. c) Not more than 1 wheel is acting on ring width (2.0 m), that means along the tunnel wagon wheels are on distance that is more than 2.0 m. As a result, "Trailer Wheel Loads" might be taken as unfavourable one and as criteria for definition of reinforcement (41 tons/wheel/ring width) regarding service load conditions.

4.2.2.4 Grouting Loads Primary grouting pressure applied to fill up the tail void behind the TBM is believed to govern both deformations and internal lining forces, as well as affect surface settlements. The grouting pressure acting on the outer surface (extrados) when the ring leaves the shield. For normal conditions, when a highly flowable mortar is used, the grouting pressure can be calculated constant around the ring. The annular grouting of the ring, with a grouting pressure minimum one bar (1 bar) higher than the surrounding water pressure, prestresses the ring and the enclosing ground. Secondary grouting pressure is an extending regular grout pressure. These transienttype loads result from a localized increase in grouting pressure ("local pumping thrust") directly behind the segment grouting holes (Figure 70).

Figure 70: Secondary grouting pressure

91

As an example, the influence of gap grouting can be analyzed with accidental extent of grouting pressure of 4 bars (400 kN/m2 or 0.4 MPa). In cross-section load will be distributed in the form of triangular plane, acting in the range of 0 - 400 kN/m wide strip of ring. Value of loading is changing linearly from zero (0) to maximal pressure and back to zero (0) bar in range of middle angle of 45 degrees. In longitudinal section the extent of loading is continuous and has extent of 4 bars, or is acting also as a type of triangular loading (used in two-, three-dimensional model). Influences from the geostatical and backfill grouting load cases need to be combined to give adverse resultant moment and shear forces.

4.2.2.5 Storage Loads After mould stripping, segments are set down and stacked on supports. Timber blocks are usually placed between segments taking care that they are aligned with the supports (Figure 71). Storage and handling (e.g. turning, packing and then loading-out operations, supply to the workface) influence the bending moment.

Figure 71: Loads for storage of segments

As an example, two situations could be considered for the segments self weight on stock of the BEG Lot H3-4 tunnel: a) Storage of segments as simple beam: moment is calculated by Ma = 0.125 x x b x d x l2 = 0.125 x 25 x 1.00 x 0.50 x 3.52 = 19.14 kNm/m width b) Erection of segment - middle support: moment is calculated by Mb = 0.5 x x b x d x l2 = 0.5 x 25 x 1.00 x 0.50 x 2.72 = 45.56 kNm/m width

Because Mb > Ma, therefore Mb will be taken into account as maximal influence during this loading case.

92

4.2.2.6 Erection Loads During erection, the lining is subjected to a number of loads such as: compressive (possibly eccentric) loads from the longitudinal thrust of the TBM; shear forces due to differential deformations between adjacent rings; forces resulting from segments overhanging during ring assembly; possible bumping impact loads; loads applied by the assembly systems retained (bolts, anchor bolts or plugs). i) Eccentric Loads during Shield Drive + Geometric Eccentricity of Thrust Shoe Action For BEG project tunnel, loadings equal to maximal possible thrust jacking pressure over ring circumference of 10345 tons may have to be applied on segments. Thrust pressures are defined in same way as for case of thrust jacking loads. Eccentricity varies between 0 / + 5 cm toward tunnel center and 0 / -2 cm outward tunnel center and applied on segment edge as distributed moment over thrust pressure length. + Variable Thrust Jacking Pressure The same loadings as for the case of regular Thrust Jacking Loads are to be applied on the segment edge but the level of forces as well as the location of application is varied (see Fig. "Eccentric Loads during Shield Drive"). ii) Eccentric Loads During Installation Loading of 15 tons per only one thrust shoe can be applied on the segment edge. This load is equal to force necessary to properly install segment with plastic dowels (plastic dowel insertion force).

4.2.2.7 Fire Loads Additionally, when fireproof concrete is used, an important aspect for the construction phase is that each segmental ring must be provided with adequate fire protection immediately following assembly. According to Haack (2003), the considerable effects of the machine fire during the construction of the rail tunnel below the Great Belt in Denmark would not have occurred, namely causing splintering up to 27 cm deep given 40 cm thick segments, if such resistant segment concrete had existed at the time. However, loading of tunnel fire in construction is not further considered in this report.

4.2.2.8 Other Loads Evaluation of the surrounding ground dynamic characteristics may be necessary in high seismic risk areas. Temperature influence in the final state is avoided by the shielding effect of the fire protection lining layer. However, in the case of certain structures (very deep tunnels, energy conveyance tunnels, etc.), climatic temperature-induced actions (such as uniform temperature variations and temperature gradients) must be considered.

93

Concrete shrinkage is neglected. Other unusual load cases/accidental actions such as impact by railway vehicles, explosions, fire loads, disaster load case of flooding inside the tunnel, buoyancy or "waterhammer" etc. will also not be dealt with further.

4.3 Concept of Interface

In a specific tunnelling project, the main Client requirements are related to asset performance, asset management, capital delivery (cost & time), and environmentally sustainable. To fulfill these requirements, a strategy (to set up a proper organization) is needed, that is to clearly identify the disciplines to be dealt with, the specific studies to be developed, the interfaces, the sources of risk. By and large, the project interfaces can be shown as in Figure 63, which are characterized by the coordination of technical and functional design, aiming at minimum cost and reliable time schedule.

Equipment

Exploitation and Maintenance

Environment

Project Interfaces

Civil Works

Safety and Ventilation

Geology and Hydrogeology

Figure 72: Identification of project interfaces (after Grasso, 2006)

More specifically, the TBM/lining interface implies both contractual and physical meanings and these two fields have certain extent of overlapping. The same is the strict linking between the tunnel design and the tunnel construction. This reflects the statement by ITA (2003): Even the best TMBs cannot function efficiently if they are not properly guided and the tunnel is not properly designed and managed.

94

4.3.1 Contractual Interface

Within the contractual discussions we can talk about several aspects. Those are the interacting activities between the Parties to improve the execution of a TBM tunnel project in general, to optimize the segmental lining, and to improve the interface between TBM and lining in particular.

4.3.1.1 General Aspects Success for mechanised excavation depends upon correct planning and accurate activities monitoring. Following main issues should be regarded in consideration of the general essential interfaces (Grasso, 2006): the support of a strong political will; the public understanding; the financial pressures; the design of construction of logistics; the contract management; the risk sharing and management; the durability and serviceability; the technical feasibility the constructability

By way of illustration, tunnelling in urban environment could be mentioned, which involves the following key elements (Chiriotti, 2006): adequate excavation method consciousness of risks (all the Parties) strategy for mitigating risks competent personnel (all the Parties) organization & responsibilities (all the Parties) development and implementation of ad hoc procedures identification of key indicators and implementation of consistent controls (monitoring)

As we can see from the above list, there are three fields that represent a close interaction between all the Parties involved to enhance the general quality of the project (i.e. consciousness of risks, competent personnel, and organization & responsibilities).
95

Regarding the consciousness of risks, a common understanding will help to reduce the sources of initial risks in urban tunnelling, including: geological risk design risk construction risk financial risk, and operation risk.

Going further into some details of consciousness risks, design risk can be due to: lack of experience of the Designer, uncompleted prediction of risk scenarios insufficient definition of countermeasures non-constructability of proposed solutions design flexibility vs. actual ground conditions loading conditions of the lining definition of TBMs operational parameters inadequate monitoring controls inadequate threshold limits And construction risk can originate from: learning curve (a period at the beginning of every job for the tunnelling crew to develop into a team and learn the idiosyncrasies of the new job) incompatibility of the machine with the ground major mechanical failures inadequate logistics lack of Contractors experience lack of personnel training lack of TBMs parameter controls lack of TBMs parameter review insufficient probing ahead inadequate procedures

The construction risk is illustrated by considering this case: When tunnelling with EPB-TBMs in a non-conventional medium (i.e., no uniform granulometry, no uniform density and no uniform groundwater head), the possible failure mechanisms may
96

involve the global failure, local failure, piping, and progressive failure. Ground and surface monitoring are effective provided that they are integrated by TBM and construction data to allow a proper interpretation. Collapses may happen if: TBM data are just accessible to the Contractor The Designer is just receiving the in-ground and surface monitoring results No effective back-analysis is in place because of difficulties in accessing data

With regard to the lining of EPB-TBM tunnelling in such a non-conventional medium, one can see that, the lining should be quality controlled by a checklist of events as follows: water leakage steps between segments lips/offsets between rings cracking after installation installation of defective segments/rings connectors, and erector failure

Regarding another key element on organization and responsibilities, a standard TBM lining implementation plan must be prepared which details how the design and construction of TBM lining is to be managed. This will be adopted by all TBM lining designers and contractors.

4.3.1.2 Segmental Lining Optimization Segmental lining systems offer a number of opportunities for optimization in detail within any phase of design. Because of the huge number of segments to be produced and because of the multiplying effect of any measure carried out, any optimization becomes rather effective. Within the tender phase the fundamental features of a lining system are to be established. But within the detailed design phase for construction, when the excavation system together with the segment production and the site logistics are established, a number of opportunities to optimize the segments in detail become due again (Walter et al 2005). The segment geometry, the geometry of the lining ring, and the application of the reinforcement can be taken into the optimizing consideration. Furthermore, segments can also be modified considering an integrated invert track in the railway tunnels in order to optimize the construction logistics. The fundamental basis to enable this process is a proper contract which manifests the design key figures at the one hand but also enables optimization in detail at the other hand. And a proper collaboration between the client, the designer and the contractor to serve for the continuity of the project philosophy and to serve for the input of the construction method is as essential.
97

During all steps of design, the dialogue and the contractual basis between the people and parties involved should only manifest the main basics in view of the final result, and leave open the opportunity to optimize the segments in terms of the best technical, practical and economical results (see Table 8). In the preliminary design phase, the main decisions are taken to establish the fundamental tunnel system design, concerning the following questions: Tunnel system consisting of two single or one double track? Rescue system consisting of cross passages between two tubes or separate rescue tunnels? Lining system consisting of mono-shell or double shell, shell sealed or unsealed? Fire safety is secured by separated or integrated fire protection? What kind of dewatering and ventilation system? Use of conventional tunnelling or TBM, or both? And, segments or not? In the tender design phase, the one system which is actually tendered has to be designed further into detail. This design includes: specification of the external loads and load combinations to be considered; dimensioning of a suitable segment system to derive a typical section including all details concerning the intermediate and final load bearing system; definition of all salient lining features concerning the sealing system, dewatering, fire protection, niches, cross passage interfaces; bill of quantities (BOQ) as a basis for tendering.

Table 8: Options of segmental lining optimization in relation to the individual design phases (Walter et al 2005) Design phases
Preliminary design Purpose of tunnel, construction method Tender design Tunnel diameter, lining system, ground, related requirements Best practice solutions, final dimensioning Detailed design for construction Typical section, lining system, quality requirements, loads and load combination Segment geometry, installation related features, reinforcement design, logistics related features

Basics

Options for optimization

Tunnel diameter, lining system, preliminary dimensioning

In the phase of detailed design for construction, the contractor shall be allowed to consider the following: i) Final segment geometry, including segment width (ring width) and segment length (ring intersection), taking into account all belongings of handling, manipulation and installation; ii) Logistics related features concerning the backup- and transport system as well as the construction method and construction sequence; iii) Reinforcement detailed design which must cover the loadings defined within the tender and the loadings related to production, transport, installation and TBM drive, but still suit the best practice of reinforcement manufacturing and of reinforcement cage suitability.
98

So, it is the contract that should manifest completely the clients requirements but should not restrict the contractors opportunity to fulfill such requirements in the most economical way and in a way which is most suitable to the TBM and backup system available.

Structural concept on segment/lining Statically speaking, the shell segmental lining should be dimensioned to maintain for the required support during excavation and over the entire lifetime load bearing. Kinematically, it should be designed to cope with all handling loads during each stage of segments production, manipulation, storage, transport and installation. The statically concept and dimensioning of the segment lining system involve the following tasks (generic segment design): ground characteristics; segment partition (ring division); segment weight; outer and inner convergence of the lining; joint connection (tongue and groove design, or advanced design of plain joint, connectors, gaskets); minimum segment thickness; segment manufacturing (finishing, chamfer in the corners, allowable tolerances on lining and on moulds, ensured by QA/QC measures - i.e. quality assurance and quality control, etc.); external loads and load combinations; non-cracked segments installation precision; thrust forces undertaking. during installation (uncracked concrete);

Looking at the relation of the lining and high flexible (universal) machines, it is worthwhile to refer to the universal tapered ring (unified ring, Figure 73). Each project is specific, but as a general rule one can say that unified ring can be used on a systematic basis for every alignment/situation. The tapers of one ring compensate for those of another ring, thereby cancelling out the overall tapering effect. In regard to kinematic control of the segments, unified ring makes it easier to install owing to its longitudinal and ring joints. The ovalisation tolerances of 1/1000 1/2000 can be ensured.

99

Figure 73: Unified ring types (Wagner, 2006)

The unified ring is usually associated with the single gasket system for waterproofing. Here single gasket system is preferable to the double system, since the later has strong compression characteristics. For example, the double sealing gaskets used for the Elbe Tunnel (Germany) did not show real advantages and the costs are very high.

Reinforcement design The reinforcement design can be completely open for the contractor to be established within the detailed design phase. In view of the huge number of the segments (e.g. about several thousands of segment rings, and dozen thousands of pre-cast segments and steel cages), a significant potential to reach an optimum between minimum overall reinforcement weight, efficient reinforcement cage production, and sufficient overall reinforcement cage quality is really worth to be taken full advantage of. Three main components of the cages need to be kept an eye on (not to mention the specific local reinforcement such as additional bars, stirrups, and helix elements): principal inner and outer field reinforcement; tensile splitting reinforcement in the longitudinal joints; tensile splitting reinforcement in the radial joints.

Potential for geometry optimization For tunnel construction, the effort of optimizations by designer and contractor in proper and innovative cooperation with the client is an open end matter. For that reason, the optimization process has to be followed up in all design phases. Within a specific railway tunnel project, for instance, apart from above improvements three other considerations for geometry modification can be mentioned:
100

intersection of the segment ring (span width of the roof segment); thickness reduction of the invert segment; and integrated track in the invert segment.

As a final aim, optimization process must successfully realize the benefits in terms of the segments geometry, the segment and reinforcement production and the best site logistics.

Remarks It is well recognized, that if the cost of construction and operation are increased and the rate of return on investment is inevitably reduced, then the viability of the tunnelling project will be threatened. Following statement by Wagner (2006) is quoted as concluding remarks for the said contractual interface: It has been proven that, the length of a tunnel driven by TBM normally should exceed approx. 2,000 m for the sake of economy. The TBM designer and the segment designer should join their effort to fix the common concepts before the start of design work. Both designers may need to show minimum 10 years of relevant experience. The Client should specify generic design of TBM and lining; and responsibilities should be specified at interface of TBM and lining.

4.3.2 Physical Interface

4.3.2.1

General

TBM/ground interaction Interaction between TBM and ground is first briefed, because the type of excavation method and support that is used dictates the behaviour of the ground. Upon excavation of the face, the need for self-bearing time and the stability of unsupported span (free span) are of our concerns. Concerning this, a general rule to remember is that, "the tunnel is built at the face". The concept of unsupported span and self-bearing time are valid both for conventional and full face mechanized excavations. In Figure 74, the free span coincides with the last stretch of the tunnel where the rock is being excavated and the supports cannot be installed yet (very variable length, depending on the TBM type: open, shielded, with or without face counter-pressure). The advance is performed in two steps: Active stroke of the cutting head for the excavation (and the spoil extraction) usually 1.4 1.8 (2.0) m; Advance of the whole machine when the cutting head has been stopped.

101

Figure 74: Free span in full face mechanized tunnelling (continuous rock excavation and spoil/muck extraction) (Pelizza, 2006)

Apart from the unsupported span, other concerns include the disturbance of the ground ahead of the face due to the cutting wheel rotation, and the shear failure of the ground along the TBM skin due to friction when the machine is advancing.

Lining/ground interaction As it is well known, the lining of the TBM tunnel has a number of functions, which put varying demands on the form and the material of the lining. The following two requirements need to be fulfilled with very little deformation over the whole working life of the tunnel: securing the inside of the tunnel against the surrounding ground and watertightness in either direction taking up permanent or mobile loads resulting from installation and traffic.

The main parameters influencing behaviour of the tunnel lining in contact with the ground (soil-structure interaction) are: lining/back grouting material and ground/back grouting material contact conditions; environment: nearness of existing or planned structures underground or at the surface; and superimposed loads.

Lining/TBM interaction The segmental ring is erected in the shielded TBM and during the advance, the rams act on the ring; therefore, the ring never can be seen independent from the TBM (Gruebl, 2006). The interaction between machine, thruster configuration and geometry
102

of the segments will influence the kinematics of erector and segments. The design of the TBM and the segmental ring must be harmonized in terms of: Rams must act on prepared sections of the ring (in correspondence with the longitudinal joints and in regions where special tensile splitting reinforcement is placed). Rolling of the tunnel shield and the ring must be taken into account. The ram axis should be identical with the ring axis. The rams axis radius may be slightly smaller (<10% of the segment thickness) than the middle ring radius. This gives the tendency to close the ring during advance. The segmental rings must be able to follow the TBM. The ring taper should be designed according to the TBM curve drive capabilities and not only according to the designed tunnel axis. The tail gap should be more than 30 mm to avoid ring constrains in the region of the tail sealing. No steel sheets or timbers should be inserted in the tail of the tunnel shield. The new ring should be erected free on the last erected ring

It is also necessary to list the following general factors which play an important role for a successful TBM advance: Correctly working TBM with all devices and aggregates Guidance system, showing the correct position in correspondence with the theoretical alignment/the designed tunnel axis (DTA) Segmental ring, interacting with the TBM, the structural requirements and the DTA Well organized infrastructure (supply and disposal) Tunnelling team, knowing and controlling the equipment Tunnel management and supervision, assisted by DAT (Decision Aids in Tunnelling)

In the following paragraphs, detailed discussions will be given on: machine operation, TBM guidance system, ring building, backfill grouting, back-up system, and lining instrumentation. The purpose is to highlight the TBM/lining physical interface in every process/operation.

4.3.2.2

Machine Operation

As already mentioned, important devices of the TBM are: Excavation system and excavation tools Systems for face stabilization according to ground conditions Transport devices for excavated ground

103

Steering equipment (rams, articulation cylinders) Ring erector Grouting system

The experience and technical skills of tunnelling machine operators as well as the availability and use of experienced foremen are important factors in the reduction of risks. If the ground quality is insufficient, the subsoil must be grouted, vibrator compression, injections or freezing to adapt it to the characteristics of the selected TBM. TBM mechanical excavation is rather continuous, generates less dust, noise or vibrations, and provides superior protection. The profile accuracy of the cavity cross section is particularly high provided the TBM is driven within its operating tolerances. Problems concerning advance control include: Inaccurate survey of the TBM before start of tunnelling Incorrect definition of tunnel axis in reference to rail axis (consider of cant) Mistake during input of DTA (designed tunnel axis); snaky advancing; Problems with control of direction (refraction at tunnel wall, laser near lining) Incorrect driving back to the DTA after a deviation

Problem concerning lining is related to a particular limit condition of boreability in hard rocks. It is because of the fact, that in order to overcome the boreability problem, there is a tendency to optimize the ratio between cutterhead revolution and the thrust (lower speed higher thrust), which can give better penetration and less cutter consumption (as a result of less vibration). However, higher thrust can adversely affect the lining bearing capacity (Figure 75).

Figure 75: TBM's thrust system affecting segments

104

Driving the bore through the critical curves may create detrimental contact between the shield tail and the newly-built segments. Problem concerning subsidence in very critical areas, i.e. under sensitive buildings with low cover, may be mitigated by filling the conical void around the machine with bentonite slurry, which is continuously injected through ports located along the shield. This provided additional help to further reduce volume loss due to the TBMs passage. Hopefully, it is expected that the modern-day control cabin full of machine data plus Controlled Boring Process (CBP) could facilitate the operators' work and achieve ever better performance. Such an advance is one of many efforts to comply with production- and automation/robotization-related demands imposed by the concept of precast concrete segments tunnel linings.

4.3.2.3

Guidance System

Guidance of full-face TBMs is vital. Guidance system is to help showing the correct position of TBM in correspondence with the designed tunnel axis. Here are some basic definitions of terms: TBM survey: Measuring and calculation of the real TBM-position and direction in relation to the designed tunnel axis. A coordinated site system with fixed survey points should be set up on the field. In practice, only a small direction fault in the starting pit can result in big error in the far end shaft Steering: Corrections of the TBM-drive resulting from real deviation Monitoring and control: Record of the TBM-drive history (graphic and numeric) and calculation, storing and analysis of all TBM and ring data

Main components of TBM guidance system are shown in Figure 76 below.

Figure 76: Components of TBM guidance system (Gruebl, 2006)

105

It is highly recommended (for the double shield TBM is indispensable) that TBMs are equipped with a guidance system that provides in real time information on the actual alignment to the TBM operator (Concilia, 2006). Apart from calculating the exact position and tendency (rolling and inclination) of the TBM for proper monitoring of tunnel alignment, the guidance system also measures the segmental ring for correct assembly (Figure 77).

Figure 77: Guidance system - measuring of the segmental ring

For any kind of the guidance system, it is necessary to know how quickly (speed and distance) the TBM can react to modifications to the trajectory it is on.

4.3.2.4

Lining Ring Building

To add to the previous subsection dealing with the segment optimization, hereinafter are given some additional points in view of lining's interaction with TBM, in terms of segment selection, erection and bolting.

The tunnel lining is discontinuous and its structural properties depend on those of the segments and contact joints between lining parts. Segment structural properties include: sectional area, inertia; modulus of deformation; and Poisson's ratio. Intersegment contact joint structural properties include: sectional area, inertia; and ring composition.

106

The choice of the adequate segment width (measured in the direction of tunnel axis) of the segment/ring is influenced by different and sometimes opposite exigencies: time of excavation; weight of the single segment; risk of damaging the segments while handling; geometrical compatibility with the rear of the TBM while the ring is coming out in the condition of the minimum radius of curvature.

The average width is varying between 1.00 m and 1.25 m (even 2.0 m) for big rings (diameter > 10 m) and between 1.25 m and 1.60 m for smaller rings. Large precast concrete segments may weigh from 1 ton up to 15-19 ton.

Segment height is the height of main girders of segments measuring in the direction of tunnel radius, for the flat type segment, it is also called "Segment thickness. Segment length is the arc length of segment measured in the transverse direction of tunnel axis. Distinction is made between the external arc length (A), the arc length at bolt pitch circle (B), and the internal arc length (C). The geometry (length, width, shape, inner conicity, etc) of the segments is defined by the interface between the TBM tail configuration and ring parameters. The choice of the number of segments per ring is dependent on the tunnel diameter (Figure 78), and is also influenced by almost the same factors listed for the length of the ring: time of excavation; weight of the single segment; risk of damaging the segments while handling; available space for handling in the back-up.

Figure 78: Choice of number of segments versus tunnel diameter (Pescara 2006)

Examples of lining in some actual projects are given in the Table 9 below.

107

Waterproofing of the ring is assured by the following factors which must be considered all with the same degree of importance: very good quality of the concrete (compressive strength fck is generally greater than 40 MPa) and of the curing process; careful handling to avoid any damage or formation of latent cracks; choice and positioning of the gasket; careful assembling of the ring; filling of the tail void with proper material in due time, volume and pressure.

Table 9: Ring and segment geometry (Wagner, 2006)

Ring Diameter Thickne Width division [m] ss [cm] [m] [-] Groene Hart NL - Railway Plabutsch A - Highway SOCATOP F - Highway Seattle USA - Subway Wanjiazhai PRC - Water Form/ shape [-] Length Weight [m] [ton]

13.3 11.1 10.4 9.42 4.3

60 40 42 40 25

2 2 2 1.6 1.4

9+1 6+0 7+1 8+0 4+0

Rectangular + key Trapezoidal/ Rhomboidal Rectangular + key Trapezoidal/ Rhomboidal Hexagonal

4.37 6.02 4.45 3.86 3.57

12 12 9.4 6.2 3.1

With regard to the ring assembling, the following should be noted: The order of arrival of the segments near the erector must respect the order of handling for the assembling process. The dimensions of the segment and of the backup are inter-related with respect to the movements (rotation and translation) that the segment itself must undertake. The segments can arrive to the erector both in the upper and in the lower part.

Ring erector may be of several types, such as vacuum-sucking erector, mobile erector (mechanical system) with remote control panel, etc. having many active movements necessary for ring assembly (Figure 79). It is recognized that TBM erector capabilities are closely related to the learning-curve phase as well as sensitivity and assembly precision.

108

Figure 79: Erector arm of TBM, 6 necessary active movements for ring assembly (Gruebl, 2006)

The quality of lining depends not only on the erection of regular segments but also on the installation of the key block (K segment). With all the segments in the ring as a whole, deformations such as offset, ovalization can happen as in Figure 80. In soft soil conditions, lining failure can occur before or even after ring closure, including: shear failure compression failure combined bending and thrust punching failure watertightness of segments

For K segment, there are two types in terms of the direction of insertion. One is to insert from the inside of tunnel, in which the longitudinal side faces of a K segment are tapered in the direction of funnel radius ("K segment inserted in radial direction"). The other is to insert in the direction of tunnel axis, in which the longitudinal side faces of a K segment are tapered in the direction of tunnel axis ("K segment inserted in longitudinal direction") (Figure 81). With trapezoidal tapered ring, key position selection in curved tunnels is of the essence; Keystone condition during the assembly: The k-segment is the last to be inserted and thus the precision of the movements must be higher. The movement of the k-segment during the insertion and its geometrical characteristics form the basis to define the length of the TBMs thrust system.

109

Figure 80: Schematic view of ring ovalisation (Wagner, 2006)

Figure 81: Insertion of the k-segment (Pescara, 2006)

110

The following Table 10 quoted from AFTES recommendations (1997) considers an example of building a lining ring comprising rectangular- (standard) and trapezoidalshaped (key and counter) segments. It details: the successive segment erection stages; recommendations associated with the operating environment for each erection stage; remarks of a general nature for each erection stage.

Finally, robotic control of tunnel lining installation is an actual need for mega tunnel projects, in view of lining quality and tunnel excavation progress rate. Computer controlled precast segments tunnel lining erection will harmonize segment geometry and machine configuration, aiming at automation, high quality and economy. Figure 82 shows the moduli of the operational program, using a recording, warn and alarm system for higher advance rates and lower running cost. The trend toward fully automated installation of segmental lining associates with dowel connectors. Future development of TBMs for large tunnel diameters leads toward new products particularly larger dowels for automation of segmental lining.

DATABASE LINING

DATABASE TBM

PROCESS CONTROL

TUNNEL/RING DATA CALCULATED STATE

"KNOWLEDGE" OF SYSTEM ABOUT TUNNELLING/ TBM

SURVEYING

CENTRAL PROCESSING

PRESENT STATE SENSOR DATA

NEXT STATE MACHINE INSTRUCTIONS FOR OPTIMIZED TUNNEL OPERATION

WARNING & ALARMS

RECORDING

MONITORING

Figure 82: SW robotic control system for segment erection, basic structure (Schulter, 1996)

111

Table 10: Lining ring building stages (AFTES, 1997)

SCHEDULE OF OPERATIONS 1) Supply of first segment to erector. 2) First segment pick-up. 3) Retraction of thrust cylinders corresponding to placement of first segment. 4) Positioning of first segment by rotating erector. 5) Radial approach of first segment. 6) Final approach with rotational, longitudinal and transverse balance adjustment. Control of approach speeds by selection of proportioning hydraulic controls. Pads of other thrust cylinders remain under pressure in contact with other segments to safely ensure: - segments holding and assembly - compression of waterproofing gaskets and prevention of their decompression. -stability of the machine under the confinement pressure. TBM main cylinder thrust on the other segments must prevent any forward displacement of the machine. At this time, the segment is simultaneously held by the erector and the thrust from the main cylinders. By ring/ring (longitudinal), segment/segment (transverse) connection. Same recommendation as for the first segment. Provide alternate installation of segments in each ring to minimize tube roll effects. Use of template to calibrate gap between counter segments. Same remark as for the first segment. Detailed analysis of loads in each pick-up system position and of indirect loads on segments. RECOMMENDATIONS REMARKS Supply possible from: - upper level; - lower level. Pick-up possible using suction pads, grippers, bolts. Light ray guidance systems can facilitate approach and final positioning of segment.

7) Holding of first segment on ring.

8) Fixing of first segment

9) Installation and fixing of standard segments. 10) Installation of counter segments.

Same remark as for first segment. It should be noted that on completion of erection, the ring is stabilized by the prestress between the erection jacks and the previously installed ring. The only contact between the shield tail and 1 the segmental lining is the shield tail seal.

11) Key segment installation.

Use of template prevents: - tearing of waterproofing gaskets. - concrete chipping. - greasing of waterproofing gaskets.

112

4.3.2.5

Backfill Grouting

Backfill grouting is the grouting work to fill the annular void between segments and the ground (tail void) by grout injection. In the case of segmental TBMs, the lining and its backgrouting are inseparable from the operation of the machine. Because of their interfaces with the machine, they must be designed in parallel and in interdependence with the TBM. Mix alternatives could be: Cement water Cement mortar (sand) Slurry (cement - bentonite) Inert mortar Pea-gravel and cement Clay (Russia)

In the early days backfilling consisted of either pea gravel or fast-setting or fasthardening cement slurry or mortar that was injected intermittently through holes in the segments. There has been a constant trend to continuously and directly inject products with retarded set and low compressive strength into the annular space directly behind the TBM tailskin by means of grout pipes routed through the tailskin (Figure 83). Given the shield wall/can thickness of 45 60 mm, grouting pipes diameter may be in the range of 20 25 mm. Consideration should be given to preventing backflow of soil and grout or water into the injection piping as grouting takes place; grease is so injected between tail seals as a tail packing material. Tail seals should be replaceable in case tail seals are damaged and grout material intrudes into the tail shield.

Figure 83: Longitudinal grout injection through the tail of slurry shield (Gruebl, 2006)

113

Requirements for the filling material may be as follows: Good pumpable (high content of fine parts, bentonite) Enough strength (> 10 MN/m2) but lower than concrete strength and to keep volume under loads Mortar must loose parts of mixing water to get an early setting behaviour (to avoid flowing around the tail skin to the front) Early hardening to avoid movements of the invert segment, when the first trailer arrives Support of the segment sealing to make the ring watertight

Machine advance is only possible if grouting mortar is available in the required amount and at the specified pressure. Carefully controlled grouting will help to avoid almost entirely significant settlements at the ground surface in the region of the TBM. For slurry shields, both slurry and grout pressures have impact on ground control during construction. Mansour (1996) has found that, the grout pressure is responsible for the control of final ground settlements. Although these settlements are always attributed to low grout pressure, extremely high grout pressure that causes local yielding for soil at the crown level, may cause ground surface settlements as well. Therefore, optimization of grout pressure value is essential to efficiently control the final ground settlements. The slurry pressure is also responsible for controlling excessive deformations at the tunnel face and maintaining its stability against plastic failure. From experience, it is found that 80 90% damage to the segment was due to grouting (e.g. excessive grouting pressure). So control of the grouting pump as well as assembly of tailsealing (wire brush) is very important (Figure 84). Because of the permeation to the ground, penetration to the ground by the injection pressure, dehydration, and over cut, the injection volume often becomes 150 - 200 % of the theoretical void volume (sectional area of the shield machine minus sectional area of the segment ring). It is desirable to control backfill grouting using both pressure and volume. In the absence of annular grouting, bending tensile cracks of the segments may happen. Incorrect grouting can cause steps in annular joints between the segments. It is common that after lining is grouted on the shield tail (primary grouting), grouting volume will settle, opening a gap on the upper part of cross section. This gap space should be grouted again, it is called secondary grouting. Secondary grouting may have bad effects on the tunnel lining segments, therefore this must be taken into account. Secondary grouting is usually undertaken within 20 m of the last ring built. Usually voids will be found in the crown but other areas also need to be checked particularly if fast setting grouts are used.

114

Figure 84: Key segment is pushed out by excessive grouting pressure (Dal Negro, 2006)

Tertiary grouting is sometimes necessary too. It is carried out following evidence of voids in secondary grouting. Other evidences include tunnel leaks, lining movement, previous high amounts pumped in, grout washed out, grout shrinks. Tertiary grouting is also decided based on the result of drilling through the lining to check for voids.

4.3.2.6

Back-up System

The equipment needed to enable the TBM to perform the excavation are located on mobile platforms which follow the machine and as a whole are called back-up. Some of the back-up equipment and plants installed include: Mucking out system (conveyors) Muck cars movers or muck conveyor extension system Segmental lining handling and erection system or supports erection system Equipment for rails and service lines extension

One of the functions of the lining is to support the back-up equipment and construction plant required for carrying out the work. The railway track needs to be placed in the working area on the back-up, or a temporary roadway in the case of transport of muck by truck. The track should be properly anchored to the invert segment in order to minimize the derailment hazard. The following operations within the back-up which may result in potential risks to the lining must be taken into account: segment delivering to erection machine; erector movement; segment positioning and erection;
115

back-up maintenance; mucking train arriving/leaving;

4.3.2.7

Monitoring and Instrumentation

Monitoring consists of a set of several in-situ measurements of the displacements, the stresses and the strains variations which occur in the ground and in the tunnel supports during the excavation procedures. Monitoring is the only mean, in a context of great variability and uncertainty, of allowing the excavation of a tunnel to be adjusted in an objective way during work procedures. By employing this kind of observational method, safety in underground construction could be ensured at the highest level. Therefore monitoring has an essential role in tunnelling. A tunnel design and the tunnel construction cannot exist without monitoring. Monitoring is the mirror that reflects the soul of a tunnel (Pelizza, 2006). Conditions of the works during shield driving shall be monitored or measured to secure the safety of construction. In other words, during construction of tunnel tubes, the interaction between tunnel segments, and the influence of the surrounding soil on the tunnel lining and vice versa, should be registered. This can be done by incorporating measuring systems in some tunnel rings of the tube. The monitoring and measurement in shield tunneling works include the following (JSCE, 1996): 1) Monitoring: i) For a closed-face type shield: Earth pressure in the cutter chamber, slurry pressure at the face, characteristics of the slurry For an open-face type shield: Conditions of the face, the amount and quality of water inflow ii) Hydraulic pressure of the jacks, torque of the cutter, meandering and balance of the shield machine, control of the volume and pressure of backfill grout, control of the volume of excavated soil discharge iii) Deformation of the shield tunnel and the deviation of the centerline from the designed alignment iv) Deformation of the structures and/or underground facilities, displacement of the ground and variation of groundwater level 2) Measurement: Stress and displacement occurring within the shield machine or lining, and earth pressure and water pressure acting upon the shield machine or lining Tunnel lining monitoring is now being challenged by automatic data acquisition, which offers many advantages. But, several drawbacks inherent in the use of automatic data acquisition should also be mentioned. These include the danger of systematic loss of data, the recording box spatial requirements, and possible problem of energy supply and independence. Further, all monitoring schedules must specify for the interpretation

116

of results; if this interpretation stage is neglected, the validity of the monitoring system must be questioned.

Concluding Notes The installation and long term stability of the segmental lining are possible only because the lining is installed with the use of a TBM. The lining is loaded in such a way which is strictly linked to the excavation and installation process (thrust of the jacks, longitudinal grout injection, waterproofing). This observation has an apparent implication if a final and optimized design is to be achieved. Gruebl (2006) emphasized that, in a circular bedded segmental ring, deformations further than about 50 m behind the TBM hardly ever occur. Virtually all the significant damage occurs during ring installation and advance of the TBM in the last 3 to 4, at most 10, rings. The bearing strength of the segmental ring is rarely exceeded, but damage can occur to the lining due to incorrect ring erection or insufficient grouting. The cost of repairing the damage may rather high and make a well calculated construction project into an economic failure. We may conclude that none of the above presented contractual and physical interfaces can be underestimated and overlooked.

117

Chapter 5

5. Information for Settlement Study

5.1 Ground Conditions

Lot H3-4 has a total length of 5.8 km, with a minimum overburden of approx. 8.5 m. Figure 85 shows the geometry of the typical cross section with the details of every soil layers. Maximum calculation water level is 28.4 m; minimum calculation water level is 17.5 m

Figure 85: Typical cross-section RQ2, homogeneous zone Nr.4 Km 33+480 to 33+800

118

In Table 11 the ground layers and the corresponding soil characteristic values are listed.

Table 11: Geomechanical parameters of Lot H3-4

Soil type / Parameters Wet density Submerged density Dry density Voids Friction angle Cohesion [kN/m3] [kN/m3] d [kN/m3] [] [] c [kN/m2] E [MN/m2] [-] GA_T1 Coarsegrained Gravel 21 GA_T2 Mixed sand gravel 21 GA_T3 Fine sand - medium sand 21 GA_T4 Silt fine sand GA_T6 Valley sand and silt 20

20

12

12

11

10

10

18,64 0,43 37

18,64 0,43 35

18,64 0,43 33

17,66 0,54 30

17,0 0,54 30

E-Modulus Poisson's ratio

f(z)

f(z)

f(z)

22,5

10

0,3

0,3

0,3

0,3

0,3

The depth-dependent moduli of elasticity of soil layers GA_T1, GA_T2, and GA_T3 are provided by the formula: f(z) = 0.1 v (0.13 z)w [MN/m2] determined with v = 373 and w = 0,65 and will be given in Appendix 4.

119

5.2 Excavation and Support

5.2.1 Shield Machine

The tunnel has an outside diameter of 12.63 m, to be driven by a Herrenknecht HydroShield TBM with a diameter of approx. 12.98 m, a length of 11.35 m. The total excavation area is approx. 132 m2 (Figure 86).

Figure 86: TBM intended for use in Lot H3-4 of the BEG tunnel project

The TBM will be designed, manufactured and operated in accordance with the Contract Specification which will define best practice for the project. The time required to procure, manufacture and deliver a TBM is more than one year. Delivery of the TBM to the Lot H3-4 worksite can be expected in the middle of 2007.

The hydraulic jacks push the TBM forward against the lining of the tunnel behind the machine while excavation is in progress. There are 21 groups (pairs) of hydraulic jacks, over 21 pads, equally distributed over the circumference of the shield. Each pair consists of two thrust jacks. The feet of the jacks are placed onto the rings. When the jacks are extended, the TBM is pushed forward. After the length of one ring is bored, the jacks are relaxed and the next ring installed.

Excavation diameter Outer diameter of structural lining Inner diameter of structural lining Inner diameter of fireproof lining

: : : :

12.98 m 12.63 m 11.63 m 11.23 m

120

Weight of TBM Total thrust pressure No. of thrusts No. of pads Pressure/pad Pad geometry (supposed only) Pressure of the support fluid Grouting pressure (regular) Grouting pressure (accidental)

: : : : : : : : :

2,350 T 10,345 kN 42 21 4,926 kN 197 mm x 950 mm 3.5 bars (350 kN/m2) 2.5 bars (250 kN/m2) 4 bars (400 kN/m2)

Allowable space for segment installation is about 2.5 m, including the width of segment (2.0 m). Grouting is provided on the shield tail.

5.2.2 Ring Configuration

The linings will be designed to withstand temporary and permanent loading including loads from the surrounding ground, groundwater, a surcharge/traffic live load of about 100 kN/m2 from the Highway A12, and to meet fire and durability requirements. The lining's internal diameter of 11.23 meters is sufficient to accommodate the cross section of the two trains plus lateral movement tolerances, overhead power supply, evacuation and access walkways, resilient and floating track slab, signalling equipment, cables and cable brackets and construction tolerances. The pre-cast concrete segments reinforced with traditional steel reinforcement will be designed to provide a robust solution capable of dealing with the handling loads from construction and the permanent loads from the ground. Segments are to be bolted together with suitable gaskets, and the annulus between the lining and the excavated ground is to be filled with grout. The 0.5 m thick outer structural lining comprises of 7 main segments (with 5 rectangular segments and 2 trapezoidal segments) and one keystone segment. Lining geometry is defined through segment types A1-A5, B, C and K. Segment length is 5.44 m, segment width is 2.0 m. The inner fire protection lining has a thickness of 0.2 m, made of cast-in-place concrete. Between the two lining layers is a thin layer of shotcrete for the levelling purpose. The tunnel inner radius is 5,615 m (diameter 11.23 m), and the opening area is 99.00 m2. Figure 87 shows the interior fittings inside a cross-section of the Lot H3-4; the rescue tunnel and escape shaft as shown in Figure 88 will not be considered here.

121

Figure 87: Typical cross-section of Lot H3-4, Mnster - Wiesing

Figure 88: Cross section with rescue tunnel and escape shaft. Lot H3-4 Mnster Wiesing

122

5.2.3 Lining Material

Concrete material is specified to be Class C 50/60. The steel grade is considered with class BSt 550. Concrete cover for reinforcement is basically 4 cm thick. Effective height h1 is determined in accordance with Austrian Standard Norm B 4700 depending on the used reinforcement. Dead specific weight of the concrete tunnel lining of 25 kN/m3 will be used for computation. Gasket will be of Phnix M 385 73 Type (Figure 89). Material properties are defined as follows:

Concrete class Compressive strength Specific weight Concrete elasticity modulus Steel elasticity modulus Gasket type Gasket width Elastic capacity Max. tested closed gap fc28 Eb Es

: : : : : : : : :

C 50/60 50/58 MPa (MN/m2) 25 kN/m3 37,000 MPa 200,000 MPa Phnix M 385 73 Typ "Wesertunnel" 618g/m 44 mm 46 kN/m 15 mm

Figure 89: Gasket dimensions for Lot H3-4

123

5.3 Numerical Analysis Tool

The tunnel of the BEG Lot H3-4 crosses the Inn River with low overburden, as well as the motorway A12 and the existing tracks of the Austrian Railway. Some information on the statical calculations of the tunnel is presented in the following. For the purpose of reference only, below are given the software actually utilized by the Designer for the BEG project: Z_SOIL.PC 2003 V 6.24 Professional software: "Soil, rock and Structural Mechanics in dry or partially saturated media" (Zace services Ltd.) RSTAB V 5.12.058: "statics of general structures". The designing firm made calculations of the lining by this RSTAB program (Ing. - software Dlubal GmbH) RFEM 2 V 2.01.135: "spatial lining units according to the method of the finite elements" (Ing. - software Dlubal GmbH) ConDim 5 concrete calculation, version: 5.04 (DI Dr. Lorenz)

However, in this work, the author will make use of PLAXIS, a finite element code for soil and rock analyses of Plaxis B.V. Netherlands, to perform independent parameter studies. The parametric calculations are to partly illustrate the theoretical points that have been presented in the previous parts of this report. Calculations focused on the Austrian BEG project's TBM Lot H3-4.

5.3.1 Soil Models in Plaxis

Plaxis offers a variety of soil models in addition to the Mohr-Coulomb model. The logarithmic compression behaviour of normally consolidated soft soils can be accurately analysed by Cam-Clay type model (Soft Soil model), or by an improved version of it for secondary compression (creep). For stiffer soils, such as overconsolidated clays and sand, an elastoplastic type of hyperbolic model is available, which is called the Hardening Soil model. Short descriptions of the available models are given below.

Linear elastic model: This model represents Hooke's law of isotropic linear elasticity. The model involves two elastic stiffness parameters, namely Young's modulus, E, and Poisson's ratio, . The linear elastic model is very limited for the simulation of soil behaviour. It is primarily used for stiff massive structures in the soil.

124

What is stiffness? The link between strain increments and stress increments will be required for performance of numerical analysis of geotechnical systems; then a stiffness matrix D is formally required to be developed and populated: = D, and this incremental link between stress and strain is the most useful definition of stiffness.

Mohr-Coulomb model: This well known model is used as a first approximation of soil behaviour in general. The model involves five parameters, namely Young's modulus, E, Poisson's ratio, , the cohesion, c, the friction angle, , and the dilatancy angle, . What is dilatancy? If the arrangement of the soil particles is disturbed by distorting the boundary of the soil sample then rearrangement will be accompanied by some change in the volumetric packing: this is dilatancy. All different soil layers can be modelled by the simple Mohr-Coulomb model. For settlement analyses, the Hardening Soil model may be preferred, but for tunnel heading stability the focus is on soil strength and not on soil stiffness

Hardening Soil (HS) model: This is an elastoplastic type of hyperbolic model, formulated in the framework of friction hardening plasticity. This second-order model can be used to simulate the behaviour of sands, gravel and overconsolidated clays.

Soft Soil model: This is a Cam-Clay type model which can be used to simulate the behaviour of soft soils like normally consolidated clays and peat. The model performs best in situations of primary compression.

Soft Soil creep model: This is a second order model formulated in the framework of viscoplasticity. The model can be used to simulate the time-dependent behaviour of soft soils. In the following parts, only Hardening Soil model will be utilized for the purpose of this report's parametric studies.

5.3.2 Hardening Soil Model

The Hardening-Soil model (isotropic hardening) is used for simulating the behaviour of different types of soil, both soft soils and stiff soils. This model uses the theory of plasticity rather than the theory of elasticity; it also includes soil dilatancy and introduces a yield cap.

125

The Hardening-Soil model represents a much more advanced model than the MohrCoulomb model. As for the Mohr-Coulomb model, limiting states of stress are described by means of the friction angle, , the cohesion, c, and the dilatancy angle, . Soil stiffness is described much more accurately by using three different input stiffnesses: the triaxial loading stiffness, E50 (characteristic of plastic straining due to primary deviatoric loading); the triaxial unloading stiffness, Eur (characteristic of elastic unloading / reloading, Figure 90); and the oedometer loading stiffness, Eoed (characteristic of plastic straining due to primary compression, Figure 91).

Figure 90: Hyperbolic stress-strain relation in primary loading for a standard drained triaxial test. qa is the asymptotic value of the shear strength, qf is ultimate deviatoric stress

ref Figure 91: Definition of tangent stiffness modulus Eoed in oedometer test results

126

As average values for various soil types, we have Eur 3 E50 and Eoed E50, but both very soft and very stiff soils tend to give other ratios of Eoed / E50. In contrast to the Mohr-Coulomb model, the Hardening-Soil model also accounts for stress-dependency of stiffness moduli. This means that all stiffnesses increase with pressure. Hence, all three input stiffnesses relate to a reference stress, pref or ref, being usually taken as 100 kPa (1 bar). For virgin oedometer loading, soil behaviour simulated by the HS-Model implies an increasing tangent stiffness modulus according to
ref Eoed = Eoed . ( 1 + a ) /( ref + a ) m

ref is the reference tangent where a = c.cot, 1 is the major principal stress, and Eoed stiffness for primary oedometer loading. We adopted the exponent m = 0.5 (stress dependent stiffness according to a power law). Within the HS Model unloadingreloading is described on the basis of Hookes law. Youngs unloading-reloading modulus for increments of stress and strain is:

ref 3 + c cot Eur = Eur . ref + c cot

ref where 3 is the minor principal stress, Eur is the reference Young's modulus for unloading and reloading, corresponding to the reference pressure ref. In many ref ref practical cases it is appropriate to set Eur equal to 3 E50 .

For many problems, especially excavation problems, there is a preference to use the Hardening-Soil model (HS-model) rather than the Soft-Soil model (SS-model).

5.4

Flowchart of Calculation

Surface settlements will be checked both with analytical and numerical methods. Variation of input parameters based on greenfield condition will be performed in order to check the respective output results, using the Hardening Soil model (symmetrical). Parameter studies for the greenfield cross-section will be carried out as follows: 1) 2) Using semi-empirical methods to calculate the surface settlement, with the actual geological and lining information of the BEG project. By using Plaxis code, the factor of contraction (volume loss) will be gradually varied, starting with a value of 0.5 %, then stepwise increasing with an increment of 0.25% until convergence is achieved. The number of elements of the FE net (the mesh) will be varied starting from about 500 elements up to approx. 1,000 elements, with an increment of around 200 elements for each step. The
127

combination of the two varying variables (the contraction and the mesh coarseness) will involve many computation cases. 3) Variation of the lining thickness will be carried out in the range of 0.30 m 0.60 m, stepwise thickness increase = 0.10 m (i.e. t = 30, 40, 50, and 60 cm). The mesh coarseness is also varied accordingly. Next, final settlements will be calculated, using the design lining thickness of 50 cm and with the converged contraction value attained at the step 1. Compare the calculated results of settlements by both analytical and numerical methods for the case of actual project input data. 3D modelling of the shield tunnel face stability is performed.

4) 5)

Such variations help to check the sensitivity and reliability of the results of face stability and ground settlements corresponding to the input parameters associated with the various analysis methods. These parameter studies supplement some interface aspects concerning the tunnel design and construction presented in the previous chapters.

128

Chapter 6

6. Tunnel Induced Ground Deformation

6.1

Settlement Induced by Tunnelling

6.1.1 Volume Loss and Settlement

- Ground movements and volume loss due to tunnelling All sub-surface excavations give rise to ground movement. In other words, ground movements are an inevitable consequence of constructing a tunnel. These movements manifest themselves, in particular, as settlement. It is not possible to create a void instantaneously and provide an infinitely stiff lining to fill it exactly. In the time taken to excavate, the ground around the tunnel is able to displace inwards as the stress relief is taking place (Figure 92). Thus it will always be necessary to remove a larger volume of ground than the volume of the finished void. This extra volume excavated is termed the "Volume Loss" (or "ground loss", "soil loss") VL (Chiriotti, 2006).

Figure 92: Inward displacement of the ground around the tunnel due to stress relief (Chiriotti, 2006)

129

For the construction of tunnel, the overall volume loss VL consists of two components: face loss and radial loss (radial displacements around the shield section and around the lining section, Figure 93) [m3/meter advance of the tunnel drive].

RADIAL LOSS - annulus

RADIAL LOSS - shield

RADIAL INJECTION SHIELD INSTALLED RINGS SCREW CONVEYOR

BEAD FACE LOSS

EXCAVATION CHAMBER

Figure 93: Components of the volume loss or the convergence generated by tunnelling work (Chiriotti, 2006)

In a properly supported non-TBM tunnel, 70-80% of total surface settlement is due to deformations ahead of tunnel face. In a shield-driven excavation, the fraction varies significantly (<< 70%) depending on the method. As an example, until a recent date, the following distribution of settlements to the surface was observed: - 10 to 20 % caused by the face; - 40 to 50 % caused by the void along the shield; - 30 to 50 % caused at the end of the tail seal. But thanks to the current technological and methodological evolutions, these are changing and settlements at the tail seal exit may only stand for a small part of the total settlements (AFTES, 1995). The net volume of the surface settlement trough will be approximately equal to the volume loss at the tunnel in most ground conditions. If the ground response is at constant volume (i.e. undrained), the relationship will be exact. The hypothesis will be checked especially if the ground is clayish and the overburden is thin. Otherwise, relationship between the displacement in the tunnel crown (Urcrown) and the middle surface settlement (Smax) can be referred to Figure 94 as an example, where C is overburden thickness above the crown, and D is excavation diameter.

130

Figure 94: Softening coefficient according to the geometry of the tunnel bored with a shield (AFTES, 1995)

VL is normally expressed as a percentage VL% of the gross area of the finished tunnel. Assuming a circular tunnel of outside diameter d, then
VL % = VL 100% VL 100% = Vtunnel d2 4

The magnitude of the volume loss VL depends on many different factors: soil type tunnelling method rate of tunnel advance tunnel size form of temporary and primary support

Before the magnitude of ground movements can be predicted it is necessary to estimate the expected ground loss. This estimate will be based on case history data and should include an engineering appraisal that takes into account the proposed tunnelling method and site conditions. NATM: London Clay VL% = 0.5% - 1.5% which compares favourably with controlled shield tunnelling

131

Open face tunnelling Stiff clay VL% = 1% and 2% Closed face tunnelling (EPB or slurry shields): A high degree of settlement control can be achieved. Sands VL% < 0.5% (0.35% can be achieved with slurry shield and EPB TBM tunneling). Soft clays VL% = 1% - 2% (excluding consolidation settlements)

- Short, medium and long term movements SHORT-TERM ground movements are identified to occur during the excavation, at a timescale that is comparable with the time taken by the advance of the tunnel heading that is the cause of ground movements. MEDIUM and LONG-TERM settlements are thought to be the result of creep, ageing and consolidation, i.e. alterations in the properties of the soil at constant load. The timescale over which they occur depends on the ground conditions, ranging from weeks or months (sands and soft clays) to years (stiff clays). The magnitude of long-term movements is hard to generalise. The long-term settlement trough widths are observed to have tendency to be wider than that of the short-term settlement trough widths. This means that the curvature of the trough, the factor most likely to cause damage to the structures on the surface, is smoother. In addition, surface structures are more able to accommodate long-term settlements by creep and stress redistribution. Thus it is the short-term movements that remain the chief issue of concern for engineers, and are the subject of this study.

- Rock-tunnelling movements Rock tunnels may not be immune from causing unacceptable ground movements where water inflow may, by way of the jointing system, cause draw-down in overlaying fine sediments (Peila, 2006).

6.1.2 Settlement Calculation Approaches

Short literature review Leca and Dormieux (1990) have used the upper and lower bound theorems of plasticity to estimate the pressure, which can be provided by compressed air, slurry or an earth pressure balance (EPB) tunnelling machine, needed either to ensure stability or obviate blow-out of the tunnel heading in soft ground.

132

Suwansawat (2002) and Suwansawat & Einstein (2006) have attempted to predict ground response and the maximum surface settlement caused by EPB shield tunnelling using artificial neural networks (ANN), at the same time to evaluate the potential as well as the limitations of ANN for that purposes. The empirical method have been proposed by a number of authors, such as New & O'Reilly (1982, 1991), Attewell and Woodman (1982), etc. to predict the settlement for green-field site conditions induced by bored tunnels. Semi-empirical method uses the parameters for ground loss determined from case histories and take into account the method of tunnelling and ground conditions; they are still being widely used today. Recently, a generic area-wide assessment of settlement identifies zones in which buildings might be at risk of sustaining damage in excess of acceptable levels based on correlation with the calculated maximum tensile strain values (Franzius 2003, 2005). The potential for damage in this area-wide assessment can be defined using the procedure described by Mair et al (1996). On the other hand, numerical methods, mainly the Finite Element Method (FEM), provide a flexible tool for a prediction of surface settlement, which have been adopted by many authors. FEM could be performed in two-dimensional (2D) or threedimensional (3D). In the past, Selby (1988) recommended that the empirical equations should be used for predictive purposes as they compared well with field measurements and were much easier to use than the finite element model. Clough and Leca (1989) have pointed out that the soil tunnelling problem has proved resistant to finite element (FE) modelling because it is complex, often involves parameters that are not well defined and is unforgiving if the analyst does not properly model both the soil and the tunnel supports, as well as the construction process. The sensitivity of the FE method to these factors has meant that it is a less reliable method for ground movement prediction than the empirical approach. However, New and Reilly (1991) deemed that the flexibility of FE models can be exploited when back analyzing ground movements and can assist in understanding the movements at particular sites and by extending conventional design techniques. Indeed, the numerical methods (e.g. FEM) have been being more and more powerful and reliable as a consequence of advances in computer technology.

Selected Calculation Approach In the following, the author will try to apply the two main settlement prediction approaches: (i) analytical/semi-empirical, based on empirical formulas derived from past observations; and (ii) finite element analysis, which is now rather popular method.

133

6.1.3 Settlement Control Approach

After prediction of settlements is completed, the other considerations to minimize settlement affecting buildings above include: Monitoring; Protective works; Defects surveys; and Repairs

i) Monitoring Design for a plan of ground and surface monitoring plays a significant role in managing settlement and damage. The scope of monitoring should be set up for the two main categories: ground and buildings. In the monitoring plan, the governing parameters shall be defined, and the necessary instruments shall be selected. The criteria for the attention and alarm limits (threshold values) shall also be established. Then, based on monitoring, if any critical scenario is detected, counter-measures will be triggered accordingly. Relation between the three main elements: design hypotheses, monitoring, and countermeasures, is given in Figure 95.

Design hypotheses

Countermeasures Monitoring

Figure 95: Monitoring in relation with other procedures (after Chiriotti, 2006)

ii) Protective measures Tunnelling-induced subsidence can be mitigated and controlled by means of (Cross London, 2005):

134

Good tunnelling practice (including continuous working, erecting linings immediately after excavation and providing tight control of the tunnelling process to reduce the magnitude of settlement); At-source measures (including all actions taken from within the tunnel during its construction to reduce the magnitude of ground movements generated at source, such as face stability, backfill grouting at shield tail, etc.); Ground treatment measures (including compensation grouting, permeation or jet grouting, control of ground water, etc.); Structural measures (to reduce the impact of ground movements by increasing the capacity of a building or structure, typically including underpinning or jacking/shoring).

iii) Defect surveys (condition of properties) Defect surveys are typically undertaken 1 month before construction starts in the area to capture the condition of all properties immediately prior to tunnel construction. It is necessary to use a reliable damage classification system for masonry structures with the concept of limiting tensile strain. A staged process of assessing risk may be adopted, including preliminary assessment, second stage assessment, and detailed evaluation. In this process, buildings are eliminated from further stages depending on the potential degree of damage predicted (Mair et al, 1996).

iv) Repairs If the damage is caused by the nominated undertakers works, the nominated undertaker has to reimburse property owners for the reasonable cost they incur in remedying material physical damage arising from ground settlement caused by the authorized works.

6.2 Empirical Calculation for Settlement

6.2.1 Formulae

Single tunnels: Usually, the first stage of settlement assessment is based on "green-field" site conditions. This means that there is no existing surface building or underground structure, or the effect of building foundations on the pattern of settlement is ignored.

135

Peck (1969) described settlement data from over 20 case histories available to him at that time, and was able to deduce that the short-term transverse settlement trough in the "greenfield" could be approximated by a Gaussian curve (same as OReilly, 1982) (Figure 96):
y2 S = S max exp 2 2i

(OReilly, 1982)

where, S = theoretical surface settlement (the Gauss error function, or normal probability curve) (m) Smax = maximum surface settlement (over tunnel axis, i.e. the settlement trough depth) (m)
Smax = VS 2 i = VL (%).

.D 2
4

1 0.313VL (%) D 2 = i 2 i

(New & OReilly, 1991/ Mair et al, 1996)

S max

D2 = 0.785. ( .Z 0 + Ps ) . i.E

(Herzog, 1985)

This equation for Smax was derived from both shield excavated and NATM tunnel data (Arioglu, 1992, quoted by Ercelebi, 2005). y i = transverse horizontal distance from the tunnel centerline (m) = standard deviation of the curve (point of inflexion of the curve) (m); a trough width parameter which can be calculated after OReilly and New (1982) or Arioglu (1992) as in the following section = (average) natural unit weight of formation (ton/m3) = tunnel axis depth (m) = total surcharge load (ton/m2) = equivalent tunnel excavation diameter (m) = (average) elasticity modulus of formation (ton/m2)

z0 Ps D E

The settlement ordinate at distance i is, according to the properties of the probability curve, equal to 0.61 Smax.

136

Figure 96: Gaussian distribution curve of the short-term transverse settlement trough in the greenfield

* Note: According to Martins (2001), another closed-form analytical solution to predict surface settlements has also been proposed:
S z =0 = 4 0 (1 ) R 2 1.38 y 2 H exp ( H + R )2 H 2 + y2

Loganathan and Poulos (1998)

where 0 is the ground loss (ratio), is the Poisson's ratio of the soil above the tunnel, R is the tunnel radius, H the tunnel depth and y is the lateral distance from the tunnel centre-line. On the basis of centrifuge testing Loganathan, Poulos and Stewart (1999) claim that this equation gives better results than the Peck's equation.

The Peck/O'Reilly's equations are based on the assumption that the settlement profile above a single tunnel is of normal probability or Gaussian form. Ground deformation is assumed to take place at constant volume. The Gaussian curve shown above is used at all levels in the ground above the tunnel. For the combined effect of multiple tunnels, the movements induced by each tunnel are simply added. The width of the settlement trough perpendicular to the tunnel is defined in terms of distance 'i' in meters from the tunnel centre-line to the point of inflexion on the curve. Peck noticed that soils of different classes - e.g. cohesionless or cohesive - gave distinct ratios of trough width parameter i to tunnel depth z0.
137

Following from this, OReilly and New (1982) expressed the trough width parameter i in the form:
i = K zo

where, K = dimensionless empirical constant, depending on the soil type z0 = depth of the tunnel axis below ground level Based on data from both cohesive and cohesionless locations, all in the UK, OReilly and New (1982) proposed the empirical relationships: i = 0.43z0 + 1.1m (for cohesive soils) and i = 0.28z0 - 0.1m (for non-cohesive soils) The data used covered a wide range of tunnel axis depths. It thus appeared justified to take K as a constant value, independent of both tunnel depth and diameter. Later work by other researchers has confirmed that K is usually in the range: K = 0.4 0.5 for cohesive soils, K = 0.25 0.35 for cohesionless soils

K values when tunnelling in stratified soils: If we assume that, all the strata influence the settlement phenomena to the same extent, then keq is the average weighted value of the various ki: z0 < 1.5D

keq =

z1k1 + ... + zn kn ztot

z1 z2 zn

k1 k2 kn D ztot<1.5D

Figure 97: K values when tunnelling in stratified soils with z0 < 1.5D

138

If we assume that, the strata within 1.5D influence are the ones that mainly influence the settlement phenomena, then ki can be weighted according to their distance from the tunnel. z0 >1.5D

keq =
z1 z2 zm zm+1

0.35( z1k1 + ... + zm km ) + 0.65( zm +1km +1 + ... + zn kn ) 0.35( z1 + ... + zm ) + 0.65( zm +1 + ... + zn )
k1 k2 km+1 km 1.5D D ztot1.5D

zn

kn

Figure 98: K values when tunnelling in stratified soils with z0 > 1.5D

Otherwise, i can be calculated by following formulae:

i= i1 + i2 + i3 3

i1 = 0.386. Z 0 + 2.84 i2 = 0.5. Z 0

D Z i3 = 1.392. . 0 2D
0.704

(Arioglu, 1992) (Glossop, 1978) (Arioglu, 1992)

Under undrained conditions, e.g. in materials with a low permeability such as stiff clay, the volume of the surface settlement trough is equal to the volume of soil which is excavated in excess of the theoretical volume of the tunnel (constant volume), we have:
Vs = VL VL (%) =
Vs .D 2 VS = VL (%). Vtunnel 4

where Vs = volume of the settlement trough (per metre length of tunnel), so, the loss at the tunnel level is completely transferred to the surface. Vs can also be evaluated as the integral of the Gaussian distribution curve:
139

 - y2 S = Smax exp 2 Vs = 2 iSmax = 2.51iSmax 2i

Smax = VS 2 i
= VL (%).

.D 2
4

1 0.313VL (%) D 2 = i 2 i

The expression for the total vertical settlement due to tunnelling S may be rewritten, substituting for Smax, as:
- y2 - y2 VL S = Smax exp 2 = exp 2 2i i 2 2i

NOTE: Under drained conditions (e.g. dense sands) Vs is usually less than VL because of dilation. Sometimes (e.g. loose granular soil) Vs could be greater than VL in case of negative dilation. The trough width widely depends on the ground characteristics and the project geometry (C/D, i.e. overburden thickness C above the crown/excavation diameter D) and much less on deconfinement which, on the contrary, strongly influences Smax.

It is also conceived that, settlements are a 3D problem. We can be interested in evaluating (i) the transversal settlements trough in a certain section while it is developing; or (ii) the longitudinal settlement trough (Figure 99 and Fig. 100).

Figure 99: 3D view of settlement trough due to tunnelling (after Attewell et al., 1986)
140

Figure 100: Evolution of settlements along a shield (AFTES, 1995)

Attewell and Woodman (1982) extended this model to derive a settlement trough in the longitudinal direction, as presented in Figure 101. According to Attewell and Woodman (1982), the generalized expression for surface settlements can be written:

S=
where:

Vs 2 i

y2 2i 2

x xi x xF G G i i

S = surface vertical settlement at a location defined by the coordinates (x,y) [m]; y = transverse distance of the considered surface point from the tunnel centerline [m]; x = longitudinal position of the considered surface point [m]; Vs = volume of the settlement trough per meter of tunnel advance [m3/m], defined as a percentage VL of Vtunnel; xi = initial position or starting section of the tunnel [m]; xf = considered position of the tunnel face [m]; G is a function defined as:
1 G ( ) = e 2 d , with = (x-xi)/i 2
2

G(0) = 0.5 when x = xf (point above the tunnel face) G(1) = 1.0 when (x-xi)

141

Initial section xi

Considered section for subsidence calculation in a cross-section

Tunnel face xf

Tunnel axis

G1=1 and G0= 0

xi x

G1=1 and G0 0
xf X

Smax

Smax

LONGITUDINAL SETTLEMENT
S

TRANSVERSAL SETTLEMENT
Smax

Figure 101: Settlement trough in the longitudinal and transversal direction

g[(x-xf)/i] G[(x-xf)/i] (x-xf)/i 0 Figure 102: Definition of G function 3

142

Values of the G function have been already calculated for different values of (x-xi)/i and they are available in the format of table. The subsidence profile in a longitudinal section is evaluated on the basis of the above general equation. Being y = 0 along the tunnel axis the expression becomes:
S= Vs x xi x xF G G = Smax ( G1 G2 ) 2 i i i

If the starting tunnel position xi and the position of the face xF are known, then it is possible to calculate the vertical displacement for different points located ahead (x > xF) or behind (x < xF) the tunnel face. When G1 = 1 and G0 0 the longitudinal displacement is a percentage of Smax, being the difference G1 - G2 < 1. The settlement directly above the tunnel face corresponds to 0.5Smax (Figure 56).

For completeness, we like to mention a personal opinion of Prof. Swoboda (personal communication), that for Smax one has to use FEM; he found out only with 3D models one can get good displacement results for TBM; for NATM one can do it with a 2D model.

* Horizontal surface displacements: Damages to above ground structures could also result from horizontal ground deformations induced by tunneling. It is assumed (Mair et al., 1996) that the horizontal surface displacement sh(y) at a distance y from the tunnel center-plane can be reasonably expressed as:

sh ( y ) =

y S ( y) H

where H is the depth to tunnel axis and S(y) the settlement at a distance y from the tunnel center-plane.

Twin tunnels: The construction of twin tunnels is a common requirement for underground railways and there exist useful equations for making preliminary ground movement predictions. In practice the tunnels will rarely be driven simultaneously and one tunnel is likely to have been excavated significantly before the other. In some cases this will give rise to an asymmetry which is not modelled by the equations. It is generally assumed that the predicted ground movements for each tunnel can be superimposed. For twin tunnels with reduced transversal distance between the axes this assumption may be unconservative. Disturbance due to the first tunnel drive can be

143

simulated by assuming a greater volume loss for the second bore and superimposing the resulting ground movements. For the configuration shown in Figure 103 the settlement resulting from the combined effects of the twin tunnels is given by
S( y , z ) = ( y D )2 y2 exp 2 + exp 2 i 2i 2 i 2 VS

(New and O'Reilly, 1991)

Figure 103: Surface settlement profile for twin tunnels (New and O'Reilly, 1991)

If two or more shield tunnels are constructed side by side or on top of one another, ground movement and tunnel movement shall be carefully observed. If necessary, auxiliary methods shall be taken in order to prevent ground relaxation and tunnel deformation.

6.2.2 Calculated Results

TBM machine has an excavation diameter of De = 12.98 m, and tunnel lining has an outside diameter of Dl = 12.63 m. Therefore, the ground loss due to overcut (in the worst case possible, i.e. without control by annular grouting) is given by:
De2 Dl2 .D 2 VL % = x100% = 5.62 % VS = VL (%). = 7.04 m3/m 2 Dl 4

144

De (m) 12.98 Area (m2) 132.32

Dl (m) 12.63 125.28

Settlement Ground loss Annular trough volume VL (%) gap (cm) 3 VS (m /m) 5.62 7.04 7.04 17.5

Using the empirical formula of Herzog (1985)

D2 S max = 0.785. ( .Z 0 + Ps ) . i.E

The maximum settlements Smax are shown in Table 12. In this calculation, average modulus of elasticity of the overburden E is taken from Appendix 4 as 32.4 MN/m2 or 3,240 ton/m2 and average specific weight is taken as 2.06 T/m3. The maximum settlement Smax = 148 mm seem to be rather large. The deduced volume of the settlement trough Vs = 4.35 m3/m, and the ground loss VL = 0.035 %. Although the volume loss is small, but the maximum settlement is in practice large; therefore, the settlement must still be controlled by high quality backfill grouting during construction. Figure 104 shows the shape of settlement trough.

Table 12: Estimation of settlement by empirical formula of Herzog (1985)

Input D = Estimations (m) 12.63m 21.8m 2.06tonne/m3
2 3,240tonne/m 2 0.0tonne/m 15.5m

i1 i2 i3 i Smax VS (m3/m)

= = = = = = =

11.25 10.90 12.91 11.69 0.148 4.35 0.035

Z0 E Ps
C

= = = = =

C/D = z /D =
0

1.23shallow tunnel V % L 1.73

145

Distance from Tunnel Center Line (y), m

-3 -3 5.1 1 -2 .7 9 -2 .2 5 -2 .9 3 -1 .4 9 -1 .2 6 -1 .7 4 -1 .2 1 -9 .7 . -6 2 . -4 7 . -1 2 . 0 7 Settlement (S), mm 1.0 .7 4. 2 6. 7 9. 2 11 14 .7 16 .2 19 .7 23 .2 25 .4 29 .9 31 .2 35 .7 .1
0 20 40 60 80 100 120 140 160

"Greenfield" Cross Section E = 3240 tonne/m2 = 2.06 tonne/m3 Ps = 0 tonne/m2 Z0 = 21.8 m

Empirically predicted Smax = 148 mm

Empirically predicted Settlement Curve, volume loss = 0.035%

Figure 104: Settlement prediction for BEG Lot H3-4 tunnel by Herzog formula

Table 13 represents the dependence of the maximum settlements and volume loss on the modulus of elasticity of soils used in the Herzog formula. Modulus of elasticity E is varied from 10 MPa to 100 MPa, while keeping surcharge Ps as zero. As a result, the maximum settlements Smax range from 481 mm to 48 mm, and the ground losses VL range from 0.113 % to 0.011 %.

Table 13: Dependence of the maximum settlements and volume loss on the modulus of elasticity of soils used in the Herzog formula, without surcharge (Ps = 0)
E (MPa) Ps (MPa) Smax (mm) Vs (m3/m) VL (%) 10
20

0
0

481
241

14.10
7.05

0.113
0.056

25.1 30
32.4

0 0
0

192 160
148

5.62 4.70
4.35

0.045 0.038
0.035

40 50 60 70 80 90 100

0 0 0 0 0 0 0

120 96 80 69 60 53 48

3.52 2.82 2.35 2.01 1.76 1.57 1.41

0.028 0.023 0.019 0.016 0.014 0.013 0.011

Next, we impose a surcharge of 0.1 MPa on the surface. As a result, the maximum settlements Smax range from 588 mm to 59 mm, and the ground losses VL range from
146

0.138 % to 0.014 %. From Table 14, it is found that the maximum settlements Smax are too large, but the associated volume losses VL are too small. Considering the actual database achieved in practice by the slurry TBM technology, we can conclude that Herzog formula does not give a good compatibility between the maximum settlements and the ground losses.

Table 14: Dependence of the maximum settlements and volume losses on the elastic modulus of soils, Herzog formula, with surcharge Ps = 0.1 MPa
E (MPa) Ps (MPa) Smax (mm) Vs (m3/m) VL (%) 10 20 24.47 30 30.65 32.4 40 50 60 70 80 90 100 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 588 294 240 196 192 182 147 118 98 84 74 65 59 17.23 8.62 7.04 5.74 5.62 5.32 4.31 3.45 2.87 2.46 2.15 1.91 1.72 0.138 0.069 0.056 0.046 0.045 0.042 0.034 0.028 0.023 0.020 0.017 0.015 0.014

Using O'Reilly & New (1982)/Mair et al (1996) formula Because Lot H3-4 tunnel is driven through a stratified soil, and because z0 = 21.8 m > 1.5D = 18.945 m, so if we choose individual k factor for each layer, after that use the weighted formula of O'Reilly & New (1982) for equivalent k factor, then we have keq = 0.338 (see Table 15).

Table 15: The equivalent factor "K" and the trough width parameter "i" by using O'Reilly & New (1982) formula
Layer 1 GA_T6 Layer 2 GA_T1 Layer 3 GA_T2
GA_T3

zi (m) 1.80

ki zi x ki 0.40 0.72

keq

i (m)

4.70 6.00 4.30 5.00 21.80

0.35 0.32 0.34 0.34

1.65 1.92 1.46 1.70

Layer 4 Layer 5 GA_T3

z0 =

0.338

7.37

147

Since i = Kz0 therefore we have i = 0.34 x 21.8 m = 7.37 m, much smaller than that of Herzog (1985) formula (i = 11.69 m).

0.313VL D 2 and varying the values of factor K = 0.25 K z0 0.5 (K = 0.25 for tunnels in sands or gravels, and K = 0.5 for tunnels in clay) and volume losses VL = 0.2 % 5 %, we can obtain the maximum settlements Smax as shown in Table 16.

Using the expression S max =

Table 16: Maximum settlements for various volume loss and K factor values. New and O'Reilly (1991)/Mair et al (1996) empirical formula
VL D (m) Z0 (m) (%) 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 12.63 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 21.8 0.20 0.25 0.30 0.35 0.40 0.45 0.50 1.00 1.30 1.50 2.00 2.50 3.00 3.50 4.00 4.35 4.50 5.00 5.62
Smax (mm) K 0.30 0.40 0.338

0.25

0.45 1 1 2 2 2 2 3 5 7 8 10 13 15 18 20 22 23 25 28

0.50 1 1 1 2 2 2 2 5 6 7 9 11 14 16 18 20 20 23 25

2 2 3 3 4 4 5 9 12 14 18 23 27 32 36 39 41 45 51

2 2 2 3 3 3 4 8 10 11 15 19 23 26 30 33 34 38 42

1 2 2 2 3 3 3 7 9 10 14 17 20 23 27 29 30 33 38

1 1 2 2 2 3 3 6 7 9 11 14 17 20 23 25 26 28 32

Taking an average value of 0.338 for K factor, we have four settlement prediction curves corresponding to volume losses of 0.5%, 1 %, 1.3 %, and 2 % as shown in Figure 105.

148

Distance from Tunnel Center Line (y), m

-2 2. 1 -2 0. 9 -1 8. 4 -1 7. 2 -1 4. 7 -1 2. 4 -9 .9 -7 .4 -4 .9 -2 .4 0. Settlement (S), mm 0 2. 4 4. 9 7. 4 9. 9 12 .4 14 .7 17 .2 18 .4 20 .9 22 .1
0 5 10 15

Volume loss = 0.5 % Volume loss = 1 % Volume loss = 1.3 % Volume loss = 2%

Figure 105: Settlement predictions for BEG Lot H3-4 tunnel by formula of New and O'Reilly (1991)/Mair et al (1996)

The prediction method of O'Reilly & New (1982)/Mair et al (1996) does not account for the modulus of elasticity of the soil E as that of Herzog (1985). Therefore, the maximum settlements Smax obtained from two methods are quite different, and they can not be reasonably compared with one another, because the volumes losses deduced from the Herzog formula are always too small.

It should be recalled that, with closed face tunnelling (EPB or slurry shields), settlement is well controlled and volume loss VL is only < 0.5% in sands and VL = 1% 2% in soft clays. The tunnel at the BEG project Lot H3-4 will be bored by slurry TBM in stratified soils ranging from silt-sand to sand and sand-gravel. Therefore, assuming a practically maximum volume loss of 2.0 % is reasonable. Moreover, the theoretical volume loss due to overcut is 5.62 % corresponding with the maximum surface settlement of 38 mm (if K = 0.338 is taken), or 51 mm (if K = 0.25 is assumed). If the tunnelling crews carefully control the backfill grouting work at the shield tail to achieve a volume loss less than 2 %, then settlement can be reduced further.

6.3 Finite Element Modelling

6.3.1 Introduction New and O'Reilly (1991) already commented that: "Desktop computers can quickly provide detailed predictions of ground movements due to the most complex underground excavations. (But) considerable care must be taken in the application of

149

these models and their apparent numerical precision should never be confused for accuracy in the field. The complex interaction of ground conditions and tunnelling method demands that engineering judgement and experience will always be required to determine the appropriate data input to the models and to evaluate the predictions provided". In practice, numerical analysis technique must include credible representation of the tunnelling process, in particular the volume loss occurring, including: gap between lining and excavated surface; internal forces progressively reduced; excavated material with reduced strength and deformation properties. It must also use an appropriate constitutive model for the soil. Linear elastic soil models usually give trough widths that are too wide. The heave of the tunnel invert due to the stress relief may be encountered, which has a dramatic effect in reducing the resulting settlements. Models based on hardening plasticity involving kinematic hardening are appropriate, as they can model small strain behaviour, the effect of stress history and cyclic loading (Chiriotti, 2006). Numerical modelling is often carried out to get back-analysis for experimental data, to verify the input parameters that had been assumed in the planning and design phases of the works. According to Ercelebi et al (2005), 3D model is preferable because for 2D FE models, it is not so easy to estimate pre-relaxation factors (sometimes called stress reduction factors), which is fraction of load effecting on tunnels, and purely based on practical experience. With the 3D model, estimation of pre-relaxation factor is no longer required when excavation stages can be modelled not only in cross-section but also in the longitudinal section. Through oral communication, Prof. Swoboda suggested the author, that to calculate the displacement of TBM with a 2D model is not recommended. TBM is a really 3D problem, but 2D can give good results for the lining forces. On the other hand, Vermeer and Mller (2001, 2003) has proposed a smart use of FEM in tunnelling, in which the results from a full 3D analysis can match with that from a 2D analysis, with the use of a so-called -value (unloading factor) according to the Load Reduction Method. In this report, Plaxis code will be used to model and predict the development of surface settlements.

6.3.2 FE Analysis by Plaxis 2D Professional

Plaxis 2D program will be used to simulate the volume loss by applying a contraction to the shield tunnel lining (uniform 'shrinkage' of the lining elements in a plane normal to the tunnelling centreline). This contraction is defined during the creation of the tunnel in the input program. A contraction can be specified here because the Lot H3-4
150

tunnel is a circular tunnel (all sections having the same radius) with a homogeneous tunnel lining.

6.3.2.1 Geometry

Geometry Since the situation here is more or less symmetric, only one symmetric half (the right half) of the cross-section is taken into account in the plane strain model. From the center of the tunnel the model extends for 40 m in horizontal direction (Figure 106). The 15-node element is adopted for this analysis. Beam elements are used in Plaxis software to model the bending of tunnel lining. The behaviour of these beam elements is defined using a flexural rigidity, a normal stiffness and an ultimate bending moment. A plastic hinge may develop for elastoplastic beams, as soon as the ultimate moment is mobilized. Interfaces are joint elements which are needed for calculations involving soil-structure interaction. They are used to simulate the thin zone of intensely shearing material at the contact of tunnel lining.

Figure 106: Geometry of model, right half of the cross-section

151

Material Properties The depth-dependent moduli of elasticity of soil layers GA_T1, GA_T2 and GA_T3 are calculated by the formula f(z) = 0.1 v (0.13 z)w [MN/m2] with v = 373 and w = 0,65. Results are given in Appendix 4, together with other parameters needed for the numerical model. Average moduli of elasticity of these three soil layers are 23.9 MN/m2, 42.6 MN/m2, and 63.2 MN/m2, respectively. Properties of other soil layers GA_T4 and GA_T6 are given in Table 11 of Section 5.1. Two cases will be tested: wet density, and dry density. Properties of the precast concrete segments lining are given in Table 17.

Table 17: Material properties of support system

Parameters Name/ Symbol Value Unit Formula

Type of behaviour Thickness Area of the cross section Inertia of cross section Specific weight of concrete Specific soil weight Young's modulus Axial/Normal stiffness Flexural rigidity (bending stiffness) Equivalent beam thickness Equivalent beam weight Poisson's ratio Maximum moment Maximum axial force

Material type tc A I c s E EA EI d

Elastic 0.5 0.5 0.0104 25 20 37,000,000.00 18,500,000.00 385,416.67 0.5

m m2 m4 T/m3 T/m3 kPa kN/m kNm2/m m

d eq = 12 EI EA

tc 3 .1 I= 12

(37 kN/mm2)

w Mp Np N/A N/A

7.5 0.2

kN/m/m kNm/m kN/m

w = c .tc s .

tc 2

(1x105 units by default)

(4.33x106 units by default)

152

2D mesh generation The 15-node element is used as the basic element type. The global coarseness will be varied from coarse to very fine, i.e. from around 500 to 1000 elements for the right half of the given cross section.

Initial conditions Before the generation of the initial stresses the tunnel lining is to be deactivated. The initial stress generation (K0-procedure) can be used to generate the initial effective stresses with the appropriate values of K0. The initial stresses in a soil body are influenced by the weight of the material and the history of its formation. This stress state is usually characterised by an initial vertical stress v,0 which is related by the coefficient of lateral earth pressure K0 (h,0 = K0.v,0). In reality, the coefficient, K0, represents the ratio of the horizontal and vertical effective stresses:
K0 =
' xx ' yy

In practice, the default K0-value is based on Jaky (1944) formula (K0 = 1 - sin) as an empirical expression for a normally consolidated soil. In PLAXIS initial stresses may be generated by specifying K0 or by using Gravity loading. The K0-procedure should be used in cases with a horizontal surface and with any soil layers and phreatic lines parallel to the surface. For all other cases Gravity loading shall be used. Very low or very high K0-values may cause initial plasticity. Using K0-values which differ substantially from unity may sometimes lead to an initial stress state which violates Coulomb's failure criterion.

6.3.2.2 Calculations

The construction of the tunnel is stimulated by a staged construction calculation in which the tunnel lining is activated and the soil clusters inside the tunnel are deactivated. Deactivating the soil inside the tunnel only affects the soil stiffness and strength and the effective stresses. Following are the calculation steps in Plaxis. The first calculation phase is a plasticity calculation, applying load advancement ultimate level. For the loading input, Staged construction is selected. Within the staged construction mode, activate the tunnel lining and deactivate the two soil clusters inside the tunnel. In addition to the installation of the tunnel lining, the excavation of the soil and the dewatering of the tunnel, the volume loss is simulated by applying a contraction to the shield tunnel lining.

153

The Contraction parameter is defined as the reduction of the tunnel area as a percentage of the original tunnel area. Activation of the contraction procedure results in a homogeneous 'shrinkage' of the tunnel lining, which reduces the cross section area of the tunnel. In order to activate this contraction, the procedure is: Select a plastic calculation, load advancement ultimate level and select Total multipliers as loading input. Enter a contraction value of 0.5 2.5 for the parameter McontrA. This is the multiplier that controls the contraction of the tunnel referred to as 'A' in the geometry model. Select some characteristic points for load-displacement curves (for example the corner point at the ground surface above the tunnel, and the point on top of lining). Start the calculations.

It is noted that, the contraction of the shield tunnel lining by itself does not introduce forces in the tunnel lining. Eventual changes in lining forces as a result of the contraction procedure are due to stress redistributions in the surrounding soil or to changing external forces.

Through many runs with variations of lining thickness (from 30 to 60 cm), mesh coarseness (500 1000), and contraction value (0.5 2.5), it is observed that the soil body collapses at different contraction values, and settlements as well as lining member forces also fluctuate accordingly. Finally, plane strain computation is made for the actual structure as follows: lining thickness of = 50 cm; number of elements = 959; number of nodes = 7868; number of stress points = 11508

Computed results: contraction converged (soil body collapses) at 1.3 % for the case of using wet density of soil; extreme total displacement is 82 mm (Figure 107) contraction converged at 1.01 % if the dry density of soil is used; extreme total displacement is 64 mm (Figure 108)

154

Figure 107: Deformed mesh (displacement scaled up 50 times)

Figure 108: Shadings of vertical displacements

155

It can be seen from Figure 109 that, the vertical displacements on the surface (52 mm) are smaller that that on the tunnel crown (64 mm). But the trough width at the tunnel crown level is narrower and steeper than at ground level.

a)

b) Figure 109: Settlement troughs on the surface a) and at crown level b)

Figure 110 shows the development of settlements of four different points in the ground when the contraction is increased. When the contraction reaches 1.01 %, soil body collapses, the settlement of Point A at the tunnel bottom attains 10 mm, the settlement of Point B on top of tunnel (crown) is 64 mm, the settlement of Point C on the side of tunnel (haunch) is 40 mm, and the settlement of Point D near the ground surface is 52 mm.

156

Figure 110: Development of vertical displacements of different points In Table 18 and 19 are compared the settlements obtained from the empirical analysis and from 2D modelling, at the same values of volume loss of 1.3 %, and 1.0 %, respectively. The settlement given by 2D modelling is greater than that of empirical formula.

Table 18: Comparison of settlement results between empirical methods and 2D analysis, VL = 1.3 % (using wet density of soil)
O'Reilly & New 2D Plaxis (1982)/Mair et al modelling (1996) formula Max settlement (mm) Volume loss (%) 9 1.30 (k=0.338) 82 1.3

157

Table 19: Comparison of settlement results between empirical methods and 2D analysis, VL = 1.0 % (using dry density of soil)
Mair et al (1996) 2D Plaxis formula modelling Max settlement (mm) 7 64 Volume loss (%) 1.00 1.01 (k=0.338)

Lining forces Normal force and bending moment in tunnel lining are shown in Figure 111. Extreme axial force is -947.08 kN/m, extreme shear force is 110.80 kN/m, and extreme bending moment is -329.78 kN/m. These member forces can be used to design the reinforcement for the lining.

a) Axial forces

b) Shear forces

158

-2.50

0.00

2.50

5.00

7.50

10.00

12.50

15.00

17.50

7.50

5.00

2.50

0.00

-2.50

-5.00

c)

Bending moment Extreme bending moment -329.78 kNm/m

Figure 111: Axial forces a), shear forces b), and ending moment c) in the lining

Stress paths A stress path represents the development of the stress state at a local point of the geometry. The visualization of stress paths provides a valuable insight into local soil behaviour. The stress paths of four points are shown in Figure 112: Point F at the tunnel bottom, Point G on top of tunnel (crown), Point H on the side of tunnel (haunch), and Point I near the ground surface.

Figure 112: Stress paths of different points

159

6.3.3 Face Stability by Plaxis 3D Tunnel

Tunnel heading stability, ground deformation/surface settlements, and loads on lining are the three main focuses of tunnel analyses (Vermeer, 2001). The shield tunnel construction can be modelled as a stepwise process. During the erection of the tunnel lining the tunnel boring machine (TBM) remains stationary. Once a tunnel lining ring has been fully erected, excavation is resumed, until enough soil has been excavated to erect the next lining ring. As a result, the construction process can be divided in construction stages (slices) with a length approximately equal to a tunnel ring. The excavation process in each staged construction phase is: the support pressure at the tunnel face needed to prevent active failure at the face, the conical shape of the TBM shield, the excavation of the soil and pore water within the TBM, the installation of the tunnel lining and the grouting of the gap between the soil and the newly installed lining (Figure 113).

contraction of shield

final lining

grout pressure

TBM
face pressure

Figure 113: Construction stages of a shield tunnel

Although runs of the 3D phased excavation of a shield tunnel with limited excavation steps can only be completed in approximately several hours, it is worth performing, considering the value of the information it is able to provide and the complexity of the model. However, in this subsection only the stability of the tunnel face will be investigated. Only the TBM is included in the model and the tunnel lining is not modelled. The purpose is to search for the minimum face pressure that is required to keep the tunnel heading stable by lowering the original face pressure until collapse occurs.

160

6.3.3.1

Geometry

In the model, again only one symmetric half is included, but it is the left half in this case. The model is 40.0 m wide, it extends 55.0 m in the z-direction and it is 52.0 m deep. With these dimensions (over three times of tunnel diameter in three directions), the model is sufficiently large to allow for any possible collapse mechanism to develop and to avoid any influence from the model boundaries. The tunnel excavation process is simulated in one excavation stage. The interaction between the TBM and the soil is modelled by means of an interface. The tunnel face pressure is modelled by means of a z-load, which is applied in the excavation stage.

Boundary conditions Standard fixities function of the program will generate full fixities at the bottom, vertical rollers at the vertical sides and rotation fixities at the ends of the tunnel. The geometry model is shown in Figure 114.

Figure 114: Geometry model in the Input window

Material properties The material properties of the soil clusters and other geometry objects are entered in data sets. Interface properties are included in the data sets for soil. In addition to the material data sets for soil and interfaces, a data set of the plate type is created for the TBM, with the properties as given in Table 20.

161

Table 20: Material properties of the TBM

TBM Parameters Type of behaviour Axial/Normal stiffness Flexural rigidity (bending stiffness) Equivalent thickness Weight Poisson's ratio Name Material type EA Value Elastic 12,600,000 Unit kN/m

EI d w

85,000 0.285 50.77 0.00

kNm2/m m kN/m/m -

Mesh generation The 2D mesh should be made fully satisfactory before proceeding to the 3D mesh extension. The basic volume elements of the 3D finite element mesh are the 15-node wedge elements. In addition to the basic volume elements, there are special elements for structural behaviour (plates, geogrids and anchors). PLAXIS allows for a fully automatic generation of 2D finite element meshes and a semi-automatic generation of 3D meshes. The standard very coarse mesh is used first. 2D mesh is shown in Figure 116. Model: plane strain; Elements: 6-noded; number of elements: 68; number of nodes: 170; number of stress points: 204.

Figure 115: 2D finite element mesh

162

3D mesh is shown in Figure 116. Model: 3D parallel planes; Elements: 15-noded wedge; number of elements: 612; number of nodes: 2168; number of stress points: 3672.

Figure 116: 3D finite element mesh

Initial conditions The initial conditions of the current project require the generation of water pressures and the generation of initial stresses. The generation of water pressures (i.e. pore pressures and water pressures on external boundaries) is based on the input of phreatic levels of 17.5 m (Figure 117). The initial stresses are generated by means of the K0procedure (Figure 118).

163

Figure 117: Active pore pressures (initial)

Figure 118: Effective mean stresses (initial)

6.3.3.2

Calculations

We concentrate on the tunnel heading stability and consider that the TBM has already advanced its own length (11.35 m) into the soil. The first construction phase will consist of the excavation of the soil to allow the installation of the TBM, the

164

application of the TBM itself, the lowering of the water level in the TBM, the application of the tunnel face pressure and the application of contraction to simulate the fact that the TBM is conical towards its tail (0.5 %). The adapted material sets (with reduced interface friction and adhesion) are assigned to the first slice in which the tunnel is excavated. The tunnel face pressure needs to be applied to the face of the TBM, and is maintained by a fluid (bentonite) with a unit weight of 14.0 kN/m3. The tunnel face pressure is 120.0 kN/m2 in the negative z-direction at the top of the tunnel (+6.49 m) and 302.0 kN/m2 at the bottom (-6.49 m). The pressure gradient is 14.0 kN/m2/m. A reference ordinate yref of +6.49 m (corresponding to the top of the tunnel), a reference pressure pref of -120.0 kN/m2 and a pressure increment pinc of -14.0 kN/m2/m are introduced. A contraction value will be introduced to model a shortening of the tunnel shell and thus a reduction of the tunnel radius during the calculation, i.e., to simulate the soil volume loss around the tunnel due to overcutting, conicity of the TBM, or any other cause. The value of contraction defines the cross section area reduction as a percentage of the whole tunnel cross section area. Here, the tail of the TBM will be given a contraction of 0.5% to simulate the conicity of the TBM. Inside of the tunnel will be set dry in the first excavation phase. Active pore pressures are given in Figure 119.

Figure 119: Active pore pressures in the first excavation phase

165

The minimum required tunnel face pressure can be found by reducing the tunnel face pressure until the tunnel heading collapses. Calculation type is 3D plastic. Calculation phase 2 starts from Phase 1. All loads defined as load system A (in this case only the Z-Load representing the tunnel face pressure), will gradually be reduced to 0. Several nodes and stress points will be selected for a later generation of loaddisplacement curves and stress and strain diagrams. On the TBM workface plane are chosen: Nodes for load-displacement curves: Point A is at the bottom, point B at tunnel haunch, point C on top of tunnel and point D at the ground surface right above the tunnel. Stress points for Stress/Strain curves: Point E is at the bottom, point F at tunnel haunch, point G on top of tunnel and point H near the ground surface right above the tunnel. The first calculation phase should successfully finish (Figure 120). It can be seen that the original face pressure is sufficiently high to keep the tunnel face stable. The displacements at the tunnel face are very small. The largest deformations 30 mm occur above the tail of the TBM. This is due to the applied contraction.

Figure 120: Deformed mesh at the end of phase 1

The total and incremental values of the realised contraction can be seen from Figure 121. The total realised contraction of 0.52 % almost corresponds to the input value of 0.5 %.

166

Figure 121: Realized value of contraction, at Front Plane (left) and Face Plane (right)

The second calculation phase should not successfully finish, because the Prescribed ultimate state not reached, and Soil body collapses (Figure 122).

Figure 122: Finish of the calculation phases

167

The Multipliers parameter -MloadA has reached a value of 0.9058, so the minimum tunnel face pressure required to prevent failure is 0.9058 x 120.0 = 108.7 kN/m2 at the top and 0.9058 x 302.0 = 273.3 kN/m2 at the bottom of the tunnel. This gives an idea about the safety of the tunnel heading against active failure. The total displacement is 31 mm.

Safety Analysis It is important to consider not only the final stability, but also the stability during construction. The stability against failure can be defined by means of a safety factor. A safety factor can be defined as the ratio of the available shear strength to the computed minimum strength required for equilibrium:
Safety factor = Savailable Sneeded for equilibrium c + n tan cr + n tan r

By introducing the standard Coulomb condition, the safety factor is obtained as:
Safety factor =

Where c and are the input strength parameters and n is the actual normal stress component. The parameters cr and r are reduced strength parameters that are just large enough to maintain equilibrium. The principle described above is the basis of the method of Phi-c reduction that can be used in PLAXIS to calculate a global safety factor. In this approach the cohesion and the tangent of the friction angle are reduced in the same proportion:
c tan = = Msf cr tan r

The reduction of strength parameters is controlled by the total multiplier Msf. This parameter is increased in a step-by-step procedure until failure occurs (calculation phase 3). The safety factor is then defined as the value of Msf at failure. To calculate the global safety factor for the situation of the original face pressure, the first increment of the multiplier for strength reduction (Msf ) is preset to 0.1. The deformations obtained for the second and the third phase will be shown below for comparison. Both the second and the third phase represent a collapse situation. The second phase (face pressure reduction) shows the soil locally moving inwards (Fig. 123), whereas the third phase (phi-c reduction) shows a chimney-like failure mechanism reaching to the ground level (Fig. 124).

168

Figure 123: Displacement increments at the end of Phase 2 (face pressure reduction)

Figure 124: Displacement increments at the end of Phase 3 (Phi-c reduction) The development of the -MloadA multiplier and the development of the -Msf multiplier can be viewed in Figure 125 and Figure 126. Figure 125 shows that, in the second phase, MloadA reaches a value of 0.9058, at which large inward movements of the tunnel face occur. Figure 126 shows that, in the third phase, Msf (may be regarded as a global safety factor) reaches a value of 2.027, at which large inward movements of the tunnel face occur.

169

However, in this type of application the procedure of phi-c reduction does not give a realistic safety factor. This is because the problem is very much dominated by the tunnel face pressure, which is not reduced in the phi-c reduction procedure. The method of phi-c reduction is much more applicable for embankment or slope stability problems, and does give a realistic safety factor in such cases.

Figure 125: Development of Sum-MloadA as a function of the displacements

170

Figure 126: Development of Sum-Msf as a function of the displacements

Concluding notes It should be noted, that this type of tunnel face stability can only be justified if the material properties of the TBM are correctly given by the machine manufacturer. In order to avoid mistakes in the modelling of face stability as well as phased excavation of a shield tunnel with the use of Plaxis 3D Tunnel program, it is desirable to: prescribe enough displacements on active mesh; avoid very slender elements; refine around tunnel(s); and reduce stiffness differences.

6.4 Summary
Surface settlements were predicted for the tunnel to be excavated in the section of Km 33+480 to 33+800 of BEG project line using both empirical and numerical prediction methods.

171

With the analytical methods, different empirical formulae result in different settlements with the same volume loss. 2D analysis presented the bigger values of settlement with the same volume loss. However, these may give some useful indications before construction begins. Although 3D analysis of face stability gave out a total settlement of 31 mm, it is unsuitable to make a comparison with the aforementioned results, since it did not account for staged excavation and grouting pressure as well as lining installation.

Limitations of the analytical/empirical methods are that, they are specific to soil type and unable to account for soil-lining interaction. Empirical methods may be useless for the complex structural configurations. Actual construction works commonly comprise a variety of intersecting excavations where tunnels may change diameter (e.g. stations) and where cross connecting adits occur. It is often difficult to readily calculate important 3D ground movements at these complex locations. Limitations of the numerical method are that, when modelling the shield tunnelling problem in 2D, the extrusion of the face may hardly be modelled directly, and the users may have to rely on some "tricks" during generating the mesh and base on approximations of the lining behaviour. Phased construction also has never been exactly modelled. For 2D FE models, it is not so easy to estimate pre-relaxation factors (sometimes called stress reduction factors), which is fraction of load effecting on tunnels, and purely based on practical experience. The Plaxis software's assumption that the contraction of the tunnel lining (applied to simulate the ground loss) by itself does not introduce forces in the lining is somewhat artificial. It could be interesting if some correlation is made to the load reduction method and the stiffness reduction method used for taking into account 3D effects in 2D analyses (Schweiger et al, 1997). Also, one can have recourse to the explanation of Augarde et al, in that the hoop shrinkage is achieved by the application of a suitable set of radial forces within the tunnel liner. The process leads to fictitious stresses within the liner, but the liner is elastic and so this does not affect the way in which the ground and the liner interact. The two methods can be used together for a given project to cross-check one another. The FEM is a strong tool but it still depends on the qualification of the users, not including specific adaptations or approximations. And it is vital that the output from the analysis is checked carefully. It is hoped that the future developments and enhancement of the presently available specialist codes will help to address the outstanding problems.

172

Chapter 7 7. Conclusions and Future Work

Conclusions In preparation of this final project, the author has investigated both theoretical and practical aspects of interactions within TBM excavations as well as interfaces between TBM and lining. The theoretical part attempted to make clear a number of critical points that have to be paid due attention in the design and construction of tunnels. Numeral calculations have been conducted to illustrate the theoretical part. In review of the report, a few points are recapped below. A rather comprehensive investigation has been performed on the interface between TBM and lining, from a practical engineering standpoint, which helps to make clear many outstanding issues relating to TBM tunnels design, construction and project management. High performance TBMs are essential for the successful construction of tunnel projects. The TBMs will be purpose built machines using proven state of the art technology and designed specifically for the project to a minimum specification to ensure their reliability in terms of performance and settlement control. They should be designed to cater for the range of ground conditions anticipated. Given the advances in tunnelling technology today, tunnels can be excavated in virtually all types of soil and environment, but there are still numerous unknown parameters. One of the most important is the knowledge of the ground through which the tunnel will be routed. Therefore, geotechnical design should be based on adequate geotechnical investigation, then data evaluation provided, and monitoring program elaborated. Risk management procedures should be provided for to cover all the possible risks; prepare measures to deal with that risks including corresponding cost estimate. Risks have to be considered from the first steps of the tunnel project, through installation process to its operation. Budget for risk management should be allocated and defined in tender documents and contract requirements. For the purpose of understanding and managing the surface settlement, the problem of face stability has been reviewed. Countermeasures to ground failure due to urban tunnelling are also given, with the particular use of grouted bodies. A case history of Metro Torino has been presented to show the application of those grout-consolidated slabs.
173

 Ground settlements caused by shielded tunnelling operations have been examined with both analytical and numerical approaches, then results compared. Face heading stability has also been analyzed by 3D numerical tool. Numerous parameter studies have helped the author to gain a critical view on the use of the available approaches. Semi-empirical methods must be applied with caution, and finite element analysis with geomechanical software must be used toward an effective way.

Recommendations for future studies Simulation of the subsidence induced by the excavation of single, twin tunnels or even more complex configurations, taking into account the presence of surface structures or structures within the zone of influence of the tunnels. A clear presentation of the effects of tunnelling on overlying structures is obviously the next step to pursue. Design of segmental concrete linings for tunnels in soft soils using 3-D numerical modelling, with the reasonable consideration of soil-structure (ground-lining) interaction. It would be useful if different commercial packages such as PLAXIS, FLAC, Phase2, etc. are used together with analytical methods to compare and validate the respective results. Otherwise, study in the feasible alternatives for the realization of deep and long tunnels coming from various markets, such as the railway link beneath the Strait of Gibraltar, which have an essential need for implementation of flexible responses due to observations and definitely involves the discussed critical interfaces, both contractually and technically.

174

List of Acronyms
ENGLISH ABI ACI ANN AR BOQ BS BTS CBP CEVP CHD CLI COE CSI D&B DAT DRB DSC DSU DTA ELS EPB EPBM/EPBS EPDM EPFL EU FEM FER FIR FLAC GBR : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : Association of British Insurers American Concrete Institute Artificial Neural Networks Advance Rate Bill Of Quantities British Standards British Tunnelling Society Controlled Boring Process Cost Estimate Validation Process Cutter Head Drive Cutter Life Index U.S. Army Corps of Engineers Computers and Structures, Inc. Drill & Blast / Drilling and Blasting method Decision Aids in Tunnelling Disputes Review Board Differing Site Condition Double Shield Universal Designed Tunnel Axis Electronic Laser System Earth Pressure Balance Earth Pressure Balance Machine/Shield Ethylen-Propylen-Dien Material Ecole Politechnique Fdrale de Lausanne European Union Finite Element Method Foam Expansion Ratio Foam Injection Ratio/Rate Fast Lagrangian Analysis of Continua Geotechnical Baseline Report

GIS GTS HM HS ICE IMIA ITA-AITES ITIG JSCE MFS MIT MT NATM NMT PLAXIS PR QA/QC RMP RPM RQD SCL SFRC SM SRF SS TBM TEN UCS VAL WSDOT

: : : : : : : : : : : : : : : : : : : : : : : : : : : : : :

Geographical Information System Gyro Tunnelling System Hand Mining Hardening Soil model Institution of Civil Engineers International Association of Engineering Insurers International Tunnelling Association - Association Internationale des Travaux en Souterrain International Tunnelling Insurance Group Japan Society of Civil Engineers Multi-circular Face Shield Massachusetts Institute of Technology Mechanized Tunnelling New Austrian Tunnelling Method Norwegian Method of Tunnelling PLasticity AXISymmetry Penetration Rate Quality Assurance/Quality Control Risk Management Plan Revolutions Per Minute Rock Quality Designation Sprayed Concrete Lining Steel Fiber Reinforced Concrete Shield Machine Stress Reduction Factor Soft Soil model Tunnel Boring Machine Trans-European Transport Network Uniaxial Compressive Strength Automated Light Vehicle Washington Department of Transportation

FRENCH AFTES FIDIC : : Association Franaise des Travaux en Souterrain (French Association of Tunnels and Underground Space) Fdration Internationale des Ingniers-Conseils (International Federation of Consulting Engineers) GERMAN AVN BEG DAUB DIN BB NORM : : : : : : Automatischer Vortrieb Na (Automatic/remote controlled Wet/slurry Driving/tunnelling machine) Brenner Eisenbahn GmbH (Brenner Railway Ltd.) Deutscher Ausschu fr unterirdisches Bauen (German Committee for Underground Construction) Deutsches Institut fr Normung e.V. (German Institute for Standardadization, registered society) sterreichisches Bundesbahn (Austrian Federal Railway) sterreichisches Institute) Normungsinstitut (Austrian Standards

ADECO-RS COREP GTT RFI

: : : : : : :

ITALIAN Analisi delle Deformazioni Controllate nelle Rocce e nei Suoli (Analysis of Controlled Deformation in Rocks and Soils) Consorzio per la Ricerca a lEducazione Permanente (Consortium for the Research and Permanent Education) Gruppo Torinese Trasporti (Turin Transportation Group) Rete Ferroviaria Italiana SpA (Italian Railway Network; or National Railway Infrastructure Administration) Societ Esecuzione Lavori Idraulici S.p.A. (Hydraulic Works Construction Company) Societ Italiana Gallerie (Italian Tunnelling Society) Societ per Azioni (Joint stock company) RUSSIAN (Construction standards and regulations)

SELI SIG S.p.A

(SNiP)

References
[1]

AFTES (1995) Settlements induced by tunnelling. Association Franaise des Travaux en Souterrain (The French Association of Tunnels and Underground Space). AFTES (1999). Recommendation for the design, sizing and construction of precast concrete segments installed at the rear of a tunnel boring machine (TBM). The French Association of Tunnels and Underground Space. AFTES 2000. New recommendations on Choosing mechanized tunnelling techniques. French Tunnelling and Underground Engineering Association. Anagnostou G., Kovari K. (1994). The Face Stability of Slurry-shield-driven Tunnels. Tunnelling and Underground Space Technology, Volume 9, Issue 2, 165-174. Andreossi E. 2001. Costs - The Insurers Point of View. International Centre for Geotechnics and Underground Construction (CUC). BEMA, Switzerland. Arrigoni G.A. 2006. Project management - Contractual and legislative aspects. Lecture at Master Course in Tunnelling and TBMs, Edition V 2005-2006, Politecnico di Torino, Italia. Assis A. P. 2006 - Examples of Underground Excavations in Rock (Part 1); Gibraltar Strait Crossing (Part 4). Lecture series at Master Course in Tunnelling and TBMs, Edition V 2005-2006, Politecnico di Torino, Italia. Augarde C.E., Burd H.J. and Houlsby G.T. Some experiences of modelling tunnelling in soft ground using three-dimensional finite elements. 4th European Conference on Numerical Methods in Geotechnical Engineering. Barla G. 1998. Tunnelling under squeezing rock conditions. Department of Structural and Geotechnical Engineering, Politecnico di Torino, Italia.

[2]

[3] [4]

[5] [6]

[7]

[8]

[9]

[10] Barla G. & Pelizza S. 2000. TBM tunnelling in difficult ground conditions. Proc. GeoEng2000, Melbourne. [11] Barton N., 1999b: TBM performance estimation in rock using QTBM. Tunnels & Tunnelling, September 1999, pp. 30-34. [12] Barton N. TBM tunnelling in jointed and faulted rock. Taylor & Francis; 1 edition (January 1, 2000). [13] Barton N. 2006. Rock Engineering with Q. Series of 5 Lectures at Master Course in Tunnelling and TBMs, Edition V 2005-2006, Politecnico di Torino, Italia. [14] BEG (Brenner Eisenbahn GmbH - Ein Unternehmen der BB). Brenner Railway Axis Tunnels, Bau-Information 2005. http://www.beg.co.at/. [15] Beil A., Horny U. (2005). Design of Final Lining on BEG Railway Project - high speed railway in Inn valley. Seminar on Tunnel Final Lining - Design, Construction, Maintenance, Praha, Czech Republic.

[16] Bickel J. O., Kuesel T. R., and King E. H. (1996) Tunnel Engineering Handbook. Second Edition, Chapman & Hall, ITP 1996. [17] Brinkgreve R.B.J., Vermeer P.A. et al (1998). Plaxis, Version 7, Complete set of manuals, Finite element code for soil and rock analyses. PLAXIS B.V. A.A. Balkema Publisher. [18] Brinkgreve R.B.J., Vermeer P.A. (2001). Plaxis 3D Tunnel, Version 1, Complete set of manuals, Finite element code for soil and rock analyses. PLAXIS B.V. A.A. Balkema Publishers. [19] Chiriotti E. & Romano M. 2006. Ground and surface monitoring in urban environment. PART 3 & 7 - Methods for predicting settlements and deformations due to tunnelling. Turin-based GEODATA Spa. Lectures at Master Course in Tunnelling and TBMs, Edition V, Politecnico di Torino, Italia. [20] COE (U.S. Army Corps of Engineers). Tunnels and shafts in rock, 1997. [21] Concilia M. 2005. Construction of Lot 5 of the Turin Metro Line 2 Tunnel. "Injection of fines into the cutterhead chamber extends the ground range of application of EPB TBMs". FELSBAU_2005. [22] Concilia M. 2006. Hard Rock Tunnelling - TBM types, Facilities and work sites organization. SELI S.p.A, Lectures at the Master Course in Tunnelling and TBMs, Edition V, Politecnico di Torino, Italia, Feb. [23] Cross London Rail Links Ltd. 2005. D12 - Ground Settlement. Crossrail Bill supporting documents. Available on the websites http://www.crossrail.co.uk/; http://billdocuments.crossrail.co.uk/ [24] Crova R. 2006. Example of Turin Metro Line 1. Gruppo Torinese Trasporti (GTT), Lectures at the Master Course in Tunnelling and TBMs, Edition V, Politecnico di Torino, Italia, April. [25] Dal Negro E. 2006. Backfill grout. UTT Mapei, Lecture at the Master Course in Tunnelling and TBMs, Edition V, Politecnico di Torino, Italia, Feb. [26] Duddeck H. (ed.) (1988). Guidelines for the design of tunnels. ITA Working Group on General Approaches to the Design of Tunnels. Tunneling and Underground Space Technology 3, pp 237-249. [27] Eisenstein Z. The Future of Mechanized Tunnelling: Introduction to the ITA Open Session at the Toronto Congress 1989. Tust_Vol_6_2_167-189. [28] Ercelebi S.G., Copur H., Bilgin N. & Feridunoglu C. Surface settlement prediction for Istanbul metro tunnels via 3D FE and empirical methods. Underground Space Use: Analysis of the Past and Lessons for the Future Erdem & Solak (eds) 2005 Taylor & Francis Group, London, ISBN 04 1537 452 9. [29] Franzius J. N. 2003. Behaviour of buildings due to tunnel induced subsidence. Doctoral thesis. University of London. [30] Garshol K.F. 2003. Pre-Excavation Grouting in Rock Tunneling. MBT International Underground Construction Group, Division of MBT (Switzerland) Ltd.

[31] Grandori, R., De Biase, A., Busillo, A., Ciamei, A. 2003. Turin Metro System Design and Operation of EPB-TBMs beyond the Limits of this Technology. Felsbau 21(6): 34-42. [32] Grasso P. 2001. Risk analysis for long tunnels at great depth - Works Planning & Financing Engineering - Design of the Safety and Smoke Control System. ITA-WG17. [33] Grasso, P., Mahtab, M., Kalamaras, G. & Einstein, H. 2002. On the Development of a Risk Management Plan for Tunnelling. Proc. AITES-ITA Downunder 2002, World Tunnel Congress, Sydney, March. [34] Grasso P.G. 2006. Requirements for Deep and Long Tunnels. International Seminar on Tunnels and Underground Works. LNEC, Lisbon, 29-30 June 2006. [35] Gruebl F. 2006. Driving of full face machines. Lecture at Master Course in Tunnelling and TBMs, Edition V, Politecnico di Torino, Italia. [36] Gruebl F. 2006. Modern design aspects of segmental lining. International Seminar on TUNNELS AND UNDERGROUND WORKS. LNEC, Lisbon, 29 30 of June 2006. [37] Haack A. 1996. Comparison between conventional tunnel driving method and TBM drives. Worldwide demand of tunnel constructions. Tunnel Boring Machines - Trends in Design and Construction of Mechanized Tunnelling. Hagenberg Austria 14-15 Dec. 1995, Wagner & Schulter (eds), pp. 89-98. Balkema/Rotterdam 1996. [38] Haack A. 2003. Technical options for fireproof tunnel linings - Limits, advantages and disadvantages of the various solutions. 1 Congresso Brasileiro De Tneis E Estruturas Subterrneas. Seminriointernacional South American Tunnelling. [39] IMIA-wgp18. Tunnel Boring Machine (TBM) applications in soft ground conditions. IMIA (International Association of Engineering Insurers) Meeting 2001, Sydney. [40] ITA-WG14. 2000. Recommendations and Guidelines for Tunnel Boring Machines (TBMs). International Tunnelling Association, Working Group No. 14 - Mechanized Tunnelling. [41] ITA-WG2. 2000. Guidelines for the Design of Shield Tunnel Lining. International Tunnelling Association, Working Group No. 2 - Research. [42] ITA-WG2. 2004. Guidelines for tunnelling risk management. International Tunnelling Association, Working Group No. 2 - Research. [43] ITIG - EP24. A Code of Practice for Risk Management of Tunnel Works. International Tunnelling Insurance Group. Jan. 2006. [44] JSCE 1996. Japanese standard for shield tunnelling. Japan Society of Civil Engineers. The third edition. [45] Kochen R. 1992. Pre - Supported Soft Ground Tunnels. Canadian Tunnelling. An Annual Publication of the Tunnelling Association of Canada.

[46] Kovari K., Anagnostou G. 1996. Face stability in slurry and EPB shield tunnelling. Proc Int. Symp. on Geotechnical Aspects of Underground Construction in Soft Ground, City University, Balkema, 453-458. [47] Kovari K. 1998. Tunnelbau in druckhaftem Gebirge / Tunneling in Squeezing Rock. Tunnel 5/1998. [48] Kovari K., Ramoni M. 2004. Urban tunnelling in soft ground using TBM's. Key note lecture at International Congress on Mechanized Tunnelling: Challenging Case Histories. Politecnico di Torino, Italy 16-19 November 2004. [49] Leca E., Dormieux L. 1990. Upper and lower bound solutions for the face stability of shallow circular tunnels in frictional material. Gotechnique, Volume 40, No. 4, pp. 581-606. [50] Leca E., Leblais Y. & Kuhnhenn K. 2001. Underground works in soils and soft rock tunneling. [51] Maidl B., Herrenknecht M., Anheuser L. Mechanized Shield Tunnelling. Ernst & Sons, 1996. [52] Mair R.J., Taylor R.N. and Burland J.B. (1996). Prediction of ground movements and assessment of risk of building damage due to bored tunnelling. In: Proc. of the Int. Symp. on Geotech. Aspects of Underground Construction in Soft Ground, 713-718, Balkema, Rotterdam. [53] Mansour M.A.M. Three-dimensional numerical modelling of hydroshield tunnelling. Ph.D. Thesis, University of Innsbruck, 1996. [54] Martins J.B. 2001. A Short Survey on Construction Problems and Numerical Modelling of Shallow Tunnels. Depart. Eng. Civil, Universidade do Minho, 4800 Guimares, Portugal. Engenharia Civil - UM, Nmero 11, 2001. [55] Mayreder Consult (D2 Consult). Paris EOLE - Est Ouest Liaison Express - Lot 35 B. CONEX - Tunnel Lining System. Structural analysis. 1993. [56] Muir Wood A. M., Fischer H. C., Girnau G., Lemley J., Broch E., Kirkland C., Eisenstein Z., Pelizza S., Haack A. Past, Present and Future of Tunnelling. Published in "1974 - 1999 ITA - AITES 25 Years", pp. 1 - 50, Year 1999 by ITA - AITES, www.ita-aites.org. [57] Munich Reinsurance Company 2004. Underground transportation systems, Chances and risks from the reinsurers point of view. [58] New B.M, O'Reilly M.P. 1991. Tunnelling induced ground movements; predicting their magnitude and effects. J.D. Geddes Ground movements and structures, Proc. of 4th International Conference, University of Wales College of Cardiff 1991, London. Pentech Press, 1992. pp. 671-697. [59] Oggeri C. 2006. Tunnel monitoring. Lecture at Postgraduate master course in Tunnelling and Tunnel Boring Machines, V Edition 2005-06, Politecnico di Torino, Italia. [60] Oreste P.P., Peila D. and Poma A. 1999. Numerical Study of Low Depth Tunnel Behaviour. Proc. Int. World Tunnel Congress 99, Oslo, pp. 155-162.

[61] Palmstrm A. & Broch E. 2006. Use and misuse of rock mass classification systems with particular reference to the Q-system. (Unpublished). [62] Parker H. W. 1999. Geotechnical investigations for tunnels. AFTES - Journes dtudes internationales de PARIS - 25 au 28 octobre 1999. [63] Parker H. W. 2005. Risks and Choices are plentiful - George Fox Seminar [64] Peck R.B. (1969). Deep excavations and tunneling in soft ground. Proceedings 7th International Conference. Soil Mechanics and Foundation Engineering, Mexico, State-of-the-Art Volume, pp. 225-290. [65] Peila D. 2006. Guidelines for tunnelling design and construction. Postgraduate master course in Tunnelling and Tunnel Boring Machines, V Edition 2005-06, Politecnico di Torino, Italia. [66] Peila D. 2006. Ground reinforcing in tunnelling. Postgraduate master course in Tunnelling and Tunnel Boring Machines, V Edition 2005-06, Politecnico di Torino, Italia. [67] Pelizza S. 2006. General aspects of tunnel construction. Postgraduate master course in Tunnelling and Tunnel Boring Machines, V Edition 2005-06, Politecnico di Torino, Italia. [68] Pelizza S. and Peila D. 2006. Rock TBM Tunnelling. Postgraduate master course in Tunnelling and Tunnel Boring Machines, V Edition 2005-06, Politecnico di Torino, Italia. [69] Pescara M. (2005). Dimensioning of segmental lining. Geodata Spa. (Turinbased). Postgraduate master course in Tunnelling and Tunnel Boring Machines, V Edition 2005-06, Politecnico di Torino, Italia. [70] Plizzari G. A., Tiberti G. 2006. Fibers for Concrete Reinforcement. Lectures at Master Course in Tunnelling and TBMs, Edition V 2005-2006, Politecnico di Torino, Italia. [71] Rehm U. 2006. Global Overview of Shields Machines. Postgraduate master course in Tunnelling and Tunnel Boring Machines, V Edition 2005-06, Politecnico di Torino, Italia. [72] Reilly J.J. 2005. Cost estimating and risk management for underground projects. Proc. International Tunneling Association Conference, Istanbul, May 2005. [73] Schulter A. (1996). The importance of geometry for computer controlled segment erection. Tunnel Boring Machines - Trends in Design and Construction of Mechanized Tunnelling. Hagenberg Austria 1995, Wagner & Schulter (eds), pp. 89-98. Balkema/Rotterdam. [74] Schweiger H.F., Schuller H. & Pttler R. 1997. Some remarks on 2-D-models for numerical simulation of underground constructions with complex cross section. Computer methods and Advances in Geomechanics, Yuan (ed.) Balkema, Rotterdam. [75] Suwansawat S., Einstein H.H., 2006. Artificial neural networks for predicting the maximum surface settlement caused by EPB shield tunneling. Tunnelling and Underground Space Technology 21 (2006) 133150.

[76] Swoboda G. and Mansour M. (1996). Three-dimensional numerical modelling of slurry shield tunnelling. Tunnel Boring Machines - Trends in Design and Construction of Mechanized Tunnelling. Hagenberg Austria 1995, Wagner & Schulter (eds), pp. 89-98. Balkema/Rotterdam. [77] Swoboda G. and Abu-Krisha A. (1999). Three-dimensional numerical modelling for TBM tunnelling in consolidated clay. Tunnelling and Underground Space Technology 14(3), 327 - 333. [78] Vermeer P.A. (2001). On a smart use of 3D-FEM in tunnelling. Plaxis Bulletin (11), 2-7. [79] Wagner H. 2003. Design aspects of mechanical excavation. Scavo Meccanizzato di Gallerie, Master Universitario di II Livello, Politecnico di Torino, Italia. [80] Wagner H. 2005. Tunnels for Tomorrow in Soils, Rock and Water. Documentation of Publications on Tunnel Design and Construction. (Tunnels fr Morgen in Boden, Fels und Wasser. Eine Dokumentation von Publikationen ber Planung und Bau von Tunnels von Harald Wagner). Jan. 2005. [81] Wagner H. 2006. Contractual aspects and work site management. Lecture at Master Course in Tunnelling and TBMs, Edition V, Politecnico di Torino, Italia. [82] Wagner H. 2006. Examples of Lining Design in Mechanized Tunnelling. Lecture at Master Course in Tunnelling and TBMs, Edition V, Politecnico di Torino, Italia. [83] Walter A., Brasser J.P., Lemmerer J. and Lutz H. 2005. Optimization of the lining segments of a double shell railway tunnel. Felsbau - Rock and Soil Engineering No.6/Nov. 2005.

Appendix 1A EOLE Lot 35B Project

Paris / France Years of construction: 1992 - 1995. The subway line (EOLE) is running 35 m deep under the city center of Paris. Job description Detail design and construction consultancy for an innerurban railway tunnel which connects the two stations of St. Lazare - Condorcet and Nord - Est. The tunnel has a length of 2 x 1.670 m, an inner diameter of 6.4 m, a segment thickness of 35 cm and an overburden between 22 m and 28 m. Services provided by D2-Consult for DG Construction Paris - Detailed design of tunnel lining - Geometry of regular cross section - Alignment study by ring rotations - Segment installation sequences - Joint Configurations - Segment dimension - "Mould" drawings - Segment reinforcement - Structural analysis - Thrust jacking loads - Trailer loads - Secondary grouting - Eccentricity during segment installation - Construction consultancy on site - Analysis of segment quality - Twisting of segments - Offsets in circumferential joints - Observation of segment installation and recommendations Back analysis of the segmental ring elements of one pass lining for 2 TBM connecting tunnels, 1.700 m length and 6.4 m inner diameter. Tunnels are excavated 30 m beneath the surface. Transition structure is located on the level of Saint-Ouen limestone as well as water saturated Beauchamp sands. Connecting tunnels are crossing successive horizons of marl, gravel and heavy limestone beneath the ground water table.

Appendix 1B
EOLE LOT 35 B / FRANCE
INNERURBAN RAILWAY TUNNEL
OWNER: SNCF

PROJECT DESCRIPTION
DETAIL
DESIGN AND CONSTRUCTION CONSULTANCY FOR AN INNERURBAN RAILWAY TUNNEL WHICH CONNECTS THE TWO STATIONS OF ST.LAZARE - CONDORCET AND NORD - EST. THE TUNNEL HAS A LENGTH OF 2X1.670M, AN INNER DIAMETER OF 6,4 M, A SEGMENT THICKNESS OF 35CM AND AN OVERBURDEN BETWEEN 22M AND 28M.

PARIS

SERVICES PROVIDED
- DETAILED DESIGN OF TUNNEL LINING
GEOMETRY FO REGULAR CROSS SECTION ALIGNMENT STUDY BY RING ROTATIONS SEGMENT INSTALLATION SEQUENCES JOINT CONFIGURATIONS SEGMENT DIMENSION MOULD DRAWINGS SEGMENT REINFORCEMENT

LOCATION

- STRUCTURAL ANALYSIS - THRUST JACKING LOADS - TRAILER LOADS - SECONDARY GROUTING - ECCENTRICITY DURING SEGMENT INSTALLATION - CONSTRUCTION CONSULTANCY ON SITE - ANALYSIS OF SEGMENT QUALITY - TWISTING OF SEGMENTS - OFFSETS IN CIRCUMFERENTIAL JOINTS - OBSERVATION OF SEGMENT INSTALLATION AND
RECOMMENDATIONS

PERIOD OF WORK: VOLUME OF WORK:

1992 - 1997 206.150 EURO

INSTALLATION OF A RHOMBOIDAL SEGMENT

TBM POLYSHIELD

TBM TUNNEL CROSS SECTION

Appendix 2A Boston Outfall Tunnel Project

Boston / USA
Owner Client Water Resources Authority Massachusetts Sehulster Tunnels Inc. Concrete Systems Inc. JV Contact: Mr. Joseph P. Sehulster, Tel. +1 603 8894-163 The tunnel underneath Boston Harbour and brings cleaned waste water from the Deer Island treatment plant over discharger pipes to the open sea. The 55 outfall tunnel diffusers, located at regular intervals over the final 2km, were drilled from the sea-bed and connections made by probe drilling from the tunnel. The main tunnel was driven by TBM through hard rock Cambridge Argillite and lined with pre-cast concrete segments Project Details Inner diameter Tunnel length Overburden Segment thickness Period of Work Volume of Work 7.4 m 14 km 45 m 25 cm

Project

Services

Detailed lining design (Conex System) Engineering for in-situ testing Construction consultancy

10/1990 12/1995 EUR 223,000

TBM Tunnel segment supply

Appendix 2B
BOSTON EFFLUENT OUTFALL TUNNEL / USA
DETAILED DESIGN OF LINING SEGMENTS
OWNER: WATER RESOURCES AUTHORITY MASSACHUSETTS

PROJECT DESCRIPTION
ROCK UNDERNEATH BOSTON HARBOUR.

TBM DRIVEN AND SEGMENTAL LINED WATER TUNNEL IN ARGELITIC INNER DIAMETER LENGTH OVERBURDEN SEGMENT THICKNESS 7,40 M 14 KM 45 M 25 CM

BOSTON

SERVICES PROVIDED
DETAILED LINING DESIGN CONEX SYSTEM ENGINEERING FOR IN-SITU TESTING CONSTRUCTION CONSULTANCY 10/1990 12/1995 223,000 EURO

PERIOD OF WORK: VOLUME OF WORK:

LOCATION

CROSS SECTION

LONGITUDINAL SECTION

SEGMENT SUPPLY

SEGMENT WITH TIMBER DOWELS IN RING JOINT

Appendix 3

Type of Pressure

[-]

JACKING THRUST PRESSURE ON LINING, BEG TUNNEL 8 Segment Rings (5 rectangular, 2 Trapezoids, 1 Keystone) Concrete Safety Factor in Force per Pressure Thrust Shoe Thrust Force per No. of Necessary C Thrust on Thrust Compression Dimentions Shoe Area Ring Shoes Safety Factor 50/60 fc = 500 Shoe Shoe [kg/cm2] [cm] [t] [-] [t] [-] [-] [cm2] [kg/cm2] 21 21 21 492.60 431.02 258.61 95 95 95 x x x 17 17 17 1615 1615 1615 305.0 266.9 160.1 1.64 1.87 3.12 1.5 1.5 2

Installed advance pressure 10344.6 (max. machine) Excavation in Sand-Gravel (max. force) Excavation in Coarse Gravel (approx. 60% of max. force) Excavation in Fine Gravel (approx. 33% of max. force) Average Installation (15 ton/dowel) 9051.5 5430.9

2987.0 315

21 21

142.24 15.00

95 95

x x

17 17

1615 1615

88.1 9.3

5.68 53.83

2 2

Remarks: Thrust shoe dimensions are supposed only as machine has not yet been officially delivered.

Appendix 4

BEG LOT H3-4. INPUT PARAMETERS FOR HARDENING SOIL MODEL "GREEN-FIELD" CROS-SECTION Volumetric Friction Dilactancy Poison's Real depth Cohesion c weight gamma angle angle Soil Layer ratio ur [m] [kN/m2] 3 [deg] [deg] [kN/m ] GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T2 GA_T2 GA_T2 GA_T2 GA_T2 GA_T2 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21 21.8 21.3 20.8 20.3 19.8 19.3 18.8 18.3 17.8 17.3 16.8 16.3 15.8 15.3 14.3 13.3 12.3 11.3 10.3 9.3 8.8 8.3 7.8 7.3 5.8 4.3 2.8 1.3 -0.2 -1.2 -2.2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 37 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20 E-Modul Coefficient of 2 [MN/m ], lateral stress f(z) = w Ko=1-sin 0.1v(0.13z) 0.000 6.311 9.903 12.889 15.540 17.965 20.226 22.357 24.384 26.325 28.190 29.992 31.737 33.432 36.691 39.801 42.785 45.661 48.442 51.140 52.461 53.764 55.050 56.320 60.041 63.642 67.137 70.536 73.849 76.014 78.146 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398 0.398

Soil Layer GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T1 GA_T2 GA_T2 GA_T2 GA_T2 GA_T2 GA_T2 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3 GA_T3

Reference pressure pref [kN/m2] 0.000 4.181 8.362 12.543 16.724 20.905 25.086 29.267 33.448 37.628 41.809 45.990 50.171 54.352 62.714 71.076 79.438 87.800 96.162 104.524 108.704 112.885 117.066 121.247 133.790 146.333 158.876 171.419 183.961 192.323 200.685

BEG LOT H3-4. INPUT PARAMETERS FOR HARDENING SOIL MODEL (continued) "GREEN-FIELD" CROS-SECTION Unloading Secant Tangent Primary loading /reloading Power in Pref E50,ref stiffness Failure stiffness stiffness E50 oedometer stiffness (average) (average) E,z [kN/m2] Eoed,ref ratio Rf E50,ref 2 stiffness Eur,ref [kN/m2] laws m [kN/m2] [kN/m ] [kN/m2] [kN/m2] 2 [kN/m ] 0.5 0.9 #DIV/0! 0.000 0.000 0.000 0.000 6.271 7,275.885 0.5 0.9 6,311.063 6,311.063 18,933.189 6,311.063 6,311.063 0.5 0.9 9,903.119 9,903.119 29,709.358 9,903.119 9,903.119 0.5 0.9 12,889.359 12,889.359 38,668.078 12,889.359 12,889.359 0.5 0.9 15,539.660 15,539.660 46,618.980 15,539.660 15,539.660 0.5 0.9 17,965.240 17,965.240 53,895.719 17,965.240 17,965.240 0.5 0.9 20,225.573 20,225.573 60,676.718 20,225.573 20,225.573 0.5 0.9 22,357.140 22,357.140 67,071.419 22,357.140 22,357.140 33.448 23,912.549 0.5 0.9 24,384.340 24,384.340 73,153.020 24,384.340 24,384.340 0.5 0.9 26,324.501 26,324.501 78,973.502 26,324.501 26,324.501 0.5 0.9 28,190.483 28,190.483 84,571.448 28,190.483 28,190.483 0.5 0.9 29,992.161 29,992.161 89,976.483 29,992.161 29,992.161 0.5 0.9 31,737.325 31,737.325 95,211.976 31,737.325 31,737.325 0.5 0.9 33,432.257 33,432.257 100,296.770 33,432.257 33,432.257 0.5 0.9 36,691.193 36,691.193 110,073.579 36,691.193 36,691.193 79.438 42,564.749 0.5 0.9 39,801.030 39,801.030 119,403.090 39,801.030 39,801.030 0.5 0.9 42,785.081 42,785.081 128,355.242 42,785.081 42,785.081 0.5 0.9 45,660.962 45,660.962 136,982.885 45,660.962 45,660.962 0.5 0.9 48,442.395 48,442.395 145,327.185 48,442.395 48,442.395 0.5 0.9 51,140.329 51,140.329 153,420.988 51,140.329 51,140.329 0.5 0.9 52,460.834 52,460.834 157,382.502 52,460.834 52,460.834 145.985 63,175.115 0.5 0.9 53,763.677 53,763.677 161,291.030 53,763.677 53,763.677 0.5 0.9 55,049.735 55,049.735 165,149.205 55,049.735 55,049.735 0.5 0.9 56,319.814 56,319.814 168,959.442 56,319.814 56,319.814 0.5 0.9 60,041.288 60,041.288 180,123.864 60,041.288 60,041.288 0.5 0.9 63,642.428 63,642.428 190,927.285 63,642.428 63,642.428 0.5 0.9 67,136.988 67,136.988 201,410.963 67,136.988 67,136.988 0.5 0.9 70,536.194 70,536.194 211,608.581 70,536.194 70,536.194 0.5 0.9 73,849.363 73,849.363 221,548.089 73,849.363 73,849.363 0.5 0.9 76,014.266 76,014.266 228,042.797 76,014.266 76,014.266 0.5 0.9 78,146.462 78,146.462 234,439.385 78,146.462 78,146.462

Curriculum Vitae
Name 15.9.1974 June 1991 1991-1996 Nguyen Duc Toan Born in Hanoi, Vietnam Graduated from Dan Phuong secondary school, Hanoi, Vietnam Enrolled at Bridge and Tunnel Engineering Section, Faculty of Engineering, University of Communication and Transport, Hanoi, Vietnam Bachelor Degree of Civil Engineering from University of Communication and Transport, Vietnam Employed as an Assistant Bridge Engineer by Taisei-Rotec JointVenture, Hanoi, Vietnam Enrolled at Pedagogic English Section, Faculty of Continuing Education, College of Foreign Language - Vietnam National University, Hanoi Bachelor Degree of Foreign Language from College of Foreign Language - Vietnam National University, Hanoi Employed as a Bridge Engineer by Louis Berger Inc., Hanoi, Vietnam Employed as a Tunnel Inspector by Transport Engineering Design Inc. (TEDI), Hanoi, Vietnam Consulting Engineer Certificate (Construction Supervision) by Ministry of Transport of Vietnam Employed as a Civil Engineer by Institute of Transport Science and Technology (ITST), Hanoi, Vietnam Study for Master Degree at University of Technology of Turin, in partnership with Consortium of Research and Permanent Education of Turin, Italy ngdtoanhanoi@yahoo.com; toan_ngd@itst.gov.vn

June 1996 1996-1998 1998-2000

April 2000 1998-2001 2001-2004 March 2004 Since 2004 Since 2005

Contact

Post Graduate Master Course by Politecnico di Torino (University of Technology) TUNNELLING AND TUNNEL BORING MACHINES V EDITION 2005-06

Master endorsed by ITA/AITES

Under the patronage of SIG ITALIAN TUNNELLING SOCIETY

in partnership with TUSC

Sponsored by: