High Performance Concrete Specifications and Practices for Bridges (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

6 The concept of HPC was first introduced in the 1980s and was often associated with high strength concrete for columns of high-rise buildings with measured compressive strengths as high as 19.0 ksi. During the same period, high strength concrete was used in limited bridge applications in the United States such as the East Huntington Bridge across the Ohio River (8.0 ksi design strength) and Tower Road Bridge in Washing- ton State (9.0 ksi design strength) (ACI Committee 363 2010). In the mid- to late 1980s, greater interest developed in the potential use of HPC in bridges to extend the service life of concrete bridge decks and the capacities of prestressed concrete beams. The interest at the national level began with the first Strategic Highway Research Program (SHRP). STRATEGIC HIGHWAY RESEARCH PROGRAM SHRP was a five-year national research program initiated in 1987 to develop and evaluate techniques and technolo- gies to combat the deteriorating conditions of the nation’s highways. One of the four program areas of SHRP was Concrete and Structures, which included project C-205 titled “Mechanical Behavior of High Performance Concrete.” The SHRP project included both bridge and pavement applications. For purposes of the SHRP project, HPC was initially defined by the following three requirements (Zia et al. 1991): 1. Maximum w/cm ratio of 0.35, 2. Minimum durability factor of 80% as determined by ASTM C666 Method A (AASHTO T 161, Method A), and 3. Minimum compressive strength of a. 3.0 ksi within 4 hours after placement, b. 5.0 ksi within 24 hours, or c. 10.0 ksi within 28 days. This definition incorporated criteria related to both durability and strength. A subsequent report cautions against confus- ing high performance concrete with high strength concrete, as there are many factors that may be more important than strength in a given application (Zia et al. 1997). By completion of the project, the criteria for HPC had been refined to those shown in Table 1. SHRP’s research on the mechanical behavior of HPC had three general objectives: 1. Obtain information needed to fill gaps in existing knowledge; 2. Develop new, significantly improved engineering criteria for the mechanical properties and behavior of HPC; and 3. Provide recommendations and guidelines for using HPC in highway applications according to the intended use, required properties, environment, and service. Both plain and fiber-reinforced concretes were included in the study. The first task of the project was a literature search and review to define the existing knowledge about the mechani- cal properties of HPC. An annotated bibliography containing 830 references published between 1974 and 1989 was com- piled and published (Leming et al. 1990). About 150 refer- ences from the bibliography were selected for critical review leading to a state-of-the-art report (Zia et al. 1991). Subsequent tasks addressed the properties for the various types of concrete listed in Table 1. The results were published in six related reports: Volume 1 Summary Report (Zia et al. 1993a) Volume 2 Production of High Performance Concrete (Zia et al. 1993b) Volume 3 Very Early Strength (VES) Concrete (Zia et al. 1993c) Volume 4 High Early Strength (HES) Concrete (Zia et al. 1993d) Volume 5 Very High Strength (VHS) Concrete (Zia et al. 1993e) Volume 6 High Early Strength Fiber-Reinforced Concrete (HESFRC) (Naaman et al. 1993). The project involved extensive testing of concrete mixes to develop information about compressive strength, flexural tensile strength, splitting tensile strength, drying shrinkage, creep, freeze-thaw resistance, rapid chloride permeability, AC impedance, concrete to concrete bonding, and concrete to steel bonding. The research involved laboratory experiments as well as field studies. chapter two EVOLUTION OF HIGH PERFORMANCE CONCRETE FOR BRIDGES

7 The research determined that the specifications at that time had been formulated primarily from the knowledge of conventional materials. As such, some requirements may not be applicable to HPC and would serve as barriers to the acceptance of HPC for highway applications (Zia et al. 1993a). Although no barriers were identified in the codes and specifi- cations of AASHTO, ASTM, and ACI, several barriers were identified in the specifications of state highway agencies. The authors of the six reports encouraged state agencies to update their specifications to accommodate the latest infor- mation on concrete technology. In 1996, the same authors of the SHRP state-of-the-art report published two sequel reports, which covered the period from 1989 to 1994 (Zia et al. 1996, 1997). The authors reported a phenomenal growth in the amount of research and applications of HPC in this five-year period, with increasing emphasis being placed on concrete durability rather than strength. FEDERAL HIGHWAY ADMINISTRATION DEMONSTRATION PROJECTS In 1993, the FHWA initiated a national program to implement the use of HPC in bridges. The program included the con- struction of demonstration bridges by state DOTs in each of the FHWA regions and dissemination of the technology and results at showcase workshops. Eighteen bridges in 13 states were included in the national program. In addition to the joint state-FHWA HPC initiative, other states have implemented the use of HPC in various bridge elements (Russell et al. 2006a). The bridges are located in different climatic regions of the United States and use different types of superstructures as listed in Table 2. The bridges demonstrate practical applica- tions of HPCs. In addition, construction of these bridges provided oppor- tunities to learn more about the placement and actual behavior of HPC in bridges. Consequently, many of the bridges were instrumented to monitor their short- and long-term perfor- mance. Also, concrete material properties were measured for most of the bridges. All bridges used precast, prestressed concrete beams with specified concrete compressive strengths ranging from 5.5 to 8.8 ksi at strand release and 8.0 to 14.0 ksi for design as listed in Table 2. Rapid chloride permeability was specified for the beams of 11 bridges with values ranging from 1000 to 3000 coulombs at 56 days or 1000 to 2500 coulombs at 28 days using accelerated curing. All bridges except the one in Ohio used a CIP deck with thicknesses ranging from 7.0 to 9.0 in. The Ohio bridge used a 3-inch-thick asphalt overlay on top of adjacent box beams. Specified concrete compressive strengths for the deck concretes ranged from 4.0 to 8.0 ksi with most values being in the 4.0 to 6.0 ksi range as shown in Table 3. Rapid chloride permeability was specified for nine bridge decks with values ranging from 1000 to 2000 coulombs at 56 days or 1500 to 2500 coulombs at 28 days using accelerated curing. The concrete mixes for the precast, prestressed concrete beams and CIP concrete decks contained various combina- tions of cement and fly ash; cement, fly ash, and silica fume; and cement and silica fume. Only one bridge included slag cement, which was used in combination with cement and silica fume for the precast, prestressed concrete beams and in combination with cement only for the CIP concrete deck (Russell et al. 2006a). As part of the initiative to implement the use of HPC in bridges, FHWA introduced eight performance characteristics, shown in Table 4, to encompass both durability and structural design. The four characteristics for durability were freeze- thaw resistance, scaling resistance, abrasion resistance, and chloride ion penetration. The four structural design character- istics were compressive strength, modulus of elasticity, drying shrinkage, and creep. For each characteristic, an ASTM or AASHTO standard test method was selected. When the test methods offered alternative procedures, such as specimen size, the alternative to be used was specified. A range of two to four performance grades was also selected for each characteristic. With this approach, it was Category of HPC Minimum Compressive Strength Maximum Water/ Cement Ratio Minimum Frost Durability Factor Very Early Strength (VES) Option A Option B 2.0 ksi in 6 hours 2.5 ksi in 4 hours 0.40 0.29 80% 80% High Early Strength (HES) 5.0 ksi in 24 hours 0.35 80% Very High Strength (VHS) 10.0 ksi in 28 days 0.35 80% Based on Zia et al. (1993a). TABLE 1 CRITERIA FOR HIGH PERFORMANCE CONCRETE

8 not necessary to specify every characteristic or to specify the same grade for different characteristics. The intent was to select the characteristics and grades to match the intended application and its environment. Each state selected the characteristics for its demonstration bridges. Estimates of relationships between each performance grade and severity of field conditions were provided to assist designers in selecting the grade of HPC for a particular project (Goodspeed et al. 1996). Following completion of the demonstration projects, information about each bridge was collected and compiled into a single source compact disc. Details about the char- acteristics specified and measured on each demonstration project are given in Russell et al. (2006a). An analysis of the specifications used for the demonstration bridges indicated that the primary characteristic specified for durability was rapid chloride permeability with values ranging from 1000 to 3000 coulombs. Approximately three-quarters of the specified values were between 800 and 2000 coulombs. Freeze-thaw resistance was specified for one bridge, whereas scaling resistance and abrasion resistance were not specified for any bridges. For strength characteristics, compressive strength was the only characteristic specified for the girders and decks of all bridges. For the majority of the bridges, the specified com- pressive strength for the girder concrete was 10.0 ksi. This corresponds with the upper limit in the AASHTO LRFD Specifications (AASHTO 2010b). The majority of measured compressive strengths were in the range of 10 to 14 ksi. For the deck concrete, the majority of the specified strengths ranged from 4 to 6 ksi. This is to be expected because there is no reason to specify higher strengths for the concrete deck in most slab and girder bridges. Durability properties are more important for bridge decks and a high strength concrete does not guarantee a durable concrete (Russell et al. 2006a). State Bridge Name Superstructure Type Specified Concrete Design Strength for Beams, ksi Open to Traffic (year) Alabama Highway 199 BT-54 10.0 at 28 days 2000 Colorado Yale Avenue Box Beam 10.0 at 56 days 1998 Georgia SR-920 AASHTO Type II, IV 10.0 at 56 days 2002 Louisiana Charenton CanalBridge AASHTO Type III 10.0 at 56 days 1999 Nebraska 120th Street NU1100 12.0 at 56 days 1996 New Hampshire Route 104, Bristol AASHTO Type III 8.0 at 28 days 1996 New Hampshire Route 3A, Bristol NE 1000 8.0 at 28 days 1999 New Mexico Rio Puerco BT1600 10.0 at 56 days 2000 North Carolina US-401 AASHTO Type IV, III 10.0 at 28 days 2000/2002 Ohio U.S. Route 22 nearCambridge Box Beam B42-48 10.0 at 56 days 1998 South Dakota I-29 Northbound AASHTO Type II 9.9 at 28 days 1999 South Dakota I-29 Southbound AASHTO Type II 9.9 at 28 days 2000 Tennessee Porter Road BT-72 10.0 at 28 days 2000 Tennessee Hickman Road BT-72 10.0 at 28 days 2000 Texas Louetta Road Texas U 54 13.1 at 56 days1 1998 Texas San Angelo AASHTO Type IV 14.0 at 56 days1 1998 Virginia Route 40, Brookneal AASHTO Type IV 8.0 at 28 days 1996 Virginia Virginia Avenue,Richlands AASHTO Type III 10.0 at 28 days 1997 Washington State Route 18 Washington W 74G 10.0 at 56 days 1998 Based on Russell et al. (2006a). 1For the Texas bridges, different concrete strengths were specified for different girder span lengths. Listed strengths are the largest values. TABLE 2 HPC DEMONSTRATION BRIDGES AND SUPERSTRUCTURE TYPES

9 State Bridge Name Specified Properties for Deck Concrete Compressive Strength, ksi Permeability, coulombs Alabama Highway 199 6.0 @ 28 days — Colorado Yale Avenue 5.1 @ 28 days — Georgia SR-920 7.3 @ 56 days 2000 @ 56 days Louisiana Charenton Canal Bridge 4.2 @ 28 days 2000 @ 56 days Nebraska 120th Street 8.0 @ 56 days 1800 @ 56 days New Hampshire Route 104, Bristol 6.0 @ 28 days 1000 @ 56 days New Hampshire Route 3A, Bristol 6.0 @ 28 days 1000 @ 56 days New Mexico Rio Puerco 6.0 @ 28 days — North Carolina US-401 6.0 @ 28 days — South Dakota I-29 Northbound 4.5 @ 28 days — South Dakota I-29 Southbound 4.5 @ 28 days — Tennessee Porter Road 5.0 @ 28 days 1500 @ 28 days1 Tennessee Hickman Road 5.0 @ 28 days 1500 @ 28 days1 Texas Louetta Road 4.0 and 8.0 @ 28 days — Texas San Angelo 6.0 and 4.0 @ 28 days — Virginia Route 40, Brookneal 4.0 @ 28 days 2500 @ 28 days1 Virginia Virginia Avenue,Richlands 5.0 @ 28 days 2500 @ 28 days 1 Washington State Route 18 4.0 @ 28 days — Based on Russell et al. (2006a). 1Includes curing for 21 days at 100°F. — = not specified. TABLE 3 HPC DEMONSTRATION BRIDGES AND SPECIFIED DECK CONCRETE PROPERTIES Based on a review of the FHWA characteristics, grades, and test methods, Russell et al. (2006a) proposed the following revisions: • Each characteristic should have three grades. • Grades should always be considered minimum perfor- mance levels. • The addition of alkali-silica reactivity, sulfate resistance, and flowability to the eight previous characteristics. • Modifications to some of the test procedures. • Requirement to specify a characteristic only when it is necessary for the intended application. The 11 characteristics and the three grades of performance are shown in Table 5. After the demonstration bridges described above had been in service for five to 10 years, the decks were inspected and their performance evaluated using relevant information pertain- ing to the construction of each bridge deck (Mokarem et al. 2009). The construction data included concrete mix design, construction practices during and after concrete placement, average daily traffic on the bridge, and maintenance performed. The deck inspections included a detailed visual inspection of the top surface, as well as the preparation of detailed crack maps. Concrete cores were acquired from selected locations for petrographic analysis. The field surveys found that the HPC was generally performing well with no indication of any significant dete- rioration resulting from material properties. There were no indications of alkali-silica reaction, sulfate attack, or other deleterious reactions. There was also no significant spalling or delamination observed on the bridge decks. There was some spalling along the edges of some cracks, but this was not considered significant. When the structural system of the bridge included conti- nuity over the supports, negative moment transverse cracks

10 Performance Characteristic Standard Test Method FHWA HPC Performance Grade 1 2 3 4 Freeze-thaw Durability (x = relative dynamic modulus of elasticity after 300 cycles) AASHTO T 161 (ASTM C666) Proc. A 60% < x < 80% 80% < x Scaling Resistance (x = visual rating of the surface after 50 cycles) ASTM C672 x = 4,5 x = 2,3 x = 0,1 Abrasion Resistance (x = avgerage depth of wear in mm) ASTM C944 2.0 > x > 1.0 1.0 > x > 0.5 0.5 > x Chloride Penetration (x = coulombs) AASHTO T 277 (ASTM C1202) 3000 > x > 2000 2000 > x > 800 800 > x Strength (x = compressive strength) AASHTO T 22 (ASTM C39) 6 < x < 8 ksi 8 < x < 10 ksi 10 < x < 14 ksi x > 14 ksi Elasticity (x = modulus of elasticity) ASTM C469 4 < x < 6 x 10 psi 6 6 < x < 7.5 x 10 psi6 x > 7.5 x 106 psi Drying Shrinkage (x = microstrain) ASTM C157 800 > x > 600 600 > x > 400 400 > x Creep (x = microstrain/pressure unit) ASTM C512 0.52 > x > 0.41/psi 0.41 > x > 0.31/psi 0.31 > x > 0.21/psi 0.21/psi > x Adapted from Russell et al. (2006a). TABLE 4 GRADES OF EIGHT PERFORMANCE CHARACTERISTICS FOR HIGH PERFORMANCE STRUCTURAL CONCRETE TABLE 5 REVISED GRADES OF PERFORMANCE CHARACTERISTICS FOR HIGH PERFORMANCE STRUCTURAL CONCRETE Performance Characteristic Standard Test Method FHWA HPC Performance Grade 1 2 3 Freeze-thaw Durability (F/T = relative dynamic modulus of elasticity after 300 cycles) AASHTO T 161 (ASTM C666) Proc. A 70% < F/T < 80% 80% < F/T < 90% 90% < F/T Scaling Resistance (SR = visual rating of the surface after 50 cycles) ASTM C672 3.0 > SR > 2.0 2.0 > SR > 1.0 1.0 > SR > 0.0 Abrasion Resistance (AR = average depth of wear in mm) ASTM C944 2.0 > AR > 1.0 1.0 > AR > 0.5 0.5 > AR Chloride Penetration (CP = coulombs) AASHTO T 277 (ASTM C1202) 2500 > CP > 1500 1500 > CP > 500 500 > CP Alkali-silica Reactivity (ASR = expansion at 56 d) (%) ASTM C441 0.20 > ASR > 0.15 0.15 > ASR > 0.10 0.10 > ASR Sulfate Resistance (SR = expansion) (%) ASTM C1012 SR < 0.10 at 6 months SR < 0.10 at 12 months SR < 0.10 at 18 months Flowability (SL = slump, S F = slump flow) AASHTO T 119 (ASTM C143) and proposed slump flow test SL > 7-1/2 in. and SF < 20 in. 20 < SF < 24 in. 24 in. < SF Strength (f'c = compressive strength) AASHTO T 22 (ASTM C39) 8 < f'c < 10 ksi 10 < f'c < 14 ksi 14 ksi < f'c Elasticity (Ec=modulus of elasticity) ASTM C469 5 < Ec < 6 x 10 psi 6 6 < Ec < 7 x 10 psi6 7 x 106 psi < Ec Drying Shrinkage (S = microstrain) AASHTO T 160 (ASTM C157) 800 > S > 600 600 > S > 400 400 > S Creep (C = microstrain/pressure unit) ASTM C512 0.52 > C > 0.38/psi 0.38 > C > 0.21/psi 0.21/psi > C Adapted from Russell et al. (2006a).

11 occurred. The bridge geometry influenced the amount of con- crete cracking. For example, bridges with skewed supports exhibited more cracking than rectangular bridges because of the torsional stresses. Using crack survey data from each bridge, the length of the transverse, diagonal, and longitudinal cracks on each deck were calculated. For all bridge decks, the average crack lengths were 0.073 ft/ft2 transversely; 0.008 ft/ft2 diagonally; 0.042 ft/ft2 longitudinally; and 0.123 ft/ft2 totally. However, the total crack lengths on each bridge ranged from 0.003 to 0.741 ft/ft2. In comparison, Browning and Darwin (2007) reported crack lengths ranging from 0.0 to 0.31 ft/ft2 for bridge decks in Kansas. The results also indicated that in some cases, the use of HPC reduced bridge deck cracking, whereas in other cases the crack lengths were greater. By comparing the crack lengths with the corresponding concrete mix proportions, Mokarem et al. (2009) concluded that a high performance concrete mix with a w/cm between 0.35 and 0.40 and a cementitious materials content between 600 and 700 lb/yd3, used with construction practices such as seven-day wet curing, should result in shorter crack lengths. RESEARCH PROJECTS Each demonstration project included a research component. On some projects, the research focused on concrete material properties. Measurements were made on different projects to determine compressive strength, modulus of elasticity, tensile strength, creep, drying shrinkage, chloride perme- ability, freeze-thaw resistance, deicer scaling resistance, and abrasion resistance. Concrete temperatures were measured during curing to determine the heat of hydration of the pre- stressed concrete beams. The use of match-cured cylinders for measurements of concrete compressive strengths com- pared with the use of conventionally cured cylinders was also investigated. On other projects, the research was used to determine prestress losses, temperature gradients in the deck and girders resulting from daily and seasonal temperature changes, strand transfer length, long-term camber, and load distribution. Information from the showcase bridges was col- lected by the FHWA and compiled onto a compact disc for easy retrieval and viewing (Russell et al. 2006a). AMERICAN ASSOCIATION OF STATE HIGHWAY AND TRANSPORTATION OFFICIALS LEAD STATES TEAM FOR HIGH PERFORMANCE CONCRETE IMPLEMENTATION To implement the results of the SHRP program, AASHTO created a task force consisting of multiple teams (Moore 1999). The AASHTO Lead States Team for HPC implementation consisted of representatives of industry, FHWA, and states. Its mission was to promote the implementation of HPC technology for use in pavements and bridges and to share knowledge, benefits, and challenges with the states and their customers. The team began its work in 1996. When the HPC lead states team’s mission ended in 2000, a majority of states were using conventional strength HPC for bridge decks, almost half were using high strength HPC for bridge beams, and others were using HPC for superstructures and substructures (Moore and Ralls 2000). The team’s out- reach initiatives included HPC bridge showcases, international symposia, conference and meeting presentations, articles in various publications, and establishment of points of contact in each state to champion HPC technology. FEDERAL HIGHWAY ADMINISTRATION HIGH PERFORMANCE CONCRETE TECHNOLOGY DELIVERY TEAM The FHWA HPC Technology Delivery Team (TDT) was established in 1997 through funding in the Intermodal Surface Transportation Efficiency Act (ISTEA). The TDT helped 13 states build HPC bridges and host or participate in tech- nology transfer forums such as showcases and workshops. Working with the AASHTO Lead States Team, the TDT influ- enced many additional DOTs to try HPC in their highway structures. When the ISTEA funding ended in 1997, about 25 states had used HPC (Halkyard 2002). The mission of the TDT was renewed in 2002 with a focus on field delivery of HPC technology. A new commu- nity of practice website for HPC was developed, allowing users to post questions, participate in discussions, and review work in progress. This website is archived at https://www. transportationresearch.gov/dot/fhwa/hpc/Lists/aReferences/ AllItems.aspx [August 11, 2012]. One publication by the TDT is the High Performance Concrete Structural Designers Guide, a source of information to structural designers for the design and construction of highway bridges and related structures using HPC. The guide includes all aspects of developing and producing HPC with desirable and beneficial characteristics (Triandafilou et al. 2005). The TDT also conducted surveys of state highway agen- cies in 2003 –2004 and 2006–2007 to determine the usage of HPC in bridges. Figures 1 and 2 summarize the results of the two surveys. These two figures indicate that every state except two has used HPC in bridge decks or beams. In the 2006–2007 survey, Montana and Mississippi reported that they had not used HPC in beams and bridge decks. However, both states reported using HPC in overlays and Mississippi reported using HPC in substructures. Therefore, every state has used HPC in at least one bridge component. In the 2003–04 survey, 37 respondents reported using of HPC for its low permeability, 30 for its high strength, and 26 for both performance characteristics (Triandafilou 2004). Asked why HPC was being used, respondents ranked deck cracking at less than five years as the most common reason,

12 FIGURE 1 States’ use of HPC in 2003–2004 survey. FIGURE 2 States’ use of HPC in 2006–2007 survey. followed by corrosion of reinforcing steel, cracking of girders and substructure elements, and freeze-thaw damage. The survey indicated that, over the past 10 years, 77% of the respondents had made changes in their bridge deck cur- ing requirements, 72% had made changes in their specified concrete strengths, and 64% had made changes in testing and acceptance requirements. The 2006–07 survey asked about the usage since 2003 of HPC for deck overlays, deck slabs, superstructures, and substructures (Triandafilou 2009). Sixty-four percent of the states reported using HPC in deck overlays, 81% in decks, 62% in superstructures, and 55% in substructures. Only three states reported not using HPC in any of those four components. Rapid chloride permeability values in the range of 1001 to 2000 coulombs were most commonly specified. Compressive strength of 4.0 to 5.0 ksi was most commonly specified for bridge decks and substructures, with 3.0 to 4.0 ksi being the next most common range. For superstructures, the compressive strength range of 8.0 to 10.0 ksi was the most frequently speci- fied, followed by the 4.0 to 5.0 ksi range.

13 The most common procedure of specifying HPC was either by special provisions for a particular project (22 agencies) or a combination of special provisions and general specifications (22 agencies). Only eight agencies used only general speci- fications. Only one agency, Virginia, had made substantial progress with end-result, performance-based specifications. AMERICAN ASSOCIATION OF STATE HIGHWAY AND TRANSPORTATION OFFICIALS SPECIFICATIONS The current AASHTO specifications related to the design and construction of the bridges with HPC include the AASHTO LRFD Bridge Design Specifications, the AASHTO LRFD Bridge Construction Specifications, and the AASHTO Stan- dard Specifications for Transportation Materials and Methods of Sample and Testing. A review of these documents was made as part of the FHWA project: Compilation and Eval- uation of Results from High Performance Concrete Bridge Projects (Russell et al. 2006a). Where sufficient information existed, proposed revisions were developed to facilitate the implementation of HPC (Russell et al. 2006b). The proposed revisions included 17 articles of the LRFD design specifica- tions, 16 articles of the LRFD construction specifications, 15 material specifications, and 14 test methods. In addition, two new material specifications were proposed. Most of the proposed revisions resulted in changes to the specifications and test methods to remove the barriers to the use of HPC. As a result, the AASHTO LRFD Bridge Design Specifications and the AASHTO LRFD Bridge Construction Specifications specifically address the use of HPC as described in the next two sections. AMERICAN ASSOCIATION OF STATE HIGHWAY AND TRANSPORTATION OFFICIALS LOAD RESISTANCE FACTOR DESIGN (LRFD) BRIDGE DESIGN SPECIFICATIONS The Commentary in Article C5.4.2.1—Compressive Strength includes a table of concrete mix characteristics by class (AASHTO 2010b). Class P and Class P(HPC) concretes are intended to be used when compressive strengths in excess of 4.0 ksi are required. A maximum w/cm ratio of 0.49 is specified along with a minimum cementitious content of 564 lb/yd3. For Class P concrete used in or over salt water, the w/cm ratio shall not exceed 0.45. Class P(HPC) concrete is permitted to have a total cementitious materials content up to 1,000 lb/yd3 compared to 800 lb/yd3 for other classes of concrete. When the LRFD Bridge Design Specifications was first developed, its applicability was limited to a maximum design compressive strength of 10.0 ksi for normal weight concrete unless physical tests are made to establish the relationships between concrete strength and other properties. The current specifications allow higher strengths to be used when allowed by specific articles. This has permitted changes to be made as the results of ongoing research on higher strength concrete become available. AMERICAN ASSOCIATION OF STATE HIGHWAY AND TRANSPORTATION OFFICIALS LOAD RESISTANCE FACTOR DESIGN (LRFD) BRIDGE CONSTRUCTION SPECIFICATIONS Article 8.2.2 of the AASHTO LRFD Bridge Construction Specifications defines two classes of HPC (AASHTO 2010a). Class P(HPC) is intended for use in prestressed concrete mem- bers with a specified concrete strength greater than 6.0 ksi. Class A(HPC) is intended for use in CIP construction where performance criteria and concrete compressive strength are specified, or where the concrete is exposed to salt or brackish water or sulfates in soils or water. Minimum cementitious materials content is not included for either class because this should be selected by the pro- ducer based on the specified performance criteria. However, a maximum cementitious materials content of 1,000 lb/yd3 is specified. Maximum w/cm ratios of 0.40 and 0.45 for Class P(HPC) and Class A(HPC), respectively, have been included. For Class P(HPC) concrete, a coarse aggregate maximum size of 0.75 in. is specified, since it is difficult to achieve the higher concrete compressive strengths with larger size aggregates. For Class A(HPC) concrete, the maximum aggregate size is selected by the producer based on the speci- fied performance criteria. Air content for Class P(HPC) and A(HPC) concretes is to be determined from trial tests, but a minimum of 2% is recom- mended in the specifications. For both classes of concrete, trial batches using all the intended constituent materials are required prior to concrete placement to ensure that the specified properties can be achieved and the cementitious materials and admixtures are compatible. For Class P(HPC) and Class A(HPC) concretes, the speci- fications permit any combination of cement and SCMs as long as the properties of the freshly mixed and hardened concrete comply with the specified values. For acceptance of Class P(HPC) and Class A(HPC) con- cretes, the specifications state that any concrete represented by a test that indicates a strength less than the specified strength will be rejected and shall be removed and replaced with acceptable concrete. It also states that the concrete age when the specified strength is to be achieved must be shown on the contract documents, because 56 days or longer may be more appropriate for HPC. For precast concrete with specified concrete compressive strength greater than 6.0 ksi, the specifications require the use of match curing for the test cylinders. The procedures for match curing are described in AASHTO PP-54 (2006). For Class A(HPC) concrete, wet curing is required to commence

14 immediately after finishing operations are complete on any portion of the placement. For bridge decks, the specification requires water curing for a minimum period of seven days irrespective of concrete strength. These curing procedures are required because HPC tends to have very little bleed water. CURRENT DEFINITIONS OF HIGH PERFORMANCE CONCRETE As part of the survey for this synthesis, agencies were asked if they had a definition, formal or informal, for high performance concrete. Twenty agencies supplied a definition. Just under half related to one or more performance characteristic such as per- meability; a similar number related to a prescriptive approach such as minimum amount of SCMs. The remaining defini- tions were more general, such as concrete with a life cycle of 75 years or more based on durability. The individual responses are provided in the answer to Question 1 in Appendix B. From the responses to this one question, it is clear that transportation agencies define or perceive HPC in many different ways. Nevertheless, many responses to other ques- tions indicated that agencies are looking for improved per- formance through the use of lower permeability concrete and reduced deck cracking. The lower permeability is being achieved through the use of SCMs (fly ash, silica fume, slag cement, or a combination). This follows the trend that was started during the demonstration projects where permeability was the most common durability characteristic that was speci- fied. (Russell et al. 2006a.) SPECIFICATION TYPES In general, specifications can be classified as prescriptive, performance or performance-based, or a combination. A pre- scriptive specification for concrete is one in which the recipe to produce the concrete is detailed, listing the ingredients and their relative proportions and generally stated in lb/yd3 of concrete. The specific ingredients are generally required to sat- isfy AASHTO or ASTM material standards, although generic names may be used if a material standard does not exist. A performance or performance-based specification is one in which the requirements are stated in terms of the desired end results rather than specific composition (ACI 2010). Performance specifications are sometimes called end result specifications. The most obvious example of a performance specification is concrete compressive strength. In reality, most concrete specifications are a combination of prescriptive and performance requirements, such as a specified minimum cementitious materials content in combination with a speci- fied minimum concrete compressive strength. The use of performance specifications involves a shift in roles, responsibilities, and risks between owners, designers, and contractors (Chrzanowski 2011). The owner or design engineer becomes responsible for defining the quality char- acteristics and acceptance criteria of the concrete for the intended application. This requires being specific about sam- pling frequency, test methods, acceptance criteria, and pay factors for acceptable and inferior quality concrete. A performance specification provides the concrete supplier with the freedom to design mixes without any restrictions in proportioning except for using constituent materials comply- ing with the specifications (Ozyildirim 2011). However, the concrete supplier may not have the ability or interest in developing the mix proportions. In addition, the cost of the concrete may increase because the risk of not meeting the specification requirements is transferred to the contractor and concrete supplier. With prescriptive specifications, the owner assumes the primary share of the risk (Taylor 2004). The use of a perfor- mance specification, however, does not preclude the use of some prescriptive requirements if this approach is more practical (Carino 2011). Agency specifications consist of several documents. The primary document is often called the “Standard Specifications” and is applicable to all projects. This document is usually updated every few years. Some agencies issue “Supplemental Specifications” to augment the standard specifications. These are issued more frequently than the standard specifications. The third document is called “Special Provisions,’ which is implemented on a project-by-project basis. The special pro- visions generally modify the standard specifications. In the survey for this synthesis, agencies were asked whether they used standard specifications or special provisions to specify HPC. The results are shown in Figure 3. The most frequent response was special provisions for specific projects. The agencies were also asked if their standard specifications and special provisions for HPC were prescriptive, performance based, or a combination. As illustrated in Figure 4, the most frequent response from agencies for which the question was applicable was a combination. Only three states reported using only performance requirements for HPC in their standard specifications and special provisions. New Hampshire and New Mexico use performance requirements in their standard specifications and Maine and New Mexico in their special provisions. LESSONS LEARNED FROM PREVIOUS APPLICATIONS In the survey for this synthesis, agencies were asked how the HPC was performing in comparison with conventional concrete in an effort to determine if the use of HPC was beneficial. The results are shown in Figure 5 for CIP concrete, precast concrete girders, and precast concrete deck panels. It can be noted that only a limited number of agencies have experience with precast concrete deck panels.

15 10 6 15 0 7 0 5 10 15 20 Combination of standard specifications and special provisions for specific projects Combination of standard specifications and special provisions for all projects Special provisions for specific projects Special provisions for all projects Standard specifications only No. of Agencies FIGURE 3 Number of agencies using standard specifications or special provisions for HPC. 15 12 2 8 8 17 2 11 0 5 10 15 20 None Combination Performance Prescriptive No. of Agencies Special Provisions Standard Specifications FIGURE 4 Number of agencies using prescriptive or performance criteria for HPC. 22 16 89 6 5 2 0 0 4 14 23 0 5 10 15 20 25 30 CIP Decks No. of Agencies Better Same Worse Not Applicable Precast Girders Precast Decks FIGURE 5 Relative performance of HPC versus conventional concrete.

16 Overall, 67% of the agencies that reported on performance stated that the CIP HPC was performing better that the con- ventional CIP concrete. Only two agencies reported worse performance. For precast concrete girders and deck panels, all respondents indicated the same or better performance with HPC. Agencies also identified performance issues in using HPC in CIP concrete decks, precast concrete beams, and precast concrete deck panels. For CIP concrete decks, 23 states pointed to specific issues; the dominant one being drying shrinkage cracking. It appears that the use of HPC has resulted in better performance, as shown in Figure 5, but has not eliminated concerns about deck cracking. Other issues mentioned for CIP concrete were workability, regional availability of aggregates, ASR, corrosion, mix proportions, heat of hydration, effect of fly ash on air entrainment, and autogenous shrinkage. For precast concrete, seven states reported performance issues related to corrosion, ASR, shrinkage, slump consistency, consolidation, and cracking. On the positive side, agencies reported many changes in specifications and practices that have resulted in improved concrete performance. These included the following: • Developing special provisions for specific projects, • Using performance-based specifications for bridge decks, • Implementing a specification addressing ASR, • Using high-strength concrete in precast girders to improve durability, • Providing multiple options for concrete constituent materials, • Specifying the amount of drying shrinkage, • Specifying permeability limits, • Starting wet curing immediately after concrete placement, • Specifying and ensuring a longer wet curing period, • Testing for more concrete properties than previously, • Specifying limits on rate of strength gain, • Using lower cement contents, • Increasing the use of SCMs (fly ash, silica fume, and slag cement), • Optimizing aggregate gradation, • Using a corrosion inhibitor, • Using self-consolidating concrete (SCC), • Placing concrete at nighttime, • Controlling evaporation rates, and • Strictly enforcing air content requirements. Increasing the use of SCMs, specifying and ensuring a longer wet curing period, and specifying permeability were listed more frequently than other factors. Agencies were also asked to identify the specifications and practices that were unsuccessful. The following were listed: • High early strength concretes were more prone to deck cracking. • High concrete strengths have led to increased cracking in bridge decks. • The use of silica fume resulted in workability issues and cracking in new decks. • The use of shrinkage-reducing admixtures was success- ful in the laboratory, but the specified air content could not be maintained in the field. • The use of shrinkage-reducing admixtures has helped reduce cracking but not to a satisfactory degree. • Increasing the cement content to obtain lower perme- ability resulted in more cracks in the deck. • Use of an evaporation retarder between final pass of the screed and start of wet curing resulted in excessive crack- ing in decks. • Use of fly ash in bridge decks has resulted in increased cracking. • Limited success was achieved with Class F fly ash content greater than 30%. • The use of 14-day wet cure for decks still results in deck cracking. The responses did not show any general consensus. The one practice that was not successful for three states was the use of silica fume. A comparison of the successful and unsuccessful practices listed above indicates that different states have had conflicting experiences with the same practices. The use of lower cement content in combination with the use of SCMs has resulted in lower permeability concrete but may result in increased deck cracking. Specifying a maximum value for permeability can be beneficial in reducing chloride penetration, but a low value can lead to increased deck cracking. The use of fly ash has been beneficial in some states but detrimental in others. One observation that emerges from the answers to these questions is the need to reduce concrete bridge deck cracking while providing a concrete with low permeability. According to NCHRP Synthesis 333 (Russell 2004), the use of concretes with lower w/cm ratios and SCMs has resulted in concretes with higher concrete compressive strengths, higher tensile strength, higher moduli of elasticity, and lower creep. Although the tensile strength is higher, the higher moduli of elasticity and lower creep have led to an increase in the amount of cracking. This provides the chlorides with an easier path to the reinforcement, thus eliminating the benefits of the low permeability concrete between the cracks. SUMMARY OF THE EVOLUTION OF HIGH PERFORMANCE CONCRETE FOR BRIDGES The first activities towards the use of HPC for bridges began in 1987 with the SHRP project titled “Mechanical Behavior of High Performance Concrete.” This was followed by demonstration projects in 13 states, the formation of an AASHTO Lead States Team for HPC Implementation, and an FHWA HPC Technology Delivery Team. At the same

17 time, revisions were made to the AASHTO LRFD Bridge Design Specifications and the AASHTO LRFD Bridge Con- struction Specifications to remove barriers to the use of HPC. As a result, state DOTs began to move away from the use of prescriptive specifications to the greater use of performance specifications and the use of HPC in bridges. Overall, the use of HPC in bridge decks, precast girders, and precast decks has resulted in better or the same perfor- mance compared with conventional concrete. In the survey for this synthesis, state agencies identified many changes in specifications and practices that led to the improvement in performance. States also identified unsuccessful practices. The major issue with the performance of concrete bridges is the need to reduce deck cracking while providing a con- crete with low permeability. Relatively few issues were identified regarding the use of HPC in precast girders or deck panels.