sign in sign up
<<
Home , Tunnel, Tunnel construction

INTERNATIONAL SOCIETY FOR MECHANICS AND GEOTECHNICAL

This paper was downloaded from the Online Library of the International Society for and (ISSMGE). The library is available here: https://www.issmge.org/publications/online-library

This is an open-access database that archives thousands of papers published under the Auspices of the ISSMGE and maintained by the and Development Committee of ISSMGE. Geotechnical engineering considerations for the analytical design of an adequate support system

Joan Ongodia, Denis Kalumba, , University of Cape Town, Graduate student & Senior Lecturer, South Africa, [email protected]

Harrison Mutikanga, Brajesh Ojha , Uganda Electricity Generation Company Limited, Uganda & Civil Engineer, Navayuga Engineering Company, India

ABSTRACT: Tunnel excavation impacts the ground both at the surface and underneath thereby affecting the natural in-situ stability. Failure results from redistribution, stress relaxation and deformations around the tunnel. Ultimately, tunnel failure occurs when the loads exceed the value of the bearing resistance. Therefore, a support system comprising unique structural components is needed to resist the weight of unstable wedges thereby ensuring that the tunnel does not fail. Nevertheless, stability depends on an effective rock- support interaction, specifically with the rock bolt which is the main support member. In order to assess and improve stability conventional methods to design are borrowed from which assesses material properties by visual observation, physical identification, using rudimentary geological handy , charts and graphs to estimate tunnel support ranges. For stability, the specific value of the required capacity to prevent failure should be established. This paper presents the background to the engineering design of an adequate tunnel support based on the peak rock load which must be resisted to ensure the tunnel remains stable.

RÉSUMÉ: L’excavation de impact le sol au niveau de la surface et en profondeur, en affectant ainsi la stabilité in situ naturelle. Des défaillances surviennent du fait de la redistribution des contraintes, le relâchement des contraintes, et des déformations autour du tunnel. En fin de compte, la défaillance du tunnel se produit lorsque les charges dépassent la valeur de résistance de soutènement. Donc, un système de support comportant des composantes structurales uniques est nécessaire pour résister au poids des rocs instables pour ainsi s’assurer contre l’écroulement du tunnel. Néanmoins, la stabilité dépend de l’interaction adéquate entre la roche et le support, particulièrement avec les boulons d’ancrage dans le roc qui est l’élément de support principal. Afin d’évaluer et d’améliorer la stabilité, les méthodes de concept conventionnelles sont empruntées à la géologie qui évalue les propriétés des matériaux par l’observation visuelle, l’identification d’aspect physique, en utilisant des outils géologiques rudimentaires, et des tableaux et graphiques pour estimer les portées souterraines. Concernant la stabilité, une valeur spécifique de la capacité requise doit être établie pour éviter la défaillance. Cet article présent la connaissance technique en concept d’ingénierie pour un support adéquat de tunnel basé sur les charges maximales des rocs instables à soutenir, et ainsi d’assurer l’intégrité du tunnel. KEYWORDS: Tunnels, Geotechnical engineering design, theory, Rock load, Support

1 INTRODUCTION Presently, the accuracy of a tunnel design mainly depends on experience and geological methods used (Mohammed 2015). The methods employed by tunnel practitioners including geotechnical engineers often involve visual observation, physical identification, rudimentary hand-held tools and estimation of corresponding support parameter ranges from charts (Hoek et al. 1995). Geological materials identified are described, assessed and categorised, in order of reducing strength, as Class I, II, III, IV and V based on the Rock Mass Figure 1. Continuum scale (adapted from Kanji 2014) Rating (RMR), Q and Rock Quality Designation (RQD) (Palmström 1995, Brown et al. 1983). RMR, Q and RQD systems give an indication of the level of deformation and (1) ability of the rock to support loads. Alternatively, rock can be classified on a continuum scale as shown in Figure 1 although where σ'1 and σ'3 are major and minor effective principal the methods are generally subjective thereby giving ranges of stresses at failure, σci is the UCS for the intact rock pieces, mb is the support parameters (Kanji 2014, Ongodia 2017). the Hoek-Brown rock mass constant, s and a are rock mass Guiding standards include the tunnel lining design guide characteristic constants. (TLDG 2004), the New Australian Tunnelling Method (NATM), the Chinese HC standard for tunneling, the Indian tunneling standard and the adapted Czech NATM (Spackova 2012, Sousa 2 KEY CONSIDERATIONS FOR TUNNEL SUPPORT 2010, PRC 2008). Further, underground stability problems are The main geotechnical engineering considerations for analysed using mostly the non-linear Hoek-Brown failure estimation of rock loads and the corresponding tunnel supports criterion (see Eq. 1) which approximates jointed rock comprise 1) geology, 2) material properties, 3) geotechnical characteristics better than the linear Mohr-Coulomb failure parameters, and 4) the process (Palmström 1995). criterion because it approximates isotropic behaviour for intact Stability of the surrounding rock mass depends on geological rock (Greer 2012). conditions, residual strength, in-situ stresses, burial depth, tunnel dimensions, excavation disturbance factor and material properties which influence stress magnitudes and deformation

- 1737 - th Proceedings of the 19 International Conference on Soil Mechanics and Geotechnical Engineering, Seoul 2017

such as the Poisson’s ratio, Young’s modulus, shear modulus interactions (Sharp 2007). Discontinuities provide and stiffness (Greer 2012). Figure 2 illustrates the network of void spaces through which water permeates the rock mass. significant factors which control stability. Porewater pressures widen the discontinuities in the structure thereby aiding physical disintegration and eventually failure. Therefore, provision is an important aspect of the support system (Bondarchuk et al. 2012). The rate of hydrogeological flows and extent of widening of the discontinuities is controlled by the minerals present both in the rock mass structure and void space infilling (Wahlstrom 1973). minerals have a high affinity for water and cause swelling and squeezing of the mineral structure when wet and dry, respectively (Kanji 2014). Infillings have strong anisotropic mechanical properties thus they significantly reduce rock strength (Palmström 1995). Typical unstable wedges and potential tunnel failure Figure 2. Web of ground factors influencing stability (Jones 1989) resulting from peak rock loads in the , sidewalls and floor are illustrated in Figure 4. Rockbursts or popping conditions 2 .1 Geology usually occur in the roof and sidewalls resulting in collapse or The ground type and its integral structure are important to for slide, respectively, whereas unstable floor wedges usually heave stability as they directly provide anchorage for the support. Key (Viggiani 2012). Furthermore, squeezing is a severe tunnel features include the rock mass structure, type, degree of construction challenge besides mudflows and face collapses weathering, interlocking, types and orientation of discontinuity (Sousa 2010, TLDG 2004, Hoek et al. 1995). sets, persistence, roughness and width of the discontinuity Tunneling strain and squeezing behaviour can be predicted (Jones 1989). Intact rock is strong although as weathering, using charts (see Figure 5 below) in which strains above 2.7% which is a significant natural earth process, takes effect the rock indicate mandatory support (Hoek et al. 1995). An equivalent mass begins to disintegrate thereby forming discontinuities such value of the laboratory measured oedometer swell pressure is as joints and lines (Brown et al. 1983). Discontinuities aid applied onto the rock in the field to ensure stability the agents of weathering including erosion such that the rock (Mohammed 2015). mass eventually degenerates into soil over geological periods. The five rock classes correspond with the varying degrees of Roof wedge collapses weathering (Palmström 1995). Weak discontinuous rock is generally unstable and requires mandatory support (Bieniawski 1992). Sidewall wedge slides or displaces Geological prediction gives insight on the material condition, state, properties and responses of the ground to the construction process such that advance support is provided where necessary and the excavation sequence adjusted timely to avoid excessive Floor wedge heaves damage of the overall rock mass structure (Hoek et al. 1995). Figure 3 illustrates how probes are used to predict geology ahead of tunnel advance such that sudden failure during Figure 4. Typical unstable wedges and failure (Ongodia 2017) excavation is avoided. However, actual material characteristics and engineering parameters of the excavation should be evaluated to determine the resistance forces to be counteracted (Mohammed 2015).

Figure 5. Strain and squeezing (TLDG 2004 after Hoek et al. 1995)

2.3 Geotechnical parameters Mechanical rock strength and in-situ stresses are the most important parameters which determine the required support capacities to aid the intrinsic structurally controlled support (RTM 2009). Other essential parameters include stratigraphy, Figure 3. Predicting rock mass deformations (Hoek et al. 1995) engineering and index material properties (Panthi 2006, Brown et al. 1983). Strength is a function of geology and the material 2.2 Material properties properties (Palmström 1995). Main geological material properties include and Although discontinuities structurally control deformation, they void space, permeability, pore pressures and mineralogy. break the overall rock mass structure into smaller isolated Hydrogeology encompasses geological and subsurface wedges thereby resulting in each block having unique

- 1738 - Technical Committee 204 & Technical Committee 2 / Comité technique 204 & Comité technique conjoint 2

characteristics (Wahlstrom 1973). Consequently, the strength of rock mass whereas values less than unity indicate overstressed discontinuous rock is an average value of the individual wedge rock mass thus failure may be likely (Greer 2012). properties (Thomas-Lepine, 2012). Usually measured strength S F = (1f -  3 ) ( 1 - 3 ) (2) is the unconfined compressive strength (UCS). The laboratory where σ1f is a major principal stress at failure, σ1 and σ3 are UCS of the rock gives a measure of its resistance to major and minor principal stresses, (σ1f - σ3) is the rock mass deformation among other tests. strength and (σ1 - σ3) is the induced/deviator stress.

2.4 Construction process Geological data, properties, rock parameters and measured field responses (see Figure 6), are essential to estimate the required tunnel support, influence the excavation geometry, construction method, process and support sequence including drainage (Bondarchuk et al. 2012). Field responses should be measured using installed instrumentation during construction to maintain stability and prevent failure by ensuring secure rock-support interactions.

Figure 7. Effect of tunnel burial depth on failure zone (Hoek et al. 1995) Figure 6. Surrounding ground responses (Hoek et al. 1995) 2.4.3 Rock-support interaction 2.4.1 Stability The purpose of a support system is to reinforce the ground, The weight of rock overburden increases with the span and prevent deformation and support rock loads (Mohammed 2015). controls stability of shallow tunnels while wedge gravity falls The roof is the most critical part of the tunnel because it along weak discontinuous shear planes influences stability of experiences the peak load (Terzaghi 1946). As such, the jointed blocky rock masses (Perri 2007). Closer to the ground required roof resistance is the upper limit for stability and hence surface, rock confinement is not significant such that cracks the minimum adequate support requirement (Adhikary and propagate easily while at greater depths, the rock is confined Dyskin 1997). significantly so that ground stresses contribute towards stability Tunnel support systems typically comprise of mainly rock (Greer 2012). Figure 7 is a schematic representation of the bolts, wire mesh reinforced shotcrete and a reinforced potential effect of tunnels at the ground surface, the influence of lining as shown in Figure 8 (Hoek et al. 1995, TLDG 2004). burial depth on the stability of tunnel structures and the The main support members are rock bolts therefore the rock- corresponding tensile and shear failure zones. From the figure, bolt interaction is discussed (Kristjánsson 2014, Thomas- the failure zones are: (a) Most significant at shallow depths Lepine 2012). Rock bolts are installed either systematically in a below the surface, (b) Smaller at relatively increased depth and pattern (see Figure 8) or at specific locations called spot bolting (c) Smallest therefore barely significant at greater tunnel burial to limit movement and prevent unravelling by anchoring depth below the ground. structurally unstable rock wedges as as load transfer from an unstable excavated face to the more stable confined interior 2.4.2 Failure rock mass (Kristjánsson 2014). Based on laboratory research, Tunnels generally fail in compression and the types of failure He (2014) found that the rock-bolt support mechanism is induced at depth include rock bursts, sliding and heaving along enforced by bolt strengthening alongside the coupling which weak discontinuous planes (Zhang et al. 2015, Liu and Zhang maximizes and minimizes stiffness of the 2003, Hoek et al. 1995). Discontinuities constitute a system of surrounding rock mass thereby ensuring stability. failure mechanism thus control rock mass stability and determine the likely type of failure (Ongodia 2017). Rock bursts and slip are wedge and planar failure mechanisms, 3 ROCK LOAD ESTIMATION respectively (Thomas-Lepine 2012). Therefore, the ultimate Earth pressures resulting from rock loads, imposed by the load of unstable rock wedges likely to fail must be thoroughly weight of surrounding rock mass, are assumed to be hydrostatic investigated in order to provide adequate stability (Nielsen 2009, (Terzaghi 1946). The weight of the rock mass is mainly a Palmström 1995). function of the earth material’s geomechanical characteristics On the other hand, Mohammed (2015) noted that spalling and dimensions of the tunnel excavation. Peak loads act at the and slabbing are minor failure types which can be estimated centres of both the tunnel roof and invert as well as on the from the rock’s Strength Factor (SF) against shear failure (see sidewall at a distance about two-thirds from the level of the Eq. 2). SF above unity indicates that the rock strength is greater invert (Nielsen 2009). According to Nielsen (2009), the rock than the induced stress such that there is no overstress in the loads at the invert and sidewall are approximately ½ and ⅓ of

- 1739 - th Proceedings of the 19 International Conference on Soil Mechanics and Geotechnical Engineering, Seoul 2017

the peak load, respectively. Formulae to calculate the loads are Jones A. P. 1989. Civil Engineers Reference Book. Rock Mechanics generally a continuation of Terzaghi (1946)’s study of shallow and Rock Engineering. Edited by L S Blake, Reed Educational and soil tunnels and the loads are factored to analyse (USACE Professional Publishing Ltd. The Bath Press, Bath pp.10/3-10/39 1997). Peak loads can be calculated using Eq 3. Further Kanji M. A. 2014. Critical Issues in Soft Rocks. Journal of Rock research should be undertaken to modify the equations in order Mechanics and Geotechnical Engineering, 6(3), 186-195 Kristjánsson G. 2014. Rock Bolting and Pull-out Test on Rebar Bolts. to adapt them for deep rock tunnels since they were mainly Thesis Norwegian University of Science and developed based on soil mechanics (Barton et al. 1974). tan tan2 (45   0.5  )H Liu Q. S. and Zhang H. 2003. Study on Stability and Support of Rock PHg(1[v  ]) (3) Masses Surrounding Deep -Mine Roadway. Journal of b 2m tan(45 0.5 ) Rock Mechanics and Engineering 22(S1): pp. 2195–2220 Roof wedge Mohammed J. 2015. Underground structures support, stress and strain of tunnel Nielsen Y. 2009. Loads on tunnels. Course notes tunnel design and Rock bolt construction. Middle East Technical University (METU). Ongodia J.E. Kalumba D. and Mutikanga H.E. 2016. An Account of Tunnel Support Systems for Soft Rock Mass Conditions. Reinforced Proceedings of the first Southern African Geotechnical Conference concrete lining held in Sun , South Africa, March- Jacobsz (Ed.) Taylor & Francis Group, , pp. 205-211 Ongodia J.E. 2017. Geotechnical engineering design of tunnel support Wire mesh systems - a case study of Karuma (600MW) hydropower in Final tunnel reinforced Uganda. Masters dissertation, University of Cape Town, South Sidewall surface shotcrete lining Africa. wedge Palmström A. 1995. Rock Masses as Construction Materials. Chapter Two: PhD thesis. RMi – a rock mass characterization system for rock engineering purposes. Oslo University, Norway, 1995, 400 p. Panthi K. K. 2006. Analysis of Engineering Geological Uncertainties Floor wedge Excavation Related to Tunneling in Himalayan Rock Mass Conditions. boundary Norwegian University of Science and Technology.

Figure 8. Typical tunnel support system (adapted from Ongodia 2017) Perri G. 2007. Behavior Category and Design Loads for Conventionally Excavated Tunnels Retrieved on 2016 October 5 from http://www.gianfrancoperri.com/Documents/90- 4 CONCLUSION 2007Behaviorcategoryanddesignloadsforconventionallyexcavatedtu nnels.pdf Main engineering consideration factors for tunneling include PRC 2008. Specification for Design of Hydraulic Tunnels. State geology, material properties, geotechnical parameters and the Development and Reform Commission, Electric Power construction process. The geological data, properties and Standards. China Electric Power Press. People’s Republic of China parameters are used as inputs to calculate the peak rock load of (PRC) T0017830 the unstable rock wedge which could cause ultimate failure. RTM 2009. Technical Manual for Design and Construction of Failure of the unstable loose wedges is aided by the weak Tunnels -Civil Elements Road Tunnel Manual (RTM) FHWA- discontinuity planes especially when the rock-support NHI-09-010 U.S. Department of Transportation Federal interaction is not secure enough or when the bolt capacity is less Administration Publication No. FHWA-NHI-10-034 than the peak load of the surrounding rock mass. Secondary Sousa L.R. 2010. Risk analysis for tunneling projects PhD thesis. Massecheutes Institute of Technology. support is provided by the wire mesh reinforced shotcrete lining Sharp J. M Jr. 2007. A Glossary of Hydrogeological Terms. The and reinforced concrete lining. University of Texas, Austin, Texas Spackova O. 2012. Risk manageent of projects. PhD_Dissertation Czech Technical University in . 5 ACKNOWLEDGEMENTS Terzaghi K. 1946. Rock deffects and loads on tunnel supports. Financial and institutional support by the Julian Baring Reprinted from rock tunnelling with supports. The Scholarship Fund and the Uganda Electricity Generation commercial shearing and stamping co. Soil mechanics series no. 25. Company Limited, respectively is greatly appreciated. Thomas-Lepine C. 2012. Rock Bolts - Improved Design and Possibilities. Masters Dissertation of the Civil Engineering school of ENTPE, in partnership with NTNU, Norway. Retrieved REFERENCES on 2016 October 5 from http://www.diva- Adhikary D.P. and Dyskin A.V. 1997. portal.org/smash/get/diva2:566190/FULLTEXT01.pdf Modelling_the_Deformation_of_Underground Excavations in TLDG 2004. Specification for Tunneling. Tunnel Lining Design Guide Layered Rock Masses. International Journal of Rock Mechanics & (TLDG). The British Tunneling Society and the Institution of Civil Mineral Science Volume 34:3-4 Elsevier Science Ltd Engineers. London: Thomas Telford Publishing Barton N. Lien R. and Lunde J. 1974. Engineering Classification of USACE 1997. Tunnels and Shafts in Rock Engineering and Design, Rock Masses for the Design of Tunnel Support. Journal of Rock United States Army Corps of Engineers (USACE) EM1110-2-2901 Mechanics Volume 6 pp. 189-236 Springer-Verlag Retrieved on 2016 August 23 from Bieniawski Z.T. 1992. Design Methodology in Rock Engineering. http://www.publications.usace.army.mil/USACEpublications/Engin Balkema, Rotterdam eer-Manuals/udt_43544_param_page/2/ Bondarchuk A. Ask M. V. S. Dahlström L. O. and Nordlund E. 2012. Viggiani G. 2012. Geotechnical Aspects of Rock Mass Behavior Under Hydropower : A in Soft Ground. Technology and Engineering 7th International Two-Dimensional Numerical Study. Rock Mechanics and Rock Symposium on Geotechnical Aspects of Underground in Soft Ground, Rome, . Brown E. T. Bray J. W. Ladanyi. B. and Hoek. E. 1983. Ground Wahlstrom E. E. 1973. Tunneling in Rock. Developments in response curves for rock tunnels. Journal of geotechnical Geotechnical Engineering 3. Elsevier Scientific Publishing engineering, Volume 109, no. 1. ASCE Company, Amsterdam. Greer A. J. 2012. Finite Element Modeling and Stress Analysis of Zhang C. Wang Z. and Wang Q. 2015. Deformation and Failure Underground Rock Caverns. University of California, Irvine, Characteristics of the Rock Masses in Tunnels. Mathematical ProQuest Dissertations Publishing, 2012 Problems in Engineering. Hindawi Publishing Corporation Hoek E. Kaiser P. K. and Bawden W. F. 1995. Support of underground excavations in hard rock. Rotterdam, Netherlands: A. A. Balkema Publishers

- 1740 -