A hypothetical tunnel project was discussed in the BTSYM workshop ILLUSTRATIONS COURTESY OF SI SHEN

INTRODUCTION

On 15 May, a technical workshop titled ‘Connecting the Dots: A Tunnel Engineering Knowledge Framework’ was delivered to the British Tunnelling Society Young Members (BTSYM) at the Institution of Civil Engineers (ICE), in London.

The topic of the workshop was aimed at early-career tunnel engineers, and personally chosen by the author, had the intention of highlighting the distinctive role tunnel engineers play within the broader field of civil engineering. The session was conducted in an open, interactive, and free-form style, allowing the audience to shape the content based on their individual skill sets, interests, and professional backgrounds.

Several key insights – shared either through prepared module or emerging from Q&A discussions – proved to be universally relevant for all tunnel engineers. A summary of these messages is provided below.

A ‘KNOWLEDGE TREE’

Despite the title of the workshop, the author believes that ‘Knowledge Tree’ is a more fitting term than ‘Knowledge Framework’, as a tree symbolises growth – reflecting the continuous accumulation of knowledge throughout one’s career.

In his early career, author acquired knowledge in an unstructured manner. However, over time, he began to understand the importance of grasping the ‘trunk’ of the knowledge tree – that is, the fundamental principles. Once these core concepts were clear, the ‘branches’ of more specialised knowledge became much easier to develop.

Photo from the BTSYM workshop ‘Connecting the Dots: A Tunnel Engineering Knowledge Framework’ held at ICE, in London PHOTO COURTESY OF BTSYM

New and existing knowledge can be placed on this ‘tree’ in appropriate locations, forming connections with other related areas. This concept also highlights the importance of maintaining coherence with the ‘big picture’, helping engineers avoid critical mistakes caused by “not seeing the wood for the trees.”

As the saying goes, “Give a man a fish, and you feed him for a day; teach a man to fish, and you feed him for a lifetime.” The author therefore explained that he did not want to offer a ready-made ‘knowledge tree.’ Instead, his aim was to draw attention to this concept and provide attendees with guidance and tips on how to build their own ‘knowledge trees.’

Engineers, in the early stages of their careers – regardless of their specific field, are encouraged to visit construction sites and engage directly with practical challenges. It is only through first-hand experience that a tunnel engineer can truly appreciate the value and application of different knowledge modules. This will be further expanded upon below.

Moreover, irrespective of their academic or technical background, all tunnel engineers must develop at least a solid understanding of four key areas: engineering geology; geotechnical engineering: structural engineering; and, structural materials. This foundational knowledge enables effective collaboration within the inherently multi-disciplinary environment of tunnel engineering.

While it is natural to specialise in one of these disciplines, a significant knowledge gap in any of them may become a ‘short plank’ in one’s career – ultimately limiting growth and effectiveness as a tunnel engineer.

‘REVERSE ENGINEERING’ OF TUNNELS

Through university courses and Continuing Professional Development (CPD) events, we are often taught theorybased modules – such as soil mechanics, concrete structure design, and advanced structural analysis – prior to fully understanding their intended practical application.

As practising tunnel engineers, it is often more effective to start from the ‘end product’ and work our way backward to understand ‘why’ we were taught these concepts in the first place.

In this context, ‘reverse engineering’ means approaching the engineering process from the practical challenges, grounding our understanding in actual construction and project needs, then tracing that back to the underlying theories required to support the solutions.

During the workshop, we held highly interactive discussions around a hypothetical tunnel project. Participants were challenged to consider the following:

  • How many technical challenges presented by the project can you identify?
  • Based on these challenges, what tunnelling method would you choose?
  • What are the key technical issues associated with the selected tunnelling method—from both construction and design perspectives?

Firstly, look at the ground. It helps for a tunnel engineer to have a high level appreciation of the key characteristics and hazards associated with each stratum. As a few examples:

  • The Alluvium layer: often exists in a river bed and typically consists of clay/silt/sand; there can be patches of peat, which can be extremely soft with high water content; and, there can be flammable gas hazards.
  • The Chalk layer: generally a water-bearing weak rock; it’s dissolvable which leads to dissolution hazards, such as the presence of ‘drift-filled hollows’ under a river bed; and, the presence of nodes of flint can add to the wears of TBM cutting tools.
  • The Glacial Till layer: is a highly variable and nonhomogeneous layer, with possible grain sizes ranging all the way from clay to boulders; usually of good strength and stiffness due to its well-graded and over-consolidated nature; the possible presence of boulders presents great challenges to TBM operations; this stratum is also a lot less well-studied in general.
Stress-Strain graphs representing three principal construction materials – concrete, steel, and soil – for the key, vital, differences between the apparently similar-looking models to be discussed in the workshop

Secondly, look at the surroundings, which can have a major influence over the choice of tunnelling method. For example:

  • The existence of metro and city usually will impose very tight ground movement control (volume loss) limits to tunnelling.
  • The presence of a nature reserve is likely to impose restrictions on groundwater draw-down and contamination.
  • The tunnels have very low ground cover next to the portals; this leads to relatively higher distortion of the tunnel lining and may require ground treatment.

Thirdly, a tunnel engineer should bear in mind the reasons behind popular choices of tunnelling methods and how they differ from each other. For example:

  • TBMs are used for their efficiency for long tunnels and capabilities to deal with ground hazards in themselves; in soft ground conditions tunnel shield and segments are likely to be used; in dry and stable grounds roadheaders may be suitable.
  • Pipejacking differs from segmentally-lined tunnels in that the entire tunnel linings are pushed along the whole route rather than, like segment, be placed stationary in set locations.
  • Sprayed concrete lined (SCL) tunnels (for sequential excavations and supports) offer flexibility in geometry and are more suitable for short tunnels and those with variable alignment and shapes.
  • Drill & Blast is a flexible excavation method that tends to be chosen more for tunnels in more remote, unpopulated mountainous areas with high strength rocks.
  • Shafts – to be break up tunnelling routes, intermediate shafts can be adopted; there are many options available to choose from (such as precast segments, sprayed concrete, diaphragm walls (D-Walls), rock bolts, etc) depending on the required geometry and ground conditions.
Stress-Strain graphs representing three principal construction materials – concrete, steel, and soil – for the key, vital, differences between the apparently similar-looking models to be discussed in the workshop
Stress-Strain graphs representing three principal construction materials – concrete, steel, and soil – for the key, vital, differences between the apparently similar-looking models to be discussed in the workshop

LEARNING IN A ‘CON-FUSION’ WAY

Given the inherently multi-disciplinary nature of tunnel engineering, the author’s personal experience shows that an engineer can go through periods of confusion in the learning process, during which one feels even less clear or confident than before. This is reckoned to be a natural (even essential) step in the whole learning and growth process, which can unfold in the following stages:

  • ‘Conceptualisation’: You learn the specifics on an isolated knowledge node; you think you have mastered it, but the knowledge is actually gained in a vacuum, disconnected from the broader context.
  • ‘Confusion’: You have collected many isolated nodes of knowledge and started to sense that they are somehow related, but sometimes apparently conflicting; you start attempting to put these nodes together but only to find these muddled up in your mind and difficult to separate.
  • ‘Fusion’: Eventually, you start to internalise the knowledge pieces as a coherent framework. The once-disconnected nodes form an integrated structure, allowing you to apply them fluidly and purposefully toward a common goal in a single combined strand.

As an example, to illustrate the point, the author attempted to confuse the workshop attendees with a commonly used ‘linear-elastic-perfectly-plastic’ stressstrain relationship for three different materials: concrete; steel; and, soil.

The constitutive models for these materials are typically taught in different modules or classes, and bear extremely similar resemblance in appearance, which can give rise to confusions. Below are some key points on how these materials are very different to each other in meaning and context, such as their stress-strain relationships being derived from quite different types of forces:

  • Reinforced concrete: the stress-strain curve applies to uni-axial compressive stresses only; the concrete must also be reinforced with at least minimum code required amount of reinforcement to demonstrate the ductility as depicted. Behaviour under tension is distinctively different – unreinforced concrete in tension is extremely brittle; also, the shape of the curve differs between bar-reinforced and fibrereinforced concrete.
  • Steel: The stress-strain is for uni-axial tension only. When plotted on the same diagram as concrete, it shows on the opposite (negative) quadrant. The stress-strain curve for steel in compression does not exist, as the failure mechanism is different. The ultimate strain is an order of magnitude higher than that of concrete, which means steel is a lot more ductile than concrete. The composite action between steel and concrete in reinforced concrete in tension makes the line non-linear and much more complicated.
  • Soil: The stress-strain curve applies to shear only – the relative sliding between particles in contact along an interface plane. The position of the failure line (horizontal) is typically defined by the Mohr-Coulomb criteria, dependent on the compressive stress perpendicular to the interface (the confining stress); it is therefore not a set value but changes in accordance with the confining stress.
  • Soil: It should also be noted that the volumetric behaviour (change in size) of soil is distinctively different from its shear behaviour (change in shape) – the drained bulk stiffness can increase as the soil gets compressed more, especially for the first time, and the undrained bulk stiffness (with water virtually incompressible, so all volumetric compression is resisted) approaches infinity in relative terms.

‘GROUND-STRUCTURE’ INTERACTION

It’s uniquely imperative for a tunnel engineer to appreciate that the ground is both loading and resistance. The extent to which it plays each role is determined by ‘ground–structure interaction.’

In civil engineering courses, the structural engineering and geotechnical engineering modules are often taught separately. ‘Soil-structure’ interaction, which is where these two meet together and work in a combined strand, is usually less well taught in classes – and where done so it usually related to surface structures and foundations. This ties exactly into my point on ‘con-fusion’ and the developing ‘knowledge tree’ as civil engineers find ways to understand how these work together.

A further challenge in appreciating their interaction can arise outside civil engineering.

In the workshop, we discussed ‘structure’ as referring to those man-made elements such as tunnel linings and props, and while covering both primary and secondary linings it was mainly the former (as it is usually the one that interacts with the ground). The following was explained in essence:

  • The ground can provide resistance to its own loading;
  • The structure shares responsibilities in providing resistance with the ground;
  • The groundwater offers no resistance (in terms of shear capacity);
  • The ground deforms when providing resistance:
    • When prevented from deformation, the ground cannot provide resistance; and,
    • When limited to less deformation, the ground provides less resistance;
    • Ground ‘relaxation’ is a term referring to a form of ground deformation, where the ground is unloaded, expands and moves, ‘activating’ a new internal structural arrangement.
  • The later (in space or time) the structure is installed, or the less stiff the structure is:
    • The more the ground moves; and,
    • The less loading the structure attracts.

The graph that follows – ‘Ground pressure on support – v – Convergence’ – relates only to experience and effects where tunnelling is (principally) horizontal (i.e., shafts are different).

Also, loading conditions are static (seismic or other dynamic loadings were not considered in the graph, or workshop), and such tunnels are not excavated at exceptional depths (hundreds of metres).

It does not, therefore, cover potential conditions for tunnels that are very deep (e.g., approaching depths of 1000m or even greater), where overburden is so high that upon a void being created by excavation the rock’s inherent strength cannot contain it intact. In such instances, and sometimes even after excavation support has been put in place, as the rock mass self-adjusts its natural structure to seek stress-energy alleviation, the consequence can be explosive. These ‘rockbursts’ are a specialist area of tunnelling, and thus is part of the ‘Knowledge Tree’; but the topic, though, was not addressed in the workshop.

One of the diagrams used at the workshop: Support load -v- Convergence

SUMMARY

In the workshop titled ‘Connecting the Dots: A Tunnel Engineering Knowledge Framework’, the author, a tunnel engineer, mainly conveyed the following messages to tunnel engineers at their early career stage:

  • Establish a ‘Knowledge Tree’ and continue to develop it throughout a career;
  • Approach engineering processes from the practical challenges first and then trace back to the theories;
  • Don’t be afraid to confuse yourself – this is progress, ultimately, rather than a setback. It is questioning, in a ‘growth stage’; and,
  • Understand the principles of ‘ground-structure interaction’.

ACKNOWLEDGEMENTS

Si Shen would like to thank BTSYM Chair, Arabel Vilas, for her encouragement in this initiative and the whole Committee for facilitating the workshop.