ACI STRUCTURAL JOURNAL TECHNICAL PAPER Title no. 105-S09

Behavior of Concrete Bridge Decks Reinforced with High-Performance Steel by Hatem M. Seliem, Gregory Lucier, Sami H. Rizkalla; and Paul Zia

This paper describes the behavior of concrete bridge decks reinforced with newly developed high-perfonnance (HPJ steel that is charac­terized by its high strength mId enhanced corrosion-resistance in comparison with conventionaL ASTM A6JS-06 Grade 60 steel. The study presented herein included testing of three full-scale bridge decks with a span-depth ratio of 12.5. The first and second decks were constructed with the same reinforcement ratio using HP and Grade 60 steel, respectively. The third deck was reinforced with HP steel using 33% less reinforcement in an attempt to use its high strength. A nonlinear finite element model was used to predict the mode offailure alldfailure loads. Test results demollstrate that the use of HP steel at a reduced reinforcement ratio is viable as flexural reinforcemellt in concrete bridge decks. The paper also presents the test results of specially-designed specimens to study the effect of bending of HP steel bars on their tensile strength.

Keywords: bent bars; bridge decks; flexural-shear; punching.

INTRODUCTION Bridge decks are frequently subjected to severe environmental

conditions that often lead to serious corrosion problems. The use of high-performance (HP) steel could help ' to mitigate corrosion problems due to its enhanced corrosion resistance. In addition, HP steel has hiyher strength compared with conventional ASTM A615-06 Grade 60 steel. Therefore, by using HP steel, the amount of required reinforcement could be considerably reduced. Reducing the amount of steel will alleviate reinforcement congestion and improve concrete placement. Steel that conforms to ASTM A 1035-072 was selected for this study because of its high-strength and enhanced corrosion resistance in comparison to conventional ASTM A615-06 1 Grade 60 steel. This paper is a part of a comprehensive study to investigate the structural behavior of HP steel for bridges. The work presented in this paper examined the behavior of bridge deck slabs and the strength of bent bars required for certain details. The experimental program presented in this paper consisted of two phases. In the first phase, three full-scale bridge decks with a span-depth ratio of 12.5 were tested to evaluate the structural perfonnance of bridge decks reinforced with HP steel as main flexu ral reinforcement in comparison with the use of conventional Grade 60 steel. In the second phase, four specially-designed specimens were tested to assess the effect of bending on the tensile strength of HP steel bars.

RESEARCH SIGNIFICANCE Recently, many state transportation departments have

begun to use HP steel as a direct replacement for conventional Grade 60 steel in concrete bridge decks.3 However, the behavior of concrete bridge decks reinforced with this novel steel is not well defined. This study is an attempt to use the high strength characteristics of HP steel in concrete bridge

78

decks. In addition, the study evaluates the effect of bending on the tensile strength of HP steel bars.

PHASE I: CONCRETE BRIDGE DECKS Test specimens

A total of three full-scale bridge decks were considered in this study to examine the flexural limit state behavior, including the mode offailure. The three decks were designed to be identical in all aspects except for the type and amount of steel used in each. All three bridge decks consisted of two spans and double canti1evers, supported in composite action by three precast, post-tensioned concrete girders having cross-sectional dimensions of 24 x 10 in. (610 x 254 mm). The overall nominal dimensions of the bridge decks were 21 ft-IO in. X 13 ft-2 in. x 8-5/8 in. (6655 x 4013 X 220 mm) with a span-depth ratio of 12.5. The supporting girders were post­tensioned using deformed prestressing bars of I in. (25 mm) diameter with an ultimate strength of 150 ksi (1034 MPa). Each girder was prestressed by four bars resulting in a total prestressing force of 360 kips (1601 kN) per girder. Post­tensioning was used to prevent the girders from torsional cracking so as to maintain their torsional stiffness throughout the test. The girders were designed so that their torsional stiffness was similar to that of the steel bridge girders of an actual bridge that was built in Johnston County, NC, in 2004 using HP steel. 3

The first and third bridge decks were reinforced with HP steel, whereas the second bridge deck was reinforced with conventional Grade 60 steel for comparison purposes. The test matrix is given in Table I , and the reinforcement details for the three bridge decks are shown in Fig. I. It should be noted that the reinforcement ratio p is calculated using the total slab thickness. The first and second bridge decks were constructed with the same reinforcement ratio using HP and conventional Grade 60 steel similar to that used in the bridge bu ilt in Johnston County, NC, in 20043 The th ird bridge deck, however, was reinforced with HP steel using only 2/3 of the reinforcement ratio used for the first two decks. The reduction in the amount of steel is based on a selected yield strength of 90 ksi (621 MPa), which is within the linear behavior of the HP steel and less than the yield strength of 120 ksi (827 MPa) determined according to the 0.2% offset method specified by ASTM A370-074 It should be noted that only the transverse steel was reduced because the deck is continuous in this direction where primary bending occurs.

ACt Stnlctural Journal, V. 105. No. I , January~February 2008. MS No. S-2006-378 received September 15, 2006, and reviewed under Institute

publicillion policies. Copyright C> 2008. American Concrete Institute. All rights reserved, lIlcluding the making of copies unless pennission is obtained from the copyright proprietors. Pertinent discussion including aum·s closure, if any. will be published ill the November­December 2008 ACi Structural Journal if the discussion is received by July 1.2008.

ACI Structural Journal/January-February 2008

ACI member Hatem M. Seliem is a PhD Candidate ill tile Deparlmelll of Civil. COllstruction, and Environmental Engineerillg at Norlll Cora/ilia State Universit)'. Raleigh. Ne. He received his BSc {Jl/d MSc from Cairo University. Cairo, Egypt. in 2()()() and 2002. respectively.

ACI member Gregory Lucier is a Research Engineer at the Constructed Facilities wboratory. Nonh Carolina Slate Vniversily. He received his BSc and MScfrom North Carolina State University in 2004 and 2006, respectively.

Sami H. RizkalJa. FACI, ;s Distinguished Professor of Cil,jl and COllstruction Engineering in the Department of Civil, Construction and Environmental Engineering, North Carolina State University. He is also Ih e Director of the Constructed Facilities Laborotory alld NSF IIUCRC in Repair of Structures (/lid Bridges at North Carolina State University. He is a member of ACI Committees 118. Use of Computers; 440, Fiber Reinforced Polymer Reinforcement: E803. Faculty Network Coordinating Committee: and Joint ACI·ASCE CommitlteS 423. Prestressed Concrete, and 550. Precast Concrete Structures.

ACI Honorary Member Paul Zia is Distinguished University Professor Emeritus at Nonh Carolina State University. He served as ACI President in 1989 and is a member of ACI Committees 363, f/igh.Stwzgth Concrete. and 440, Fiber Reinforced Polymer Reinforcement. and Joint ACI·ASCE Committees 423, Prestressed Concrete, and 445. Shear and Torsion.

Material properties The three bridge decks were constructed using normal­

weight concrete with average compressive strengths at the day of testing for the three bridge decks of 4500, 5300, and 7000 psi (31, 36, and 48 MPa). The concrete compressive strength was determined using 4 x 8 in. (102 x 204 mm) concrete cylinders cast for each deck and cured under the same conditions as the deck. Tension coupons of HP and Grade 60 steels were tested according to ASTM A370-074

The stress-strain characteristics of the HP and Grade 60 steel are shown in Fig. 2. The HP steel reinforcing bars exhibited a linear stress-strain relationship up to 100 ksi (689 MPa) followed by a nonlinear behavior up to an ultimate strength of 173 ksi (1193 MPa). According to the ASTM A370-074

,.",

offset method (0.2% offset), the yield strength was determined to be 120 ksi (827 MPa). The initial modulus of elasticity was 29,000 ksi (200 GPa). followed by a nonlinear behavior and reduction in the modulus of elasticity when the stress exceeded 100 ksi (689 MPa). The yield strength of the Grade 60 steel was determined to be 68 ksi (469 MPa).

Test setup and instrumentation Two 440 kips (1957 kN) capacity hydraulic actuators were

used to simultaneously apply a concentrated load to each span to simulate the effect of truck wheel loads. Two 10 x 20 in. (254 x 508 mm) steel plates of I in. (25 mm) thickness were used to transfer the loads from the actuators to comply with the AASHTO Specificationss for tire contact area. A 112 in. (13 mm) thick neoprene pad was placed under each loading plate to prevent possible local crushing of the concrete. The supporting girders rested on concrete blocks to transfer the applied loads to the strong floor resulting in a clear span of 96 in. (2438 mm). The clear span of supporting girders was determined based on the equivalency of the torsional stiffness of the supporting girders to that of the steel girders used in the actual bridge. Figure 3 shows an isometric view of the test setup and a photograph of the first bridge deck before testing.

A total of 72 channels of instrumentation were used on each bridge deck. A 550 kips (2447 kN) capacity load cell was mounted to each actuator to measure the applied load. Twenty-four string potentiometers (string pots) were used to measure the deflection profiles of the bridge deck along the longitudinal and transverse directions. In addition, six linear potentiometers were used to measure the deflections and rotations at the midspan of each girder. Twenty wire arch strain gauges (refer to Fig. 4(a)) were used to measure the concrete stra.in at various locations, Twenty electrical resistance

H{IIl)

..J , .. ".\ L. ___ -, • • ___ --'I ".". I ISM DU111 - 12143 mmj 11743 ...... 1 [S841D1J1) I L-------------------------,t.J~~lr------------------------~

Reinforcement Details of the First and Second Bridge Decks

Fig. I-Reinforcement details for three bridge decks.

Table 1--Bridge deck test matrix

Bridge Steel Bottom reinforcement Top reinforcement

deck type Transverse Longitudinal Transverse Longimdinai

First HP No.5 al6.75 in. No.5 at]O in. NO.5 at 6.75 in. No.4 at 14 in. (No. 16at 170 mm) p =0.54% (No. 16 at 250 mm) p = O.36~Q (No. 16at 170mm)p = O.54% (No. t3 al360 mm) p = O. t7%

Second Grade No.5 at 6.75 in. NO.5 at 10 in. NO.5 at 6.75 in. No.4 at 14 in. 60 (No. 16 at 170 mm) p = 0.54% (No. 16 a1250 mm) p = 0.36% (No. 16 at 170 mm) p = 0.54% (No. 13 at 360 mm) p = 0.17%

Third HP NO.5 at 10 in. No.5 at 10 in. No.5 at 10 in. No.4 at 14 in. (No. 16 at 250 mm) p = 0.36% (No. 16 at 250 mm) p = 0.361'° (No. 16 at 250 mm) p = 0.36% (No. 13 at 360 mrn) p = 0.17%

ACI Structural Journal/January-February 2008 79

strain gauges of 120 ohm and 0.24 in. (6 mm) gauge length were attached to selected reinforcing bars to detennine the strains in these bars. Data were electronically recorded by a data acquisition system. Figure 4 shows the locations of the wire arch strain gauges used and establishes the notation used hereafter.

A i

"" "" "" .,

'" '" , ,

/' -/

~

Strain (%)

1.200 ~&h. rfonn.nc.

'."" .. Grad,60 ~

~ .. j • ."

""

" Fig. 2-Stress-strain characteristics of HP and Grade 60 steel bars.

Fig. 3-(a) Test setup; and (b) first bridge deck before testing. (Note: I in. = 25.4 mm; 1ft = 304.8 mm.)

T8&B'pr !. ~L

",. ....L

17&B,'jV"" + E 32S

E ,...,-+ M

" .. HaM T6&610 14" <5

"" JL -rJL JL -v- -v- ... -v-

" 1'" ,J\!!& 83 TS& 61 .. " ~ i

)2.5'

1 t -,,-TIS 82 Notation

r & 8: Top & Bottom

321" TI8Bl 6: Bottom

E 2"·10" (6655mm) )

(a) Locations a/wire arch slrain gauges

left Span Righi Span

. ~ ~ • '""

'I " Transverse Direction . •

~ I I . d 1> - " -" I i .. load Plate ~ load Plate

<3 (20" x 10') "

(20" x 10"') " = .. " .3 .~ ~ :> ~

T

1 E 2,'·10·C6655nYn) )

(b) Schematic pial!

Fig. 4-Locations of wire arch strain gauges and notation for three bridge decks. (Note: I in. = 25.4 mm; I ji = 304.8 mm.J

80

Test results Test results were analyzed to critically examine the

performance of bridge decks reinforced with HP steel as main reinforcement compared with the behavior of bridge decks reinforced with Grade 60 steel. Detailed test results can be found elsewhere3

Load-deflection behavior-The three bridge decks were subjected to loading and unloading to load levels of 50. 100, and 150 kips (222, 445, and 667 kN) per span, and then to failure. The load-deflection envelopes up to failure for the three tested bridge decks are shown in Fig. 5. It should be noted that the deflections shown in Fig. 5 are measured at the center of the respective deck span directly under the applied load. It can be seen from Fig. 5 that the first bridge deck reinforced with HP steel using the same reinforcement ratio as used for the actual bridge exhibited smaller deflections than that of the other two bridge decks at the same load level. The slightly higher stiffness of the first deck is likely due to the higher concrete compressive strength and to the higher strength of HP steel. Despite the lower reinforcement ratio used for the third bridge deck (33% less than that of the first two decks), it was capable of achieving the same ultimate load-carrying capacity as the second bridge deck reinforced with the Grade 60 steel. This behavior is attributed to the higher tensile strength of HP steel. The slight increase of the deflection measured for the third bridge deck compared with the second deck is due to the use of less steel and to the slight reduction of the modulus of elasticity of HP steel at high stress levels.

The deflection profiles along the transverse direction for the right span of the second and third bridge decks are shown in Fig. 6. It should be nored that the deflection profiles are plotted for the last loading cycle only. Therefore, residual deflections are shown at the beginning of the loading cycle

250

200

50

o

250

200

'[ ;§.' 50 ~ .3

o

0

0

Mid-Span Vertical Deflection (mml 13 25 38 51

.. , -.'

64

:~ f'\.' .. ;.

:

I '- ~""""'_-1 , : / : ' ,

f /~ ..... "''' ~o SeMctI t.o.d r .. Fnt Bridge Oeck-MMFX

- $«ond Bridpe DedI-G60 _I - -Trinl BridGe "DeI:JI-MMFX

0.5 1.5 2 2.5 Mid-Span Vertical Deflection (in.)

(a) Leji span Mid-Span Vertical Deflection (mm) 13 25 38 51 64

-'. .' .- .' -

// ,{' ,

/' \ 1 . , : / I

c: 100 • ~ '" IF \ I

1

50

o o

AASHTO L FO Se~ic. Load I: .. Fif$l &id{Ie Ooek·MMFX -5l'!l:md Bridlili Oqck.G~.~

1 - Third Br Deck-MMFX

0.5 1 1.5 2 Mid-Span Vertical Deflection (In.)

(b) RighI span

2.5

1000

800 Z ".

600 ~ .3 c

400 ~

'" 200

o

1000

800 Z ".

600 'g .3 c

400 a. '"

200

o

Fig. 5- Load-deflection envelopes of three bridge decks.

ACI Structural Journal/January-February 2008

(zero load). The deflection profiles indicate that the maximum deflection occurred at midspan under the applied load. The deflection profiles also show that the deflection behavior of the deck reinforced with reduced amount of HP steel is very similar to that of the deck reinforced with conventional Grade 60 steel.

...,.0:5 t---------------t 13-

.!.

-2.5 ·64 , o 15 ~ ~ M n 00 1M lW 1~

Olstanc. from longitudInal CL of Bridge Deck (in.)

(a) Second bridge deck

c · 20" l' I

15 ~ 45 ~ H 90 1M 1m 1~ Distance from Longitudinal CL of Bridge DIck (in.)

(b) Thirdbridgedeck

Fig. 6--Transverse deflection profiles for right spans of second and third bridge decks. (Note: J in. = 25.4 mm.)

• -2.5

_. -~~

__ SpanL.o.xl " O~I~

__ Sp,an lc.3d "50~1pos(222 ~N)

_Sp.an Load_ lQO klp$ (~5 kN) --Spa.n I.00<I - 150 1tipoI\687 kN)

_~lo.3Ol . HIOt1ps(M911H)

s~ LOItCI .. :tIN q,.(901 i<l'I}

Distance From Transverse Cl of Bridge Olck (In.)

(a) Second bridge deck

32.5" 3U' ¥* e

6 1 20 40 60

Olliance From Transverse CL of Bridge Deck (In.,

(b) Third bridge deck

Fig. 7-Longitudinal deflection profiles for righT spans of second and third bridge deck. (Note: J in. = 25.4 mm.)

ACI Structural Journal/January-February 2008

Deflection profiles in the longitudinal direction for the right span of the second and third bridge decks are given in Fig. 7. It should be noted that the deflection profiles are plotted for the final loading cycle only. Accordingly, the deflections shown in Fig. 7 include the residual deflections from previous loading cycles. The longitudinal deflection profiles demonstrate the curvature in the longitudinal direction, implying the two-way flexural behavior of typical bridge decks under concentrated loads. In addition, the deflection profiles illustrate that the deflection at the edge of the decks was very small. This indicates that selection of the length of the test models is adequate for carrying the total load and, therefore, representative to typical bridge decks.

Crack pattern- No cracks were observed up to a load level of 50 kips (222 kN) for any of the three bridge decks. The first visible top cracks occurred at a load level of approximately 60 kips (267 kN) for each deck. According to the AASHTO Specifications,5 an axle of the design truck consists of a pair of 16 kip (71 kN) wheel loads spaced 6 ft (183 mm) apart. Therefore, at a load level of21 kips (93 kN), which includes the dynamic allowance, the three bridge decks remained uncracked and the deflection at the service load level was identical for the three bridge decks. Therefore, reducing the amount of HP steel used in the third bridge deck did not alter the serviceability behavior.

As expected, negative flexural cracks formed before positive flexural cracks due to the higher induced negative moments at the middle support in comparison to the positive moment values at midspan. Positive moment flexural cracks were observed at load levels of 100 kips (445 kN) and became radial at a load of 150 kips (667 kN), as shown for the first bridge deck in Fig. 8(a) and (b), respectively. Formation of

J ,

·1 4i /

:1 <

~ ... .....,...

/

j

(b) AT load level of 150 kips (667 kN)

Fig. 8- Positiveflexural cracksforfirst bridge deck.

81

the radial crack pattern confirms the two-way mechanism under the effect of the concentrated applied loads. Further loading led to spreading and widening of the flexural cracks until the formation of flexural-shear cracks at the top surface of the deck close to the middle girder. The formation of the flexural-shear cracks led to a sudden drop in the load, as shown in Fig. 5. The flexural-shear cracks, however, formed symmetrically on both sides of the middle girder of the first bridge deck and allowed an increase of the load and finally failure of the deck by punching shear in both spans. For the second and third bridge decks, the flexural-shear crack occurred only on the left side of the middle girder. This allowed the load to increase until failure occurred due to punching shear in one of the spans only.

Mode aftailure- Due to the selection of sufficient length of the test model, the behavior of the bridge decks under the effect of the concentrated loads was two-way flexural mechanism followed by development of an arching action supported by membrane forces developed in the bottom layer of the reinforcement. Due to the continuity used in the test models, at the measured first peak of load for the first bridge deck, a sudden drop in the load resistance occurred due to the formation of the flexural-shear cracks along the top surface of the bridge deck on both sides of the middle girder. Further loading caused widening of those cracks associated with a slight increase in the load resistance until punching failure occurred under the applied concentrated loads. Punching failure of both spans occurred almost simultaneously at load levels of 229 kips (1019 leN) and 216 kips (961 leN), and corresponding deflections of 1.8 in. (46 mm) and 1.6 in. (41 mm) for the left and right spans, respectively. Figure 9 shows the first bridge deck at the conclusion of the test, where the punching areas under the loads and the shear cone at the bottom of the left span can be seen.

The behavior of the second bridge deck, reinforced with Grade 60 steel using the same reinforcement ratio, was similar to the first deck. At the measured peak load, a sudden drop in the applied load occurred due to the formation of the flexural-shear crack on the top surface of the bridge deck to the left side of the middle girder only. Failure of the left span was due to flexure-shear failure that prevented an increase in load sufficient to cause punching shear. The maximum measured load for the left span was 185 kips (823 leN) and the maximum deflection was 2.2 in. (56 mm). Failure of the right span was due to punching shear at :i load level of 204 kips (907 leN) and a corresponding deflection of 0.7 in. (18 mm).

For the third bridge deck, the right span failed in punching shear at a .load of 203 kips (903 leN) and a corresponding deflection of 1.0 in. (25 mm) before the failure of the left span. Formation of the flexural-shear crack in the left span caused a sudden drop in the applied load and prevented the capability to increase the applied load to induce punching shear failure in the left span. Failure of the left span was due to flexural-shear failure at a load level of 181 kips (805 leN) and the maximum measured deflection was 1.9 in. (48 mm). Figure 10 shows the second and third bridge decks at failure, where the punching area under the applied concentrated load at midspan and the flexural-shear crack formed within the vicinity of the middle girder are clearly visible.

Strain in concrete and steel-Based on the deformations measured by the wire arch strain gauges, concrete compressive strains were determined. Concrete strain is plotted for the final loading cycle only and hence includes the residual strain developed in previous loading cycles. The strain

82

obtained from the wire arch strain gauges located in the right span of the three bridge decks 14 in. (356 mm) from the centerline of the deck (T6 in Fig. 4(a)) are shown in Fig. 11. The measured strain from the second deck indicates that the compressive strain of the concrete in the vicinity of the punching area reached the limiting value of 0.002 typically

(b) Punching cone in left span

Fig. 9- Failure of first bridge deck.

(b) Third bridge deck

Fig. 10-Failure of second and third bridge decks.

ACI Structural Journal/January-February 2008

observed by others to cause punching shear failure6 -8 The first and third bridge deck compressive strains exceeded this value, however, reaching values of 0.0031 and 0.0036, respectively. To explain th is behavior, thc strain in the reinforcing steel has to be investigated.

The steel strains were measured using conventional electrical strain gauges attached to the reinforcing bars before casting. The strains measured in the bottom transverse steel bars of the right span of the first and second bridge decks are shown in Fig. 12. The steel strains are shown for the final loading cycle only, and therefore include residual strains developed in previous loading cycles. It can be seen from Fig. I2(a) that the steel bar at midspan of the first bridge deck did not reach the yielding strain of HP steel bars, as defined previously. On the other hand, Fig. 12(b) shows that the strain in the bar located at midspan of the second bridge deck exceeded the

StraIn (Inlln) ~.003 -0.002 -0.001

"'00

50 -First BlIdge'ii-~_";;"iF:FX-;--l--l---~~\--J 200 -SIlCOlld Bridge o.ck,.Grade 60

""'" .""" Strain (h\lcro-StralnJ

Fig. ii-Concrete compressive strain in vicinity o!punched area for three bridge decks.

250

200

~ 6150 ~

~ ~100

~ 50

o

250

200 . ~

~150 ~ .3 c WO a .,

50

o

o

o

Steel Strain (Inffn)

0.001 0.002 0.003 0.004 0.005 0.006

Edge Qu~rtl:rSpan C .....

I /) / I W/ I V / U# V

::y') Mld-Sp,n

/-

/ Mid·Span

[-aUrlerSPlMl -E<lga

1000

800 Z ;,:.

600 -g .3

400 ;

200

o

~ .,

1000 2000 3000 4000 5000 6000 ' Steel Strain (Micro-Strain)

(a) First bridge deck

Steel Strain (lnlln)

a.C)()1 0.002 0.003 0.004 0.005 0.006

Edge U~r1~r spL Mld.SpL

II // /j 1// II /

'I;; '/L Mld.Span

1: au.lter SpIn -Edge

1000

800 ~

600 ~ ~ c

400 ~ ., 200 ;;

o 0

o 1000 2000 3000 4000 5000 6000 Steel Strain (Micro-Suain)

(b) Second bridge deck

Fig. 12-Straill in bottom transverse steel for right spans of first and second bridge decks.

ACI Structural Journal/January-February 2008

yielding strain of Grade 60 steel. Collected data from the various strain measurements indicate that yielding was very localized in the vicinity of the concentrated load3

It is well established that the concrete in the vicinity of the shear cone is under a triaxial state of stress.6-8 For the second bridge deck, the steel in the vicinity of the punching cone had yielded before failure. Therefore, the steel bars were no longer restraining the concrete, and failure occurred when the concrete reached its peak stress at a corresponding strain of 0.002. For the first and third bridge decks, the transverse and longitudinal compressive stresses were continuously increasing up to failure because the steel in the vicinity of the punching cone did not yield. Therefore, the steel bars were still restraining the concrete and the concrete was still intact, thus maintaining the aggregate interlock across the shear crack. Consequently, the measured concrete strain was still increasing until the concrete crushed in the vicinity of the shear crack at a strain level much higher than 0.002, reaching values of 0.0031 and 0.0036 for the first and third bridge decks, respectively.

Predicted punching capacity-The predicted punching shear strengths for the three bridge decks according to two different design codes are summarized along with the measured values in Table 2. The design codes presented in Table 2 are the AASHTO SpecificationsS and ACI 318-05 .9

The design equations used for the predictions are Units in kips and in.

Units in Ib and in.

(2)

where Ve is the punching shear capacity of bridge deck; Pe is the ratio of long side to short side of loading plate; fe' is the concrete compressive strength; bo is the perimeter of critical section at a distance of d/2 from loading plate; d is the effective section depth; and r:J.s is the constant.

It is clear from Table 2 that the predicted values according to the AASHTO and ACI 318-05 design codes compare very well to the measured values for bridge decks reinforced with HP and Grade 60 steel.

ANALYTICAL MODELING The three bridge decks tested in this study were analytically

modeled using a finite element analysis program. 10 The concrete material model is based on the smeared cracking model. II As stated in the program manual, this model is a mechanics­based formulation using plasticity theory that permits incorporation of cracking and other concrete response,

Table 2-Predicted and experimental punching shear strength

Transverse Punching shear strength , kips (kN) Bridge Sleel reinforcement deck type rmio, % Experimental AASHTO ACI

First HP 0.54 229 (1019) 229 ( 1019) 230 (1023)

Second Grade 60 0.54 204 (907) 184 (818) 184 (818)

Third HP 0.36 203 (903) 199 (885) 200 (890)

83

characteristics. to Within the concrete constitutive model, cracking and all other forms of material nonlinearity are treated at the finite element integration points. Cracks are assumed to form perpendicular to the principal tensile strain direction in which the criterion is exceeded and they are allowed to form at each material point. When cracking

" ~

o

(a) Mesh usedfar analytical model

Mld-$pan Vertical O,f1ect!on (mm) 12.1 25.4 38.1 50.8 63.5

2~r---~~--r----r----r---~

1000

200

:; 150 - /~t~~jI"i [ : :::: \ . , \ : ,

BOO ~

600 ~ 3 ~

~

• 100 -~ --- - - .. -- --.' -. ~ -:~ ---- - 400 :i ! 0

~

0 0

o

, \ ' It -----: ... -- : -- -- -:---l:::~~~ 200

, '-::~~

0.5 1.5 2 Mld-$pln Vertical Defteetlon (In.)

(b) First bridge deck

Mld,Spln V,rHeal Defleetlon (mm) 13 25 38 51

2.5

64 250 t-- -t----+---+---+-- --j

, 1000 200 oJ _ .. , .• L _ __ . _ ' _ _ _ _ _ L _ __ _ ••

)~ , , , '!'-; -~-\i~' ~'~ :.-~ ::~'-' :7, ~. __ t:: i ---

, \ ' , , .-. ---~-- - 1 -~--- - .... -----~.:- -- - - 400 :i

, I I ," , It

800 ~

50 --- - -:----\.~-----:-- -..: --·t..ftSpan 200 --Rlghj~

o o

o

- ","""" 0

,

0.' 1.5 2 MId...sPiln Vertical OenKtion (In.)

(C) Second bridge deck

Mld...sp,a.n Vertical Den,ctlon (mm) 13 25 38 51

, ,

2.5

64

1000 200 __ _ -' ___ _ .L '"""" __ '- ___ .J _ ___ L ___ _

, \ " I ."/ • • , ' \ I , 800 ~

-- :)7- - -~::.~~.:.::-. ~ ::.i:.;._ -f ---- 600 l I I I I : , _ --0 -- -- ... -- ., - '--- - ... -- '71" .. ---- 400 a.

: :. \ : : /.'~ : I/)

50 ----:---- t- --t-.. ::.---),('- ... ~$pMl 200

o o

I I ' -- R.IOhI~n

- ""-CAP 0

0.4 0.8 1.2 1.6 Mld-$pan V&rtlcal o.flKtion (In.)

(d) Third bridge deck

2.4

Fig. 13- Mesh, analytical, and experimental load-deflection envelopes for three bridge decks.

84

occurs, the normal stress across the crack is reduced to zero and distribution of stresses around the crack is recalculated. Concrete modeling also included residual tension stiffness for the gradual transfer of load to the reinforcement during crack formation. In addition, the program accounts for the reduction in shear stiffness due to cracking and further decay as the crack opens. The reinforcement is modeled as individual sub-elements within the concrete elements. The stiffness of the bar sub-element is superimposed on the concrete element stiffness in which the bar resides. The anchorage loss is modeled as an effective stiffness degradation of the bar as a function of the concrete strain normal to the bar.

A three-dimensional analysis was conducted for the three bridge decks using 20-node hexahedral continuum elements with quadratic isoparametric displacement interpolation. Only 114 of the deck was modeled due to its symmetry about both axes. A convergence study was conducted including mesh size (number and size of elements) and loading increment. The depth of the deck was divided into five layers within its thickness with a total number of elements of 1040 for the deck and supporting girders, as shown in Fig. l3(a).

Analytical results The predicted and experimental load-deflection envelopes

for the three bridge decks are compared in Fig. 13. It can be seen that the predicted load-deflection behaviors of the three bridge decks compared very well with the measured values. The initial and post-cracking stiffness were accurately predicted by the analytical model. Tn addition, the ultimate load was also reasonably predicted considering the fact that the two spans of the second and third bridge decks failed in two different modes. The predicted ultimate deflections, however, were slightly less than the experimental values, which is due to the fact that the program was terminated when the concrete strain in the compression zone reached the value of 0.003. For validation purposes, the portion of the first bridge deck that failed due to punching shear was cut using a concrete saw to reveal the failure mode, as shown in Fig. 14(a).

. ..- , - . , .' -~~ ,

'. (a) Experimented

(bjANACAP

Fig. 14- Principal strain conlOurs at failure for first bridge de ck.

ACI Structural Journal/January-February 2008

The strain contours shown in Fig. 14(b) depicting the punching shear cone match very well to the actual failure mode.

PHASE 11: TENSILE STRENGTH OF HIGH-PERFORMANCE STEEL BENT BARS

Test specimens and test setup A specially-designed specimen was used to evaluate the

effect of bending HP steel bars on their tensile strength; the specimen used is shown in Fig. 15(a) for NO.5 (No. 16) bars. The specimen consisted of two concrete blocks used to anchor the two ends of the bent bar in the sbape of a closed stirrup. The two sizes of steel reinforcement used for this

(a) Schematic plan view v.f henl bars fest setup

(b) Test setup

Fig. I5-Schematic details and test sel!tp for bent bar specimens.

(b) Bonded

Fig. 16- Location offailure of No. 4 (No. 13) debonded and bonded bent bar specimens. .

ACI Structural Journal/January-February 2008

phase were NO.4 and No.5 (No. 13 and No. 16) bars with two specimens for each size. The bend was 90 degrees according to ACI 318-05 ,9 and the lengths of the HP steel stirrups were selected based on the dimensions of the concrete blocks, the dimensions of the hydraulic jack, and the load cell placed between the concrete blocks. The concrete blocks were heavily reinforced with conventional Grade 60 stirrups to prevent premature failure . The blocks were cast using wooden forms that were special1y designed to accommodate the anchored ends and temporarily braced to prevent bending of the exposed portion of the bar between the two blocks before testing.

Two different configurations were used to debond the steel from the concrete within the same specimen. In the first configuration, the stirrup was completely debonded in the left concrete block by using thick rubber tape. In the second configuration, only the straight portion of the stirrup was debonded in the right concrete block to transfer the tension force directly to the bent portion of the bar. This study is a continuation of a previous study 12 that was conducted in 2002 at North Carolina State University using the same specimen. In the previous study, only the straight portions of the stirrups were debonded rather than debonding the entire U-shaped section of the bar for the specimens reported in this paper. Debonding of the entire U-shaped portion allows relative movement of the bar with respect'to the concrete. This movement allowed pure testing of the bent bars rather than representing typically bonded bars and sti rrups used for concrete structures. The results obtained from the previous study are compared with those obtained from the work reported herein to demonstrate the effect of bond to the concrete on the tensile strength of bent bars.

The test setup, shown in Fig. 15(b), consists of a 120 kips (534 kN) capacity hydraulic jack, a 150 kips (667 kN) capacity load cell, and four linear potentiometers to measure

160

150

_ 120

~ ~ 90

'so

30

o o

160

150

120 .. '" . 90

~ 0

60

30

0 0

( I

,I

f

....-('" \ A! ExIe~sL.oll!r 1Iofl<i8(l _ R""",yed _

"""""'"

2 5 6

Strain f%)

(a) No.4 (No. 13) bars

--\ "'."".,

2

1 ~~

3

Strain f%)

r ,--R( fnO'o'ed

5 6

(b) No.5 (No. 16) bars

1,200

1.000

BOO .. ~

~ 600 ~

~ 400 0

200

o

1.200

1.000

600 ~

!!. 600 . g 400 0

200

o

Fig. 17-Stress-strain relationship for debonded and bonded bent bars.

85

the relative displacement between the two blocks. The hydraulic jack and the load cell were centered between the two branches of the stirrup to ensure equal distribution of forces in each branch. An axial mechanical extensometer of 2 in. (S 1 mm) gauge length was mounted on the exposed portion of the stirrup to measure the elongation during loading. A data acquisition system was used to electronically record the readings of the load cell , the potentiometers, and the extensometer.

Test results Failure of all the specimens for both bar sizes occurred at

the bent portion of the stirrup that was debonded from the concrete in Ule left block, as shown in Fig. 16(a). This failure location was confinned by visual inspection after sawing the concrete blocks at the location of failure as shown in Fig. 16(a). It should be noted that failure of the specimens tested in the previous study l2 (with bonded bends) occurred at the exposed straight portion of the stirrup, as shown in Fig. 16(b). Typical stress-strain characteristics of No.4 and No. S (No. 13 and No. 16) debonded bent bars along with those of bonded bars, are shown in Fig. 17. The stress-strain characteristics of the bonded and debonded bent bars indicate that their behavior is sintilar to the bonded bars, including the linear and nonlinear behavior up to a strain value of 1.5%. Testing of the debonded bent bars, however, emphasizes the induced residual strains due to bending of the bars that reduced both the strength and the strain at ultimate. This behavior reflects the well-established phenomenon of the stress concentration at the bend location due to the bending process. It should be mentioned that in actual structures, bent bars are always bonded to the concrete, enabling them to reach the ultimate stress and strain of straight bars as reported previously.12 Based on these tests, the results suggest that HP steel bars can be bent up to 90 degrees without affecting their ultimate strength or strain, provided the bend is fully encased and bonded to concrete.

CONCLUSIONS AND RECOMMENDATIONS In light of the test results, the following conclusions can be

drawn: I. The ultimate load-carrying capacity of the three bridge

decks investigated in this study was on the order of 10 times the service load prescribed by the AASHTO Specifications5 ;

2. Punching shear was the primary mode of failure for the three bridge decks. Due to continuity used in the test models, flexural-shear failure was observed as a secondary mode of failure;

3. The cracking load of the three tested bridge decks was more than twice the service load prescribed by the AASHTO Specifications5 Hence, under service load level, the three bridge decks behaved as uncracked sections. Therefore, using 33% less HP steel should not alter the serviceability behavior of concrete bridge decks;

4. The bridge deck reinforced with 33% less HP steel developed the same ultimate load-carrying capacity as that reinforced with Grade 60 steel. This performance is attributed to the higher strength of HP steel compared with Grade 60 steel; and

S. Behavior of bonded HP steel bent bars is sintilar to the behavior of straight bars. Debonded bent bars exhibit sintilar behavior to straight bars, including the linear and the nonlinear behavior up to a strain of 1.5%. Its ultimate

86

strength, however, is reduced by 6% and its ultimate strain by 70%.

Based on Ihe research findings, the following design guidelines can be recommended:

1. Substituting HP steel directly for conventional Grade 60 steel in a design, as was done for the actual bridge built in Johnston County, NC, in 2004 and the first test specimen is a conservative approach;

2. HP steel can be used as the main flexural reinforcement for cast-in-place concrete bridge decks at a reinforcement ratio corresponding to 33% less than that required for Grdde 60 steel. Therefore, design of reinforced concrete bridge decks using HP steel can use a yield stress of 90 ksi (621 MPa) for the HP steel bars;

3. Reduced reinforcement ratio of HP steel shall satisfy all ntinimum reinforcement ratios prescribed by the AASHTO Specifications.5 In addition, the reduced reinforcement ratio of HP steel must comply with the crack control requirement of the AASHTO Specifications5; and

4. HP steel bars can be bent up to 90 degrees without reducing their ultimate strength or strain provided that the bend is fully encased and bonded to concrete.

ACKNOWLEDGMENTS The authors gratefully acknowledge the support of the North Carolina

Department of Transportation (NCDOT) for sponsoring this research project. Thanks are due to MMFX Tectmologies Corporation for supplying the MMFX steel; to CC Mangum Inc. of Raleigh, NC, for helping to cast the three bridge decks; and to StcclFab of Charlotte, NC, for donating the steel sections used in the testing frames. Special thanks are extended to C. Walter for taking part in the second phase of the experimental program. The authors would also like to thank J. Atkinson and B. Dunleavy at the Constructed Facilities Laboratory for their help.

REFERENCES 1. ASTM A615/A615M-06, "Standard Specification for Deformed and

Plain Carbon-Steel for Concrete Reinforcement." ASTM International, West Conshohocken, PAt 2006. 6 pp.

2. ASTM AI035/A1035M-07. "Standard Specification for Deformed and Plain, Low-Curbon, Chromium, Steel Bars for Concrete Reinforcement," ASTM International, West Conshohocken, PA, 2007, 5 pp.

3. Rizkalla, S. H.; Zia, P. ; Seliem, H. M.; and Lucier, G., "Evaluation of MMFX Steel for NCDOT Concrete Bridges," Research Report, North Carolina State University, Constructed Facilities Laboratory, Raleigh, NC, Dec. 2005. 109 pp.

4. ASTM A370-07. "Standard Test Methods and Definitions for Mechanical Testing of Steel Products," ASTM International, West Conshohocken, PA, 2003, 47 pp.

5. American Association of State Highway and Transportation Officials, "AASHTO LRFD Bridge Design Specifications." 3rd Edition, Washington, DC. 2004.

6. Kinnunen, S .. and Nylander, H.. "Punching of Concrete Slabs without Shear Reinforcement," Transactions of the Royal Institute of Stockholm, Stockholm. Sweden. 1960, 11 2 pp.

7. Marzouk, H., and Hussein, A., "Punching Shear Analysis of Reinforced High-Strength Concrete Slabs," CalUldian Joumal of Civil Engineering, V. 18, 1991. pp. 954-963.

8. Mufti. A. , and Newhook, J., "Punching Shear Strength of Restrained Concrete Bridge Deck Slabs," ACI Structural Journal, V. 95, No.4, July-Aug. 1998. pp. 375-381.

9. ACI Committee 318, "Building Code Requirements for Structural Concrete (ACI 318-05) and Commentary (318R-05)," American Concrete Institute. Farmington Hills, MI, 2005. 430 pp.

10. James, R. G., "ANACAP Concrete Analysis Program Theory Manual," .Version 3.0, Anateeh Corporation, San Diego, CA, 2004.

II. Rashid, Y. R., "Ultimate Strength Analysis of Prestressed Concrete Pressure Vessels," Nuclear Engineering and Design, 1968, pp. 334-344.

12. EI·Hacha, R., and Rizkalla, S. H., "Fundamental Material Properties of MMFX Steel Bars," Research Report 02·04, North Carolina State University, Consuucted Facilities Laboratory, Raleigh, NC, July 2002, 60 pp.

ACI Structural Journal/January-February 2008