Development of shrinkage limit specification for high performance concrete used in bridge decks

Accepted Manuscript Development of shrinkage limit specification for high performance concrete used in bridge decks Tengfei Fu, Tyler Deboodt, Jason H. Ideker PII:

S0958-9465(16)30170-6

DOI:

10.1016/j.cemconcomp.2016.05.015

Reference:

CECO 2651

To appear in:

Cement and Concrete Composites

Received Date: 21 August 2015 Revised Date:

16 March 2016

Accepted Date: 19 May 2016

Please cite this article as: T. Fu, T. Deboodt, J.H. Ideker, Development of shrinkage limit specification for high performance concrete used in bridge decks, Cement and Concrete Composites (2016), doi: 10.1016/j.cemconcomp.2016.05.015. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Development of Shrinkage Limit Specification for High Performance Concrete Used in Bridge Decks

RI PT

Tengfei Fu1*, Tyler Deboodt1, Jason H. Ideker1 1. School of Civil and Construction Engineering Department, Oregon State University, Corvallis, Oregon 97331, USA Abstract:

Early-age cracking of high performance concrete (HPC) structures, in particular bridge decks,

SC

results in additional maintenance costs, burden on serviceability, and reduced long-term performance and durability. The causes behind cracking in HPC are well known and documented

M AN U

in the existing literature. However, appropriate shrinkage limits and standard laboratory/field tests are not clearly established in either the technical literature or in specifications. The purpose of this research was to provide shrinkage threshold limits for specifications which allow proper criteria to ensure crack-free or highly cracking-resistant HPC. The restrained ring test (ASTM C1581) was used to identify the cracking potential of 14 different HPC mixtures. By comparing free shrinkage (ASTM C157, 75 × 75 x 285 mm specimen) and restrained shrinkage tests results,

resistance.

TE D

a free shrinkage limit of 450 microstrain at 28 days was proposed to ensure satisfactory cracking

AC C

EP

Key Words: high performance concrete, bridge deck, drying shrinkage, cracking

1

Corresponding author: [email protected], Tel +1-5412242096, 101 Kearney Hall, Oregon State University, Corvallis, Oregon 97331

1

ACCEPTED MANUSCRIPT

1. Introduction Among 605,000 bridges across the country monitored by the United States Department of Transportation (USDOT), 26.9% of them were reported “structurally deficient” (bridge having

RI PT

major deterioration and cracks that reduce its load-carrying capacity) or “functionally obsolete” (bridge no longer meeting the current design standards) in 2010 [1]. In 2013, a grade of C+ was given to the national bridge system by the American Society of Civil Engineers (ASCE), and an annual investment of $20.5 billion was estimated to improve current bridge conditions [2]. In

SC

2003, a nationwide state DOTs survey conducted by the Michigan DOT [3] on early-age bridge deck cracking issues indicated that 78% of the 31 responding states identified transverse cracking, which indicates the presence of drying shrinkage. Cracking, especially at early age, in

M AN U

high performance concrete (HPC) may result in a significant decrease in concrete durability and service life of the structure. Concrete bridge decks demand qualities from HPC such as low permeability, high abrasion resistance, superior durability, and long design life. To meet these requirements, concrete used for bridge decks is usually produced with a low water to cementitious material ratio (w/cm), typically less than 0.40, high overall cement contents,

TE D

inclusion of supplementary cementitious materials (SCMs, e.g. silica fume, fly ash and slag), and smaller maximum aggregate size (due to reinforcement constraints). All these features in the mixture design make HPC bridge decks inherently susceptible to shrinkage and increased cracking risk [4, 5]. A comprehensive report on factors that affect shrinkage of hardened

EP

concrete can be found in literature [6].

From a concrete materials perspective, it is a significant challenge to overcome cracking risk is

AC C

to reduce the shrinkage, and ultimately the stresses generated as a result of such shrinkage. To mitigate cracking issues due to shrinkage, many methods have been studied and documented. During the last 15 years, internal curing with pre-wetted fine lightweight aggregate (FLWA) has been proven to be effective in mitigating concrete cracking potential [7-9], and has been steadily progressing from laboratory research [10-14] to field applications [11, 15-18]. Another focus over the last 20 years has been shrinkage reducing admixtures (SRAs), which have also proved to be successful in reducing shrinkage induced cracking [19-25]. Some other techniques that have proven effective in controlling cracking in concrete bridge decks are fiber reinforced concrete [26], shrinkage-compensating concrete [27], and special construction practices (i.e. 2

ACCEPTED MANUSCRIPT

extended curing duration, controlled slump, and proper environmental conditions during placement). Moreover, the type of aggregate has a significant impact on the amount of shrinkage in concrete. Research showed that sandstone aggregate concrete exhibited the highest drying shrinkage, while concrete made from limestone aggregate proved to be the most cracking-

RI PT

resistant [28, 29]. Other authors have shown that higher aggregate content (in volume or/and in maximum size) could reduce shrinkage due to relatively low cement paste content [29, 30].

The free shrinkage test specified in ASTM C157 [31] is a simple and widely used test to assess

SC

shrinkage of a given concrete mixture. Due to its simplicity, the free shrinkage limits have been set up based on ASTM C157 test by many agencies, including Unified Facilities Guide Specifications (UFGS) [32] and some state DOTs [33-36]. The Federal Highway Administration

M AN U

(FHWA) has also implemented a single value shrinkage limit in the new specifications (FP-14) [37]. . Table 1 gives a brief summary of free shrinkage limits used by different agencies in United Stated. There is also shrinkage requirement in CEB-FIB[38], New Zealand[39], Canada[40], and UK[41].

However, there is no shrinkage threshold limit commonly agreed

upon to ensure a crack-free or highly cracking-resistant concrete

TE D

Table 1 Summary of shrinkage control limit(s) by different agencies in U.S. Agency (Date)

Shrinkage limit(s)

UFGS*[32]

500 microstrain maximum. For OPC: 500 microstrain at 28 day; For HVFA**: 500 microstrain at 56 day. Varied (mixture and age specified, ranging 350 to 800 microstrain at 28 day)

EP

FHWA [37]

AC C

Virginia DOT [33, 34] New Jersey DOT [35]

450 microstrain at 56 day.

Washington DOT [36]

320 microstrain at 28 day

*UFGS – Unified Facilities Guide Specifications, for military service constructions; **HVFA – High volume fly ash, minimum 50% class F fly ash.

To assess the cracking potential of HPC, the restrained ring test has been used by many researchers [35, 36, 42-47] in the last decade. This test had been standardized as ASTM C1581 [48] and AASHTO T334 [49] (formerly known as AASHTO PP34-98). It is a practical tool to 3

ACCEPTED MANUSCRIPT

evaluate cracking potential of concrete and mortar, especially after the quantitative analysis of this test has come into existence by implementing strain gauges to quantify the stress rate development of the specimens [50]. Based on either time-to-cracking (ToC, time in days between initiation of drying and crack formation in the concrete ring) or stress rate (calculated

RI PT

from strain gauge recording), ASTM C1581 suggests a cracking potential classification, as shown in Table 2. If a connection were made between free and restrained shrinkage tests results, a shrinkage limit could be identified to assess the cracking potential of given HPC mixtures.

SC

Table 2 Cracking potential classification (Based on stress rate at time-to-cracking). [43, 48] Stress Rate at Cracking, S, MPa/Day

Potential for Cracking

0 < tcr ≤ 7 7 < tcr ≤ 14 14 < tcr ≤ 28 tcr > 28

S ≥ 0.34 0.17< S < 0.34 0.10 < S < 0.17 S < 0.10

High Moderate-High Moderate-Low Low

M AN U

Time-to-Cracking, tcr, Days

Cracking of high performance reinforced concrete structures, in particular bridge decks, is of

TE D

concern to the Oregon Department of Transportation (ODOT) and in fact most Departments of Transportation. Cracking at early ages (especially within the first year after placement) results in additional costs and a significant maintenance burden. A commonly agreed upon testing method and subsequent shrinkage threshold limit will ensure a higher degree confidence in specifying

EP

and receiving high-cracking resistant or crack-free concrete. This research is part of a comprehensive effort to reducing cracking issues in HPC bridge decks. In total, 14 mixtures were investigated, including different curing durations, shrinkage reducing strategies (internal curing,

AC C

SRAs, or synergetic effect), and different aggregate sources. By comparing free and restrained shrinkage tests results, a free shrinkage limit was proposed to ensure a satisfactory cracking resistance. The testing protocols could be used to establish shrinkage limits for bridge decks made with HPC in other locations using “local” materials. It should be noted that the main focus of the proposed study was the effect of material properties on shrinkage and cracking of HPC for bridge decks. In the field, there are many other issues that may affect cracking, including structural effects (loading and restrain conditions), temperature variations, construction practices (finishing, curing, etc.). More detailed information can be 4

ACCEPTED MANUSCRIPT

found in literature [28, 51, 52]. Results presented in this study was meant to help with materials (mixture design) selection.

2.1

RI PT

2. Experimental Materials

The cementitious materials used in this research were an ASTM Type I/II ordinary Portland

SC

cement, an ASTM C618 Class F fly ash, and an ASTM C1240 silica fume. The oxide analysis of the cementitious materials is shown in Table 3.

Cement

SiO2, %

20.51

Al2O3 , %

4.72

Fe2O3 , % CaO, % MgO, % K2O, %

3.23 64.21 0.8 0.29

SO3, % LOI, % C3S, % C2S, %

AC C

C3A, % C4AF, %

Silica Fume

55.24

90.2

15.77

0.5

6.27 10.2 3.64 2.08

1.5 1.0 2.6 0.8

0.49

1.51

-

2.7 2.62 61.51

0.7 0.23 -

0.1 2.9 -

12.4

-

-

7.03

-

-

9.84

-

-

EP

Na2 O Eq., %

Fly Ash

TE D

Composition

M AN U

Table 3 Cement and fly ash oxide analysis (wt %)

An ASTM C494 Type F polycarboxylate-based high-range water reducer was used to achieve consistent workability (target 150 mm slump). An air-entraining admixture was also added to achieve a target air content of 5±1.5 % to ensure proper freeze/thaw resistance. One SRA (hexylene glycol type), which is compatible with the air entrainer, was used in some mixtures at a dosage rate of 2 % of the total cementitious materials by mass.

5

ACCEPTED MANUSCRIPT

The coarse and fine aggregate used in this study were from several different sources. Four local siliceous aggregate sources were used. Three (Local A, B, and C) were the local river gravels and river sands from different areas in the state of Oregon. Another (Local D) was manufactured local siliceous gravel and sand, known as high strength aggregate. A siliceous limestone was also

RI PT

used. The maximum size of all aggregates was 19 mm. Petrographic study was done and was not presented for brevity reasons. In addition, in some of the mixtures, a fine lightweight aggregate (FLWA) of expanded shale was used as a partial replacement of the normal sand to provide internal curing. Determination of the absorption capacity and desorption of the FLWA, as well

SC

as the replacement level can be found in [53]. The replacement level of FLWA was based on the Bentz Equation [54] and the calculation can be found in [55]. The properties of the aggregates

M AN U

are shown in Table 4.

In addition, a proprietary mortar mixture (MasterEmaco S 440MC Repair Mortar, formerly LA Repair Mortar) was evaluated. This particular mortar was used in the field as crack sealing mortar by Oregon DOT.

Absorption Capacity (%)

Desorption Capacity (%) Fineness Modulus at 84% RH

Local sand A

2.41

3.08

-

3.0

Local gravel A

2.44

2.58

-

7.1

2.54

2.58

-

2.9

2.59

2.27

-

7.5

2.48

3.46

-

2.6

Local gravel C

2.53

3.17

-

7.2

Local sand D

2.58

2.74

-

3.3

Local gravel D

2.62

2.04

-

6.7

Limestone

2.68

0.58

-

6.5

Expanded shale

1.55

17.50

16.0

2.7

Local sand B Local gravel B

AC C

Local sand C

TE D

Specific Gravity

EP

Table 4 Aggregates properties (as received)

2.2

Methods

6

ACCEPTED MANUSCRIPT

Fresh properties (slump, air content, unit weight, and temperature) were measured for quality control purposes. The target slump was 150 mm, and the target air content was 5±1.5 %. A pressure air meter was used for concrete without lightweight aggregate (pressure method, ASTM C231), and a roll-a-meter was used for concrete with FLWA (volumetric method, ASTM C173).

RI PT

Fresh concrete temperature was measured at the end of each mixing using an infrared thermometer. Mechanical properties, including compressive strength, splitting tensile strength, and modulus of elasticity were tested on ∅100 mm × 200 mm concrete cylinders, which were

SC

cured in a moist room (23 ± 2 °C and 100 % RH) until testing at an age of 28 days.

Free drying shrinkage was monitored using the ASTM C157 test on concrete prisms (75 mm × 75 mm × 285 mm), as shown in Fig. 1 (a). The specimens were removed from the mold 24 hours

M AN U

after casting. The specimens were transferred in a moist room for the desired curing duration (i.e. 3 days or 14 days in this study). Upon the end of curing duration, the specimens were moved to a drying environment of 23 ± 2 °C and 50 ± 4 % RH. During drying, length change of the prisms was monitored up to 180 days. The mass change was also recorded.

(b)

AC C

EP

TE D

(a)

Fig. 1. (a) Free shrinkage test (ASTM C157), (b) restrained shrinkage test (ASTM C1581)

All restrained shrinkage ring specimens were prepared according to ASTM C1581, as shown in Fig. 1(b). Detailed information on specimen dimension can be found in literature [53]. Four strain gauges were attached 90° apart to the inner surface of the steel sing at mid-height. The strain was recorded using a data acquisition system. After concrete placement, all specimens were immediately moved to the environmental chamber and covered with wet burlap and

7

ACCEPTED MANUSCRIPT

polyethylene sheeting until 24 hours of age. The outer PVC rings were then removed and all concrete surfaces were covered with wet burlap and the polyethylene sheeting for the entire desired curing duration. During the curing duration, to maintain the moisture condition, the burlap was re-wetted every 48 hours to maintain a 100% RH within the polyethylene sheeting

RI PT

cover. At the end of the desired curing period, the burlap and polyethylene sheeting were removed, and all concrete rings were exposed to the controlled drying condition (23 ± 2 °C and 50 ± 4 % RH). Next, the top concrete surface was sealed with a waterproof silicone sealant to ensure drying only circumferentially. All rings specimens were inspected every 24 hours until

SC

cracks were observed in all three replicates. By examining the strain gauge recording, the exact time of cracking can be determined. The time between exposure to drying and cracking is termed

M AN U

time-to-cracking (ToC, days), which is an important parameter to evaluate the cracking resistance of the tested concrete. According to the strain gauge reading, an average stress rate (MPa/day) at cracking in the concrete could also be calculated as per ASTM C1581, and then used as another parameter in cracking evaluation. For practical reasons, all tests were terminated at 60 days regardless of whether cracks were observed or not. For each mixture design, three rings specimen were tested following the recommendation of ASTM C1581, except FLWA-2

TE D

where one ring specimen was lost due to a malfunction of the concrete mold. For all three ring specimens, concrete materials were from the same mix to minimize variation. 2.3

Mixture Design

EP

In total, 14 concrete mixtures with different wet cure durations, cementitious content, aggregate sources, shrinkage reducing methods, and w/cm were cast. All concrete mixtures in this project were based on a HPC mixture design for bridge decks used by the Oregon DOT. The target

AC C

compressive strength was 34.5 MPa with minimum strength of 27.6 MPa. A w/cm of 0.37 was used in most of the mixtures, except for an ordinary portland cement (no SCMs) where a w/cm of 0.42 was used. The w/cm was raised 0.05 to examine the effect of higher w/cm on drying and cracking performance for this mixture design. The total cementitious materials content was 375 kg/m3, containing 30% class F fly ash and 4% silica fume as mass replacement. To compare results from different aggregate sources, the coarse and fine aggregate content were set to 44.0% and 27.3% for all mixtures (except the mortar mixture). The high-range water reducer and air entrainer dosages were adjusted to achieve similar workability and air content for all mixtures.

8

ACCEPTED MANUSCRIPT

This mixture design was applied as a baseline with necessary modifications. When doing the mixture modifications, one principle was to keep all materials the same as the base line in terms of volume proportions. In addition, a proprietary mortar mixture was used. Table 5 shows the

RI PT

detailed mixture proportioning and descriptions. In total, 14 concrete mixtures with different wet cure durations, cementitious content, aggregate sources, shrinkage reducing methods, and w/cm were cast. All concrete mixtures in this project were based on a HPC mixture design for bridge decks used by the Oregon DOT, This HPC is

SC

designed to meet strength and durability requirement outlined by ODOT in the 2009 Oregon Standard Specification for Construction [56]. The target compressive strength was 34.5 MPa with minimum strength of 27.6 MPa. A w/cm of 0.37 was used in most of the mixtures, except

M AN U

for an ordinary portland cement (no SCMs), where a w/cm of 0.42 was used to investigate the effect of higher w/cm on drying and cracking performance. The total cementitious materials content was 375 kg/m3, containing 30% class F fly ash and 4% silica fume as mass replacement. To compare results from different aggregate sources, the coarse and fine aggregate content were set to 44.0% and 27.3% for all mixtures (except the mortar mixture). The high-range water reducer and air entraining admixture dosages were adjusted to achieve similar workability and

TE D

air content for all mixtures. This mixture design was applied as a baseline with necessary modifications. During mixture modifications, one principle was to keep the volume of all materials the same as the base line. In addition, a proprietary mortar mixture was used according

descriptions.

EP

to the manufacture’s instruction. Table 5 shows the detailed mixture proportioning and

AC C

Table 5 Concrete mixture proportioning (kg/m3) Mixture ID

Mixture descriptions

Wet curing duration (days)

Cement

Fly ash

Silica fume

Water

Coarse aggregate (source)

Sand (source)

FLWA

SRA

HPC-A1

Control HPC

3

249

112

15

139

1071 (A)

659 (A)

-

-

HPC-A2

Control HPC

14

249

112

15

139

1071 (A)

659 (A)

-

-

SRA-1

2% SRA

3

249

112

15

131

1071 (A)

659 (A)

-

7.5

SRA-2

2% SRA

14

249

112

15

131

1071 (A)

659 (A)

-

7.5

FLWA-1

25% FLWA

3

249

112

15

139

1071 (A)

400 (A)

164

-

9

ACCEPTED MANUSCRIPT

25% FLWA

14

249

112

15

139

1071 (A)

400 (A)

164

-

SYN

2% SRA + 25% FLWA

14

249

112

15

131

1071 (A)

400 (A)

164

7.5

OPC-1

OPC (no SCMS)

14

375

-

-

139

1071 (A)

659 (A)

-

-

OPC-2

OPC (with higher w/cm)

14

375

-

-

158

1071 (A)

659 (A)

-

-

HPC-LS

Limestone

14

249

112

15

139

1176 (limestone)

659 (A)

-

-

HPC-B

HPC, local B

14

249

112

15

139

1140 (B)

695 (B)

-

-

HPC-C

HPC, local C

14

249

112

15

139

1114 (C)

678 (C)

-

-

HPC-D

HPC, local D

3

249

112

15

139

1153 (D)

705 (D)

-

-

RM

Repair mortar

3

SC

RI PT

FLWA-2

3.

Results and discussions

3.1

Fresh and hardened properties

M AN U

Proprietary mortar material, mixing according to manufacturer’s instruction

Table 6 shows the summary of fresh properties and 28 day hardened properties for all the mixes.

TE D

All samples were wet cured for 28 days in a standard moisture room prior to mechanical property testing. Hardened properties represent the average of three samples for compressive strength and splitting tensile strength. Elastic modulus was the average of two samples. Effect of mechanical property on shrinkage cracking is a complicated issue, and often omitted or misinterpreted in

EP

cracking evaluation. A following manuscript will attempt to provide a comprehensive tool to evaluate shrinkage cracking issue by combining free shrinkage, compressive/tensile strength, and

AC C

modulus of elasticity.

Table 6 Fresh and hardened properties 28 day hardened properties

Fresh properties

Mixture ID

HPC-A1 HPC-A2

Slump (mm)

Air content (%)

Temperature (°C)

Unit weight (kg/m3)

Compressive strength (MPa)

Tensile strength (MPa)

Modulus of elasticity (GPa)

125 125

6.0 4.5

21.4 23.0

2306 2344

28.8 35.4

3.42 4.06

22.9 28.7

10

ACCEPTED MANUSCRIPT

SRA-2 FLWA-1 FLWA-2 SYN OPC-1 OPC-2 HPC-LS HPC-B HPC-C HPC-D

230 140 215 205 150 205 55 55 50 95 95 -

5.5 4.5 7.5 3.0 6.0 3.0 3.5 7.0 5.0 7.0 5.5 -

21.6 20.8 20.4 22.0 20.0 23.8 25.4 19.8 23.0 25.4 22.8 25.0

2262 2326 2214 2302 2182 2418 2374 2213 2286 2270 2304 -

33.2 36.4 36.6 45.4 26.1 44.7 34.5 34.2 24.5 27.1 34.6 61.2*

3.97 3.78 3.72 5.17 2.76 3.67 3.42 3.90 2.68 2.91 3.92 5.86*

28.0 29.3 24.2 29.6 22.0 32.2 30.0 32.4 31.2 32.0 31.1 29.4*

RI PT

SRA-1

M AN U

3.2

SC

RM *Cylinder samples were only wet cured for 3 days, then moved to the drying chamber then tested at 28 day.

Free shrinkage

Table 7 gives a summary of free shrinkage measurements of all mixtures at different ages up to 180 days. HPC-A2 is the control mixture for all free shrinkage tests. Generally, the free shrinkage at the early age was reduced for mixtures using all mitigation methods (SRA, FLWA, or synergy of both). The presence of SRA was intended to reduce drying shrinkage, while the

TE D

addition of FLWA was to minimize autogenous shrinkage[57]. The synergy of SRA and FLWA most significantly reduced the free shrinkage. In addition, different aggregate sources had a great impact on the magnitude of drying shrinkage. By using limestone as the coarse aggregate, the

EP

shrinkage was significantly reduced compared to all siliceous aggregate mixtures. Ultimate shrinkage prediction for each mixture is also given in Table 7. The prediction is based on ACI 209 model (as shown in Eq. 1) with a curve fitting method. A non-linear Levenberg-Marquardt

AC C

least squares fitting tool was used to fit the experimental drying shrinkage curve. More details about this method can be found in [58].

Equation 1

Where εsh(t, tc) = shrinkage strain at concrete age t since the start of drying at age tc, mm/mm; εshu = notional ultimate shrinkage strain, mm/mm; α, f = constants defining the shape of drying

shrinkage curve.

11

ACCEPTED MANUSCRIPT

Table 7 Summary of free shrinkage (microstrain) 56 day

90 day

180 day

Ultimate prediction

3 14 3 14 3 14 14 14 14 14 14 14 3 3

340 290 133 190 280 323 140 360 300 240 473 317 277 207

600 550 337 447 535 663 345 600 557 380 860 610 500 447

727 630 443 573 633 800 465 690 677 430 960 730 527 610

780 715 497 640 703 870 530 750 747 457 1033 810 617 740

863 760 570 710 737 917 620 830 837 563 1167 897 677 853

967 839 706 821 811 1019 724 991 1041 652 1234 1028 804 1191

RI PT

28 day

AC C

EP

TE D

HPC-A2 SRA-1 SRA-2 FLWA-1 FLWA-2 SYN OPC-1 OPC-2 HPC-LS HPC-B HPC-C HPC-D RM

7 day

SC

HPC-A1

Wet curing duration (days)

M AN U

Mixture ID

Fig. 2. Effect of curing time and shrinkage reduction methods on free shrinkage

By incorporating SRA in the HPC mixture as shown in Fig. 2, the shrinkage was significantly reduced for both 3-day wet cured samples and 14-day wet cured samples. For FLWA mixtures,

12

ACCEPTED MANUSCRIPT

the FLWA helped in reducing shrinkage for 3-day cured samples, but had little effect on the 14day cured samples. The combination of SRA and FLWA reduced the drying shrinkage by about 50 %, which was the most effective in reducing drying shrinkage among all 14-day wet curing

EP

TE D

M AN U

SC

RI PT

mixtures. These findings are consistent with the findings reported in [57, 59].

AC C

Fig. 3. Effect of aggregate sources on free shrinkage

Aggregate type also had a significant impact on shrinkage behavior. As shown in Fig. 3, all local siliceous aggregate (HPC-A2, HPC-B, and HPC-C) resulted in high shrinkage both at early-age and long-term. Mixture HPC-D, which used manufactured siliceous aggregates, performed better than the control, HPC-A2. By using limestone instead of siliceous river gravel as coarse aggregate, free shrinkage was reduced by 31% at 28 days compared to the control, HPC-A2. The RM mortar mixture, showed relatively lower shrinkage in the early age, likely due to the high volume of quartz sand in the aggregate, which is believed to perform best among aggregates

13

ACCEPTED MANUSCRIPT

from a shrinkage point of view[29, 60]. Nevertheless, the long-term shrinkage (> 800 microstrain at 180 day) was still considered high likely due to high paste content in this mixture. Given a closer look at the physical properties of different aggregates, it seems that the absorption

RI PT

capacity could affect the shrinkage performance. Among HPC-A2, HPC-B, HPC-C, HPC-LS, and HPC-D, the limestone (aggregate in HPC-LS) had the lowest absorption capacity (0.58 %), followed by local type D (manufactured siliceous aggregate in HPC-D, 2.04 % for rock and 2.74% for sand), local type B (local siliceous aggregate in HPC-B, 2.27 % for rock and 2.58 %

SC

for sand), local type A (local siliceous aggregate in HPC-A2, 2.58 % for rock and 3.08 % for sand), and, local type C (local siliceous aggregate in HPC-C, 3.17 % for rock and 3.46 % for sand). This correlates well with the shrinkage values (HPC-LS < HPC-D < HPC-B < HPC-A2 <

M AN U

HPC-C). This is likely due to the absorption capacity correlating to the modulus of elasticity of the aggregate [61, 62]. The lower the absorption capacity indicates fewer pores in the aggregate particles and therefore likely a higher modulus of elasticity. Higher modulus of elasticity could better resist volume change when the cement paste shrinks due to drying.

However, no

conclusion could be drawn due to many other possible variations such as aggregate gradation, sand equivalency, shape, and possibly aggregate mineralogy. More discussion can be found in

AC C

EP

TE D

discussion of restrained shrinkage results.

14

ACCEPTED MANUSCRIPT

Fig. 4. Effect of w/cm, SCMs, and synergy of LWFA and SRA on free shrinkage To investigate the impact of w/cm and SCMs on shrinkage, two mixtures were modified from the control HPC mixture design. OPC-1 and OPC-2 are 100 % portland cement mixtures without

RI PT

SCMs. Additionally, OPC-2 had a higher w/cm of 0.42. As shown in Fig. 4, both OPC mixtures showed over 800 microstrain shrinkage at 180 days of age, which is similar to HPC-A2, and considered high shrinkage. Changing the w/cm from 0.37 (OPC-1) to 0.42 (OPC-2), the impact on shrinkage was insignificant. This result confirmed that the presence of SCMs (30% fly ash

SC

and 4% silica fume) and a low w/cm (0.37) did not contribute to the high shrinkage of the control

3.3

Restrained shrinkage test

3.3.1

ASTM C1581

M AN U

HPC mixtures.

Table 8 gives a summary of the ASTM C1581 ring results, including ToC and the corresponding stress rate. Upon cracking, a sudden change was observed in two or more strain gauges, which

TE D

was be confirmed by visual inspection. The stress rate at ToC was calculated according to ASTM C1581. Based on ToC or stress rate, a cracking potential was assigned to each mixture. Fig. 5 shows a good relationship between ToC and stress rate, with a correlation coefficient of 0.89. The power-law relationship indicates that with a decrease of stress rate, the ToC should be

EP

exponentially prolonged. When determining the cracking potential classification, the authors believe that high priority should be given to stress rate at cracking. This is because the stress rate directly quantifies the stress in the concrete ring specimen. In addition, ToC is involved in the

AC C

stress rate calculation. In other words, stress rate provides a more comprehensive evaluation. Table 8 Summary of time-to-cracking and stress rate of ASTM ring tests Mixture

Curing duration

Time-to-cracking, Days

Cracking potential classification *

Stress rate, MPa/Day

(days)

A

B

C

Ave.

A

B

C

Ave.

4.0

5.5

5.2

4.9

0.380

0.315

0.338

0.344

H

HPC-A1

3

HPC-A2

14

4.4

4.6

3.6

4.2

0.343

0.281

0.482

0.369

H

SRA-1 SRA-2

3 14

13.9 16.1

18.4 14.9

18.8 11.6

17.0 14.2

0.094 0.104

0.073 0.093

0.094 0.139

0.087 0.112

L ML

15

ACCEPTED MANUSCRIPT

3 14

6.5 7.4

7.0 7.9

7.3 -

6.9 7.7

0.238 0.245

0.213 0.263

0.284 n/a

0.245 0.254

MH MH

SYN

14

19.7

14.0

14.0

15.9

0.115

0.070

0.060

0.081

L

OPC-1 OPC-2

14 14

4.0 4.2

5.6 4.6

5.3 3.6

5.0 4.1

0.340 0.257

0.275 0.266

0.329 0.238

0.315 0.254

MH MH

HPC-LS

14

40.9

no crack at 60 day

23.1

>41

0.045

≈0.10

0.099

0.082

L

HPC-B

14

3.5

7.1

8.4

6.3

0.410

0.305

0.197

0.304

MH

HPC-C

14

6.3

4.0

1.9

4.1

0.279

0.208

0.283

0.257

MH

HPC-D RM

3 3

11.2 28.0

8.4 33.0

11.4 23.0

10.3 28.0

0.205 0.072

0.243 0.063

0.227 0.084

0.225 0.073

MH L

EP

TE D

M AN U

SC

* H – High; ML – Moderate High; ML – Moderate Low; L – Low

RI PT

FLWA-1 FLWA-2

Fig. 5. Time-to-cracking versus stress rate

AC C

It is noted that SRA significantly prolonged the ToC, and decreased the stress rate. Comparing HPC to SRA, the ToC was prolonged from around 5 days to more than 14 days, which lowered the cracking potential from “high” to “moderate low” or even “low”. FLWA also prolonged the ToC and decreased the stress rate, but not as effectively as SRA. SYN showed the lowest free shrinkage and a similar ToC to that of SRA. Mixture SYN was also able to reduce the cracking potential from “high” to “low”.

16

ACCEPTED MANUSCRIPT

By comparing OPC-1 and OPC-2 to the control HPC-A mixtures, they showed similar cracking resistance and high cracking potential. This means for the given mixture design and local type A siliceous aggregate, the incorporation of SCMs and variation of w/cm between 0.37 and 0.42 did

RI PT

not significantly affect (either improve or worsen) cracking performance.

By switching the coarse aggregate from the local siliceous aggregate to a limestone aggregate while all other mixture design components remained the same, HPC-LS resulted in an average ToC of 41 days comparing to about 5 days ToC for HPC-A2. Similarly, the manufactured

SC

siliceous aggregate (HPC-D) also outperformed the local natural aggregates (type A, B, and C), extending the ToC to about 10 days. Among many aggregate types, concrete made with

M AN U

limestone usually exhibits significant lower shrinkage than concrete made with sandstone [29, 60]. One explanation is that the rough surface of limestone and other crushed aggregates provide better mechanical bonding which results in higher resistant against shrinkage from cement paste. Another theory states that aggregates with a higher modulus of elasticity allow for more restraint against shrinkage in the paste fraction.

The higher modulus of elasticity result in less

compressibility caused by stress generation in the paste [60]. However, there were case studies

TE D

[61] that showed aggregate with absorption capacity below 1.0% (correlated to high modulus of elasticity) exhibited low cracking pentagonal in the field. This is likely due to higher local internal stress concentration is generated due to higher modulus of elasticity even with relatively smaller shrinkage strain. Nonetheless, more recent research proved that the drying shrinkage of

EP

aggregate itself affects shrinkage much more than the modulus of elasticity [63]. The authors believe the effect of the aggregate on drying shrinkage and cracking merits further investigation.

AC C

Stress relaxation is another possible reason why the limestone mixture lasted for a significant longer time before cracking. In fact one HPC-LS ring specimen showed no cracking when the test was terminated at 60 days after initiation of drying. One set of strain gauge data is presented in Fig. 6, showing the strain development in three individual ring specimens of mixture LS. Ring C showed a typical response as most tests, which consisted of an increase in strain and a sharp decrease toward zero strain at the end. However, ring B did not show this sudden change, but rather a slow decrease in strain indicating stress relaxation in the ring due to prolonged testing period. This can be seen in the later age of ring A as well. After a certain period of time, about 28 days in this case, the effect of stress relaxation started to impact the cracking behavior of the 17

ACCEPTED MANUSCRIPT

ASTM ring specimens. This phenomenon was also overserved in AASHTO ring specimens[44, 51]. In other words, if a concrete mixture survived 28 days or longer in the ASTM ring, the cracking potential could be further lowered due to stress relaxation. However, ring A and C did not sustain the ring test as long as ring B, which was likely due to materials properties variability.

TE D

M AN U

SC

RI PT

This phenomenon is not unusual in restrained ring tests [36, 43, 51, 64].

Fig. 6. Strain development versus time, three individual rings of mixture LS

EP

In addition, the repair mortar (RM) also exhibited superior cracking-resistance (ToC around 28 days), despite that RM showed high free shrinkage in the later age. The result indicates that

AC C

drying shrinkage related cracking are unlikely a concern when repairing an HPC bridge deck using this material.

3.4

Shrinkage limits

Generally speaking, the higher the free shrinkage is, the higher the cracking potential is. In Fig. 7(a), the points can be distinguished into two groups. The group of points in the lower left (in the circle) represents mixtures with lower shrinkage (less than 450 microstrain at 28 day) and lower stress rate (approximately 0.1 MPa/day at cracking). Similarly, in Fig. 7(b), the results tend to

18

ACCEPTED MANUSCRIPT

form two groups. One group consists of mixtures with relatively higher shrinkage (500 microstrain or higher at 28 day) and lower ToC (less than 10 days), while the other group (in the box) includes mixtures with relatively lower shrinkage (between 280 to 450 microstrain at 28

AC C

EP

TE D

M AN U

SC

7(b), including SRA-1, SRA-2, SYN, HPC-LS, and, RM.

RI PT

day) and higher ToC (more than 13 days). It is the same group of mixtures in Fig. 7(a) and Fig.

19

ACCEPTED MANUSCRIPT

(b)

AC C

EP

TE D

M AN U

SC

RI PT

(a)

Fig. 7. Comparison between (a) free shrinkage and stress rate, (b) free shrinkage and ToC.

In Fig. 8, all 16 mixtures were listed from the lowest shrinkage mixture (SRA-1) to the highest (HPC-B), with the recommended shrinkage limit (450 microstrain) in the middle. It shows that

20

ACCEPTED MANUSCRIPT

all mixtures with lower than 450 microstrain free shrinkage at 28 days from initiation of drying have greater than or approximately equal to 14-day ToC in the ASTM ring test, which would be considered moderate low cracking potential. Mixture SYN, SRA-1, SRA-2, HPC-LS, and RM would be expected to perform reasonably well in the field with respect to cracking related to

RI PT

drying shrinkage. Therefore, a 450 microstrain free shrinkage at 28 day from initiation of drying

AC C

EP

TE D

M AN U

SC

seems to be a reasonable shrinkage limit for bridge deck concrete mixture designs.

Fig. 8. Ranking of free shrinkage (microstrain) at 28 day, ToC (day) in parentheses

4.

Conclusions

21

ACCEPTED MANUSCRIPT

The research investigated 14 HPC mixtures comparing effects of different curing duration, shrinkage reducing strategies, and different aggregate sources on shrinkage and cracking performance of resulting concretes. This research focused on free shrinkage (ASTM C157) and restrained shrinkage test (ASTM C1581). The goals was to identify shrinkage limit and testing

•

RI PT

protocols to ensure highly cracking-resistance HPC. Several conclusions are listed as follows: According to the results of free and restrained shrinkage tests, by incorporating SRA alone or a synergistic mixture of SRA and FLWA, the cracking potential of HPC was

•

SC

significantly reduced.

The source of the aggregate has a significant impact on shrinkage tendency and cracking

M AN U

potential. The results showed that the HPC mixture with the local siliceous river gravel had much higher drying shrinkage and cracking risk compared to the HPC mixture with a coarse limestone aggregate. •

Based on the results of this study, shrinkage of 450 microstrain at 28 day from initiation of drying based on ASTM C157 test was proposed as a limit to achieve satisfactory

Acknowledgements

TE D

cracking resistance for HPC concrete used for bridge decks.

The authors would like to acknowledge the Oregon Department of Transportation for supporting

EP

the research project (SPR 728). Grateful acknowledgement is also made to the Portland Cement

AC C

Association Education Foundations who partially funded this research project.

22

ACCEPTED MANUSCRIPT

References

AC C

EP

TE D

M AN U

SC

RI PT

[1] USDOT. 2010 Status of the Nation's Highways, Bridges and Transit: Conditions and Performance: Report to Congress. https://www.fhwa.dot.gov/policy/2010cpr/, 2014 (accessed March 2016). [2] ASCE. Report Card for America's Infrastructure, http://www.infrastructurereportcard.org/a/#p/bridges/overview, 2013 (accessed March 2016). [3] Aktan H, Fu G, Dekelbab W, Attanayaka U. Investigate causes & develop methods to minimize early-age deck cracking on Michigan bridge decks. Detroit, MI: Wayne State University, Civil and Environmental Engineering Dept.; 2003. [4] Holt EE. Early-Age Autogenous Shrinkage of Concrete. Technical Research Centre of Finland, VTT Publications, No 446. 2001. [5] Bentz DP, Jensen OM. Mitigation strategies for autogenous shrinkage cracking. Cem Contr Compos. 2004;26(6):677-85. [6] ACI Committee 209. Report on Factors Affecting Shrinkage and Creep of Hardened Concrete (ACI 209.1R-05). Farmington Hills, MI. : ACI; 2005 (reapproved 2013). [7] Mohr BJ, Bentz DP. ACI SP-256 Internal Curing of High Performance Concrete Lab and Field Experiences. Farmington Hills, MI. : ACI; 2008. [8] Bentz DP, Weiss WJ. Internal curing : a 2010 state-of-the-art review. February 2011: NISTIR 7765, NIST, U.S. Deptartment of Commerce; 2011. [9] Schindler AK, Grygar JG, Weiss WJ. ACI SP-290 The Economics, Performance and Sustainability of Internally Cured Concrete. Farmington Hills, MI. : ACI; 2012. [10] Lura P. Autogenous Deformation and Internal Curing of Concrete. Delft: DUP Science; 2003. [11] Friggle T, Reeves D. Internal Curing of Concrete Paving: Laboratory and Field Experience. ACI. 2008;SP-256:71-80. [12] Slatnick S, Riding KA, Folliard KJ, Juenger MCG, Schindler AK. Evaluation of Autogenous Deformation of Concrete at Early Ages. ACI Mater J. 2011;108(1):21-8. [13] Paul A, Lopez M. Assessing Lightweight Aggregate Efficiency for Maximizing Internal Curing Performance. ACI Mater J. 2011;108(4):385-93. [14] Deboodt T, Fu T, Ideker JH. Durability assessment of high-performance concrete with SRAs and FLWAs. Cem Concr Compos. 2015;57(0):94-101. [15] Villarreal VH, Crocker DA. Better Pavements through Internal Hydration. Concr Int. 2007;29(02):32-6. [16] Villarreal VH. Internal Curing - Real Word Ready Mix Production and Applications: A Practical Approach to Lightweight Modified Concrete. ACI 2008;SP-256:45-56. [17] Delatte N, Crowl D, Mack E, Cleary J. Evaluating High Apsorptive Materials to Improve Internal Curing of Concrete. ACI. 2008;SP-256:91-104. [18] Cusson D, Lounis Z, Daigle L. Benefits of internal curing on service life and life-cycle cost of high-performance concrete bridge decks – A case study. Cem Concre Compos. 2010;32(5):339-50. [19] Tazawa E-i, Miyazawa S. Influence of cement and admixture on autogenous shrinkage of cement paste. Cem Concr Res. 1995;25(2):281-7.

23

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

[20] Folliard KJ, Berke NS. Properties of high-performance concrete containing shrinkagereducing admixture. Cem Concre Res. 1997;27(9):1357-64. [21] Bentz DP, Geiker MR, Hansen KK. Shrinkage-Reducing Admixtures and Early-Age Desiccation in Cement Pastes and Mortars. Cem Concr Res. 2001;31(7):1075-85. [22] Rongbing B, Jian S. Synthesis and Evaluation of Shrinkage-Reducing Admixture for Cementitious Materials. Cem Concr Res. 2005;35(3):445-8. [23] Bentz D. Curing with Shrinkage-Reducing Admixtures. Concr Int. 2005;27(10):55-60. [24] Bentz DP. Influence of Shrinkage-Reducing Admixtures on Early-Age Properties of Cement Pastes. J Adv Concr Tech. 2006;4(3):423-9. [25] Saliba J, Rozière E, Grondin F, Loukili A. Influence of shrinkage-reducing admixtures on plastic and long-term shrinkage. Cem Concr Compos. 2011;33(2):209-17. [26] Dubey A, Banthia N. ACI SP-268 Fiber Reinforced Concrete in Practice, 2012. [27] ACI Committee 223. Guide for the use of shrinkage-compensating concrete. Farmington Hills, MI: ACI; 2010. [28] Krauss PD, Rogalla EA. Transverse Cracking in Newly Constructed Bridge Decks. NCHRP Report 380. Washington, D.C.: TRB, National Research Council; 1996. [29] Burrows RW. The Visible and Invisible Cracking of Concrete. Farmington Hills, MI: ACI; 1998. [30] Nilson AH, Darwin D, Dolan CW. Design of Concrete Structures 13th Edition. New York: McGraw Hill; 2004. [31] ASTM Standard C157. Standard Test Method for Length Change of Hardened HydraulicCement Mortar and Concrete. ASTM International. West Conshocken, PA; 2008. [32] Unified Facilities Guide Specifications. WBDG whole building design guide: Washington, D.C. : National Institute of Building Sciences.; 2012. [33] Mokarem DW, Weyers RE, Lane DS. Development of a shrinkage performance specifications and prediction model analysis for supplemental cementitious material concrete mixtures. Cem Concr Res. 2005;35(5):918-25. [34] Ramniceanu A, Weyers RE, Mokarem DW, Sprinkel MM. Bridge Deck Concrete Volume Change. The Virginia Transportation Research Council, VTRC 10-CR5, Charlottesville, VA, Feb. 2010. [35] Nassif H, Aktas K, Najm H. Concrete shrinkage analysis for bridge deck concrete. Trenton, NJ: Dept. of Transportation, the State of New Jersey, FHWA NJ-2007-007; 2007. [36] Qiao P, McLean D, Zhuang J. Mitigation Strategies for Early-Age Shrinkage Cracking in Bridge Decks. Washingon DOT Research Report No. WA-RD 747.1, Apr. 2010. [37] FHWA. Standard specifications for construction of roads and bridges on federal highway projects. Washington, D.C.: USDOT, FHWA; 2014. [38] Beverly P (Eds.), Fib Model Code for Concrete Structures 2010: Ernst & Sohn; 2013. [39] The New Zealand guide to concrete construction.: Cement & Concrete Association of New Zealand; 2010. [40] CSA A23.1 Concrete materials and methods of concrete construction : test methods and standard practices for concrete. Canadian Standards Association; 2014. [41] Harrison T. Standards for Fresh Concrete: British Standards Institution; 2004. [42] Weiss WJ. Prediction of Early-Age Shrinkage Cracking in Concrete Elements PhD Dissertation. Evanston, IL, Northwestern University; 1999. [43] See HT, Attiogbe EK, Miltenberger MA. Potential for Restrained Shrinkage Cracking of Concrete and Mortar. Cem Concr Aggreg. 2004;26(2):123-30.

24

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

[44] Brown MD, Smith CA, Sellers JG, Folliard KJ, Breen JE. Use of Alternative Materials to Reduce Shrinkage Cracking in Bridge Decks. ACI Mater J. 2007;104(6):629-37. [45] Lindquist W, Darwin D, Browning J. Development and Construction of Low Cracking High-Performance Concrete (LC-HPC) Bridge Decks: Free Shrinkage, Mixture Optimization, and Concrete Production. The University of Kansas Center for Research, Inc. Report No. SM92. Nov. 2008. [46] Ray I, Gong Z, Davalos JF, Kar A. Shrinkage and cracking studies of high performance concrete for bridge decks. Constr Build Mater. 2012;28(1):244-54. [47] Wang K, Schlorholtz SM, Sritharan S, Seneviratne H, Wang X, Hou Q. Investigation into Shrinkage of High-Performance Concrete Used for Iowa Bridge Decks and Overlays. IHRB Project TR-633, Sept 2013. [48] ASTM Standard C1581. Standard test method for determining age at cracking and induced tensile stress characteristics of mortar and concrete under restrained shrinkage. West Conshohocken, PA; 2009. [49] AASHTO T334-08. Standard method of test for estimating the cracking tendency of concrete. AASHTO; 2008. [50] ACI Committee 231. Report on early-age cracking: causes, measurement, and mitigation. Farmington Hills, MI: ACI; 2010. [51] Folliard K, Smith C, Sellers G, Brown M, Breen JE. Evaluation of Alternative Materials to Control Drying-Shrinkage Cracking in Concrete Bridge Decks. Report No. FHWA/TX-04/04098-4, 2003. [52] Hadidi R, Saadeghvaziri MA. Ransverse Cracking of Concrete Bridge Decks: State-of-theArt. J Bridge Eng. 2005;10(5):503-10. [53] Ideker JH, Fu T, Deboodt T. Internal Curing of High-Performance Concrete for Bridge Decks. Oregon Department of Transportation Research Section, SPR-711. Dec. 2013. [54] Bentz DP, Lura P, Roberts JW. Mixture proportioning for internal curing. Concr Int. 2005;27(02):35-40. [55] Fu T, Deboodt T, Ideker JH. A Simple Procedure on Determining Long-Terms Chemical Shrinkage for Cementitious Systems Using Improved Chemical Shrinkage Test. ASCE J Mater. 2012;24(8). [56] ODOT. Oregon Standard and Specifications for Construction. Oregon Department of Transportation; 2008. [57] Zhutovsky S, Kovler K, Bentur A. Effect of hybrid curing on cracking potential of highperformance concrete. Cem Concr Res. 2013;54(0):36-42. [58] Fu T, Deboodt T, Ideker JH. Prediction of Drying Shrinkage for Internally Cured HPC. ACI SP-290-10. 2012. [59] Meddah MS, Suzuki M, Sato R. Shrinkage reducing effect of a combination of internal curing and shrinkage compensating agents on high-performance concrete. ACI SP-256, pp17-32. 2009. [60] Troxell GE, Davis HE. Composition and Properties of Concrete. New York: MacGraw-Hill Book Co.; 1956. [61] Delatte N, Crowl D. Case Studies of Internal Curing of Bridge Decks in the Greater Cleveland Area. ACI SP-290-06. 2012. [62] Alexander KG. Aggregates and the Deformation Properties of Concrete. ACI Mater J. 1996;93(6):569-77.

25

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

[63] Hyodo H, Tanimura M, Sato R, Kawai K. Evaluation of Effect of Aggregate Properties on Drying Shrinkage of Concrete. 3rd International Conference on Sustainable Construction Materials and Technologies. 2013;Session T2-8, No. 308, Aug. 2013. [64] Radlinska A, Pease B, Weiss J. A Preliminary Numerical Investigation on the Influence of Material Variability in the Early-Age Cracking Behavior of Restrained Concrete. Mater Struct. 2007;40(4):375-86.

26