ANALYSIS AND OBSERVATION OF CREEP AND SHRINKAGEEfFECTS IN THE MACAU-TAIPA BRIDGE “PONTE DA AMIZADE”

José Manuel Catarino* and Chan Mun Fongi** Investigador Principal, Chefe do Dep. de Materiais de Construção do LNEC

** Chefe do Dep. de Estruturas e Edifícios do Laboratório de Engenharia Civil de Macau, LECM

ABSTRACT

The Laboratory of Civil Engineering of Macau (LECM) monitored the Macau-TazpaBridge, “Fonte da Amizade “, to carry out its long-term observation. During construction,equ4ment was installed for measurement ofstrains, linear and angular displacements. Theinterpretation ofresuits is based on a mathematicai modelfor analysis ofcreep and shrinkageeffects in prestressed reinforced concrete sfructures integrating ali available information inthe observation study, i. e., construction sequence, loads, material properties, andenvironment conditions. A summary of the observation pian of this bridge, interpretation ofmeasurements made and computational techniques usedfor this analysis are presented.

1- DESCRIPTION OF THE BRIDGE

The new Macau-Taipa bridge, “Ponteda Amizade”, has a total length of 4414 m.Starting from Areia Preta of Macau, thebridge can be divided into a North trestie, aNorth viaduct, an outer bridge, an inter-bridge viaduct, an inner bridge, a Southviaduct, and a $outh trestie landing ontoPac On of Taipa island (Figure 1). Thewidth ofthe bridge deck is 19,3 rn with four3,75 m wide traffic lanes, a central barrier,and two emergency walkways.

The construction was promoted by thePort and Bridge Office of Macau and thedesign was made by Câncio Martins. Thecontractor was a consortium of ConstruçõesTécnicas and Teixeira Duarte. Partex-CP$

and Pengest made the supervision andLECM made the quality control of theconstruction, in Macau and in precast yardsin GuangZhou. Being extinguished the Portand Bridge Office of Macau, theresponsibilities related to the operation ofthe bridge have been transferred toDS$OPT, Direction of $oils, Public Worksand Transportation.

The North and $outh tresties, withlengths of 912 m and 996 m respectively,are constructed from 12 m x 9 m precastslabs sitting on cast-in situ piers at 12 mapart. The three viaducts are 35 m spannedstructures composed by precast piers andbeams. The height of the piers varies and isachieved by mounting precast I-blocks,each with a maximum height of 4 m.

33

TAIP

A

AAAA

AAAA

AAÂA

AAAA

AAA

flà1

HA HI

RI

i99

6m

63

0m

182

m80

5m

South

Tre

stie

$o

utk

Yia

%uc

tIn

ner

Bri

dge

Inte

r—bri

dge

Via

duct

3

rMA

CA

U

FEl

A11

i’iA

àjjj

AAÂÂ

AAAÂ

AÁAÂ

AAÃA

ÃA

329

m560

m91

2rn

Oute

rB

ridge

Nort

hV

ictd

uct

Nort

hT

rest

ie

[51I<1o

a)

-oa)

‘.4

oa)

U)

Lx

o

co

(1,

c4Ci)

oCI)

00

CO

CI)

rJ)

CI)

1.4cEa)

a)

-o.t-Da)-o1

‘-4oa)-o

LI)a)

oo

c1CI)

‘O‘-4CI)

‘-4CI)

‘-4CI)

o’CI)

CI)

‘O

CO

(4

—4

co

o

II)

4,)

r) —

‘o“4

ç

o1,4-)

q4,)

o4)-‘.4

o

o4)4-)

o

o4-44-,

o4,)

4)

o4,)

4-)

figure 2 — Sections Monitored with Strain Gauges

)0

Rio1—

-‘-4-‘4a-

35

On top of the I-block is a crosscapping beam which forms the supports forsix 34,7 m span precast I-beams. The deckof the viaducts is of cast-in situ concrete.

The outer bridge is 329 m long withtwo 112 m spans over the exterior port ofMacau in order to provide the necessaryclearance for free shipping passages. Theinner bridge is 1 $2 m long and has only one112 m span over the interior port of Macaufor shipping passage. Both bridges have aheight clearance of 30 m above sea levei.The 112 m span is obtained by usingtriangular cabie stayed structure. Cables aretíed at one end to the top of cast-in-situpylons which rise to a maximum height of59 m above sea levei. The other end of eachcable is connected to precast longitudinalgirders with one end resting on a supportingpier and the other end as cantilever. Crossgirders were then placed on the cantileverend of the longitudinal girders to serve asbearing for precast I-beams similar to thoseused in the viaducts. A cast-in-situ deck isput on to complete the structure.

2- MONITORING PLAN

Upon request of the Consortium ofConstruções Técnicas and Teixeira Duarte,LECM prepared the observation plan of“Ponte da Amizade” (Dinis, 1991). Duringthe construction some adjustments had to bemade in the original proposal to arríve atthe fmal form of the observation plan(Chan, 1994). The fmal plan coveredobservation of the behaviour of the bridgeduring construction and for a period of oneyear afier its completion. Within the oneyear monitoring period, five trí-monthlydata recording have been performed. Theobservation plan can be divided into twoparts: topography survey andinstmmentation program.

The topographic survey involves theuse of theodoijte Wild TC 200 to monitorthe deflectjons at selected locations of thestmcture during construction, in order toprovide early detection of any abnormalitythat might have occurred.

The instrumentation program isdesigned to observe the performance of theouter bridge, a portion of the North viaduct,and a portion of the inter-bridge viaduct.Parameters such as strains, angularrotations, and joint displacements are ofinterest. For interpreting mademeasurements, it is important to observe itsdevelopment with time and constructionsequences.

The strain in stmctural elements ismeasured by vibrating wire strain gaugesembedded inside concrete. Four types ofstructurai elements are monitored, namely,precast I-beams, the cast-in-situ deck,precast longitudinal girders, and cast-in-situpylons. The locations of monitoredeiements are shown in Figure 2. Vibratingwire strain gauges were installed in sectionsSito S5 and $7 to $11. Details of a typicalsection are illustrated in Figure 3. In section$6, only the longitudinal girders aremonitored. The longitudinal girders are alsomonitored in sections which align with $3,$5, $7, and $9. Two gauges are put on eachmonitored section of the longitudinalgirders. In each of sections $12 to S27 inthe pylons, four gauges are installed.

for measurement of strain,compensating gauges for creep andshrinkage are used in order to obtaininformatíon conceming the behaviour of thebridge under dead and live loads. $hrinkagecompensating gauges are placed inside20 cm x 20 cm x 75 cm concrete prismsproperly protected from the sides tosimulate the actual exposure of the elementsin site. for creep compensation, a vibratingwire strain gauge is embedded in the20 cm x 20 cm x 75 cm concrete prism. Aflat jack system, with regulator to maintainconstant pressure of 10 MPa on the prism,is used. The set up is placed inside a steelframe which works as a reaction frame.Exposure condition is the sarne as that forshrinkage compensating gauge. Such acreep compensating device is designed andpatented by LNEC.

36

ouE

0cO

(1)

Figure 4 — Measured and Predicted Total and Shrinkage Strains

Figure 3 — Location of Strain Gauges

1000

900

800

700

600

500

400

300

200

100

O

O 100 200 300 400 500 600

Time (in Days)

37

plan, 4 creepused for the

2 for the pylon.

In this monitoringcompensating gauges are1-beam, 3 for the deck, andFor shrinkage compensation, 12 gauges areused for the 1-beam, 18 for the deck and 3for the pylon. The measured strains for thecreep and shrinkage prisms of 1-beam areplotted in Figure 4.

for measurement of angular rotations,63 tihplates are used to obtain bi-directionalrotation of the deck. For the pylon, 3resistance type inclinometers are installed ineach pylon of the outer bridge givíng a totalof 24 inclinometers.

for expansion joínt displacement,special made vibrating wire jointmeters areinstalled at six expansion joints in the Northviaduct, the outer bridge, and the inter-bridge viaduct.

3- CONSTRUCTION SEQUENCE

This paper deals mainly with theanalysis of development of strains alongtime in a structural element. To achievethat, the construction sequence must beconsidered, based on good record of allimportant events experienced by structuralelements.

For discussion, the measured totalstrains in 1-beam and deck of $1 and S2 areshown in Figure 5 and Figure 6. The 1-beamhas gone through the following constructionsequence:

a) casfing of the beam in GuangZhou,China;

b) in precast yard, first prestressing oftwo cables in the beam, each wiffia prestressing force of 2340 kN,minimum strength of concrete atprestressing shall be no less than70% ofthe designated B35 gradeconcrete;

c) in Macau site, the prestressing of athird cable with the sarne 2340 kNprestressing force;

d) positioning of the bearn onto crosscapping beam;

e) setting up of formwork and castingof deck;

f) removing formwork;g) placement of asphalt surface.Therefore each 1-beam has to be

compared with its corresponding anaiyticalmodel separately, in order to reflect theactual dates of the construction sequence. InFigs. 5 and 6, the measured data of onebeam are presented.

For the deck, the most importantchange in ioading takes place duringconstruction, is in placement of the asphaltsurface. The average strains measured in thedeck of sectíons Si and S2 are also piottedinfigs. 5and6.

4- STRUCTUR%L ANALYSIS ALONGTIME

In prestressed reinforced concretestructures composed of linear parts, severaitypes of rheological effects can occur inconcrete sections, which can be classified asfollows:

type 1 due to different dryingconditions of concrete;

type II due to compatibility ofconcrete, steel reinforcementand prestressing steel;

type III in sections made of severaiparts with different ages,thickness and exposureconditions, due to thecompatibility between theseparts.

Ali these rheologicai effects have, asconsequence, increases of strains andintemal coaction states of stress which, instaticaliy indeterminate structures, produceredistributions of intemal and bearingforces. In linear pieces, it is conimonlyadopted the assumption of plane crosssections.

Rheological effects along the smallerdimension of a concrete part due todifferent drying conditions (type 1) can beestimated by finite eiement analysís, takinginto consideration the evolution ofboundary conditions and concrete creep andshrinkage iaws for ali possible histories ofrelevant parameters.

3$

ooE

(1)c

a)

o1oEc

0

OL..

a)

2nd. prestressCasting of deck

OO

lst prestress

t-beam

Píacement of asphalt - - - - Deck

100 200 300 400

Time (in Days)

500

Figure 5— Strains at 1-beam and Deck of Si

600

1200

1000

800

600

400

200

o

-200

1200

1000

800

600

400

200

O

-200

2nd. prestres Pacement of asphaft

O /O

O

Casting of deck

ptesttess

O

1-beam

Deck

100 200 300 400 500

Figure 6— Strains at 1-beam and Deck of 52

600

Time (in Days)

39

There is not yet available enoughexperimental data to calibrate the behaviourof concrete in variable conditions oftemperature, humídity and applied stresses.So, only some extrapolations based on wellknown concrete response permit this type ofanalyses.

For the structural analysis along timeof “Ponte da Amizade”, a mathematicalmodel based on (Catarino, 1986) has beenused, making possibie the interpretation ofmeasured strains and dispiacements, as thismodel has the foliowing possibilities:

a) step by step analysis along timewith consideration of concrete andprestress steei rheology in crackedor uncracked sections;

b) skyline and Cholesky method forstorage of rigidity matrix andresolution of the system ofequations;

c) consideration of constructionphases;

d) composed concrete sections withparts cast in different dates;

e) change of support conditions;f) imposal of intemal forces and

strains;g) rheologic modeis of materiais

based on CEB Modei Code (CEB,1978) or other available data.

In this mathematical model, takinginto consideration the constructíonsequence, the behaviour of used materiais,ali relevant dates with variations of loadsand measurements in equipment used, timeis divided in intervals for step by stepanalysis. During each intervai of timestresses are assumed as constant, andincrements of strains are computed inconcrete sections at the end of the timeinterval, based on “superposition principie”.After this, the compatibility of materiaisinside sections is “re-established” andvariatjons of stresses in concrete sectionsare computed, as well as a system of“fictitious forces” that retum the structureto the sarne shape of the beginning of thetime interval. Finaliy, the instant analysis ofthe whole stmcture is done, based on theactual geometry, materiais parameters,

environment conditions, removal offictitious forces and appliance of otherloads.

This mathematical model estimatesrheological effects due to the compatibilityof materiais and different parts of concretesections (types II and III above described).Type 1 effects can be anaiysed separateiybecause, in most cases, results can besuperposed. However, classical methods ofmeasuring strains in situ can not detecttype 1 effects, but only mean values ofstrains along the thickness of a concretepart. Because of this, it is always necessaryto have separate analysis of the referredrheologicai effects.

In several long term observationsmade by LNEC, the interpretation ofmeasured strains and displacements hasbeen tried tbrough this computer program(TINTIN). After discovering the relevantparameters that permit the approach toexperimental data, a better understanding ofthe behaviour of the structure becomespossible, as weil as the elimination oferroneous information.

The sarne computer program has beenused for modelling the effects of creep andshrinkage in concrete sections withcomponents of different thickness.(Catarino, 1993) presents the method usedfor this analysis. Basically it consists insubdividing section in parts of sarnethiclmess, and treat it as a plane frarnestructure with rigid bars to impose lineartotal strains in each section.

5- INTERPRETATION OFMEASURED STRMNS

5.1 - Creep and Shrinkage Strains

Figure 4 includes comparisons madeof creep, shrinkage and total strains incompensators for measurements made inthe bridge deck.

The analysis of measured strains hasbeen based on the rheologic modeis ofconcrete of CEB Model Code (CEB, 197$).

40

14

‘ :.

___

0__

10 100 1000 10000 100000-2

— Time (in Days)

Figure 7— Computed Stresses of I-beams and Decks at Si and S2

0.045 T

0.04

0.035

0.03

c 0.025 ..L

0.02—

0.015 —

o

0.01 -‘-

0.005 1O

-0.005

1-beam at Si

1-beam at S2

--—DeckatSl

Deck at S2

10 100 1000 10000 100000

Time (in Days)

Figure 8 — Computed Deflection ofl-beam at S2

41

The purpose of this analysís is thecalibration of well known material modeisto a group of experimental data, so ffiat, inthe structurai analysís, the sarne modeis canbe extrapolated for ali existing situatíons.The procedure used consists of combiningthe possible values of parameters of thosemodeis which influence elastic andrheological properties, in order to optimiseffie approach bePveen measured andcomputed stTains.

5.2 - Beam and Deck Strains

Figs. 5 and 6 include comparisonsmade of sfrains measured in 1-beam andDeck of sections $1 and S2 of ffie bridge,referring ffie most important consfructionphases for this - part of the sfructure. Thefour cases included in these figures showvery good correlation, much better ffian inanalysis of this type made in otherstructures. Part ofthis achievement is due tothe simplicity offfie structural system used.

6- ESTIMATION OF OTHERPÃRÃMETERS

Afier acbieving a reasonable approachbetween measured and computed strains ina part of the bridge deck, it is very reliableto estimate other pararneters like stresses,deflections, shortenings and rotations in ifiesarne instrumented areas and, wiffi aninferior degree ofreliability, it is possible toestimate the evolution of ffie sarnepararneters for the life time ofthe bridge.

6.1 - Stresses

figure 7 shows ffie evolution ofstresses in sections Si and $2 during ffieperiod of observation and its extrapolationfor a life time of 50 years. The stresses havebeen computed at ffie sarne locations of ffieinstalled gauges.

figure 9 — Computed Shortening and Rotation of 1-beam at AI

c

•0(0a:c

co

oa:

Eco)cc0)to-ca,

0.004

0.003

0.002

0.001

0

-0.001

-0.002

-0.003

-0.004

-0.005

-0.006

1000 10000 100000

Time (in Days)

42

The estimated stresses in I-beams andDeck have adequate ranges of values inaccordance with age and class of concrete.

6.2 — Deflectíons

Figure $ shows the evolution ofdeflection in section S2 (middle span)during the period of observation and itsextrapolation for a life time of 50 years.

This figure demonstrates the verygood balance adopted in design betweenpermanent loads and prestress, importantfor the long term performance of the bridgedeck.

very reliable to estimate other parametersnot easily measurable, like stresses, strainsand displacements in some parts of thestructure. With an inferior degree ofreiiability it is certainly capable to estimatethe evolution of ali parameters for the iifetime of the bridge.

from the measurements and analysisalready made, it can be stated that thestmctural behaviour of this bridge, verifiedsince the construction and estimated for itslife time, seems quite satisfactory. Inparticular the balance adopted in designbetween permanent loads and prestressprove to be excellent in terms of long termevolution of deflections of the deck.

6.3 - Shortenings and Rotations

figure 9 shows the evolution ofshortening of the 1-beam and rotatíon insection Si, during the period of observationand its extrapolation for a life time of50 years.

The plotted shortening refers only tothe left part of the 1-beam, with a length of5 m. The comment made for deflection in$2 is also applicable to estimated rotation insection Si.

7- CONCLUSIONS

The information presented in thispaper is only a small part of a systematicanalysis of ali measurements made in theNew Macau-Taipa Bridge, ‘Ponte daAmizade”, which is being carried by theLaboratory of Civil Engineering of Macau(LECM).

Although the type of structure of thisbridge is few sensitive to concreterheological effects, even in this case isindispensable the use of a mathematicalmodei with the above describedpossibilities for the interpretation ofmeasured strains, angular and lineardisplacements.

Afler achieving a reasonable approachbetween measured and computedparameters in ali possible situations, it is

REFERENCES

Catarino, J.M. (1986), “ComportamentoReológico de Estruturas de BetãoArmado Pré-esforçado - Análise eObservação”, LNEC, thesis

Catarino, J.M. and Silveira, P. (1993),“Creep and Shrinkage Effects inConcrete Sections with Componentsof Different Thickness”, RILEMConCreep 5, Barcelona

CEB, Comité Euro-Intemational du Béton(197$), “Code Modele CEB-FIP pourles Structures en Béton”, Bulietind’Information n° 125, CEB

Chan, M.F. and Lam, C.L. (1994),“Monitoring of New Macau-TaipaBridge - Instrumentation Part, LECMreport n° 543

Dinis, V.M.S. (1991), “Proposta paraObservação do ComportamentoEstrutural da Nova PonteMacau-Taipa”, LECM report n° 18$

43