YREREPORT ~_

Guideiines for the Design of Tunnels

ITA Working Group on Genera! Approaches to the Design of Tunnels

Abstract-e- This ,e(Grui report bv thr ITA Warkmg- Group On (~eneral.ipproaches to th« Drsuin or Tunnels presents IrllernatlOl1a! desHUlprocedures iot tunnels. In most tW1nt'lIzrzç proterts. CTroundactiuel» pcructpates In prooiding stabilitv to openmg-. T heretore,t he urnerai approach la the desIgn o! tunnels includes sitei.Ilvestl,;;atwns. ground prooines and in-situ mO/uIoTln,;;, as ue]! ashe ana/vszs oi stresses and deiormntions, For the latter, the ditierent.tructural desz,;;n models' apolied at present-s-including theobseruationai method-e-are presented. Guidelines lar the structural.ieuuline ol the tunnel Immg and natzonal recommendalzons on,,,mze! design are aiso gn'en. ft Is hoped that the mlormatlOn herrin,based 011 expertences [rom a unde range of tunnel/mg- provects. uull be-tissemrnated la tunnel designers throughout ihe world. I

1 , Scope of the Guidelinesrr he International Tunnelling Associauon (ITA)\Vorking Group on General Approaches to the Design

, of Tunnels was established in 1978. As its Iirst project,the group developed aquestionnaire aimed at compilinginforrnation about structural design models used in differentcountries for tunnels eenstrucred prior to 1980. A synopsis ofrhe answers to the questionnaire was published bv theInternational Tunnelling Association in ]982 (ITA ]982).As a continuation of that Iirst report, the working group

herein presents guidelines that atternpt to condense thevarrous answers from the first report and include additionalexperiences in the general approaches to the design ol tunnelstructures, These guidelines tulfill one of the rnain objectivesof the International Tunnelling Associauon, namelv, todisperse intermation on underground use and underground

llillllia.;nuctur,esth,roughom the world bv crossing national bordersF'l"'Î ,<I language harriers.

j hose iruerested in the subject of tunnel design should alsoconsult nuhlished reporisot other ITA work ing groups. e.g,the recent IT,\ report on contractual sharing of risk (seeTijt/ST 3:2) and ITA recommendations on mairnenance ofrunn-Is (see [:rUST 2:3). Furtherrnore, a number of national~Hld international organizarions. such as the InternationalSocietv on Rock Meeharnes. have published recornmenda-uons on related subjects, such as field measurernents andlaboratorv iesungs lor rock and ground. Some ol thesepublications and reports are lisred in the Appendix.In tunne llina, most otten the ground acuveiv partir ipates

in providinz stahilitv ro the opening. Theretore. the designprocedure lor tunnels. as cornpared 10 aboveground-trucrures. is much more dependent on such factors as the SHeSI rnanon. ihe rharacter ist ics, and [he excavarion andsupport rnerhods used. Recornrnendauons on tunnel design

This report IS rditeti 0'11 Heinz Duddeck. Ammateur o! theITA Warkmg Group on (;meral Approaches m (he Design ofTunnels. Present address: Prof. Hein: Duddeck, Techrucal[!mver.\!l1l ol Braunschwezg. Beethouenstrasse 5I. ,,00 Braunschusrig,Fed eral Reoublic of (;ermanv.

{(wl'ullmQ; (Hlä {'r:JnQ'1'ound SplKt j't-,f!'lmloU;'V, Vol No pp. :"::57-2'i9. 19X~

Pnmed Hl (~i(';l.i Hnt~Hn.

Résumè-Le grouoe de tmvazi AITES SUf ie dimensionnemeni destunnels presente ici .Ion deu xteme rapport. En rassemblant loutes:nfol'matuJns, 'lUl rtaient accessibles entre les pa»: :>IU i,dimens!071nement des tunnels, nous espérons, que les expenencesgagnées sur beaucoup de protets des travaux souterrams seron/propagèes dans tout Ie monde. Parce que Ie sol parucipe d'une grarui»partie à iournir des moyens de stabilitè pour des ouverturessouterrames, des methodes de dlmenszonnement comprennent ausSIhlen les inVestlgatlOns sur Ie chantler. ies essais laboratolTes et lasurveIllance pendant Ie progres du travai/ que l'analvse descontramles et des detormations. Concemant ce dernzer ootnt, Jesmode/es de dimenswnnement diitèrents et actuellement ~ppliquéssont prèsentès, y compris auss! la methode d'obseroation,Recommendations oous les details de revêtement et queiou-,recommandatlOns natlOnaies sur Ie dimenswnnement des tunnelsachèoent ce rapport.

naturaHy are lirnited with regard to their consistencv andapplicabilitv because each tunnelling project is àffected byspecial features that must be considered in the design.Nevertheless, it is hoped that the general outline provided inthese guidelines, based on the experience gained hom manvtunnelling projects, may be of sorne help for these staTting aproject.

2. Outline of General Approaches2.1. General Procedure inDesigning a TunnelPlanning a tunnelling project requires the interdependen-

participation of the tollowing disciplines, at a minimum:

• Geology.• Geotechnical engineering.• Excavarion technology. e.g. machine tunnelling.• Design of the supporting structural elements. induding;long-term behavier of rnaterials.

• Coacracr principles and law,

Although the experts in each of these disciplines mav heresponsible only for their specific area of know ledge, thedecision on the ma in design features should be the outcomeo[the cooperative iruegration of all the disciplines. Only ihuscan it be ensured rhat the project. in all its details, has beendeveloped in unitv, and nor as the consecuuve addition of theseparate work of each of the experts.The basics doeurnerus lor tunnel design should inc!udeor

cover:

• The geologica! report presenting the resu irs ol thegeological and geophvsical survey.

!lil The hvdrogeological report.t> The geotechnical report on site invesrigarions. includingthe interpretation of the resu lts of site and laboratory testswirh respect to the tunnellins; process, soi l and mekclassificarion. etc.,

• Inforrnauon on l ine, cross-secuon. drainage, andstructural elements afiecting later use of the tunnel.

!lH>«i· 779R SM UGO + ,noPl'f~pmon Prrs, pa,

..

• Plans for and a description of the projeered excavauon ordriving procedure. including rhe different cross-secrionsreiated to different ground condinons.

• Design doeurneuts for the types of excavauon rnethodsand tunnel supports likely to he applied, considering,e.g. excavation advance and face support (tvpes andnumber of anchors. shotcrete strength, closure length,ete.

• The program Ior the in-situ monitoring of the tunnel bvfield measurernents,

• The analvsis of stresses and deforrnations rfor unlinedtunnels as wel! as for single-er double-lined tunnels I. andthe dirnensioning of the tunnel support for intermediatephases and final linings,

• The design tor waterproofing or drainage.• Structural doeurnerus for the Iinal design of the tunnelproject. including the detailing,

• During and after the excavation, reports on the fieldmeasurernents and interpretation of their results withrespect to the response of rhe ground and the structuralsafety of the tunnel.

• Doc~mentation of the problems encountered during theexcavation and measures applied, e.g. strengthening theground or changing the projeered type of support. basedon monitoring resuits.

I

The above sequence of these basic doeurnenis also providesthe general outline of the design procedure.

2.2. Elements of the StructurslDesign Model for TunnelsIn planning, designing, analvsing and detailing a

structure, engineers promise that the structure wiJl neithersuffer structurally nor collapse during its projeered lifetirne,Thus, rnodels of the realîry are necessarv for analvsis in orderto predier the behaviour of a tunnel during the excavation andduring its lifetime, Models are also needed for bidding onprojects,The following main elements involved in the design

procedure are shown as a flow-chart in Fig. I:(1) Geology and site inuestigations must confirm the line,

orientation, depth, etc.. of the opening, e.g. a cavern.(2) Ground probing and soil or rock mechanics must be

aoplied to determine the ground characteristics, e.g. primarystresses, soil or rock strength, Iaults, water conditions.

Fer N aciual StaUlontv·lnltnOWO sail/tiV ma1çm

F(gure 1. Design process (or tunnelling.

238 Tt·C'.;NELLlNG AND .• )NDERGROUND SPACE TECHNOLOGY

Experience end preliminar» estimates or calculationsare used to deterrnine the cross-senion required and the choiceof the excavation method er the tunnel driving machine to beused, as well as the methods ot dewatering the ground and theselection of the supporting structural elements,(4) After steps IlH3) are completed, the tunnelling

engineer must derive, or even invent, a structural model. Byapplving equilibrium and cornpatibility conditions to themodel. the engineer has to arrive at those criteria that arefactors in deciding whether or not the design is safe. Differentrnodels may be used for each excavation phase, Ior thepreliminarv and the Iinal tunnel lining, or for differentground behaviour, e.g. in discontinuous rock or homo-geneous soft soil. Modelling of the geometrie features rnavvarv greatlv, depending on the desired intensitv of theanalvsis.f:;) A safety concept drawn hom failure hypotheses mav be

based on criteria such as strains, stresses, deformarion. orIailure modes.The bypass in Fig. I indicates that for manv underground

structures, as in mining or in self-supporting hard rock. nodesign models at all are applied, In such cases, pastexperiences alone may he suffierent.Risk assessment by the contractor as well as by the owner is

needed at the time of contract negotiarions. Risks involvepossible structural failures of the tunnel support and lining,functional Iailures after completion of work, and Iinancialrisks. The contractual aspects also include risk sharing andrisk responsibilities.In-situ monitoring can be applied only after the tunnelling

has begun, If the displacements stop increasing over time. itgenerall y may be assumed that the structure is designed safe!y.Yet monitoring provides onlv part of the answer to thequestion of safety, Ior it does nor tell how close rhe structurernay be to sudden collapse or nonlinear failure modes. Theresults of field rneasurements and experiences duringexcavation may com pel the engineer to change the designmodel by adjusting it to real behaviour,An iterative, step-bv-step approach is characteristic of the

design of structures in the ground that ernplov theparticipating strength of the ground (see loops in Fig. I). Thedesigner may begin by applving estimated and simplebehavioural models, Adjustments based on actual experiencesduring rhe tunnelling excavation (such as excavating theinitial sectien in the same ground condinons or driving apilot tunnel) will bring the model doser to reality and refine ir(if refinement is consistent with the overall accuracyauainable). The interpretations of in-situ measurements (andsome back analyses) also rnav assist designers in making theseadjustments.All of the elements of the structural design model in Fig. I

should be considered an interacting unitv, Scattering ofparameters or inaccuracv in one part of the model wiJl affectthe accuracv of the model as a whole. Therefore, the Samedegree of simplicity or refinemenr should be providedconsistentlv through all the elements of the design model. Forexample, ft is inconsistent to apply very refined mathematicaltools simuitaneously with rough guesses of importantgroundcharaeteristics.

2.3. Different Approaches Basad onGround Conditions and Tunnelling MethodsThe response of the ground to extavation of an opening can

vary widely. Based on the type of ground in which mnnellingtakes place. four principal types of tunneHing may be defined:I) for cut-and-cover tunnelling, in most cases the ground

acts only passively as a dead load on a tunnel structure erectedlike any aboveground engineering structure.(2) In soft ground, immediate support must be provided by

a süH !ining (as. for example. in the case of shield-driventunnels with tubhings for ring support and pressurized slurry

Volume 3, Number 3. 1988

lor tace support). In such a case. the ground usuallvproviding resistance Hl outward

detorrnauon ot the lining,i. j) In medium-hard rock or in more cohesive soil, the

ground mav he streng enough to allow a eertam open secnonat the tunnel face. Here. a eertam amount of stress release maypermanentlv he vaiid befere the supporting elements and thelining acting effectively, In th is siruation onlv afraction the prirnarv ground pressure is acting on thelininz.

i. .J:l···When tunnelling in hard rock. ihe alone mavthe stabilitv of the opening sa that onlv a th in

anv, wil! be necessarv for surface proteetion. The designmodel must take imo account the rock around the tunnel inorder 1O predier and verify safety considerations anddeformations.

Especiallv in ground condirions that change along thetunnel axis, the ground may he strengthened bv injections.anchoring, draining. freezing, etc. Under these circumstances.case (2) mav he irnproved, at least temporarilv, to caseThe characteristic stress release at the tunnel face t Erdrnann

1983) is shown in Figs 2 and 3. The relative crown displace-ment w is pletred along the tunnel axis, where w/wo = l.0represems the case of an unsupported tunnel. In medium-stift ground nearlv 80%of the deforrnations have already takentplace befere the lining (shown here as shotrrete) is stiffenough to participate.

O,ll

(1,2______~.L

0,6 <,.0 \llIll1ll!1IlOdolI without

0,& / stress '.iRst

1.0 w, ,wIIMut il""'!lT-------L--r;;,o-t w/.,

Figure 2. Crown displacement w along the axis, ahead andbevond the tunnel face.

»lolm ••~Ol GoorG~I ,;; \>

, I"

-lilt-I

Figure 3. Ground stresses acttng on the lining as fmetions ofIhè pnmarv stress (Erdmann 1983J.

Volume 3. :'Jumber 3. 1988

Fm a simpliiied plane model with na stress release, wherethe Iull primary siresses are assurned to act on a lined opening,ihe displacement mav be onlv 0.4 of that occurring in theunsupported case. The corresponding stress release is shownin Fig. 3. The simplified example, considering onlv theconstant part of radial pressure, yields the values shown for aring stiffness of EaA = 15.000xO.3 =4500 MN/m and a grounddeforrnation modulus of EK = 1000M:'J/m2•

Even in the unrealistic case when the full primary stress actssimultaneousiv on the ground opening and the lining, onlv.">5%of the stress is taken by the Iining: in the case of EBA2250 MN/m, onlv 38% is taken by the lining. If an opensectien of 0.25 of the tunnel diameter is left without liningsupport. the iining takes onlv 25%of the primary stresses: IorLr = 0.5 D, it takes only 12%of the primary stresses.FOT very soft ground requiring immediate support (as in the

case of very shallow tunnels). almest 100%of the primarystresses are acting on the lining, The values change, of course.with other stiffness relationships and other stress distribu-tions than these shown in Fig. 3, with other cross-sections.and other tunnelling methods,

2.4. Site tnvestiqetions; StructurslAnalysis end In-Situ MonitoringAn adequate intensity of site exploration, from which

geological and hydrological mappings and ground profilesare derived, is most important for choosing the appropriatetunnel design and excavation rnethod. A well-documentedgeological report should provide as much intermation as isobtainable about the physical features along the tunnel axisand in the adjacent ground, The amount of informationshould be much greater than the inforrnation required Iorentering directly into a structural analvsis.The results of an analysis depend very rnuch on the

assurned model and the values of the significant parameters.The main purposes of the structural analvsis are to providethe design engineer with: (I) a better understanding of theground-structure interaction induced by the tunnellingprocess; (2) knowledge of what kinds of principal risks areinvolved and where they are located: and (3) a tooi forinterpretating the site observations and the in-siturneasurements,The available mathematica 1methods of analysis are much

more refined than are the properties that constitute thestructural model. Hence, in most cases it is more appropriateto investigate alternative possible properties of the model, oreven different models, than to aim for a more refined model.FOT most cases. it is preferabie that the structural modelernploved and the parameters chosen Ior the analyses belower-Iimit cases that may prove that even for unfavourabl«assurnptions, the tunnellieg process and the Iinal tunnel aresuffici:ffit!y safe. In general. the structural design model doesnot trY to represent exactlv the verv actual condinons in thetunnel. although it covers these conditions.In-situ monitoring is important and should he an integral

part of the design procedure. especiallv in cases wherestahil ity of the tunnel depends on the ground properties.Deformanons and displacemenrs generallvcan he measuredwith much more accuracv than stresses. The geometrv of thedeformanons and their development over time are mostsignificant Ior the interpretanen of the act ual events,However, in-situ moniroring evaluates onlv the verv local andacrua I situatiou in the tunnel. Therefore, in genera I thecondinons taken into account bv the design calculations donot coincide with the condinons that are monitored. Only byrelaring rneasuremenr results and possible failure modes bvextrapolaring can rhe engineer arrivé at considerations ofsafety margins.In many cases, exploratorv tunnelling may he rewarding

because of the inforrnation it vields on the actual response ofthe ground to the proposed m~thods for drainage. excavation,

TI'NNELLlNG AND UNDERGROt'ND SPACE TECHNOLOGY 239

TBM drivinz. support. etc. In important cases a pilot tunnelmav be driven: such a tunnel mav even be eniarged to the lul!fin~l tunnel cross-secnon in the most representarive groundalong the tunnel axis, For Iarger projects, it rnav he useiul toexcavate a trial tunnel prior to commencing the actual work,:VIoreintensive In-situ monitoring of the exploratory tunnelsections should check the design approach by numericalanaiysis.

2.5. Design Criteria andEvaluating Structursl SafetyAn underground structure may lose its serviceabilitv or its

structural safety in rhe following cases:The structure loses its watertightness,The deformations are intolerably large.The tunnel is insufficiently durable for its projeered lifeand use.The material strength of the strucrural elements isexhausted locally, necessitating repair.The support rechnique (for exarnple, in erectingsegmental linings) fails or causes damage.Exhaustion of rhe material strength of the system causesstructural failure, although the conesponding deforma-tions develop in a restrained manner over time.The tunnel collapses suddenlv because of instability.

The structural design model should yield criteria related to'Iailure cases. against which the tunnel should be designedsafely. These criteria mav be:

• Deforrnanons and strains.• Stresses and utilization of plasticiry.• Cross-sectional Iining failure,• Failure of ground or rock strength.• Limit-analysis Iailure modes.

In principle, me safety margins may be chosen differently Ioreach of the failure cases lisred above. However, in realitv theevaluauon of me actual safety margins is most complex andvery much affected by the scattering of the involved propertiesof the ground and the structure and, Iurthermore, by theinteracting probabilisnc characteristics of these properties.Therefore, the results of any calculation should be subject tocrincal refleceion on their relevanee to the actual conditions.'National codes for concrete or steel structures may not

always he appropriate for the design of tunnels and thesupporting elements, Computational safetyevaluationsshould alwavs he cornplemented by overall safetyconsiderations and risk assessments employing crincalengineering judgment, which may include the Iollowingaspects:

• The ground characteristics should be considered in lightof their possible deviauons from average valnes.

• The design model itself and the values of parametersshouid he discuseed bv the design team, whieh includesall of the experts involved (see Secnon 2.1. "CeneralProcedure in Designing a Tunnel," above).

• Several and more simp Ie calculation runs withparamerrie variations may uacover rhe scattering of theresults. In general. this approach is much moreinformative than a single over-refined investiganon.

• The in-situ measurements should he used for successiveadjustment of design models.

• Long-term measuremem of ddormations via extra-poiation may reveal lO a large extent the final stabilitv olthe struemre, although sudden collapse may not blOannounced in advance.

3. Site Investigationsand Ground Probings3.1. GeoJogical Data and Ground ParametersThe appropriate amount of ground investigations on site

and in laboratories mav vary considerably from project 1O

2,10 TUNNELLlNG AND I 'NDERGROUND SPAC.E TECHNOLOGY

Because the types of ground explorations andprobmgs depend on the special features of the tunnellingproject. its purpose. excavation rnethod, ere.. thev should hechosen bv the expert team, especially in consultatien with thedesign engineer. The intensitv of the ground explorationswil] depend on the hornogeneirv of the ground, the purpose ofthe tunnelling, rhe cost of boring, e.g, for shallow or deepcover. and ether factors.The geological investigations should include the following

basic geotechnical information (see also 15RM Commissionon Classification of Rocks and Rock Masses 1981).

3.1.1. Tunnels in rock

Zoning. The ground should he divided in geotechnical unitsIor which the design characteristics mav be considereduniform. However, relevant characteristics rnav displavconsiderable variations within a geotechnical unit. TheIollowing aspects should be considered Ior the geologicaldescripnon of each zone:

• Name of the geological formation in accordance with agenetic ciassification.

• Geologie structure and Iracruring of the rock mass withstrike and dip orientauons.

• Colour, texture and mineral cornposition,• Degree of weathering,Parameters of the rock mass e.g. in five classes of intervals.

including:

• Thickness of the lavers.• Fracture intercept,• Rock classification,• Core recovery.• Uniaxial compressive strength of the rock. derived fromlaboratory tests.

• Angle of friction of the fractures (derived from laboratorvdirect shear tests).

• Strength of me ground in on-site situations.• Deforrnation properties (modulus).• Effect of water on the rock quality.• Seismie velocity.Primary stress field of the ground. For larger tunnel

projects, tests evaluating the natural stresses in the rock massmay be recommended, For usual tunnel nroierts one shou klleast estimate the stress ratto Oh! Cl" al ruunes ,cve" wnere Oh ISthe lateral ground pressure and (1" me major principal stress(usuallv in the vertical direction), Ior which the weight of theoverlving rock generally may be taken. Teetonic ssressesshould be indicared.Water conditions. Two types of inforrnation about water

condinons are required:(I) Permeabilitv, as determined bv:Coefficient k (mis) (from field tests).Lugeon unit (from tests in boreholes),(2) Water pressure:At the tunnel level (hvdraulic head),.tt piezometric levels in bereholes.Deiormabilitv of the rock mass. In-situ tests are required ro

derive the two different deformation moduli. which can bedetermined either from static medlods (dilatometer tests inboreholes. plate tests in auits. or radial jacking tests inchambers) or from dynamic methods (wave velocitv byseismic-refraction or by geophysical logging in borehoies).Engineering judgment should he exercised in ehoosing the\'alue of ,he modulus most appropriate for the design-forinstance. oy .he relevant tangent of the pressur,e-dtef(Jrl1naitioncurve ar the primary stress level in the statie method.Propenieslor which information is needed when tunnel

boring machines are to he employed indude:Abrasiveness and hardness.Mineral composites. as, e.~.quartzite contents.Homogeneity.

Volume 3. Number 3, 1988

Suielling potential of the rock. The présence of sultates,hvdroxvdes, or minerats should he nvesugated bvrnineraloaical testing. A special odeometer test mar used todeterrnine the swell test-curve of a specimen subjeered first to aload-unload-reload cycle in a dry state. and then unloadedwith water.The following ground water conditions should be given:Water levels. piezometric levels. variations over time.pore pressure measurements in confined aquifers.Water chemisuv.Water temperatures,Expected amount of water inflow,

3.1.2. Tunnels in sml

The geotechnical description should primarilv follow therecornmendations given above for rock. Additional specialfeatures for wil include:1. Soil identijication (laboratorv testing):

• Partiele size distribunon,• Atterberg Iimits WI' wp'• Unit weights. ')I, "Id. 'Yl:.• Water content ur.• Permeabilitv k.• Core recovery.2. Meehamcal properties determined bv laboratory testing:

• Friction angle dru, cP.• Cohesion e", e.• Compressibility Tnt" c.:

3. Meehaniea!:tJroperties deternined bv field testing:

• Shear strength Tl' (Vane-test).• Penetration N (Standard Penetrasion Test).• Deformabilitv E (Plate bearing, Dilatometer).4. Ground water eondition (in addition to these liseed in

3.1.1.): permeability, as determined by pumping tests.

3.2. Evaluation of Parameters by GroundProbing en« Laboratory TestsThe properties of the ground that are relevant for the tunnel

design should be evaluated as carefully as possible, In-situtests. which cover larger ground rnasses, generally are moresignificant than are laboratorv tests on srnall specimens.which often are the better preserved parts of the coring. Thenatural scattering of ground properties requires anappropriate number of parallel tests-at least three tests foreach propertv (see also the cortesponding 15RM recorn-mendanons l.Results of laboratorv tests must be adjusted to site

conditions. The size of specimen, the effects of ground water,the inhomogeneitv of the ground on site. and the effects ofscattering must he considered, The conclusions drawn fromtests also should take into consideration whether thespecimens were taken from distutbed or undisturbed ground.In many cases, the first part of the runnelling may he

interprered as a large-seale test, the experiences from whichmav he drawn upon not only Ior the subsequent excavationsbut also Ior prediering ground hehaviour. In certain cases,long horizon tal boreholes may facilitate ground prohingahead of the face, or a pilot tunnel may serve as a test tunnelthat at the same time provides drainage. The on-siteinvestigations provide valuahle resuhs for checkillg thecorrelation of in-situ tests with laboratorv tests.Special tests correspond directly to the proposed

mnnelling method may oe required. e.g. lor the sufficiempreservation of a membrane at the face of a hentonite shield.The evaluation of the parameters should indicate lheexpected scattering. From probabilistic consideratioll ofnorma!ly dislributed quantities il ean he dedun'd that a mean

Volume 3. Number 3. 1988

value or a value corresponding to a moderately conservativefractile of a Caussian distribution is more appropriate thanthe worst case value.Aset of all the parameters descrihing the ground behaviour

of one tunnel sectien with regard 10 tunnelling should he seenas a comprehensive unit and should he well-balanced inrelation 10 each ofthe parameters. For example, a small val ueof ground deîormation modulus indicates a tendency 10

plastic behaviour, to which corresponds a ratio of lateral tovertical primarv stress that is closer to LD. Hence, Ioralternative investigations some complete. balanced sets ofparameters shouid be chosen instead of considering eachparameter alone, unrelated to the ethers,The available rnethods Ior ground prohing and laboratorv

tests, their applicabilitv and accuracv are given in theAppendix.

3.3./nterpretation of TestResults end ûocumentetionThe field and laboratorv tests should be given in well-

doeurnenred reports. in the form of actual results. Based onthese reports. an interpretation of the tests that is relevant tothe actual tunnelling process and the requirements of thedesign models for Ihe structural analvsis is necessary. At thetime the tests are planned, the team of experts referred to inSectien 2.1 should decide which ground properties andground characterisucs are necessary for the generalgeotechnical descripuon of the ground and for the projeereddesign model. Thus, a closer relationship mav be achievedbelween ground investigations and tunnelling design. andbetween the amount and refinement of tests and thetunnelling risks.The doeurnerus should lay open the rationat inter-

pretational way in which design values are derived from testresults. This method has proven 10 he especially useful in theteadering process, because it condenses the relevant data forthe description of the ground and for the design of the tunnelon a band along the tunnel axis beneath a graphicalrepresentation of the runnel profile (see the examples in Figs9-13).Such condensed tables may be prepared Iirst for rendering

and the preliminarv design. and then improved throughexperience gained and incoming monitoring results,However, it should be clearlv stated, especiallv in the contractpapers. that much relevant information is lost oroversirnplified in such tab les. and that therefore thegeotechnical reports and ether complete doeurneuts should beconsidered the primary doeurneuts.

4. On Structural DesignModels for Tunnelling4.1. Alternative Design ModelsThe ëxcavation of a tunnel changes the prirnarv stress field

into a rhree-dimensional pauern at the tunnelling race.Farther from the face. the stress field eventuallv will return toan essentiallv two-dimensional svstern. Therefore, the tunneldesign may consider only two-dimensional stress-stram Iieldsas first approxirnations,The design of a tunnel should take into account the

imeraction between ground and lining. In order to do so. thelining must be placed in dosest possible hond with theground. To preserve Ïts nalural strength. Ihe ground shouldhe kept as undislurbed as possible. The deformationsrestdting from Ihe tunnelling process (see Fig. 2) reduce lheprimarv ground pressure and create stresses in the iiningeorresponding la that fractionai part of the primary stresses inthe ground which act on the suslaining lining. The stressesdepelld on the stiffness relationship of the ground lO theIining, as weil as on the shape of the tunnel cross-section. Thelatter should he selected such that an arching anion in thegroundand the !ining may develop.

TiiNNELUNGAND UNDERGROl1ND SPACE TECHNOLOGY 2·11

IGv=y h

Gv Gv Gvmn mn rrm rrmGh~$~~m~~m~~m~Ghom [[[] om rrrn

{r.::\2 ''\.V

empiricalapproach

t t t ~ .::1$::1-'+7--

Figure 4. Alternative plane-strain design models fOT different depths and ground stiitnesses.

Figure 4: presents Iour different structural rnodels for aplane-strain design analysis, The cross-secrions need not becircular. These Iour models are explained more explicitlybelow.In soft ground, irnmediate support is provided by a

relativelv stiH lining. For tunnels at shallow depth (as forunderground railways in cities) , it is agreed that a two-dimensional cross-sectien may be considered, negleering thethree-dimensional stress release at the face of the tunnelduring excavation. In cases (I) and (2) in Fig. 4, the groundpressures acting on the cross-sectien are assumed to be equalto the prirnarv stresses in the undisturbed ground. Hence, it isassumed that in the final state (some years alter theconstruction of the tunnel), the ground eventuallv will returnto nearly the sarne condition as befere the tunnelling,Changes in ground water levels, traffic vibrations, erc., mavprovoke .this "readjustment,"In case ( I), for shallow tunnels and soft ground, the full

overburden is taken as laad. Hence, no rension bedding isallowed at the crown of the tunnel. The ground reaction issimplified by radial and tangenrial springs, arriving at abedded-beam model,In case (2), for rnoderatelv stift ground, the soil stiffness is

ernploved by assuming a rwo-dirnensional continuurn modeland a complete bond between lining and ground. As in case(I l, stress release due to predeformations of the ground isneglected, Inward displacements result in a rednetion of thepressure on the lining,Case assurnes that some stress release is caused

deforrnauons that occur befere the lining participates, Inmedium-hard rock er in highlv cohesive soil, the ground mavbe streng enough 10 allow a certain unsupporred section alrhe tunnel face (see Fig. 2). Also, for tunnels having aoverburden. a rednetion of the acting crown pressure(represented in Fig. 4 bv h < H) is taken into account.In case 141,the ground stresses acung on the lining are

determined bv an empirica! approach, which rnav he based onpreviousexperiencee wirh the sarne ground and rhe sametunnelling rnethod, on in-situ observations and monitoring

242 Tt';-';7'iE:LUNG ANI) UNDERGROUND SPACE TECHNOI.OGY

of initial tunnel sections, 00 interpretation of the observeddata, and on continuous improvements of the design model.If a plane model is not justitied-e-as is the case for caverns,

for more complicated geometries of underground structures,or for an investigation directly at the tunnelling face-a three-dimensional model may be necessary (see Fig. 5). The three-dimensional model also may be conceived as consisring ofdiscontinuous masses (block theory) or a continuurn withdiscrete discontinuous fissures or faults,

Q,

Figure 5a. Three-dimensional continuum model.Figute 5b. Example of tuio-dimensional finite-elementmodel.

4,2. Continuurn orDiscontinuum ModelFor structural design models such as those in Figs Sa and b,

the ground may he modelled as homogeneous orheterogeneous. isotropie or anisotropic: as a (WO-dimensional, i.e, allowing some stress release befere thelirnng is acting, or a three-dirnensional stiff medium. TheIining may be modelled either as a beam element withbending stiffness or as a continuurn. Plasticitv, viscosirv,Iracture of the rock, non-Imear srress-strain and deforrnation

Volume 3, Nurnber 3, 1988

behaviour, etc., ma. be covered bv special assumptions formaterial laws.The design criteria are computed by numerical solutions,

Frorn rheir origins, the finite-element metbod and theboundarv-element method are basically continuurn methods.Thus, homogeneons media and stress-strara Iields areevaluated best, In general. discontinua such as rock withfissures and faults, and Iailure modes, which are initiared bvlocal rupsure. shear failure, or full collapse, cannot be coveredbv connnuum rnethods.A continuurn or discontinuurn model is appropriate for

tunnel structures where theground provides the principalstability of the opening (as in hard rock) or where thegeometrical properties of the underground opening can bemodelled onlv by numerical analysis, e.g, in the case of doselyspaeed twin tunnels.

4.3. 8edded-Beam Model(Action-Reaction Model)If the stiffness of the ground is smal! compared to the

stiffness of the lining, a design model such as that shown inFig. 6 may be emploved. In such a case, the active groundpressures are represented by given loads and the passivereaction of the ground against deformarions is simulated bvconstant bedding m~uli. 1?~model may he particularl~well-suited to the design af linings af shîeld-driven tunnels.As to applicabilitv, the stiffness ratio 13 may be smaller than200:

13 = E.R3/E] < 200,where: E. is the representative deformation stiffness

modulus af the ground,R is the radius of the tunnel cross-sectien or its

equivalent for non-eireular tunnels,E] is the bending stiffness of the lining,

A more correct soletion for the bedding is given by a non-zero stiffness matrix for al! elements with regard to radial andtangendal displacements,However, in most cases and in view of the unavoidable

approximations based on the ether assumptions, a simplerapproach may be sufficient. Such an approach considers only~dial (and, eventually, tangential) bedding, negleering theinterdependence of radial and tangenrial displacements andbeddings, For non-eireular cross-sections, the continuurnsolution reveals that bedding may be increased at cornersections of the Iining, with smaller radius of the curvature.The bedded-beam model may be adjusted to more complex

cases, e.g, by reducing the crown load in accordance withstress release at the tunnel face (see Fig. 3)or, for deep tunnels,by assuming bedding also at the crown.For articulated effmive hinges in linings the bending

rnoments are smaller; the deîormations may be larger,depending on rhe ground stiffness. For hinged linings thelimit of f3 given above is aot valid.The analvsis of the bedded beam vields ring farces. bending

mo~e~ts, and d~formations as design criteria for the lining. Ifthe lining' rmg IS completely dosed, the bending momemsmay he considered less important than the ring forces forproviding equilibrium (a smaller safety factor may be

fb''V·H --rrt+rJ Ia~'f; 8.Hi4R

\j~~GhG"••K,,'Gt:Î:tIJ Kr"conlllt.

Gv radial ground hoop bendingy ",volume w.igll1 displ. reactien tere.!> momentlll

Figure 6. Example of a bedded-beam model for shaUowtunnels.

Volume 3, Number 3. 1988

w'j"" __ usionfJI$. streu t.!M".

E '_i!lwi •••••l1lldccntlfwity

Figure 7. Characteristic curves [or the ground and the suppon[or conuergence-coniinement models (Fenner-Pacher c~rves J.

justified Ior the bending rnoments). Allowances also may bemade for a plastic rotatien capacity of the Iining segrnents.For tunnels with verv pronouneed stress release due to

inward deformations, e.g, Ior deep tunnels in rock, a simpleapproach to d~sign considerations is given by theconvergence-confinement model, which is based only on theint~ction of the radial inward displacement and the supportreaenen to ~hesedeforrnations by resisting ring Iorces and thecortesponding outward pressure (see Fig. 7).The primary stresses (Jo in the ground are released with

progressive inward displacements. The acting pressure maye~en mcrease when rock joints are opening with largerdisplacements, In self-supporting rock, the ground char-acreristic in Fig. 7 meets the w-axis; because the primarystresses are released completely, a supporting lining is notnecessary, Befere the supporting memhers are installed. it isunavoidable-c-even desirable-i-raar decompression associatedwith the predeforrnation Wo will oceur, The stiffness of thelining determines where both curves (characteristic lines) wil!interseet. At this point, equilibrium as weU as compatibilitvcondinons are fulfilled. If the ground characteristic is known,e.g., by in-situ monitoring, rhe predeformation Wo and thestiffness of the lining (including its development over timeand as tunnelling advances), and even its plastic propertiesare veI')' de~isive for the actual stresses in the lining, Bothcurves m Fig. 7 may vary considerably.In its usual analytical form, the convergence-confinement

. model assumes constant ground pressure along a circulartunn~llining. Consequently, üyields only ring Ierces and nobending moments at all. However, ie may be extended to coverground pressures that vary along the tunnel lining (Cesta1986).The model mayalso he applied as a first approximation Ior

non-~ircular tunnel cross-sections, although the supportreaenen curve is dîstinctlv different, e.g. for horseshoe-typecross sections. Therefore. it may he helptul to use theconvergence-confinement model in combination with acontimmm model and in-situ rneasurements.~lth?ugh the convergence-confinement approach is

primarily a tooi Ior the interpretation of field rneasurementsi t also may he applied in support of the empirica! approach:

4.4. Empirical ApproachThe structural elemems :md the excavation procedure,

especially for the preiiminary support of the tunnel, may beselected mainly based on experience and empiricalconsiderations that rdy more on direct observations than onnumerical ?llculations. This procedure may be especiallyreasonahle af experiences from a successful tunnelling projectcan be applied to a similar. new one yet to he designed. Such atransier of information is justified only when:

• The ground conditions, including those of the groundwater, are comparable.

• The dimensions of the tunnel and ltS cross-sectionalshape are similar.

• The depths of overburden are approximately the same.

TUNNELLING AND UNDERGROVND SPACE TECHNOLOGY 243

• The runnellina methode to he emploved are the same,• In-SItu rnonitorina yields results cornparable to those lorthe precedmg tunnelling project.

One disadvantage of prolonged application of the empirica!approach is that, lacking an incentive to applv a moreappropriate tunnelling design via a consistent safetyassessmern, the structure may he designed overconservauvelv.resulting in higher construction costs, The sirnpleapproach coutributes little to the advancement of rhe state ofthe art in tunnelling.The ernpirical approach to tunnel design mav also be

applied to larger projects in only alightlv changing ground ifprovision is made (especiallv in the tender) Ior initialexperiences to be extrapolated to the subsequent sectionsalong [he tunnel axis. Such a sinration jusrifies arneasurernent programme that is more intensive for the Iirstsections, in order to gain experience,

4.5. Observational MethodBy combining analvtical methods with the empirica!

approach and the immediate interpretations of in-situmeasurernenrs. a tunnelling design procedure that isadjustable as the tunnel excavarion proceeds rnav be applied,In this approach. the field rneasurements of groundmovements, displacements and stresses in the Iining are us~don an ongoing basis to verify or modifv the design of tfietunnel. More intensivelv instrumented sections at the earlystages of the tunnelling provide the data for these procedures.The interpretation of the rneasured data vields insight intothe ground behaviour as areaction to the tunneilingprocedure.In applving the observational method, the following

condinons must be met:

• The chosen tunnelling process must be adjustable alongthe tunnel Iine,

• Owner and contractor must agree in advance oncomractual arrangements that allow for modifications ofthe design 00 an ongoing basis during the project.

• The field measurements should be interprered on thebasis of a suitable analvtical concept relatingmessurement data to design criteria.

• The interpretation of a partienlar instrumented sectienmust be used to draw conclusions abour the othersections of the tunnel. Hence, the experiences arerestricted to these tunnel sections that are comparablewith respect to ground condinons. ground cover. etc. (seeSecnon 4.4 "Empirical Approach").

• Field measurement should be provided throughout theentire length of the tunnel in order to check its assurnedbehaviour.

4.6. Special Design FeaturesSpecial considerations may he neeessarv if unusual ground

behaviour is expected or is caused by ground improvements.Some special design features and considerations arediscussedbelow.

4.6.1. Ground impl'Ovement techniqu8sGrouting and injections. Intensive groming or injenions

of the ground may improve the ground characteristicsconsidered in the design model. Although in most casesgrouting is applied only for dosing discominuities in rock orfor strengthening soft ground. in both cases the goal is lOachieve better homogeneüy.Drainage anti. compressed air. UsuaHy the ground is

stabiiized bv dewatering it and by avoiding inf/ows Ol water.Ground failure may be avoided if the pore water pressure isminimized. The assumed ground characteristics mav be validonly if successful drainage is possible or if water inflow isprevented. as in tunnelling under compressed air.

244 TI':'<NELLlNG AND UNDERGROUND SPACE TECHNOLOGY

Ground [reezing, lrnproving the ground bv Ireezingchang-es the ground properties. The time-dependent stress-'train behaviour of Irozen ground can be significant. Freezingdraws water toward the Iining, causing an increase in watervolume and heave at the surface. Conerering on frezen grounddelavs the strength development of the concrete.

4.6.2. Unusual ground bahaviour

Suielling ground. Stress release due to tunnelling and-erground water intlux may cause swelling and a correspondingincrease in pressure on the lining. In these cases. a circularcross-sectien or at least an invert arch is recommended. Theswelling resulting from a chemica! reaction, as in anhvdrid.generally is much more pronounoed than that due to rhephvsical absorption of water. as in day.Underground erosion. mtning subsidence, and sinhhoies.

Tunnelling in ground that is subject to settlements. as in thecase of gypsum erosion or mining subsidence, requires specialdesign considerations. A flexible lining that follows theground movements by utilizing its plastic deformationcapacity is more suitable in these cases than is a too-rigid orbriule, failure-prone lining, If the ground has sinkholepotentials, a tunnel structure that can be repaired easilv mavhe more economical than a structure designed to allow thebridging of the sinkholes,

5. In-Situ Monitoring5.1. Purpose ofIn-Situ MeasurementsIn-situ monitoring during the excavation and at longer

intervals after the tunnel is cornpleted should be regarded asan integral part of the design not only for checking thestructural safety and the applied design model but also forverifying the basic conception of the response of the ground totunneiling and the effectiveness of the structural support.The main objectives of in-situ monitoring are:(I) To control the deformations of the tunnel. including

securing the open tunnel profile. The tirne-historydevelopment of displacernents and convergences mav beconsidered one safety criterion. although field measurementsdo not vield the margins the structure can endure beforefailing.(2) To verify that the appropriate tunnelling method was

selected,(3) To control the seulements at the surface. e.g, in order to

obtain information on the deforrnation pattem in the groundand on that part of settlements caused by lowering the waterlevel.(4)To measure the development of stresses in the structural

rnembers, indicating sufficient strength or the possibility ofstrength Iailure.(5) To indicate progressive deformations, which require

immediate action for ground and support strengthening,To furnish evidence for insurance claims, e.g. by

provicling resulls of levelling the settlements at the surface intown areas.

5.2. Monitoring MethodsA programme for monitoring the deformations and stresses

during the excavation may comprise the following-measuremems (see Fig. 8):(I) Levelling the crown (at the least) inside the tunnel as

soon as possible. With regard to imerpretation of the data.2 reveals that often onlv a smal! fraction of the emire

crown movemem can be m~nitored because a larger partoccurs before the bolt ean be set. For difficult tunneHing, theclistance between two crown readings may be as close as w-IS m. LeveHing of thè invert is recommended for rock havingsweHing potemials.(2) Convergence readings (in triangular settings; K in Fig.

Volume 3. Number 3, 1988

L

Illt:Th.lll9,Om'______ L

uppersaencn

[Dis t . "'SO ... l00ml

outerlining

L

...--14.0m ..•..

leveHing l grCK.Ind preSlll.G elCtensometer econvergences K ringtorces R sliding micrometer SL

8. Example of in-sim monitoring of the tunnelexceoation, the preiiminarv lining, and the surfacesettlements.

8) should be rhe standard method for earlv information.are easilv applied and are accurate ro within 1 rnm.

In a Iew cross-secuons, the linings mav he equippedwith stress cells for reading the ground pressures and ringIorces in the lining (G and R in Fig. 8).(4) Stress cells also should he installed in a tew sections of

the final secoud lining if long-term readings are desired alterthe tunnel has been compieeed.(5) Surface levelling along the tunnel axis and

perpendicular to ir vield settlements and the correlation tomeasurements inside me tunnel (see Fig. 2).(6) Extensometers. inclinometers, gliding micrometers rnav

he installed from the surface wel! ahead of the tunnelling face.•vielding deformarion measurements within me ground (seeFig. 8). Monitoring of the ground deformations is especiallyappropriate for checking and interpreting the design model,Therefore, [he installation should he combined withconvergence readings and stress cells in the samecross-sectien.The frequencv of [he readings depends on how Iar frorn the

tunnelling face the measurements are taken, and on theresults. For exarnple, readings may he performed initiallv twotimes a dav: then be reduced to one reading per week Iourdiameters behind theface: and end with one reading permonth if the time-data curves justifv this reduction inmeasuremem readings.

5.3. Interpreting Resuttsof In-Situ MonitoringThe results of in-situ monitoring should be interprered

with regard to the excavation steps, the structural supportwork, and the structural design model in conjunction withsafety considerations.The actual readings normallv show a broad scatter of

values, Expectations of reliabrlity may not be met. especiallvfor pressure cells, beeause stresses and strains are verv localcharacteristics. Detormauon and con vergen ce readings aremore reliablv obtainable because displacements registerintegtals along a larger sectien of the ground,The in-situ measurements should be interprered in

consideration of the Iollowing:

lID The results should verify whether the tunnelling methodis appropriate.

lID Grapbed time-historv charts may reveal a decreasing rateof deformauon. or unoover danger of coltapse.

• Large discrepancies between rhe theoretically predieredand actuallv observed deformauons may force revision of

model. However, measurements are valid onlvIor the acmal state at ehe time and the place where thevare taken. Long-term influences such as rising waterlevel. traffk vibrations, and long-term neep are notre~rislere~during excavation.

• readings mav promote visual understanding of thestrucmral behavior of groun<.!and support imeraction.

•• The readings may cover ooly a fraction Ol the acmalphenomena if bolts and stress cells are installed too late(see Fig. 2).

lID The tunnel may he considered stabie when all the

Volume 3, :\lumber 3. 1988

readings cease to increase, However, a safety marginagainst failure-e-especiallv sudden collapse-e-cannot bededuced Irom measurernenr, except bv extrapolation.

6. Guidelines for me StructuralDetaiiing of the üningOn design aspects with regard to maintenance the reader is

relerred to other recommendations of the ITA (see T&UST2:3). For concrete linings, [he following structuralspecifications are suggested,I) The thickness of a secend Iining of cast-in-place

concrete rnay have a lower limit of 25-30 cm to avoid concreteplacing problerns such as undercompaction or honev-combing of concrete. The following lower limits mav herecornmended:-20 cm, if lining is unreinforced:-25 cm, if lining is reinforeed.-30 cm for watertight concrete.(2) Reinforcemem mav be desirabie for crack control. even

when it is not required Ior covering inner stresses. On theother hand, reinforcemem mav cause concrete-placingproblerns or long-term durabiliry problerns due to steelcorrosion. If reinforcement in the secend lining is providedfor crack control. a closely-spaced steel rnesh reinforcementrnav have the following cross-secrions in both directions:

• At the outer surface. at least 1.5 cm-/m of steel;• At the inner surface. at least 3.0 cm-/m of steel.(3) The recommended minimum cover of reinforcement is:

3.0 cm At the outer surface if a waterproofmembrane is provided,

5.0 cm-6.0 cm At the outer surface if it is directly incontact with the ground and groundwater.

4.0 cm-5.0 cm At the inner tunnel surface.5.0 cm For the tunnel invert andwhere water is

aggressive,

(4) For lining segments, specificauons (I), (2) and (3) aboveare nor valid, especially if the segrnented tunnel ring is theouter preliminary Iining. For detailing the tunnel segments,special attention should be given to avoiding damage duringtransport and erection,(5) Sealing against water (waterproofing sheets) may be

necessarv under the following conditions:

• When aggressive water action threatens to damageconcrete and steel.

lID When the water pressure level is more than 15m abovethe Clown.

lID When there is a possibilirv of freezing of ingressing wateralong the tunnel secnon close to the portals.

• Wnen the inner installations of the tunnel must beproteered.

(6) In achieving watertighmess of concrete. specialspecifications of the concrete mixture, avoidanee of shrinkagestresses and temperarare gradients during setting, and theIinal quality of the concrete are much more important thantheoretica I cornputarions of crack widths .

Temperature effects (tension stresses) may be somewhatcontrolled by working joints (as close as 5 mat the portals)and by additional surface reinforcemem in concrete exposedto low temperatures,(8) An initial lining of shotcrete may he considered to

participate in provi.ding stabi.Iitv of the tunnel only when thelong-term durability of the shotcrete is preservecÎ.Requiremems for achieving long-termdurability indude theabsence of aggressi.ve water. the limitalion of concreteadditives for acce!erating the seuing (Iiquid accelerators), andavoiding shotcrete shadows behind steel arches andreinforcements.

TlTNNELLING AND UNDERGROt'ND 'sPACE TECHNOLOGY 245

eonsolîclGtion cfthe tunnel

I I Ii '_1011,,, j~

r'~ ,itNt (l:UGht,

1 <:=mf'Oi

I 1 pol.t_I grCMQ

=:i1

zcne

i i1 I1 H f .AH a ••••

eonvetgenee ..... :- ~ •• #

, .•Wltion Of ~ •••••••• $ •• .....II anisoll'OPYI 'I IdeeompressedI I nismmollY1 !

j roek tl!fl'lll4ll"ature

!I stre_ witllinI Ihe wall

-, "

adaPlÎOll of roek II round aboul lhe lunneI .

Figure 9. Table of measured data end encountered conditions along a tunnel in Eranee.

legend:

-.-----HttigM Uo.,. c.. Horth. Hannow.r

: QUO." Oty

ust = IIPOO"

N't 2 1'f'Jk!cn.

ls' :: \OW-115

J 1: ,,"thout inv .•"

Technica! design eonditions (TVR)tor the Hopfenberg - Tunnel (Fed. RGermany)

longiiudinal sectien withdesign characteristics

Figure 10. Predieeed ground conditions, tunnelling classes and design characteristics along a tunnel of the rapid railuiav line inGeTmany.

246 TUN'NELUNG AND UNDERGROUND SPACE TECHNOLOGY Volume 3; Number 3, 1988

<00•• He ig ntNi: Sands!one' t mudstCh!

aJtern3tmnSi: mudmolW'F : fault .tone-

IfI'1>''200,?O

I, 100

IKi IQ rJ'll"ter

,(::-41kQ <HUt 42u .ag 44lua

500 600: 800 200 400 fiOO?(JO 200.00 600 800 209 11Kl' il(M) AOf) :.DJ

Rock ""me5.5. and m.s.(s.p.)

Geolog.e.l For mat ron i N';:Nishiyama Formation Si: ShitvaForma t Ion Ni

Seism.e Veloei IJ 2.0-2.2

-.=52-56't./{ 4.5-4.8

Unconltned Compr"""vo ~Ireoglh(Compe renee Facter ) i

W.ler Inflow • I i I t Ie

Rock CI •• s IL IS IN 11.

NOl.

SqueezingPr-oper-ty

a.s.: sands tone m.s.:muds tone LR:sandstone predominates

Figure 11. Predicted ground conditions along a tunnel line iexemple submiued by [apen).

Figure 12.Documentation of geology, ground classes. support, geotech nicaliield measurements gathered duringa tunnel projectin Austria.

Volume 3. Nurnber 3. 1988 TC:'\:--:ELUNG AND UNDERGROUND SPACE TECHNOLOGY 247

7. Exampies of Presentationof Tunnel Design DataFigures 9-12 are national exarnples of tabuiated

intermation on geotechnical condinons and designcharacteristics given in condensed farm along a longitudinaltunnel sectien. This inforrnation mav be part of the renderingdocuments and should be amended with ongoing tunnelling,By gatbering the data actuallv encountered along the tunnelline in a sirnilar table, a comnarison can he made betweenpredieredand actual tunnelling conditions, 0

ReferencasErdrnann. J. !983•.Comparison of rwo-dimensional and developmentof rhree-dirnensional desig-n rnethods Ior tunnels un Oermam.

Braunschweig, West Cermanv: Berichte Instuut fur Statik.Technica! Universitv of Braunschweig.

Gesta, P. 1986. Recommendations Ior use of the ronveraence-confinemem rnethod. Tunnels Ouurages SouterrainS 73: 18-39.

International Society of Rock Mechanica Commission onClassiticauon of Rocks and Rock Masses, 1981. Int. j. Rockstechanics Mining Sci. 18: 85-110.

International Society of Rock Mechanics, 1975. ISRM Recornrnenda-uons on site invesrigation techniques,

International Tunnelling Association Worki.ng Croup on StructuralDesign of Tunnels. 1982.Advances Tunnell. Technol. SubsurfaceUse 2(3): 153-228.

Nota

'See, tor exarnple, the Swiss SlA Dokument 260 or the correspondingC.S.-ASCE Code.

Appendix. International and Natienel Recommendations on Structural Design of Tunnels.,Although the Iollowing selected list of recommendations by national and intemational organizations is not complete, itneeertheless should provide the reader with sourees of additional inforrnation regarding the design of tunnels.OrganizationiCountry PubheationInternational Tunnelling Views on structural design models for tunnelling. Adoances in TunnellingAssociation (ITA), Technology and Subsurface Use 2:3 (1982).

International Society forRock Meeharnes (ISRM)

ISRM recommendations on site investigation techniques, July 1975.

ISRM Committee on Field Tests:Document No. l-Suggested Metbod for Determining Shear Strength

Document No. 2-Suggested Methods Ior Rock Bolt Tesung

ISRM Commiuee on Laboratory Tests; ISRM ComJnittee on Suielling Rocks:Document No, I-Suggesteà Methods for Determining the UniaxialCompressive Strength of Rock Matenals and Point Load Strength Index.

Document No. 2-Suggested Methods for Determining Water Content.Porosity, Density, Absorption and Related Properties. Swelling and SlakeDurability Index Properties.

Austral ia Australian Standard 1726 - S.A.A. Site Investigation Code,

Austria

Australian Standard 1289- Methods of Testing Soils for Engineering Purposes.

ÖNORM E 2203 Untertagebaunorm, Richtlinien und Vertragsbeseimmungen,Werkvertragsnorm.

Projektierungsrichtlinien für Oeotechnische Arbeiten, RVS 9.240 u. 9.241,Forschungsges, Srassenwesen, Nov. 1977.

Federal Republic of Germany(in Cermarn

Reeommendations Ior rhe design of underground openings in rock. Tunnelbau-Teschenbucn 1980,Gluckaui-Verlag, Essen (1980), pp. 157-239.

Recornmendations for the analvsis of Tunnels in soft ground (980), Bautechnik10 (980), Berlin, pp, 349-356.

Recommendations Ior the Concrete Lining of Tunnels in soft ground (986).Bautechnik 10 (986), Berlin, pp, 331-338.

France Tunnels et Ouoreges Souterrains, Special Issue July 1982, pp, 32-123;

Rêflexions sur les methodes usuelles de calcul du revêtement des souterrains(Usual calculation methods Ior the design of tunnel linings).

248 TVNNELUNG ANO !JNDERGROPND SP.\CE TECHNOLOGY Volume 3. Nurnber 3, 1988

JapanTunnel Engineering Committee,Japan Society of Civil Engineering,Japan Tunnelling Association

Switzerland

United Kingdom

Unieed States of AmencaAmerican Society ofCivil Engineers (ASCE)

Prèsentation de la mèthode de construction des tunnel avec sourènernentimmediat par bèton projetè et boulonnage (Presentation of the tunnelconstruction method with imrnediate support by shotcrete and bolting).

Recomrnandarions SUf les condinons d'emploi du boulonnage (Recommenda-tions for condinons of the use of bolting).

Tunnels et Ouurages Souterrains 73 (jan.zFeb. 1986), pp. 18-38:Recommendations Ior use of the convergence-confinemem method,

Tunnels et Ouurages Souterrains 67 (jan.z Feb. 1985), pp. 32-43:Recommandations relatives au choix d'un type de soutenement en galerie(Recornmendations Ior the selection of tunnel support).

Tunnels et Ouutages (1984), pp, 80-97: Recommandations relativa à l'ernploides citres dans ia construction des ouvrages souterrains (Recommendations onthe use of steel arches as temporatv support in tunnel structures).

Standard Specificatiens for Tunnels:

Mountain Tunnelling Metbod. Nov. 1986.

Shield Tunnelling Method. June 1986.

Cut-aud-eover Method. june 1986.

Recommandation SlA No. 199: Etude du massif rocheux pour les travauxsouterrains. 1975. (Also in German)

Norme SlA No. 198:Travaux souterrains (avancement à l'explosif), 1975. (Alsoin German)

Recommandation SlA No. 198/1: Construction de tunnels et de galeries enroeher-au moven de tunneliers, 1985. (Also in German)

British Standard 1377. Methods of test for soils Ior civil engineering purposes.British Standards Insritution, 1975.

British Standard 5930, Code of Praenee Ior site invesugations, British StandardsInstitution, 1981.

Craig, R. N. and Muir Wood, A. M. A review of tunnel lining praenee intheUnired Kingdorn. TRRL Supplementary Report 335, 1978.

Tunnelling Waterproofing. CIRIA Report 81,1979.

Dumbleton, M. J. and West, G. A guide to site investigation procedures fortunnels. TRRL Laberatory Report 740,1976.

Guidelines fOT Tunnel Lining Design. Ed. by T. O'Rourke. ASCE Technica]Committee on Tunnel Lining Design, Teehuical Council on Research.

Volume 3, Number 3, 1988 TUNNELLlNG AND UNDERGROUND SPACE TECHNOLOGY 249