Missouri University of Science and Technology Missouri University of Science and Technology

Scholars' Mine Scholars' Mine

International Conference on Case Histories in Geotechnical Engineering

(2013) - Seventh International Conference on Case Histories in Geotechnical Engineering

02 May 2013, 4:00 pm - 6:00 pm

3D FEM Modelling of a Deep Excavation Case Considering Small-3D FEM Modelling of a Deep Excavation Case Considering Small-

Strain Stiffness of Soil and Thermal Shrinkage of Concrete Strain Stiffness of Soil and Thermal Shrinkage of Concrete

Yuepeng Dong University Of Oxford, United Kingdom

Harvey Burd University Of Oxford, United Kingdom

Guy Houlsby University Of Oxford, United Kingdom

Zhonghua Xu East China Architectural Design & Research Institute, China

Follow this and additional works at: https://scholarsmine.mst.edu/icchge

Part of the Geotechnical Engineering Commons

Recommended Citation Recommended Citation Dong, Yuepeng; Burd, Harvey; Houlsby, Guy; and Xu, Zhonghua, "3D FEM Modelling of a Deep Excavation Case Considering Small-Strain Stiffness of Soil and Thermal Shrinkage of Concrete" (2013). International Conference on Case Histories in Geotechnical Engineering. 46. https://scholarsmine.mst.edu/icchge/7icchge/session03/46

This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been accepted for inclusion in International Conference on Case Histories in Geotechnical Engineering by an authorized administrator of Scholars' Mine. This work is protected by U. S. Copyright Law. Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more information, please contact scholarsmine@mst.edu.

Pape

3D

De

ABS A tywas procsmalmultanisomatethe fthe cand t

INTR Deepprovexcastrucbuildarouare ccond In aexca(Finnet alHsievaryfor b Numdeepbecaand existmodHash

er No. 3.28b

D FEM MSMAL

Yuepeng epartment of E

STRACT

ypical top-dowback-analyse

edures. The mll-strain stiffnti-surface soil otropic linear erial and inclufield measuremcomputed excthe thermal sh

RODUCTION

p excavationsvide undergrouavations will cture deformadings and servnd deep excacomplicated aditions and the

an effort to avations, a nuno, Atmatzidi. 2000; Finno

eh et al. 2010; a lot from on

back analysis.

merical modelp excavations.ause FEM anrequires advats some relatelling (Simpshash et al. 199

MODELLILL-STRA

Dong, HarveEngineering S

Oxford, U.K

wn deep excaved with a commodel considess of the soilmodel develoelastic materi

uding the shrinments. Two secavation behavhrinkage of co

N

s are frequentund space. Hoinevitably in

ations which mvices. The pravations is noand critically e method of co

better underumber of caseis et al. 1989; o and Bryson

Liu, Jiang et ne to another, t

lling is a pow However, it

nalysis is comanced skills oted publicatioon 1992; St. J93; Hashash a

ING OF AAIN STIFF

ey Burd, GuyScience, UniveK. OX1 3PJ

vation projectmplex 3D FEdered the detal and thermal oped at Oxforial to represennkage of concets of calculatviour. The res

oncrete should

tly used in cowever, the conduce groundmay cause darediction of grot straightforw

dependent ononstruction.

rstand the bee histories hav

Ou, Liao et a2002; Liu, Nal. 2011). Altthey provide v

werful tool fois not routine

mplicated andf experienced

ons on both 2John, Potts et and Whittle 19

A DEEP EFNESS O

C

y Houlsby ersity of Oxfor

t in Shanghai EM numericalailed construc

shrinkage of rd which has bnt the effect ofcrete. The numtions were consults showed

d not be neglec

ongested citieonstruction of d movementsamage to adjaround movem

ward because n both the gro

ehaviour of ve been publial. 1998; Ou,

Ng et al. 2005;though those cvaluable resou

or the analysily used for de

d time-consumd engineers. T2D and 3D Fal. 1993; Wh

996; Ou, Chio

EXCAVAOF SOIL ACONCRE

rd,

was carefullyl model in A

ction procedurf the concrete been implemef the joints in merical resultnducted to invthat small-str

cted in deep ex

es to deep and

acent ments

they ound

deep ished Liao

; Ou, cases urces

is of esign ming, There FEM

hittle, ou et

ATION CAAND THE

ETE

East Chin

y monitored dABAQUS to res and strucbeams and sl

ented into ABthe wall. The

ts captured thevestigate the inrain soil stiffnxcavation pro

al. 1996; Zdr2011), but thmissed a numthat small-strexcavation b1993; Zdravshrinkage of temperature aprevious analBurd et al. 20 In this papeanalysed witinfluence of sis investigate CASE HISTO General Desc Shanghai Xinhigh-rise builis reinforced

ASE HISTERMAL

Zhna Architectur

Shanghai,

during the coninvestigate thtural informaabs. The sma

BAQUS. The e beam and slae main excavanfluence of so

ness is crucialoblem.

ravkovic, Potthe modelling wmber of imporain stiffnessehaviour (Sim

vkovic, Pottsf concrete struand curing plysis, which is012).

r, a top-dowth a 3D FEsoil models and.

ORY DESCR

cription

ngye bank building with a 3concrete fram

TORY COSHRINK

honghua Xu ral Design & P.R. China, 2

nstruction prohe influence oation. The calll-strain stiffndiaphragm waabs were modation behaviooil models andl to capture th

ts et al. 2005was largely siortant details. is vital imp

mpson 1992;Ws et al. 200uts due to vaprocess is selds evident and

wn deep excavEM model innd the therma

RIPTION

ilding, as show3-floor baseme

me-shear resist

ONSIDERKAGE OF

Research Inst200002

cess. This casof different mlculations inclness is modellall was model

delled as a liner and agreed d thermal shrihe excavation

; Hashash, Soimplified and Research ha

portant to caWhittle, Hash05). But theariation of thedom considernon-negligib

vation case hn ABAQUS, l shrinkage of

wn in Fig. 1, isent. The maintance wall sys

1

RING F

titute,

se history modelling luded the led with a lled as an ear elastic well with inkage on behavior

ong et al. therefore s showed

apture the hash et al.

thermal e ambient red in the le (Dong,

history is and the

f concrete

s a 82.5m n building stem, with

Paper No. 3.28b 2

pile-raft foundation (Wang and Wang 2007; Xu 2007). The piles are bored piles, 0.9m in diameter, and around 60m in length. The whole project covers an area of about 7856m , and the excavation area is around 6200m . The excavation is 14.2m deep on the west side, and 12.2m deep on the east side. The project is situated in the downtown area of Shanghai, surrounded by 15 densely distributed buildings of which 8 are historical buildings with high protection standard, as well as some aged pipelines.

Fig. 1, Plan view of the deep excavation Fig. 2 shows the section view of the excavation and the supporting structures. The excavation is retained by a 1m thick diaphragm wall which is supported by horizontal beams and slabs, as shown in Fig. 3.

Fig. 2, Section view of A-A

Fig. 3, First floor underground beams and slabs

The 60m deep piles provide vertical support to the whole structure. The excavation is constructed with top-down methods, and the above buildings can be constructed to the 3rd floor at the same time with the excavation. This excavation was carefully measured during construction. Geotechnical Conditions and Soil Properties The city of Shanghai is situated at proximately 70km from the sea shore, in the large coastal plain limited by the East China Sea and the Yangtze River which is designated as the ‘Yangtze River Delta’. The subsoil of Shanghai is composed of Quaternary sediments of the Yangtze River estuary which consist of clay, loam, silt and sand, the different deposits being the final result of the variation from an estuarine to fluviatile sedimentation process (Dassargues, Biver et al. 1991). The elevation of the ground surface is typically from 2.2m to 4.8m above sea level (Xu, Shen et al. 2009). According to the site investigation report, the site is on a flat coastal plain, with ground elevation between 4.80m to 3.87m, and ground water table 0.5 to 1m below the ground surface. Based on the differences of soil characteristics, physical and mechanical properties, the soil profile can be divided into 7 sub layers, as shown in Fig. 4, with the corresponding soil properties.

60

50

40

30

20

10

016 18 20 20 40 60 0.5 1.0 1.5 0.5 1.0 0 20 40 0 10 0 10 20 30

⑧2-2

⑤2

②

⑤1-2

⑤1-1

④

③

Soil layers

Dep

th (m

)

①

⑧2-1

γt (kN/m3)

wn

wl

wp

wn,wl,wp (%)

e

Cc

Su (kPa)

c (kPa) φ (Ο)

Note: γt＝unit weight, wn=water content, wp=plastic limit, wl=liquid limit, e＝void ratio, cc＝compressive index, su＝field vane shear strength, c=cohesive strength, φ=internal

friction angle.

Fig. 4, Geotechnical profile and soil properties Some of the soil parameters for numerical modelling are derived from Fig. 4. The unit weight generally increases with depth, but it is convenient to take its average value, roughly 18.5kN/m . For the undrained shear strength s , the data in Fig. 4 is not sufficient for numerical modelling. Therefore, a more complete S profile is collected from Dassargues, Biver et al.(1991), and Equation (1) is derived by linearizing the data. s 20 2z kPa (1) The small-stain stiffness of the soil is missing from the site investigation report, but it could be collected from publications about Shanghai clay. Stiffness at very small strain G can be measured using dynamic methods. Cai, Zhou et al.

Pape

(200clay and t

The Shi (

Basein wand i

Instr Due and the cdeepin Fi

FThe

er No. 3.28b

00) studied thand silt throuthe value of G

G

small-strain s(2008), and sh

,

ed on Equationhich both the it is believed t

G/G

0

Fig.

rumentations

to the high ecomprehensivconstruction p

p excavation. Tig. 6.

Fig. 6, Layout observed item

he dynamic cugh several typG is citied and

20 2z

tiffness propehown in Equat

11 9.2

.

n (3), a normastiffness an

that it has adv

γ/

5, Normalized

environmentalve field measuprocess to mThe detailed i

of instrumentams include:

characters of pes of laboratod generalised a

MPa

erties are takention (3).

%.

alised curve isnd shear strainvantages in ph

/(Su/G0)

d S-shaped cu

l protection surement was

monitor the peinstrumentatio

ation at the co

typical Shanory dynamic tas Equation (

(2

n from Zhang

(3

s plotted in Fin are norma

hysical meanin

urves.

standard, a cacarried out duerformance ofon layout is sh

onstruction sit

nghai tests, (2).

2)

and

3)

ig. 5, alised ng.

areful uring f the hown

te

1. Late2. Vert

diap3. Late

14. Grou5. Grou

56. Settl

pipepipeand

7. Buil

FEM MODE Model Descri The FEM structures anshows the exc

100

m

Ux=Uy

U

Fig

The mesh of in Fig. 8 and

Fig

eral wall displtical and hor

phragm wall, eral earth dis1~ 2. und water levund settlemen

5~ 8, 91lement of pip

elines ( 01elines ( 01~telephone cab

lding settleme

L AND INPU

iption

model inclund the top-dowcavation geom

400 m

y=0

Ux=Uy=0

g. 7, Soil mesh

the diaphragmFig. 9.

g. 8, Diaphrag

acement, 1~rizontal displa

1~ 32; splacement ou

el, 1~ 5; nt outside the 1~ 12, 13~pelines, inclu

1~ 12 , ca09 , gas pi

ble pipelines ents;

UT PARAMET

des the detwn constructi

metry and bou

U

h and boundar

m wall and su

gm wall geome

~ 9; acement at to

utside the ex

excavation ~ 16; uding electricast-iron watipelines ( 01

01~ 08 ;

TERS

tailed informion procedure

undary conditio

400 m

Ux

x=Uy=Uz=0

Ux=

ry conditions

upport system

etry and mesh

3

op of the

xcavation,

1~ 4,

cal power ter-supply 1~ 04 ,

mation of es. Fig. 7 ons.

x=Uy=0

x=Uy=0

is shown

Pape

The hexapopuinterprovgroumomabouin FEmodmodmodQuadwith The Conscalcu Inpu The 1999ABAconssurfahardbehasuchhistousingthe o

er No. 3.28b

Fig. 9, Sup

soil and the ahedral elemeular to model rnal force andven that it tenund settlemenment caused but the wall cenEM (Dong, Belled with 2-elled with 4el has totalldratic elementhe linear ele

excavation solidation andulations are co

ut Parameters

soil is modell9) developed AQUS througsider the smallace model, fo

dening plasticiaviour of soilh as hysteresory. Non-lineag a number oouter fixed sur

Fig.

upporting syste

diaphragm wents. Howeverthe diaphragm

d bending monds to overesnt because it by shear stresntre line due Burd et al. 20-noded 3D b

4-noded quadly 102,036 e

nts are tried, aements but it ta

is modeledd dewateringonducted on th

led with a muat Oxford wh

gh a subroutl-strain stiffne

formulated wiity theory, takl at small strais and depenarity of the smf nested yieldrface, as show

10, Nested yi

em and supers

wall are modelr, shell elemem wall becausoment directlytimate the wmisses the b

ss on the inteto its no-thick012). The pilbeam elementdrilateral shelelements andand the resultsakes much lon

d in undrag is not conshe Oxford sup

ulti-surface soihich has been tine UMAT ess of the soilithin the framkes into accouains, and alsondence of stmall-strain resd surfaces of t

wn in Fig. 10.

ield surface m

structures

lled with 8-nents are also e it can outpu

y. But it has all deflection

beneficial benerface of the kness in geomes and beamsts. The slabsll elements. d 116,756 nos are almost snger time to ru

ained conditsidered here. percomputer.

il model (Houimplemented(Dong 2011)

l. The nested ymework of wunt the non-lio includes efiffness on re

sponse is achithe same shap

model

oded very

ut the been

n and nding

wall metry s are s are This

odes. same un.

ions. The

ulsby d into ), to yield

work-inear ffects ecent ieved pe as

The stiffnessstrength in(1) and Equaparameters foFig. 5. To invsoil models model, are al The diaphragproperties to et al. 2005).

/ 0this case hist30 , In most analyare modelledthe design stiand workmanPotts et al. 19In the analymaterial but idue to tempeambient tempBoone and CThe Coefficieassociated walso based on ANALYSIS The strategy which is beliproblem and other analyseanalysis and excavation be The central aof the case hAnd in other time. The ruTable 2. Theas well as the The focus of soil models excavation binto two grou RESULTS IN There is a modelling anpresent all of

s at very smncrease linearlation (2) respor the nested yvestigate the inin ABAQUSso used.

gm wall is consider the j. The out-of-

0.1 is adoptedtory. The ela0.2.

yses from exid as linear elaiffness to consnship of the co993). Althoughyses here, thincluding the trature changeperature varia

Crawford 2000ent of Therma

with temperatun back analysi

STRATEGIE

of the analyseieved to consthe result agr

es are groupefield data to

ehaviour.

analysis is conistory to obtacalculations,

un ID and dese numerical ree field data.

f the calculatioand therma

ehaviour. Theups and the res

NTEPRETAT

large amounnd field measuf them here.

mall strain ly with depth,pectively. Theyield surface mnfluence of so, linear elasti

modelled wijoints in the w

f-plane and ind based on soastic propertie

isting publicatstic material wsider open acconstruction (Sh simple, it ha

hey are modethermal contra

e during concration(Whittle,0; Hashash, Mal Expansion (ure change ∆s.

ES AND CALC

es is to focus sider every imrees best withed and comp

o investigate

nducted basedain a group of

only one parascription are esults are show

ons is to invesal shrinkage erefore, the casults are comp

TION

nt of data urements, but Therefore, on

and the u, as shown in e small-strainmodel are derioil models, twic and the Tr

ith anisotropiwall (Zdravkon-plane stiffnome back an

es of concrete

tions, beams with 20% redcess, shrinkagSimpson 1992as no physicalelled as lineaaction which irete curing pro Hashash et

Marulanda et a(CTE) is10∆ 35

CULATIONS

on one centramportant aspeh the field datpared with ththeir influenc

d on the backf optimised paameter is chanshown in Tab

wn in the next

stigate the infof concrete

alculations arpared in each g

from both nit is not appr

nly some sele

4

undrained Equation

n stiffness ived from

wo built-in resca soil

ic elastic ovic, Potts ness ratio nalyses of e are

and slabs duction of ge, cracks, ; St. John, l meaning. ar elastic is evident ocess and al. 1993; al. 2003). 10 / , which is

S

al analysis ect of the a. All the

he central ce on the

k analysis arameters. nged each ble 1 and t sections,

fluence of e on the re divided group.

numerical opriate to

ected data

Paper No. 3.28b 5

from the field measurements is compared with numerical results, as shown in Fig. 11.

P9

P8

P3 AA9

L12

Wall 1 Wall 2

Wall 4

Wall 6

Wall 8

L07

AA12

Line 3

Line 1

Fig. 11, Field instrumentations The comparison is focused on one wall deflection at wall centre (P9) and one at wall corner (P8), and the ground settlement along Line 1, Line 2 and Line 3. And it is believed that the data can reflect a complete picture of the excavation behaviour. In this section, results from two sets of calculations are presented, in order to demonstrate the influence of soil models and thermal shrinkage of concrete on the excavation behaviour. Influence of soil models To investigate the effect of different soil models, besides the central analysis which considers the small-strain stiffness of the soil using the nested yield surface model, three other runs are conducted with simpler soil models, as shown in Table 1. The results are compared with the filed data.

Table 1 FEM Runs and Description Run ID Description Central analysis

Soil Model: Nested-yield surface model, stiffness and strength increase linearly with depth; Wall Model: anisotropic elastic, E E⁄ 0.1, Beams and Slabs: elastic, α 10 10 / , ∆T 35

SME Same as central analysis except that the soil model is linear elastic with constant soil parameters

9 , 0.49 SMTC Same as central analysis except that the soil model

is Tresca with constant soil parameters; 9 , 0.49, 50

SMTV Same as central analysis except that the soil model is Tresca and soil stiffness and strength increases linearly with depth;

180 , 0.49, 20 2 To make these analyses comparable, some assumptions of the soil parameters are adopted. For linear elastic analysis and Tresca soil model with constant soil properties, the stiffness G is adopted as G (stiffness at 50% of the shear strength) from

Fig. 5, and the strength S is taken at the depth of 15m , roughly half of the wall depth. For Tresca soil model with variable soil parameters, G 0.180G , and G S⁄ 1000 is used here, so G 180S . The results of calculations in Table 1 are shown in Fig. 12 - Fig. 16, together with the field data.

-40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90 100-32-30-28-26-24-22-20-18-16-14-12-10

-8-6-4-20

Wal

l dep

th /m

Wall deflection /mm

Field data Central analysis SME SMTC SMTV

P9

 

-40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90 100-32-30-28-26-24-22-20-18-16-14-12-10

-8-6-4-20

Wal

l dep

th /m

Wall deflection /mm

Field data Central analysis SME SMTC SMTV

P8

Fig. 12, Wall deflection at P9 and P8 From Fig. 12, it is shown that the central analysis agrees well with the field data. Tresca soil model with variable stiffness and strength soil parameters could also capture the pattern of wall deflection very well. But the linear elastic model and Tresca soil models with constant soil parameters perform rather poorly, and therefore they are not suitable to use for prediction purposes. The ground settlement along Line 1 is shown in Fig. 13. Again, the results indicate that the central analysis with the nested yield surface model captures the ground movement very well because it considers the small-strain stiffness of the soil. But the other three, even the Tresca soil model with variable soil properties which could predict the wall deflection well, fail to produce both the pattern and magnitude of the ground movement. For the linear elastic and Tresca soil model, the ground movement around the excavation is upward which contradicts with the field data. If they were used for the prediction of the adjacent infrastructures around the excavation, the result would be misleading. Therefore, in order to get reasonable results for ground movement, the small-

Paper No. 3.28b 6

strain stiffness must be considered in the analysis.

0 5 10 15 20 25 30 35 40-30

-20

-10

0

10

20

30

40

50

60

Gro

und

settl

emen

t /m

m

Horizontal distance from wall /m

Field data Central analysis SME SMTC SMTV

Fig. 13, Ground settlements along Line 1 The ground settlement along Line 2 and Line 3 is shown in Fig. 14. Again, it demonstrates that when the small-strain stiffness of soil is considered the numerical result can capture the ground settlement. Otherwise, the results are disappointing when compared with the field data.

0 10 20 30 40 50 60 70-30

-20

-10

0

10

20

30

40

50

60

70

80

Gro

und

settl

emen

t /m

m

Horizontal distance /m

Field data Central analysis SME SMTC SMTV

Line 2

  

0 10 20 30 40 50 60 70 80-30

-20

-10

0

10

20

30

40

50

60

70

80

Gro

und

settl

emen

t /m

m

Horizontal distance /m

Field data Cental analysis SME SMTC SMTV

Line 3

Fig. 14, Ground settlements along Line 2 and Line 3 Influence of concrete thermal shrinkage To investigate the influence of thermal contraction of concrete and different temperature change on the excavation behaviour, results from another two runs, as shown in Table 2, are

compared with the central analysis as well as the field data. The results are presented as below.

Table 2 FEM runs and description

Run ID Description Central analysis

Soil Model: Nested-yield surface model, stiffness and strength increase linearly with depth; Wall Model: anisotropic elastic, E E⁄ 0.1, Beams and Slabs: elastic, α 10 10 / , ∆T 35

ANE1T30 Same as central analysis except ∆T 30 ANE1T40 Same as central analysis except ∆T 40  The results of the analyses listed in Table 2 are shown in Fig. 15~Fig. 16, together with the field data.  

0 5 10 15 20 25 30 35 40 45 50-32-30-28-26-24-22-20-18-16-14-12-10-8-6-4-20

Wal

l dep

th /m

Wall deflection /mm

Field data Central analysis ANT30E1 ANT40E1

P9

 

0 5 10 15 20 25 30 35 40 45 50-32-30-28-26-24-22-20-18-16-14-12-10-8-6-4-20

Wal

l dep

th /m

Wall deflection /mm

Field data Central analysis ANT30E1 ANT40E1

P8

 

Fig. 15 Wall deflection at P9 and P8 The results show that the wall deflection is sensitive to temperature change inside the concrete during curing process. When the concrete cools down by 5 , the beams and slabs shrink and the wall deflections increase around 3mm. But the wall deflection increment at P8 is slightly smaller than that at P9 due to the corner effect. Therefore, this effect should not be neglected in the analyses. The ground settlement behind the wall along Line 1 is plotted in Fig.16, together with the field data.

Pape

 

The arouwhic The 3 are 

 

Fig. and sensichanwall

er No. 3.28b

0 5-30-28-26-24-22-20-18-16-14-12-10

-8-6-4-2024

Gro

und

settl

emen

t /m

m

Fig. 16

results showund 3mm whech is accompa

ground settleme shown in Fig

0 10-30-28-26-24-22-20-18-16-14-12-10

-8-6-4-20

Gro

und

settl

emen

t /m

m

0 10-30-28-26-24-22

-20-18-16-14-12-10-8-6-4-2

0

Gro

und

settl

emen

t /m

m

Fig. 17, Grou

17 is consisteindicates th

itive to thermnge behind the

centre.

10 15Distance

6, Ground settl

ws that the gren the temperanied by the in

ments behind g. 17, together

0 20 30Horizonta

20 30Horizont

Fiel Cen ANT ANT

Li

und settlemen

ent with the gat ground seal contractione wall corner

20 25 3e from wall /mm

lements along

round settlemrature of concncrease of the

the wall alonr with the field

40 50al distance /m

Field data Central analysis ANT30E1 ANT40E1

Line 2

40 50 60tal distance /m

d datantral analysisT30E1T40E1

ne 3

nts along Line

ground settlemettlement beh

n of concrete. is smaller tha

30 35 40

Field data Central analysis ANT30E1 ANT40E1

Line 1

g Line 1

ment increasecrete reduceswall deflectio

ng Line 2 and d data.

60 70

0 70 80

2 and Line 3

ment along Lihind the waBut the settleman that behind

s by 5 ,

on.

Line

ine 1 all is ment d the

CONTOUR D The displaceshown belowsoils and stru The ground v18.

Fig

The largest gcentre aroundaround the cowall, the largis also evidensoil displacemincreases, andzero.

As shown in at the wall cdeflection at shape of the

DISPLAY

ement contouw, from whicuctures is clear

vertical displa

g. 18, Ground

ground settlemd the excavatorner due to thger is the grount inside the ment decreasd the displace

Fig. 19, W

Fig.19, the lacentre area clothe corner is sdiaphragm w

urs from thech the displacrly seen.

acement conto

d vertical displ

ment is concention, while thehe soil archingund settlementexcavation dues as the dist

ement around

Wall deflectio

argest wall defose the excavsmaller due to

wall also affec

e central anacement distrib

our is displaye

lacement (m)

ntrated behinde settlement ig effect. The lt area. The baue to stress retance from exthe boundary

n (m)

flection is convation depth, o the corner efcts the wall d

7

alysis are bution of

ed in Fig.

d the wall is smaller longer the asal heave elief. The xcavation

y is nearly

ncentrated while the ffect. The

deflection.

Pape

The fieldat thcond

Fig Fig. usefuundeand bene CON The probnon-procof diprovcons In ordetaimod

ACK

er No. 3.28b

displacement d measuremenhe toe. This

ditions and wa

g. 20, Displac

20 shows theul to comparerstand the perbending mom

eficial for desi

NCLUSIONS

3D numericalblems, which c-linear materedures. From isplacements a

vide rather usstruction.

rder to captureiled structure el should cons

1. The smaconsideremovemenand strenwall defmovemenconstant suitable t

2. Concretechange temperatuproblems

KNOWLEDG

at the wall bont it is usuallys assumptionall depth.

cement distribu

e deformationre with the frformance of

ment can also ign purposes.

l analysis is acould considerrial propertithe numerica

and stresses isseful informa

e the main exand construc

sider the follo

all-strain stifed to get reant. Tresca soingth propertieflection, but nt is rather poproperties an

to use. e thermal contduring the ure variation s.

EMENT

ottom is not ty assumed than depends on

ution of the su

n of the suppofield measurethe structure. be obtained

a true reality inr details of irres and stag

al results, the ss clearly preseation for rese

xcavation behaction informatowing two asp

ffness of thealistic wall dil model withes could predits prediction

oor. The Trescnd linear elas

traction induccuring proceis non-neglig

trivial, althougat the wall is fn the geolo

upporting syst

orting system. ement in ordeThe internal fdirectly, whic

n deep excavaregular geomeged construcspatial distribuented, which cearch, design

aviour, besidetion, the 3D F

pects:

e soil shoulddeflection gro variable stiff

dict the pattern of the groca soil model tic model are

ed by temperaess and amb

gible in excava

gh in fixed gical

tem

It is er to force ch is

ation etries, ction ution could

and

s the FEM

d be ound fness rn of ound with

e not

ature bient ation

The scholaracknowledgecase history The calculatio REFERENCE Boone, S. J. a

TemJourEng

Cai, H., J. Zhhorizearthof T

Dassargues, ApropEng

Dong, Y. P. (MovExcaDepOxfo

Dong, Y. P., deepstiffnYouUniv

Finno, R. J., Dperfgeot

Finno, R. J. aadjaclayFaci

Hashash, Y. McorrexcaGeo

Hashash, Y. MinveclayAna1075

Hashash, Y. MmovclayEng

Houlsby, G. TundrSymTori

Liu, G. B., Rcharclay1828

rship from ed, which is c

is from Shanons were cond

ES

and A. M. Cramperature, elasrnal of Geotecineering 126(

hou, et al. (200zontally layerhquake shakin

Tongji UniversA., P. Biver, eperties of the Qineering Geol(2011). Numervement and Stavation. PRS artment of En

ford. H. J. Burd, et

p excavation cfness of soil anung Geotechniversity of LeeD. K. Atmatzi

formance of a technical engiand L. S. Brysacent to stiff exy." Journal of Pilities 16(1): 1M. A., C. Marection and struavations." JouenvironmentaM. A., H. Sonerse analyses oys." Internationalytical Metho5. M. A. and A. vement predicty." Journal of Gineering 122(6T. (1999). A mrained clay. Pr

mposium on Prino. . J. Jiang, et aracteristics of y." Canadian G8.

China Schcrucial for mynghai Jiao Toducted on Oxf

awford (2000)stic modulus, chnical and Ge10): 870-881.00). "Plastoelared sites underng." Tongji Dasity 28(2): 177et al. (1991). "Quaternary selogy 31(1): 71rical Modellintructure Deforfirst year tran

ngineering Sci

al. (2012). 3Dcase history cond thermal conical Engineerseds. idis, et al. (19deep excavatineering 115(8

son (2002). "Rxcavation supPerformance o0-20. rulanda, et al. ut loads in Ce

urnal of Geotecal Engineeringng, et al. (2011of a deep excanal Journal fords in Geomec

J. Whittle (19tion for deep eGeotechnical 6): 474-486. model for the roceedings of re-Failure Def

al. (2011). "Dea 38 M deep e

Geotechnical J

holarship Coy study in Oxong Universitford supercom

). "Braced excand strut loadeoenvironmen astic responser multi-directiaxue Xuebao/7-182. "Geotechnical diments in Sh

1-90. ng of Ground rmation Induc

nsfer report. Oience, Univers

D FEM modelonsidering smntraction of cos's Symposium

89). "Observeion in clay." J8): 1045-1064Response of bupport system inof Constructed

(2003). "Tementral Artery chnical and g 129(6): 495-1). "Three-dimavation in Chir Numerical a

chanics 35(9):

996). "Groundexcavations inand Geoenvir

variable stiffnf the Internatioformations of

eformation excavation in Journal 48(12)

8

ouncil is ford. The

ty, China. mputer

cavations: ds." ntal

of ional /Journal

hanghai."

ced by xford, sity of

lling of a mall-strain

oncrete. m 2012,

ed Journal of .

uilding n soft d

mperature

-505. mensional cago

and 1059-

d n soft ronmental

ness of onal Soil,

soft ): 1817-

Paper No. 3.28b 9

Liu, G. B., C. W. W. Ng, et al. (2005). "Observed performance of a deep multistrutted excavation in Shanghai soft clays." Journal of Geotechnical and Geoenvironmental Engineering 131(8): 1004-1013.

Ou, C. Y., D. C. Chiou, et al. (1996). "Three-dimensional finite element analysis of deep excavations." Journal of Geotechnical and Geoenvironmental Engineering 122(5): 337-345.

Ou, C. Y., P. G. Hsieh, et al. (2010). "Performance of excavations with cross walls." Journal of Geotechnical and Geoenvironmental Engineering 137(1): 94-104.

Ou, C. Y., J. T. Liao, et al. (2000). "Building response and ground movements induced by a deep excavation." Geotechnique 50(3): 209-220.

Ou, C. Y., J. T. Liao, et al. (1998). "Performance of diaphragm wall constructed using top-down method." Journal of Geotechnical and Geoenvironmental Engineering 124(9): 798-808.

Simpson, B. (1992). "Retaining structures: displacement and design." Geotechnique 42(4): 541-576.

St. John, H. D., D. M. Potts, et al. (1993). "Prediction and performance of ground response due to construction of a deep basement at 60 Victoria Embankment." Predictive soil mechanics. Proc. of the Wroth memorial symposium, Oxford, 1992: 581-608.

Wang, W. D. and J. H. Wang (2007). Design, analysis and case histories of deep excavations supported by permanent structures. Beijing, China Architecture & Building Press.

Whittle, A. J., Y. M. A. Hashash, et al. (1993). "Analysis of deep excavation in Boston." Journal of Geotechnical Engineering - ASCE 119(1): 69-90.

Xu, Y.-S., S.-L. Shen, et al. (2009). "Geological and hydrogeological environment in Shanghai with geohazards to construction and maintenance of infrastructures." Engineering Geology 109(3–4): 241-254.

Xu, Z. H. (2007). Deformation Behaviour of Deep Excavations supported by Permanent Structure in Shanghai Soft Deposit. PhD, Shanghai Jiao Tong University, China.

Zdravkovic, L., D. M. Potts, et al. (2005). "Modelling of a 3D excavation in finite element analysis." Geotechnique 55(7): 497-513.

Zhang, P. S. and J. Y. Shi (2008). "Effect of stress path circumgyration on shear modulus under small strain and initial stress state." Yantu Gongcheng Xuebao/Chinese Journal of Geotechnical Engineering 30(3): 379-383.