TBM$og$bergslagsproblema/kk$

Implikasjoner$for$frem/dige$

høyhas/ghetsbaner$$

Øyvind'Dammyr,'NTNU''$

Bruddtyper$i$laboratoriet$

Lavt$omslutningstrykk$$

(enaksiell$test)$

Høyt$omslutningsA

trykk(triaxial$test)$

Ektensjons/tensilt$brudd$(hardt/s/vt$berg)$Skjærbrudd$

I" II" II"

Axial"strain,"ε"

Klasse$1:$Sprekker$utvikler$seg$stabilt.$Arbeid$må$gjøres$

(energi$/lføres)$for$å$svekke$styrken$yMerligere$

Klasse$2:$Sprekker$utvikler$seg$ustabilt,$selvpropagerende$

(energi$frigjøres)$

Figur$basert$på$resultater$fra$Noferes/$and$Rao$(2010)$

Mikrosprekker$

ini/eres$

Mikrosprekker$

propagerer$

systema/sk$

Mikrosøyler$

former$seg$

Mikrosøyler$

kollapser$

Makroskopisk$

skjærbrudd$

I"laboratoriet:$σci(crack$ini/a/on)$=$spenningsnivå$for$ini/ering$av$

mikrosprekker$ca.$0.4"av"UCS"σcd(crack$damage)$=$spenningsnivå$for$interaksjon$

mellom$mikrosprekker$ca."0.8"av"UCS.$SeM$på$som$langA/ds$laboratoriestyrke$av$prøven$

In"situ:"Bergslag$og$spenningsytringer$skjer$ved$0.4±0.1"av"UCS,$nær$eller$reM$over$spenningen$som$

trengs$for$å$ini/ere$mikrosprekker$Diederichs$et$al.$(2004)$$

$

τ$$=$$c$$+$$σntanφ$

Kohesjon$ Friksjon$

Mohr$Coulomb$

Kohesjon$svekkes$før$friksjon$

ak/veres$ved$ekstensjonsbrudd$

aksiell$spenning$

omslutningstrykk$

Trondheim

Bergen

Stavanger

Kristiansand

OsloOslo Airport

Towards Gothenburg (Sweden)

Potential futureHSR lines

100 km

EMer$Dammyr$et$al.$(2013)$

Norsk$vestlandstopografi$

Ca.$1000m$

27MPa$(σ1)$

1.$

Gneiss:$UCS$150MPa$

Enkel$spenningsbetraktning,$k=0.33,$σ3=9MPa$

σ1"

σ3"

σ1"

σ3" σθmax$3σ1Aσ3$

σθmin$3σ3Aσ1$

σθmin$

σθmax$

UCS$=$150MPa$

σci"=$150*0.4$=$60MPa"σθmax"=$3σ1Aσ3$=$3*27A9$=$72MPa"σθmin$$=$3σ3Aσ1$=$3*9A27$=$$$0MPa$

$

1.$

2.$

2.$2.$

Mine$by$experiment$Canada$

!!! = 0.49! ±0.1 !+ !1.25!!"#$

!! !!!!!!!!!!!!!!!!

Martin et al. 143

Fig. 9. Relationship between depth of failure and the maximumtangential stress at the boundary of the opening.

Fig. 10. Comparison between the predicted depth of damageinitiation using the Hoek–Brown brittle parameters given by eq.[10] and measured depths of failure given in Table 1.

not persistent relative to the size of the opening such that thefailure process is essentially one of cohesion loss. In the nextsection the Hoek–Brown brittle parameters are applied to sev-eral well-documented case histories and are also used to assessthe effect of tunnel geometry on the depth of brittle failure.

Application of Hoek–Brown brittleparameters

In the previous section most of the analyses, using Hoek–Brown brittle parameters, were applied to near circular open-ings in fairly massive rocks. In this section the same conceptsare applied to other opening shapes and to rock masses thatare described as anisotropic. All analyses in this section werecarried out using the elastic boundary element program Exam-ine2D (Curran and Corkum 1995) or the plastic-finite elementprogramPhase2 (Curran andCorkum1997). In these programsthe stability is expressed in terms of a strength factor that isanalogous to the traditional factor of safety such that a Strength

Factor < 1 implies failure or the region that is over-stressed.Martin (1997) showed the brittle failure process initiates

near the tunnels face and hence is three-dimensional. Thus, itis not surprising, as indicated by Fig. 7, that two-dimensionalanalyses using the Hoek–Brown brittle parameters cannot beused to predict the actual shape of the notch. Nonetheless, forsupport design purposes, it is necessary to determine how deepfailure will occur and the lateral extent of failure. This can beachieved by the application of the Hoek–Brown brittle param-eters. In the following example applications, taken from doc-umented case histories, a comparison of the results with bothHoek–Brown frictional and brittle parameters are presentedto demonstrate that this approach can be used to estimate thedepth of failure.

Elastic versus plastic analysesThe theory of elasticity would suggest that the optimum

shape of a tunnel is an ellipse with the major axis parallel to thedirection of maximum in-plane stress, with the ratio of major(2a) to theminor (2b) axis of the ellipse being equal to the ratioof the maximum (�1) to minimum (�3) stresses in the planeof the excavation (Fig. 11a). This optimum shape producesuniform tangential stresses on the boundary of the excavationwith the tangential stress equal to �1 + �3. Fairhurst (1993)pointed out however, that while the tangential stress is constanton the boundary it is not constant for the regions behind theboundary of the tunnel and should failure occur the inelasticregion that develops for an elliptical shaped tunnel, is muchlarger than if the tunnel geometry were circular or an ellipseoriented parallel to the minimum stress axis (Fig. 11b).

Read and Chandler (1997) carried out an extensive studyto evaluate the effect of tunnel shape on stability by excavatinga series of ovaloid and circular openings at the UndergroundResearch Laboratory,Manitoba. Because of the extreme in situstress ratio (Ko ⇡ 6) it was not practical to excavate an ellipseof the optimum shape (e.g.,18 m by 3 m in dimension). As acompromise, they excavated an ovaloid 6.6 m wide and 3 mhigh in a rock mass with the following average properties:

Rock type GraniteIn situ stress �1, �3 59.6, 11.1 MPaIntact rock strength �c 224 MPaRock-mass rating RMR ⇡100Hoek–Brown constants m 28

s 0.16Residual parameters mr 1

sr 0.01

Figure 12 shows the results from two analyses using Exam-ine2D and the shape of the notched region that formed shortlyafter excavation (Read, personal communication). In the firstanalyses, the Hoek–Brown parameters are based on laboratorystrength tests, which gave �c = 224 MPa and m = 28, butwith the parameter s = 0.16 to reflect that failure initiates atabout 0.4�c, consistent with the findings in the section entitledBrittle rock-mass strength around tunnels . Those results areshown in Fig. 12a and indicate that the excavation is stable, i.e.,

©1999 NRC Canada

Rf/a$=$0.49$+$1.25*(72/150)$

Rf/a$=$1.09$

Rf$=$1.09$*$5$

a$=$5m$

Rf$=$5.45$

Depth$of$spalling$=$5.45$–$5$=$45cm"

Mar/n$et$al.$(1999)$

74

risons were possible in adjacent bored and blasted sec-tions. The blasted sections needed more rock support.

4.3 The Tverrelvdal tunnelThe diameter of this tunnel was 6.25 m and a totallength of 6,500 m was bored with TBM 204-215.

As a precaution, the TBM was fitted with rock bol-ting equipment on each side of the main body.

The first 1,700 m was bored virtually without stabili-ty problems. After the break-in period the net penetra-tion was approx. 1.3 m/h and the weekly progress 70-80 m. The rock material had good boreability (DRI at65-75), but the rock mass had little jointing. As boringwith high cutterloads in the massive rock resulted inheavy vibrations of the relatively light body of thisTBM, the rock bolting equipment required extensivemaintenance and was eventually removed.

After about 1,700 m of advance, increasingly intensespalling occurred. Some spalling had been anticipated,but the intensity was unexpected. On some sectionsthe rock at the contour was more or less crushed, andthe progress of tunnelling was slowed down due to:• the installation of rock support• gripping problems due to overbreak in the walls• hand clearing of debris and rock fragments

Fig. 2 Overbreak due to crushing and spalling

During the first critical phase, rock bolting was per-formed by jack-leg drilling. Later, working platformswere added making it possible to install rock bolts du-ring operation of the TBM, as originally intended.

Mechanical expansion shell anchored bolts withlength 1.5-3.0 m were used more or less systematical-ly. The bolts were installed immediately behind thecutterhead roof shield. Experience proved that it wasnecessary to install the bolts as early as possible. Inareas where the spalling removed the tunnel wall needed for gripping, time consuming use of steel beams and wooden supports became necessary to allow the TBM to move forward.

Fig. 3 Rock bolting from work platforms on the 6.25 mdiameter TBM

The spalling also created a lot of debris and rockfragments that had to be removed. In particular this hap-pened when the TBM advance was halted due to rockbolting and problems with gripping, and the spallinghad time to develop. The use of a Bobcat 400 minifront-end loader, which could be operated under theTBM body, eased the mucking. The spalled and crushedmaterial was dumped in small cars, which were pulledinto the track laying area under the bridge conveyer.The normal TBM crew of four men was reinforcedwith two for rock support etc. Seven weeks were needed to pass through the worst section of 200 m.

Further ahead, the intensity of the spalling de-creased, probably because the rock cover increased,thus counteracting the effects of the horizontal stres-ses. Still, on some sections the rock spalling could develop fast enough to occur 20-30 m behind the face,above the TBM backup where it was difficult to installrock support and clear debris. Systematic bolting be-hind the cutter head was then performed until condi-tions became more stable.

Out of the total section of 6,500 m, 650 m had exten-sive spalling requiring systematic bolting. A total of2500 bolts were installed. This increased the construc-tion time by 7-8 weeks, which could have been re-duced to half the time with better preparation.

4.4 The Reinoksvatn tunnelIn this transfer tunnel a total of 9,500 m was boredwith the 3.5 m diameter TBM, first 6.5 km south fromReinoksvatn access towards Linnajavri and then 3 kmnorth towards Lievsejavri.

σθmax$

Kobbelv$krahverk$

Rojat$et$al.$(2009)$Lötschberg$basistunnel$

Aicha-Mules Pilot Tunnel- Double Shield TBM WIRTH SH630

106,7

355,6

147,5

223,1

401,8373,0

309,8

475,2

603,2

665,2

568,8

493,6

564,4536,1

177,5

0,0 0,0 0,0

36,1

509,8

293,6

418,6

454,9438,2

462,1

407,8

221,3

520,7495,3

0

100

200

300

400

500

600

700

800

giu-0

8

lug-0

8

ago-

08

set-0

8ott

-08

nov-0

8dic

-08

gen-

09

feb-0

9

mar-09

apr-0

9

mag-0

9giu

-09

lug-0

9

ago-

09

set-0

9ott

-09

nov-0

9dic

-09

gen-

10

feb-1

0

mar-10

apr-1

0

mag-1

0giu

-10

lug-1

0

ago-

10

set-1

0ott

-10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

MONTH

bore

d tu

nnel

(m)

.

0

500

1.000

1.500

2.000

2.500

3.000

3.500

4.000

4.500

5.000

5.500

6.000

6.500

7.000

7.500

8.000

8.500

9.000

9.500

10.000

10.500

Cha

inag

e P.

k.

.

ml di Galleria nel Mese

Produzione Cumulativa

Figure 10. Monthly production at the Brenner Exploratory Tunnel

Unfortunately, on 9 August 2009 at km 6+151, the day after the TBM had passed, the left-hand sidewall lining of the tunnel partially collapsed for a length of approximately 30 meters. This incident led to a stoppage of work lasting about 5 months. Thanks to the reconstruction of the geological model, the presence of a sub-vertical fault was identified parallel to the direction of excavation, not forecast in the design phase. The interaction between deforming stresses due to the excavation and stresses due to the particular features of the fault lead to a highly asymmetrical load on the tunnel section, thus damaging the lining and causing its collapse (Figure 11).

Figure 11. Detail of lining collapse

After the TBM stopped, a number of activities were undertaken with the following objectives: 1. Ensuring the safety of the tunnel 2. Restoring the internal section so that the TBM could resume operation 3. Start up of the TBM again in the stretch involving the fault sub-parallel to the excavation 4. Defining the procedures and the instruments for ensuring greater safety for the continuation of the

advance until the Mules access tunnel was reached. The following activities were undertaken to ensure the safety of the tunnel: 1. Installation of 2.2m long steel bolts secured with resin between points km 6+103 and 6+127, to

sustain the crown of the lining

Lining$kollaps$(Grandori$et$al.$2011)$

Bergslag$og$blokkering$av$

kuMerhodet$

Brenner$piloMunnel$(AichaAMauls),$alvorlige$spenningsytringer$

UCS$200MPa:$

UCS$150MPa:$

UCS$120MPa:$

Liten$risiko$for$bergslag$ Stor$sannsynlighet$for$bergslag$ Store$ujordringer$mtp$

bergslag$forventet$

Referanser$

•  Dammyr,$Ø.,$Nilsen,$B.,$Thuro,$K.$and$Grøndal,$J.$2013.$Possible$concepts$for$waterproofing$of$Norwegian$

TBM$railway$tunnels.$Rock$Mech$Rock$Eng.$In$press.$DOI:$10.1007/s00603A013A0388A5$

•  DIEDERICHS,$M.,$KAISER,$P.,$&$EBERHARDT,$E.$(2004):$Damage$ini/a/on$and$propaga/on$in$hard$rock$

during$tunnelling$and$the$influence$of$nearAface$stress$rota/on.$–$Int.$J.$Rock.$Mech.$Min.$Sci.,$41:$785A812.$•  Grandori,$R.,$Bieniawski,$Z.$T.,$Vizzino,$D.,$Lizzadro,$L.,$Romualdi,$P.$&$Busiollo,$A.$2011.$Hard$rock$extreme$

condi/ons$in$the$first$10$km$of$the$TBM$driven$Brenner$Exploratory$Tunnel.$In$Proceedings$of$the$Rapid$

Excava/on$and$Tunneling$Conference$(RETC),$San$Francisco,$USA,$June$19A22,$2011,$pp.$667A$685.$San$

Francisco:$SOC$Mining$Metallurgy$&$Explora/on$Inc.$

•  MARTIN,$C.D.,$KAISER,$P.$K.,$&$MCCREATH,$D.$R.$(1999):$Hoek–Brown$parameters$for$predic/ng$the$depth$

of$briMle$failure$around$tunnels.$–$Can.$Geotech.$J.,$36:$136–151.$•  NOFERESTI,$H.$&$RAO,$K.$S.$(2010):$New$Observa/ons$on$the$BriMle$Failure$Process$of$Simulated$

Crystalline$Rocks.$–$Rock$Mech$Rock$Eng,$43:$135A150.$