101 Peck Hendron y Mohraz RETC (1)

7/27/2019 101 Peck Hendron y Mohraz RETC (1)

1/29

5Soft Ground Tunneling

Chairmen: Dewayne L. MisterekSam Taradash

Denver Federal Center, Denver, Col o.Commerci al Sheari ng & Stampi ng Co. , Youngstown, Ohi o


2/29

C hapter 79

STATE OF THE ART OF SOFT-GROUND TUNNELING

by R. B. Peck, A. J. Hendron, Jr., and B. Mohraz

Professor of Foundation EngineeringProfessor of Civil EngineeringAssistant Professor of Civil Engineering

University of Illinois at Urbana-Champaign

INTRODUCTION

The state of the art of soft-ground tunneling wasdiscussed in detail by the senior author at the 7thInternational Conference on Soil Mechanics and FoundationEngineering held in Mexico City in 1969. Little can begained by repeating the information assembled at thattime. Hence, in this report, only the briefest summarywill be given of the overall state of the art andattention will be concentrated on trends and developmentssince 1969.

A characteristic of recent developments is thecontinued trend toward use of shields and excavatingmachines. The implications of this trend with respectto design and construction, and particularly with respectto the treatment of unfavorable ground conditions, willbe examined.

Settlement associated with lost ground, a subject asold as tunneling itself, will be reviewed and methodswill be considered for its reduction.

Design of tunnel supports was discussed in detail in1969 with respect to essentially rigid or essentiallyflexible types of linings. In this report, a definitionof flexibility will be considered and design proceduressuggested for linings of flexibility intermediatebetween the two extremes.


3/29

260 RETC PROCEEDINGS, VOLUME 1

SHIELDS AND MACHINESToday, almost all circular and some horseshoe tunnels

are excavated within the protection of shields. Safetyis the prime consideration; the possibility of collapseof an unprotected crown or face is no longer consideredtolerable. The trend is understandable and justifiable,but it has its undesirable features. Todays workmanhas lost much of his skill in hand mining, and so hashis supervisor.

This aspect of the state of soft-ground tunneling isunfortunate because complex modern underground systemsinvolve many geometrical forms not adaptable to shieldtunneling. These include junctions of rapid transitlines, stations formed by excavating between shield-driven tubes, escalator and stairway passageways andtheir junctions with driven tunnels, and a host ofauxiliary structures for various purposes.

It is a truism in tunneling that the beginning of ajob is almost always fraught with lost ground, slowprogress, and even accidents, until the crew andsupervisory staff become acquainted with the necessarysteps of the work to be done. The period of learningmay easily be several weeks to a few months. If handexcavation is regarded as a minor adjunct to shieldtunneling, and if most of the planning is devoted toeconomical and rapid progress of the running tunnels,whatever hand work is necessary becomes a fruitful sourceof delays, accidents, and excessive loss of ground.

The lore of hand mining in unfavorable ground isalmost forgotten; men with a variety of experience indirecting hand mining are a vanishing breed; skilled,soft-ground hand miners are rare. Unsatisfactory andinept methods, discarded and replaced by better onesmany years ago, are being revived through ignorance ofthe lessons of the past. It is a matter of concern toall interested in soft-ground tunneling that ouremphasis on progress and mechanization is causing us tolose an important and useful heritage.

To remedy this situation, the tried and true tech-niques need to be restated and brought up to date forthe benefit of those who have a job to do, who wish todo it well, but who have neither the time nor theopportunity to study old and somewhat obscure descrip-tions of difficulties and how they were overcome.


4/29

STATE OF THE ART OF SOFT.GROUND TUNNELING 261

No significant improvement in shields as such can benoted in the past few years. They still tend to roll,are difficult to steer, and are difficult to keep ongrade. Shapes other than circular are occasionallyattempted but, except for roof shields, have found littleapplication. Most of the comments about shields areassociated with loss of ground and its prevention andwill be deferred until that topic is discussed.

Excavating machines are becoming increasingly popular.The excavating equipment itself remains undesirably sen-sitive to changes in the nature of the ground; thegreatest advances appear to have been in the systems forremoving and handling the muck and installing the lining.The difficulty of controlling the face under adverseground conditions has led to increased use of methodsto improve or homogenize the properties of the soil sothat progress will not be impeded and so that changes inthe type of excavating equipment will not be necessary.The same methods of improvement also aid in reducinglost ground and will be discussed further under thatheading.The investment in an excavating machine is so greatthat rapid progress is essential for recovery of profit.Choice of the best machine for given conditions dependsin part on the experience of the constructor and in parton the accuracy with which the significant characteris-tics of the soil deposits are portrayed to the bidders.The two conditions that appear to have given the greatestdifficulty in recent years are the presence of ground-water in pervious zones and the presence of larger sizesand greater quantities of boulders than anticipated.Both conditions have led to litigation, a sure indicator

of an unsatisfactory state of the art.With respect to boulders, the limitations of testborings should be appreciated. For example, in excavat-ing a tunnel of 10-ft diameter by machine, two 8-inchboulders per foot of tunnel would usually be considereda large number. Yet, statistically, it is likely that,if the boulders were uniformly distributed throughoutthe deposit, only one boulder would be encountered in aboring 100 ft long. The actual influence of the boulders

depends on several factors in addition to their frequency.If they are large compared to the size of openings orslits in the excavating machine, they may be troublesome.If, in addition, they are embedded in a hard cohesivematrix they may greatly impede the progress of even ahand-mined shield and may render completely impotent amechanical excavator of almost any type.


5/29


The detrimental effects of groundwater in imperviouszones depends to a great extent on the type of thegeological formation and the details of its structure.Whereas a waterbearing lens may drain almost harmlesslyinto the heading, a waterbearing seam connected to asource of supply may lead to instability and a run.Investigations of groundwater conditions should includean assessment of the geologic implications. There isneed for far better understanding and cooperation in thisrespect among the engineer who conducts the subsurfaceexploration, the engineering geologist who can developthe implications of the structure of the deposit, andthe prospective builder of the tunnel who should have thebackground to appreciate the implications.

Indeed, one of the outstanding shortcomings in thestate of the art of soft-ground tunneling at the presenttime is the manner in which subsurface information isobtained, presented, made available to bidders, andrelated to the contract documents. The engineer or owner,fearing claims, is strongly tempted to place no con-clusions regarding the behavior of the soil in thecontract documents, although he and his advisors areprobably the only ones having the time and facilitiesto make an adequate assessment of the subsurface condi-tions . The bidders, on the other hand, are tempted tobe optimistic to enhance their likelihood of being thelowest bidder, and to look for every apparent deviation,significant or otherwise, from the conditions they saythey have assumed on the basis of the contract documents.This mutually antagonistic relationship is unhappilygrowing worse and threatens to overshadow many of thetechnical improvements that potentially decrease thecost of tunneling.

LOSS OF GROUNDAn approximation of the settlement that must be anti-cipated above a single shield-driven tunnel, executedwith proper techniques and good workmanship, can be madeby the procedure advocated by Schmidt (1969). The shape

of the settlement curve is that of the probabilityfunction; the significant parameters are shown in Fig. 1.The maximum settlement can be estimated on the assumptionthat the volume of the settlement trough will be about1 per cent of the volume of the tunnel. Under exception-ally good condition$ and workmanship, the settlement maybe as little as half this amount; in contrast, volumesof settlement of up to 40 or 50 per cent of the volume


6/29

STATE OF THE ART OF SOFT-GROUND TUNNELING 263

of the tunnel are not unknown. Such settlements repre-sent, of course, the results of extremely poor practices.

Rotio & s unction f A Jnd ~JiVolume of trough ~ 2.5 i 8mox. conditions

Fig. 1. Settlement curve above shield-driventunnel as predicted by Schmidt (1969).

The settlements immediately associated withconstruction (exclusive of long-time consolidation) maybe conveniently separated into those associated withmovement toward the working face, invasion of thesurrounding soil into the annular space left by thetailpiece clearance and such similar features as polingplates, and inflow of material with groundwater enterinqthe tunnel at unprotected places. The movements may beaccentuated by yawing, diving or nosing of the shieldand by the necessity for negotiating curves. The varioussources of settlement are well illustrated in the paperby Hansmire and Cording.

The movement of the soil toward the working face andthe invasion of the annular spaces surrounding thetunnel lining are caused by the reduction or removal of


7/29

. . ... .


the original stresses within the soil mass. Fundamen-tally, two ways are available for preventing or reducingthe movement. Either the ground must be so stiff andstrong, or must be converted into a medium so stiff andstrong, that the reduction of stress causes negligibledeformations; or else the reduction of stress must beeliminated or restricted until the tunnel lining iscapable of sustaining the earth loads without significantdeformation.

The strength and stiffness of a granular soil belowwater table may, of course, be increased drastically bydrainage. The drainage is necessary to preventinstability; the increase in stiffness is a valuablebyproduct. Other than by drainage, improvement of theproperties of granular soils is being accomplished by theinjection of cement or chemical qrouts; the choicedepends on the nature of the formation. Althoughexpensive, grouting may be particularly attractive if itsuse in short intervals of particularly bad ground caneliminate the necessity for compressed air. Grouting mayalso reduce or eliminate the flow of groundwater.

Many misconceptions still exist concerning thebenefits of grouting and, particularly, the manner inwhich grout penetrates or permeates the soil and servesits useful purpose. It can be taken for granted that thevoids of a granular soil are rarely filled completely oruniformly by any kind of grout. Grout of a givenconsistency preferentially enters the voids of thecoarsest material from which, if it does not set tooquickly, it slowly permeates the less pervious materials.Compressible materials such as fine sand or coarse silt,or laminated silts, sands and clays, are often split bythe grout. The fluid grout takes the form of a lens orsheet from which it may penetrate remaining portions ofthe soil. Often the principal influence of the sillsand dikes of grout is the compression or consolidation ofthe intervening material while the grout pressure isstill being maintained. The peculiarities of groutpenetration are well known in some quarters, but oftenunappreciated in others. A series of investigations inthe 1940s and 1950s, of grout patterns found in rail-road roadbeds, is particularly enlightening and would beworth contemplation by those who wish to improve theirfeel for the manner in which grout carries out itsfunction. They are published in the Proceedings of theAmerican Railway Engineering Association, a jOUrna~ notwidely read by tunnel designers or constructors.


8/29


The art of chemical grouting has improved to thepoint that massive or varved silts have been successfullyimpregnated, as indicated in the paper by Anderson andMcCusker. Far more grouting is done in Europe than inthe United States. It can be anticipated that thepractice will grow in this country with increasing use ofmining machines, because grouting offers the greatestpotential for selective improvements of specific zoneslikely to be the seat of trouble in otherwise unsatis-factory ground. The most effective grouting, under thesecircumstances, is accomplished from the ground surfaceahead of tunnelinq, or perhaps from pilot drifts,because attempts to grout from inside the heading duringthe use of tunnel driving machines seriously impedesprogress of the machines.

Dewatering remains a fundamental procedure for generalimprovements of granular materials. Although thetechniques for dewatering are well established, closeenough spacing of deep or eductor wells and sufficienttime for adequate dewatering are still not alwaysprovided. Stabilization of granular material by de-watering can substantially reduce the loss of ground atthe working face and can, to some extent, increase thetime available for expansion of lining against the soilbehind the tailpiece or for filling the tailpiece void.It can rarely eliminate the latter source of lost groundunless sufficient apparent cohesion is developed toincrease the stand-up time appreciably.

In plastic cohesive soils, no satisfactory way isavailable for increasing the strength of the material,but air pressure may be used to decrease the reductionin stress due to excavation until the permanent lining isplaced. The high cost and physiological effects of airpressure place serious limitations on its utility. Itprovides a positive means in plastic soils, however, forpreventing the inward movement of the soil behind thetailpiece of the shield until appropriate measures canbe taken. Hence, where appreciable loss of ground isintolerable, compressed air still represents the mosteffective method of control.

Todays state of the art includes at least onedemonstrably successful tunneling machine,in Japan, inwhich fluid pressure is held against the working facewhile the workmen can erect the lining in free air. Theprinciple is sound, and progress is being made.Similarl~, the use of slip forms and an exotic quick-settinq strong material holds promise for being able notonly to fill the annular space behind the tailpiece


9/29


promptly, but to provide the permanent lining as well.Such a blue-sky device, described in the paper by Parkerand Semple, seems to be practicable in principle, and thevarious components of the equipment have been tested.Successful application may not be too far away. Never-theless, although the development of such semi-automatedtunneling machines is a desirable step forward, undersome circumstances less automated procedures, and evenhand mining, may be economically preferable. The bestultimate development of the state of the art is likelyto include improved methods for hand mining as well asfor machine mining, and the ability of engineers andconstructors to chose the system most suitable for thecircumstances.

DESIGN OF TUNNEL LININGSThe design procedures summarized in 1969 were dividedinto two categories: those to be used in proportioningflexible and rigid tunnel liners. A liner is said to beflexible if it interacts with the surrounding soil insuch a way that the pressure distribution on the linerand the corresponding deflected shape result in negli-gible bending moments at all points in the lining. Arigid liner is one which deflects insignificantlyunder the loads imposed by the soil; thus there is verylittle soil-structure interaction. Real linin~s, however,are neither perfectly flexible nor perfectly rigid.In present practice there is no quantitative methodto classify the stiffness of a tunnel liner in terms ofboth the structural properties of the liner and thestress-strain characteristics of the surrounding soil.A tunnel liner which may be stiff with respect to a soft

clay may behave as a flexible liner in a very stiff clay.Thus, there is a need to account for both the stress-strain properties of the soil and the flexibility of thetunnel liner. In this section a method will be presentedfor quantitatively determining the relative flexibilityof tunnel liners of stiffness intermediate betweenessentially flexible and essentially rigid.The structural engineer designs a tunnel lining for

certain combinations of thrust and moment. The magnitudeof the thrust and moment is dependent upon the stiffnessof the lining relative to that of the medium and to thedepth of the tunnel. In order to appreciate the factorsaffecting the structural design of liners of intermediateflexibility, the design procedures presently used by


10/29


structural engineers for both flexible and rigidtunnel liners are reviewed briefly.Flexible liners, which interact fully with the soil insuch a way that a nearly uniform pressure distributionultimately acts on them, do not have to be designed for

moments consistent with the initial stress distributionin the soil. But the liner must be designed to accomm-odatethe diameter changes necessary to develop a uniformpressure distribution on the liner. These diameterchanges can be estimated from experience and are usuallyin the range of about half a per cent. The structuralsection must be designed to withstand the bending momentsinduced by the estimated diameter changes. In addition,it must be designed to prevent buckling. In soft claysthis is usually accomplished by insuring that the over-burden stress, YH, is less than 3 EI/R3, where EI and Rare the flexural stiffness and the mean radius of theliner, respectively.

For rigid liners the coefficient of earth pressure atrest is usually estimated, and the moments and thrustsare calculated on the assumption of no interactionbetween the soil and the liner. Thus the soil is assumedto produce a load on the lining as shown in Fig. 2,where the maximum moment is given by

M= ~ 1/4 yH (K. - 1) R2

The thrust at the springline is given by

Ts = yHR

and the thrusts at the invert and crown are given by

CI = yH K. R

(1)

(2)

(3)

It should_be noted that ~he moment (eq. 1) is given bya constant, K, times YHR2; K is commonly referred to asthe moment coefficient. For values of coefficient ofearth pressure at rest equal to 1/2 and 2, the momentcoefficients are 12.5 per cent and 25 per cent respective-ly. These moment coefficients are too high because ofthe assumption of no interaction between the liner and


11/29


yH

Koy H

Fig. 2. Pressure distribution on a rigid liner assumingno interaction between the soil and the liner

the soil. Therefore, a more general procedure forproportioning tunnel liners of intermediate stiffnessmust take into account the soil-structure interactionand must yield design moments and design thrusts asfunctions of liner stiffness. The structural engineerthen needs only to compare the expected moments andthrusts at any point in the section with the limitingvalues of thrust and moment which can be determined fora given structural section from an interaction diagram.Definition of Stiffness Ratio for Tunnel Liners

TISestiffness of a tunnel liner-soil system isconveniently considered as being divided into twoseparate and distinct types. The first is extensionalstiffness, which is a measure of the equal all-arounduniform pressure necessary to cause a unit diametralstrain of the liner with no change in shape. The secondis flexural stiffness, which is a measure of the magni-tude of the non-uniform pressures necessary to cause aunit diametral strain which results in a change in shapeor an ovaling of the liner.

Recent analytical work by Burns and Richards (1964)and H6eg (1968), in soil-structure interaction, can beused to assess quantitatively the stiffness of a linerrelative to a soil medium. In these studies the relativestiffness of the liner and surrounding medium is char-acterized by two ratios designated as the compressibilityratio and the flexibility ratio. A definition of and aphysical interpretation ~f these ratios are given below.


12/29


The compressibility ratio is a measure of the exten-sional stiffness of the medium relative to that of theliner. The extensional stiffness of the medium can beobtained by considering a portion of the medium subjectedto a uniform external pressure, p, as shown in Fig. 3a.

P! HHHHH_

.-/0 1 \

( P;\\ //-.

-

ttHt!HH.(a)

P

P

(b)Fig. 3. Medium and liner under astate of uniform compression.

The diametral strain across an imaginary circular tunnel(shown by the dotted line) is given by

AD = ~ = g (l+V)(l-2V) (4)F mand the extensional stiffness is given by

-%-D D = (l+V)E(l-2V) (5)

where E and v are the Youngs modulus and the Poissonsratio of the medium. The extensional stiffness of theliner, which replaces the cylinder of material withinthe imaginary circle shown in Fig. 3a, can be obtainedby considering a ring subjected to a uniform pressure, p,as shown in Fig. 3b. The diametral strain is given by


13/29


AD . @&F ELt (6)

where E , R, and t are respectively the modulus ofelasticity, the radius, and the thickness of the ring.The extensional stiffness of the liner in plane strainis obtained from eq. 6 by replacing ER by Et/(1-v12) where VE is the Poissonts ratio of the linermaterial. Thus , the extensional stiffness of the lineris given byEit

+TADD= (1 - VIZ)

The compressibility ratio, C, is obtained by dividin~eq. 6 by eq. 7. -

c =

The above expressionsectional thickness,composed of built-up

(7)

E(1+V) (1-2V)Egt (8)1

(Gk2) R

is for a liner of uniform cross-t. Since most tunnel liners aresections with non-uniform thickness,similar to that sho~n in Fig. 4, eq. 8 is modified bytaking the thickness, t, as the cross-sectional area, A,of a typical element divided by the length, L, of theelement; that is, t = A/L.

The flexibility ratio is a measure of the flexuralstiffness of the medium relative to that of the liner.The flexural stiffnesses of both the medium and the liner,as defined here, are essentially measures of theresistance of each to a change in shape under a state ofpure shear. The flexural stiffness of the medium can beobtained by considering the diametral strain of theimaginary circle shown in Fig. 5a. The diametral strainis given by

(9)


14/29

STATE OF THE ART OF SOFT-GROUND - - . . .lUNNtLINti z/1

r

Cross-Sectional Area = A

Fig. 4. A typical built-up liner section

P

P+H-1mP

(a) (b)Fig. 5. Medium and liner under a state of pure shear

The flexural stiffness of the medium is taken as a ratioof the pressure, p, to the corresponding unit diametralstrain across the cylinder. The resulting expressionfor the flexural stiffness of the medium is

& = E 1+) (lo)


15/29


The diametral strain of a ring subjected to thepressure distribution shown in Fig. 5b is

(11)

where 1P is the moment of inertia of the cross-secti~nalarea ofthe ring. If E~Ip, is replaced by EIIi/(1-vL2) toaccount for the plane strain effect in the liner, theliner stiffness is given by

-?--RADD= (1 - VB*) (12)

The flexibility ratio is obtained byeq. 12. Thus , dividing eq. 11 by

E(13)

In all the expressions above, Ikis the moment ofinertia of the cross-section per u It length alon? theaxis of the tunnel. Thus , for the section shown InFig. 4, IL is the moment of inertia of the entire cross-section divided by the length L.Burns and Richard (1964) have shown that, on accountof the interaction between the soil and the structure,the resulting thrusts and moments are affected by(1) the compressibility ratio(2) the flexibility ratio(3) the slippage which takes place at the interfacebetween the structural liner and the medium.


16/29


Various solutions are given for both full and noslippage between the medium and the liner. Because ofthe existence of high shear stresses at the interfacebetween the liner and the medium for most cases, thecondition of full slippage is believed to approximatemore nearly the behavior of soft-ground tunnel liners.Although the expressions developed by Burns and Richardare for the case of a one-dimensional airblast loadingfor protective structures, the expressions may easily bemodified to give the thrusts, moments, and displacementsfor various initial values of the coefficient of earthpressure at rest as shown in Figs. 6 to 9 inclusive.The equations are given below; they are valid only fora deeply buried tunnel.For crown and invert:

T= + [(l+KO) bl -*(l- Ko) b21 YH R

M= g (1 - Ko) b2 yH R2

w= ; ~ [(1-v)(l + Ko) blCc+H3T(l-Ko)b2FJ

and for springline

T = +[(l+Ko)bl+~(l-Ko) b2]yHR

M= -:(1- Ko) b2 yH R2

1 YHR [(1-v)(l + Ko) blC= ~~ c

-:* (-Ko)b2F]

(14)

(15)

(16)

(17)

(18)

(19)


17/29

.. . . . ... . ..


wherebl= l-alb2 =l+3a2-4a3

and(1-2V) (c-1)al = (1-2V)C + 12F+1-2va2 = 2F+5-6v

2F-1a3 = 2F+5-6vY ,= unit weight of soil,H = height to center of the tunnel,R = mean radius of the liner,c = compressibility ratio,F = flexibility ratio,

and Mc is the constrainedgiven by

M=cThe displacements givenby

modulus of the soil which is

E (1-v)(l+V)(1-2V)eqs. 16 and 19 refer to aportion of the medium containing the liner and loadedby external pressures. The expressions include the dis-placements which have already taken place due to the

free-field stresses.The expressions for moments, eqs. 15 and 18, indicatethat the moment is proportional to (1 - Ko) , which is ameasure of the difference between the major and minorprincipal stresses in the free field. The momentexpressions also indicate that the moment is a functionof the flexibility ratio and not of the compressibility


18/29


o.l&

0.16

0.14

0.12

0.10

0.08

0.06

0.04

0.02

9

Ir

II~III-1III

-~I

K=0,5-- K =2.0

v = 0.4

\\\\\

---- ----1 I I {20 40 60 80 100 )Flexibility Ratio

Fig. 6. Variation of moment coefficient


19/29

_


Tn Fig. 6, the dimensionless moment, or the momentcoefficient, (M/YHR2), is given as a function of theflexibility ratio for two values of Ko. As the flexi-bility ratio increases, the moment coefficient decreases.The decrease is very substantial for a flexibility ratioless than 10 and, thereafter, the moment coefficient isless than 5 per cent. Thus, for flexibility ratiosgreater than about 10, the curves indicate that the linerbehaves relatively flexibly with respect to the medium.The moment coefficients for design could be obtainedfrom Fig. 6 or the moment expressions.Figures 7 and 8 give the thrust coefficients (T/YHR)as functions of flexibility and compressibility,respectively. The thrust coefficient is a function ofboth the flexibility and the compressibility ratios,and also of the coefficient of earth pressure at rest.Fig. 7 shows that the thrust coefficient is relativelyinsensitive to the flexibility ratio but is sensitiveto the initial value of Ko. This plot shows, however,that the thrust coefficient is practically the same for

all flexibility ratios greater than 10; this indicatesthat tunnel liners with flexibility ratios greater than10 behave as flexible liners. For flexibility ratiosgreater than 10, the thrust coefficient is nearly1/2 (1 + Ko) as given by Peck (1969) for a compressi-bility ratio of 1.0. This simplified expression isconservative for compressibility ratios greater than 1.0;for compressibility ratios less than 1.0, the expressionmay be modified as follows to give a simple relationwhich approximates the more detailed calculationsgiven by eqs. 14 and 17 for flexibility ratios greaterthan 10:

T. + (1 + Ko) [1.2 - .2C]y HR (20)

In Fig. 8 the thrust coefficient is shown as a functionof compressibility ratio for two values of K. and twoflexibility ratios. This plot indicates that the thrustcoefficient decreases as the compressibility ratioincreases. Moreoverr for a given compressibility ratio,the thrust at the crown and springline are somewhatdifferent for low values of the flexibility ratio, butapproach each other as the flexibility ratio increases.The same effect is shown in Fig. 7. It is suggestedthat eqs. 14 and 17 may be used for preliminary design.


20/29


1

I

,

Documents

101 Peck Hendron y Mohraz RETC (1)