ANALYSIS AND OBSERVATION OF CREEP AND SHRINKAGE EfFECTS IN THE - 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-Tazpa Bridge, “Fonte da Amizade“, to carry out its long-term observation. During construction, equ4ment was installed for measurement ofstrains, linear and angular displacements. The interpretation ofresuits is based on a mathematicai modelfor analysis ofcreep and shrinkage effects in prestressed reinforced concrete sfructures integrating ali available information in the observation study, i.e., construction sequence, loads, material properties, and environment conditions. A summary of the observation pian of this bridge, interpretation of measurements made and computational techniques usedfor this analysis are presented.

1- DESCRIPTION OF THE BRIDGE and Pengest made the supervision and LECM made the quality control of the The new Macau-Taipa bridge, “Ponte construction, in Macau and in precast yards da Amizade”, has a total length of 4414 m. in . Being extinguished the Starting from Areia Preta of Macau, the and Bridge Office of Macau, the bridge can be divided into a North trestie, a responsibilities related to the operation of North viaduct, an outer bridge, an inter- the bridge have been transferred to bridge viaduct, an inner bridge, a South DS$OPT, Direction of $oils, Public Works viaduct, and a $outh trestie landing onto and Transportation. Pac On of Taipa (Figure 1). The The North and $outh tresties, with width ofthe bridge deck is 19,3 rn with four lengths of 912 m and 996 m respectively, 3,75 m wide traffic lanes, a central barrier, are constructed from 12 m x 9 m precast and two emergency walkways. slabs sitting on cast-in situ piers at 12 m The construction was promoted by the apart. The three viaducts are 35 m spanned Port and Bridge Office of Macau and the structures composed by precast piers and design was made by Câncio Martins. The beams. The height of the piers varies and is contractor was a consortium of Construções achieved by mounting precast I-blocks, Técnicas and Teixeira Duarte. Partex-CP$ each with a maximum height of 4 m.

33 TAIPA

AAAAAAAAAAÂAAAAAAAA fl à 1 H A H I RI i

996 m 630 m 182 m 805 m South Trestie $outk Yia%uct Inner Bridge Inter—bridge Viaduct

3

rMACAU FEl A 11 i’i A à jj j AAÂÂAAAÂAÁAÂAAÃAÃA

329 m 560 m 912 rn Outer Bridge North Victduct North Trestie [51 o 1, I<1 4-) o o

q 4,)

o 4) -‘.4

a) a) -o -o .t-D a) -o1a) ‘.4o ‘-4o a) o -oa)

U) LI) a) Lx o o II) co c1 CI) ‘O (1, ‘-4 o CI) 4) 4-) 4,) c4 ‘o Ci) ‘-4 “4 o CI) r) — CI) ‘-4 CI) 00 ç o’ CO CI)

CI) o CI) ‘O rJ) CO (4 CI) —4 )0 co c1.4 E a)

o 4-4 4-,

o 4,) 4) -‘-4 -‘4 a- Rio o 4,) 1— 4-)

figure 2 — Sections Monitored with Strain Gauges

35 On top of the I-block is a cross The instrumentation program is capping beam which forms the supports for designed to observe the performance of the six 34,7 m span precast I-beams. The deck outer bridge, a portion of the North viaduct, of the viaducts is of cast-in situ concrete. and a portion of the inter-bridge viaduct. The outer bridge is 329 m long with Parameters such as strains, angular two 112 m spans over the exterior port of rotations, and joint displacements are of Macau in order to provide the necessary interest. For interpreting made clearance for free shipping passages. The measurements, it is important to observe its inner bridge is 1$2 m long and has only one development with time and construction 112 m span over the interior sequences. for shipping passage. Both bridges have a The strain in stmctural elements is height clearance of 30 m above sea levei. measured by vibrating wire strain gauges The 112 m span is obtained by using embedded inside concrete. Four types of triangular cabie stayed structure. Cables are structurai elements are monitored, namely, tíed at one end to the top of cast-in-situ precast I-beams, the cast-in-situ deck, pylons which rise to a maximum height of precast longitudinal girders, and cast-in-situ 59 m above sea levei. The other end of each pylons. The locations of monitored cable is connected to precast longitudinal eiements are shown in Figure 2. Vibrating girders with one end resting on a supporting wire strain gauges were installed in sections pier and the other end as cantilever. Cross Sito S5 and $7 to $11. Details of a typical girders were then placed on the cantilever section are illustrated in Figure 3. In section end of the longitudinal girders to serve as $6, only the longitudinal girders are bearing for precast I-beams similar to those monitored. The longitudinal girders are also used in the viaducts. A cast-in-situ deck is monitored in sections which align with $3, put on to complete the structure. $5, $7, and $9. Two gauges are put on each monitored section of the longitudinal 2- MONITORING PLAN girders. In each of sections $12 to S27 in the pylons, four gauges are installed. Upon request of the Consortium of for measurement of strain, Construções Técnicas and Teixeira Duarte, compensating gauges for creep and LECM prepared the observation plan of shrinkage are used in order to obtain “Ponte da Amizade” (Dinis, 1991). During informatíon conceming the behaviour of the the construction some adjustments had to be bridge under dead and live loads. $hrinkage made in the original proposal to arríve at compensating gauges are placed inside the fmal form of the observation plan 20 cm x 20 cm x 75 cm concrete prisms (Chan, 1994). The fmal plan covered properly protected from the sides to observation of the behaviour of the bridge simulate the actual exposure of the elements during construction and for a period of one in site. for creep compensation, a vibrating year afier its completion. Within the one wire strain gauge is embedded in the year monitoring period, five trí-monthly 20 cm x 20 cm x 75 cm concrete prism. A data recording have been performed. The flat jack system, with regulator to maintain observation plan can be divided into two constant pressure of 10 MPa on the prism, parts: topography survey and is used. The set up is placed inside a steel instmmentation program. frame which works as a reaction frame. The topographic survey involves the Exposure condition is the sarne as that for use of theodoijte Wild TC 200 to monitor shrinkage compensating gauge. Such a the deflectjons at selected locations of the creep compensating device is designed and stmcture during construction, in order to patented by LNEC. provide early detection of any abnormality that might have occurred.

36

(1) O c

0

E

u o

1000

200

300

100

400

500

600

700

800

900

O

O

Figure

100 4

—

Measured

Figure

200

3 and

—

Location

Predicted

37

300

of

Total

Strain

and

Gauges

400

Shrinkage

Strains

Time

500

(in

Days) 600 In this monitoring plan, 4 creep f) removing formwork; compensating gauges are used for the g) placement of asphalt surface. 1-beam, 3 for the deck, and 2 for the pylon. Therefore each 1-beam has to be For shrinkage compensation, 12 gauges are compared with its corresponding anaiytical used for the 1-beam, 18 for the deck and 3 model separately, in order to reflect the for the pylon. The measured strains for the actual dates of the construction sequence. In creep and shrinkage prisms of 1-beam are Figs. 5 and 6, the measured data of one plotted in Figure 4. beam are presented. for measurement of angular rotations, For the deck, the most important 63 tihplates are used to obtain bi-directional change in ioading takes place during rotation of the deck. For the pylon, 3 construction, is in placement of the asphalt resistance type inclinometers are installed in surface. The average strains measured in the each pylon of the outer bridge givíng a total deck of sectíons Si and S2 are also piotted of 24 inclinometers. infigs. 5and6. for expansion joínt displacement, special made vibrating wire jointmeters are 4- STRUCTUR%L ANALYSIS ALONG installed at six expansion joints in the North TIME viaduct, the outer bridge, and the inter- bridge viaduct. In prestressed reinforced concrete structures composed of linear parts, severai 3- CONSTRUCTION SEQUENCE types of rheological effects can occur in concrete sections, which can be classified as This paper deals mainly with the follows: analysis of development of strains along type 1 due to different drying time in a structural element. To achieve conditions of concrete; that, the construction sequence must be type II due to compatibility of considered, based on good record of all concrete, steel reinforcement important events experienced by structural and prestressing steel; elements. type III in sections made of severai For discussion, the measured total parts with different ages, strains in 1-beam and deck of $1 and S2 are thickness and exposure conditions, due to the shown in Figure 5 and Figure 6. The 1-beam between has gone through the following construction compatibility these parts. sequence: Ali these rheologicai effects have, as a) casfing of the beam in GuangZhou, consequence, increases of strains and ; intemal coaction states of stress which, in b) in precast yard, first prestressing of staticaliy indeterminate structures, produce two cables in the beam, each wiffi redistributions of intemal and bearing a prestressing force of 2340 kN, forces. In linear pieces, it is conimonly concrete at minimum strength of adopted the assumption of plane cross be less than prestressing shall no sections. 70% ofthe designated B35 grade Rheological effects along the smaller concrete; dimension of a concrete part due to c) a in Macau site, the prestressing of different drying conditions (type 1) can be third cable with the sarne 2340 kN estimated by finite eiement analysís, taking prestressing force; into consideration the evolution of d) positioning of the bearn onto cross boundary conditions and concrete creep and capping beam; shrinkage iaws for ali possible histories of e) setting up of formwork and casting relevant parameters. of deck;

3$ a)

a) c (1) E o o

o

0 c 1

L.. O E o

1200 1000 -200

1000

1200

-200

200

400

400 200 600 800 600

800

o

O

2nd. 2nd.

lst

prestres

prestress

ptesttess

prestress

100

100

Casting

Figure

Figure

Casting

O

200

5—

200

6—

of

deck

Strains

Strains

of

Píacement

deck

39

Pacement

O

/

at

at

300

1-beam

300

1-beam

O

of

asphalt

of

and

and

asphaft

Deck

Deck

400

400

of

of

Si

52 - -

O

O

Time -

Time -

500

500

1-beam Deck

Deck t-beam

(in

(in

Days)

Days)

O

600

600 There is not yet available enough environment conditions, removal of experimental data to calibrate the behaviour fictitious forces and appliance of other of concrete in variable conditions of loads. temperature, humídity and applied stresses. This mathematical model estimates So, only some extrapolations based on well rheological effects due to the compatibility known concrete response permit this type of of materiais and different parts of concrete analyses. sections (types II and III above described). For the structural analysis along time Type 1 effects can be anaiysed separateiy of “Ponte da Amizade”, a mathematical because, in most cases, results can be model based on (Catarino, 1986) has been superposed. However, classical methods of used, making possibie the interpretation of measuring strains in situ can not detect measured strains and dispiacements, as this type 1 effects, but only mean values of model has the foliowing possibilities: strains along the thickness of a concrete a) step by step analysis along time part. Because of this, it is always necessary with consideration of concrete and to have separate analysis of the referred prestress steei rheology in cracked rheologicai effects. or uncracked sections; In several long term observations b) skyline and Cholesky method for made by LNEC, the interpretation of storage of rigidity matrix and measured strains and displacements has resolution of the system of been tried tbrough this computer program equations; (TINTIN). After discovering the relevant c) consideration of construction parameters that permit the approach to phases; experimental data, a better understanding of d) composed concrete sections with the behaviour of the structure becomes parts cast in different dates; possible, as weil as the elimination of e) change of support conditions; erroneous information. f) imposal of intemal forces and The sarne computer program has been strains; used for modelling the effects of creep and g) rheologic modeis of materiais shrinkage in concrete sections with based on CEB Modei Code (CEB, components of different thickness. 1978) or other available data. (Catarino, 1993) presents the method used In this mathematical model, taking for this analysis. Basically it consists in into consideration the constructíon subdividing section in parts of sarne sequence, the behaviour of used materiais, thiclmess, and treat it as a plane frarne ali relevant dates with variations of loads structure with rigid bars to impose linear and measurements in equipment used, time total strains in each section. is divided in intervals for step by step analysis. During each intervai of time 5- INTERPRETATION OF stresses are assumed as constant, and MEASURED STRMNS increments of strains are computed in concrete sections at the end of the time 5.1 - Creep and Shrinkage Strains interval, based on “superposition principie”. After this, the compatibility of materiais Figure 4 includes comparisons made inside sections is “re-established” and of creep, shrinkage and total strains in variatjons of stresses in concrete sections compensators for measurements made in are computed, as well as a system of the bridge deck. “fictitious forces” that retum the structure The analysis of measured strains has to the sarne shape of the beginning of the been based on the rheologic modeis of time interval. Finaliy, the instant analysis of concrete of CEB Model Code (CEB, 197$). the whole stmcture is done, based on the actual geometry, materiais parameters,

40 ___

14 1-beam at Si 1-beam at S2 --—DeckatSl Deck at S2

‘ :.

10 100 1000 10000 100000 -2 0__— 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 1

O 10 100 1000 10000 100000 -0.005 Time (in Days)

Figure 8 — Computed Deflection ofl-beam at S2

41 be the the to modeis The calibration ffie rheological

5.2 computed made

referring Deck phases four very the analysis

structures.

a

extrapolated

structurai -

possible simplicity

group

procedure

Beam

cases

good approach

The

Figs.

of

of

a, a: a: •0 for -c t

E o) c o c c (0 c o c c o 0)

sections

which

ffie

sfrains

stTains.

of

purpose

Part

of

of

5 correlation,

included

and properties,

this

-0.001 -0.002 -0.004 -0.003 -0.005

-0.006 values

0.002 0.001 0.003 0.004 analysís,

experimental

and

well offfie most

this

used

ofthis -

Deck

for

part

bePveen

0

measured

$1

influence

6

known

of

of

type

important

structural ali

consists

in

and

of

Strains include

achievement

the

this in

parameters

much

these existing

figure

the

order

S2

sarne

made

material

data,

measured

analysís

in

sfructure.

of of

better

figures system

elastic 9

comparisons

consfruction

1-beam

to

modeis —

ffie

combining

so

situatíons.

Computed

in

optimise

of

is

ffiat, modeis

ffian

bridge,

is

due

used.

those

show

other

The and

and can

and

the

in

to in

Shortening

42

between 6- PÃRÃMETERS to a

pararneters deflections, estimate inferior sarne

6.1

period

stresses been for

installed

part

estimate

ESTIMATION

and - a

Stresses

life

Afier

computed

figure

of

Rotation

instrumented

of

degree

1000

in measured

the

gauges. time

observation

the

for acbieving

sections

shortenings

other

7 bridge

of

of

the

ofreliability,

at

evolution

shows

1-beam

50

life

pararneters

ffie

and

OF

years.

deck,

10000

Si

a

areas

Time

time

sarne

and

at

reasonable

computed

OTHER

and

and

ffie

AI

it

The

ofthe

(in

its

of

locations

is rotations

it

and,

$2

evolution

Days)

like

is

extrapolation very

stresses

ffie

possible

during

100000

bridge.

strains

approach

wiffi

stresses,

reliable

in

of

sarne

have

ifie

ffie

ffie

an

in

to

of very reliable to estimate other parameters The estimated stresses in I-beams and not easily measurable, like stresses, strains Deck have adequate ranges of values in and displacements in some parts of the accordance with age and class of concrete. structure. With an inferior degree of reiiability it is certainly capable to estimate 6.2 — Deflectíons the evolution of ali parameters for the iife time of the bridge. Figure $ shows the evolution of from the measurements and analysis deflection in section S2 (middle span) already made, it can be stated that the during the period of observation and its stmctural behaviour of this bridge, verified extrapolation for a life time of 50 years. since the construction and estimated for its This figure demonstrates the very life time, seems quite satisfactory. In good balance adopted in design between particular the balance adopted in design permanent loads and prestress, important between permanent loads and prestress for the long term performance of the bridge prove to be excellent in terms of long term deck. evolution of deflections of the deck.

6.3 - Shortenings and Rotations REFERENCES figure 9 shows the evolution of shortening of the 1-beam and rotatíon in Catarino, J.M. (1986), “Comportamento section Si, during the period of observation Reológico de Estruturas de Betão and its extrapolation for a life time of Armado Pré-esforçado - Análise e 50 years. Observação”, LNEC, thesis The plotted shortening refers only to the left part of the 1-beam, with a length of Catarino, J.M. and Silveira, P. (1993), 5 m. The comment made for deflection in “Creep and Shrinkage Effects in $2 is also applicable to estimated rotation in Concrete Sections with Components section Si. of Different Thickness”, RILEM ConCreep 5, Barcelona

7- CONCLUSIONS CEB, Comité Euro-Intemational du Béton (197$), “Code Modele CEB-FIP pour The information presented in this les Structures en Béton”, Bulietin paper is only a small part of a systematic d’Information n° 125, CEB analysis of ali measurements made in the New Macau-Taipa Bridge, ‘Ponte da Chan, M.F. and Lam, C.L. (1994), Amizade”, which is being carried by the “Monitoring of New Macau-Taipa Laboratory of Civil Engineering of Macau Bridge - Instrumentation Part, LECM (LECM). report n° 543 Although the type of structure of this bridge is few sensitive to concrete Dinis, V.M.S. (1991), “Proposta para rheological effects, even in this case is Observação do Comportamento indispensable the use of a mathematical Estrutural da Nova Ponte modei with the above described Macau-Taipa”, LECM report n° 18$ possibilities for the interpretation of measured strains, angular and linear displacements. Afler achieving a reasonable approach between measured and computed parameters in ali possible situations, it is

43