Turner Fairbank Highway Research > Structure Research > Structure Publications > Table of Contents > Chapter 5

CHAPTER 5. EVALUATION OF “HIGH-PERFORMANCE CONCRETE DEFINED FOR HIGHWAY STRUCTURES”

 

SUMMARY

A review of the FHWA definition of HPC was made to identify whether the performance characteristics, test methods, and range of grades were appropriate and to propose any modifications based on experience with the definition since it was published in 1996. Based on the review, the eight existing characteristics of freeze-thaw durability, scaling resistance, abrasion resistance, chloride penetration, compressive strength, modulus of elasticity, shrinkage, and creep are appropriate, with the addition of alkali-silica reactivity, sulfate resistance, and flowability. However, abrasion resistance and creep should only be specified for special situations. Three grades should be assigned to each characteristic and the values in each grade should be revised to reflect recent data and experience and to raise the performance level of each characteristic. Several modifications to the test methods are suggested.

INTRODUCTION

In 1993, FHWA initiated a national program to implement the use of HPC in bridges. As part of the program, FHWA produced a definition of HPC that identified a set of concrete performance characteristics sufficient to estimate long-term concrete durability and strength for highway structures. Standard laboratory tests, specimen preparation procedures, and grades of performance were suggested for each characteristic. 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.

The definition was published in 1996.⁽³¹⁾ Subsequently, numerous bridges have been built with HPC. The purpose of this article is to report on a review of the FHWA definition in light of experiences and data collected since 1996 and to address the following questions:

• Are the performance characteristics appropriate?
• Are the test methods appropriate?
• Are modifications to the test methods needed?
• Is the range of grades appropriate?
• Do the values assigned to each grade need modification?

FHWA HPC DEFINITION

Performance Characteristics

Eight performance characteristics were used in the definition of HPC to encompass both durability and structural design. The four performance characteristics related to durability are: freeze-thaw resistance, scaling resistance, abrasion resistance, and chloride ion penetration. The four structural design characteristics are: compressive strength, modulus of elasticity, shrinkage, and creep.

Test Methods

For each characteristic, there is a standard test method published by ASTM or AASHTO. The eight performance characteristics and the corresponding test methods are listed in table 12, which is reproduced from table 1 of reference 31.

Since standard test methods sometimes offer different options, the specimens and procedures listed in table 13 (table 2 of reference 31) were adopted. Unless listed otherwise, the following were also stipulated in the original paper:

• Cylinder size: 100 by 200 mm (4 by 8 inches) or 150 by 300 mm (6 by 12 inches).
• Specimen curing: For non-steam-cured products, moist cure specimens for 56 days or until the listed test age, whichever is longer, or match cure and use a maturity meter. For steam-cured products, cure specimens with the member or match cure until test age.

Footnote 2 of table 12 states that all tests are to be performed on concrete samples moist or submersion cured for 56 days, with a reference to table 13 for exceptions. The only exceptions
in table 13 are for creep and shrinkage.

Performance Grades

For each FHWA HPC performance characteristic, a range of two to four performance grades
was established as listed in table 12. A higher number for the grade indicates a higher level
of performance. An empty box in table 12 indicates that the grade is not applicable. With this approach, it is not necessary to specify every characteristic or to specify the same grade for different characteristics. The characteristics and grades should be selected to match the intended application and its environment. For example, a bridge deck supported on girders needs a specified compressive strength, but is unlikely to require specified values for modulus of elasticity and creep. It is not necessary to require all performance characteristics for a given application.

Exposure Conditions

Recommendations for the application of performance grades for the durability characteristics were provided separately as given in table 14 (table 3 of reference 31). Because there is a lack of information correlating field condition severity and laboratory test performance, the relationships shown in table 14 serve only as suggestions and local experience should receive careful consideration in selecting the grades. For scaling resistance, recommendations in table 14 are only provided for grade 1, whereas three grades are listed in table 12.

Table 12. Grades of performance characteristics for high-performance structural concrete.¹

¹ This table does not represent a comprehensive list of all characteristics that good concrete should exhibit. It does list characteristics that can quantifiably be divided into different performance groups. Other characteristics should be checked. For example, HPC aggregates should be tested for detrimental alkali-silica reactivity according to ASTM C 227, cured at 38 °C, and tested at 23 °C, and should yield less than 0.05 percent mean expansion at 3 months and less than 0.10 percent expansion at 6 months (based on Strategic Highway Research Program (SHRP) C-342, p. 83). Consideration should also be paid to (but not necessarily limited to) acidic environments and sulfate attack.

² All tests to be performed on concrete samples moist or submersion cured for 56 days. See table 13 for additional information and exceptions.

³ A given HPC mix design is specified by a grade for each desired performance characteristic. For example, a concrete may perform at grade 4 in strength and elasticity, grade 3 in shrinkage and scaling resistance, and grade 2 in all other categories.

⁴ Based on SHRP C/FR-91-103, p. 3.52.

⁵ Based on SHRP S-360.

⁶ Based on SHRP C/FR-91-103.

⁷ Based on PCA Engineering Properties of Commercially Available High-Strength Concretes.

⁸ Based on SHRP C/FR-91-103, p. 3.25.

⁹ Based on SHRP C/FR-91-103, p. 3.30.

¹⁰ Based on SHRP C/FR-91-103, p. 3.17.

* AASHTO T 2 should be AASHTO T 22.

Table 13. Details of test methods for determining HPC performance grades.

* ASTM C 799 should be ASTM C 779.

Table 14. Recommendations for the application of HPC grades.

	Recommended HPC Grade for Given Exposure Condition
Exposure condition	N/A2	Grade 1	Grade 2	Grade 3	Grade 4
Freeze/Thaw Durability Exposure (x= F/T cycles per year)1	x < 3	3 ≤ x < 50	50 ≤ x
Scaling Resistance Applied Salt3 (x = tons/lane-mile year)	x < 5.0	5.0 < x
Abrasion Resistance (x = average daily traffic, studded tires allowed)	no studs/ chains	x ≤ 50,000	50,000 < x < 100,000	100,000 ≤ x
Chloride Penetration Applied Salt3 (x = tons/lane-mile year)	x < 1	1.0 ≤ x < 3.0	3.0 ≤ x < 6.0	6.0 ≤ x
1 F/T = freeze/thaw. A freeze/thaw cycle is defined as an event where saturated concrete is subjected to an ambient temperature, which drops below -2.2 °C (28 °F), followed by a rise in temperature above freezing. 2 N/A = not applicable. NA indicates a situation in which specifications for an HPC performance grade are unnecessary. 3 As defined in SHRP S-360.

 

PERFORMANCE GRADES IN FHWA SHOWCASE BRIDGES

Information on concrete mixtures, concrete properties, research projects, girder fabrication, bridge construction, live-load tests, and specifications from 19 HPC bridge projects was compiled in task A. The information was placed on a CD-ROM for easy retrieval and viewing.⁽³²⁾

To provide a basis for evaluating the FHWA definition of HPC, data on the eight performance characteristics were extracted from the compilation and are summarized in tables 15 through 18. Tables 15 and 16 list specified and measured durability characteristics for precast, prestressed concrete girders and cast-in-place concrete decks, respectively. Tables 17 and 18 list specified and measured strength characteristics for precast, prestressed concrete girders and cast-in-place concrete decks, respectively. The data are based on a combination of measurements made as part of the quality control tests and measurements made as part of the research that was conducted on the bridges. It should be noted that not all of the tests were made exactly in accordance with the procedures defined in the FHWA HPC definition. Where the procedures were reported in sufficient detail to identify exceptions, the exceptions are noted as footnotes to the tables. A dash indicates that the characteristic was not specified or reported.

The data in tables 15 and 16 indicate that the primary characteristic specified for durability was chloride penetration, with values ranging from 1000 to 3000 coulombs (C). Approximately three-quarters of the specified values are grade 2 in the performance definition. The majority of the measured values are less than 2000 C, with half of the values less than 2000 C corresponding to grade 3 of the definition.

Table 15. Specified and measured durability characteristics
for precast, prestressed concrete girders.

State	Bridge	Freeze-Thaw, %		Scaling		Abrasion, inches		Chloride Penetration, 1C
		Spec.	Meas.	Spec.	Meas.	Spec.	Meas.	Spec.	Meas.
AL	AL 199	—	982	—	—	—	—	—	2,720
CO	Yale Ave.	—	—	—	—	—	—	—	—
GA	S.R. 920	—	—	—	—	—	—	3,000	3,963
LA	Charenton	—	—	—	—	—	—	2,000	1,355
NE	120th St.	—	—	—	—	—	0.0243 0.0404	—	292 377
NH	Route 104	—	1075	—	—	—	—	1,000	1,590
NH	Route 3A	—	—	—	—	—	—	1,500	1,100
NM	Rio Puerco	—	—	—	—	—	—	—	—
NC	U.S. 401	—	—	—	—	—	—	2,000	3,704-4,557
OH	U.S. 22	—	86	—	—	—	—	1,000	2136
SD	I-29 Nb	—	—	—	—	—	—	—	65
SD	I-29 Sb	—	—	—	—	—	—	—	88
TN	Porter	—	—	—	—	—	—	2,5007	390
TN	Hickman	—	—	—	—	—	—	2,5007	496
TX	Louetta NB	—	—	—	—	—	—	—	≤ 1,000
TX	Louetta SB	—	—	—	—	—	—	—	≤ 1,000
TX	San Angelo EB	—	—	—	—	—	—	—	≤ 1,000
TX	San Angelo WB	—	—	—	—	—	—	—	≤ 1,000
VA	Route 40	—	—	—	—	—	—	1,5007	2287
VA	VA Ave.	—	268 918	—	0-18	—	—	1,5007	1257
WA	S.R. 18	—	100	—	—	—	—	1,000	496

¹ Tested at 56 days unless noted otherwise.

² Tested at 14 days after storage in limewater.

³ After 30 min.

⁴ After 60 min.

⁵ Tested at 306 days.

⁶ At 320 days.

⁷ Tested at 28 days after 21 days at 100 °F.

⁸ At 2 months using AASHTO T 161, procedure A, with test water containing a 2% NaCl solution.

Specimens were steam cured and air dried.

1 inch = 25.4 mm

Table 16. Specified and measured durability characteristics for cast-in-place concrete decks.

State	Bridge	Freeze-Thaw, %		Scaling		Abrasion, inches		Chloride Penetration, 1 C
		Spec.	Meas.	Spec.	Meas.	Spec.	Meas.	Spec.	Meas.
AL	AL 199	—	922	—	—	—	—	—	2,873
CO	Yale Ave.	—	—	—	—	—	0.0253	—	5,597
GA	S.R. 920	—	—	—	—	—	—	2,000	198
LA	Charenton	—	—	—	—	—	—	2,000	1,390
NE	120th St.	—	—	—	—	—	0.0334 0.0595	1,800	507 671
NH	Route 104	—	976	—	0-1	—	—	1,000	753
NH	Route 3A	—	—	—	—	—	—	1,000	1,060
NM	Rio Puerco	—	—	—	—	—	—	—	—
NC	U.S. 401	—	—	—	—	—	—	—	—
OH7	U.S. 22	80	80	—	—	—	< 0.02	1,000	—
SD	I-29 Nb	—	—	—	—	—	—	—	461
SD	I-29 Sb	—	—	—	—	—	—	—	1,058
TN	Porter	—	—	—	—	—	—	1,5008	3,280
TN	Hickman	—	—	—	—	—	—	1,5008	3178
TX	Louetta NB	—	—	—	—	—	—	—	1,730
TX	Louetta SB	—	—	—	—	—	—	—	900
TX	San Angelo EB	—	97.9	—	2-3	—	0.04	—	703
TX	San Angelo WB	—	97.3	—	0	—	0.07	—	2,573
VA	Route 40	—	—	—	—	—	—	2,5008	7788
VA	VA Ave.	—	1089	—	0-19	—	—	2,5008	1,4578
WA	S.R. 18	—	—	—	—	—	—	—	2,164-3,43410

¹ Tested at 56 days unless noted otherwise.

² At 14 days.

³ ASTM C 779, procedure A, after 30 min.

⁴ After 30 min.

⁵ After 60 min.

⁶ At 140 days.

⁷ Values for abutments.

⁸ Tested at 28 days after 21 days at 100 °F.

⁹ Tested at 6 months using AASHTO T 161, procedure A, with the test water containing a 2 percent NaCl solution. Specimens were moist cured for 2 months and then air dried.

¹⁰ Between 3.5 and 6.5 months.

1 inch = 25.4 mm

Table 17. Specified and measured strength characteristics for precast, prestressed concrete girders.

State	Bridge	Compressive Strength, 1 psi		Modulus of Elasticity, ksi		Shrinkage, millionths		Creep, millionths/psi
		Spec.	Meas.	Spec.	Meas.	Spec.	Meas.	Spec.	Meas.
AL	AL 199	10,0001	8,440-11,060	—	5,200- 7,1001	—	3702	—	0.182
CO	Yale Ave.	10,000	7,800-14,000	—	5,000	—	5513	—	0.4933
GA	S.R. 920	10,000	13,300	—	4,970	—	—	—	—
LA	Charenton	10,000	10,502-12,023	—	5,976	—	4894	—	0.315
NE	120th St.	12,000	13,944	—	6,280	—	4105	—	—
NH	Route 104	80001	7,7551	—	5,350	—	—	—	—
NH	Route 3A	80001	11,2001	—	—	—	—	—	—
NM	Rio Puerco	10,000	10,151	—	—	—	—	—	—
NC	U.S. 401	10,0001	11,8001-15,000	—	3,8161-5,240	—	470 510	—	0.2146 0.2546
OH	U.S. 22	10,000	9,570-12,920	—	4,647	—	9007 1,050	—	0.2688
SD	I-29 Nb	99001	15,9001	—	7,180	—	1509	—	—
SD	I-29 Sb	99001	13,2501	—	7,03010	—	1409	—	—
TN	Porter	10,0001	9,6511	—	6,660	—	—	—
TN	Hickman	10,0001	10,5291	—	6,239	—	—	—
TX	Louetta NB	13,100	14,4401	—	6,030-6,840	—	39211	—	0.22211
TX	Louetta SB	13,100	14,5501	—		—		—
TX	San Angelo EB	14,000	15,2401	—	5,560-7,370	—	38211	—	0.20511
TX	San Angelo WB	89001	10,1301	—	—	—	29811	—	0.42011
VA	Route 40	80001	9,0601 11,4901	—	5,960-6,660	—	290-50012	—	—
VA	VA Ave.	10,0001	11,2001	—	5,900-6,30012	—	330-45012	—	—
WA	S.R. 18	10,000	12,220	—	—	—	—	—	—

¹ At 28 days. All other values at 56 days.

² 4- by 8-inch cylinders at 180 days after starting at a concrete age of 24 h.

³ At 196 days. Measurements started at a concrete age of 1 day.

⁴ 6- by 12-inch cylinders. Measurements started at a concrete age of 2 days.

⁵ 2- by 2- by 10-inch prism at 64 weeks.

⁶ 4- by 8-inch cylinder at 120 days.

⁷ At 150 days.

⁸ At 180 days after loading at 7 days.

⁹ 6- by 6- by 12-inch prism in outdoor environment.

¹⁰ At 90 days.

¹¹ Tests started at 2 days on 4- by 20-inch cylinders.

¹² At 1 year.

1000 psi = 6.895 MPa, 1000 ksi = 6.895 gigapascals (GPa), 1 millionth/psi = 145.04 millionths/MPa

Table 18. Specified and measured strength characteristics for cast-in-place concrete decks.

¹ At 56 days. All other values at 28 days.

² 2- by 2- by 10-inch prism at 64 weeks.

³ Values for abutments.

⁴ At 124 days.

⁵ 6- by 6- by 12-inch prism in outdoor environment.

⁶ 4- by 20-inch cylinders loaded at 2 days.

⁷ 4- by 20-inch cylinders.

⁸ At 1 year.

⁹ At 64 weeks.

1000 psi = 6.895 MPa, 1000 ksi = 6.895 GPa, 1 millionth/psi = 145.04 millionths/MPa.

Freeze-thaw resistance was specified for only one bridge and scaling resistance and abrasion resistance were not specified for any bridges. However, these characteristics were measured for several bridges. 

The observation that only chloride penetration was frequently specified for durability may reflect reluctance on the part of States to specify other characteristics. This reluctance may be a result of a combination of the following factors:

• Lack of familiarity with the test methods.
• Lack of in-house capability to perform the tests.
• Impact on the cost of concrete when additional characteristics are specified.
• Increased time needed to perform the tests.

For the strength characteristics, compressive strength was the only characteristic specified for girders and decks of all bridges. For the majority of the bridges, the specified strength for the girder concrete was 69 MPa (10,000 psi), corresponding to the lower limit of grade 3. The majority of the measured strengths were in the range of 69 to 97 MPa (10,000 to 14,000 psi), corresponding to grade 3. For the decks, the specified strengths ranged from 28 to 41 MPa (4000 to 6000 psi), except for Georgia, Nebraska, and one bridge in Texas. The range of 28 to 41 MPa (4000 to 6000 psi) is outside the range of the strength performance characteristics for HPC. This is to be expected since there is no reason to specify an HSC for the deck in most slab and girder bridges. For decks, the emphasis should be on durability. An HSC does not ensure a durable concrete.

The observation that compressive strength was the only strength characteristic specified for the HPC bridges probably indicates that it was not necessary to specify modulus of elasticity, shrinkage, or creep for the types of bridges that were built. These characteristics are more significant in long-span structures where control of deflections and prestress losses are more important. Most of the bridges described in this report consisted of cast-in-place concrete deck slabs on precast, prestressed concrete girders. A difficulty in specifying shrinkage and creep criteria is that the tests must be run for 180 days. For most bridges, this means a delay between the time of awarding a contract and casting the HPC members. For shrinkage of deck concrete, selection of a specific value to reduce cracking is somewhat arbitrary because there is no direct correlation between the shrinkage of a laboratory specimen and the likelihood of cracking in the deck.

For most of the bridges listed in table 17, the modulus of elasticity was measured. Values ranged from about 28 to 48 GPa (4000 to 7000 kips/inch²), corresponding to grades 1 and 2.

Although shrinkage and creep were measured on several projects, it is difficult to draw any conclusions since the specimen sizes, curing conditions, and ages at which values are reported varied considerably and often deviated from the procedures described in table 13.

REVIEW OF EXISTING CHARACTERISTICS, GRADES, AND TEST METHODS

General

Various definitions of HPC have been developed over the years.⁽³³⁾ Most of these definitions define HPC in a qualitative manner. The FHWA definition is the first to quantify the definition for a variety of characteristics and provides criteria for HPC. The word definition implies a few words or a phrase, whereas the FHWA definition provides a classification system for HPC and the word classification is suggested for future use.

The existing definition uses a maximum of four grades to classify HPC for each performance characteristic. However, all four grades are not used for each characteristic. This appears to cause confusion. For one characteristic, grade 2 is the highest and, for another, grade 4 is the highest. To simplify the system, it is recommended that each characteristic have three grades.

With the change to provide three grades for each characteristic, revisions to table 14 for freeze-thaw durability and scaling resistance are needed. These should be done by the original authors since they know the basis for the original recommendations.

Another item that has caused confusion is that each grade has an upper and lower limit. This is necessary when determining the grade for a particular test result. However, when specifying a grade, an upper or lower limit may not be appropriate. For example, concrete that is specified to have a scaling resistance of grade 2 is still acceptable if the actual scaling resistance is grade 3. Consequently, the grades should always be considered minimum performance values, even though upper and lower limits are specified. In other words, any actual grade that is higher than that specified is acceptable.

Freeze-Thaw Durability

Resistance to cycles of freezing and thawing is important for structures that can become critically saturated and are exposed to a severe freeze-thaw environment. Thus, it is only important in regions experiencing cycles of freezing and thawing. Without appropriate measures, cracking, scaling, and disintegration of the concrete can occur.

For proper resistance to freezing and thawing, concretes must have sound aggre­gates, a proper air-void system, and have matured prior to freezing. This generally means the development of a compressive strength of about 28 MPa (4000 psi).⁽³⁴⁾ The freezing and thawing damage can occur in both the cement paste and the aggregate.⁽³⁵⁾

Some rocks have pore sizes that are not large enough to expel water, so hydraulic pressures in the aggregates can occur and lead to a distress known as D-cracking. Optimizing the air-void system in the concrete cannot mitigate this distress.⁽³⁶⁾ Proper selection of the coarse aggregate or the reduction of the coarse aggregate size may be needed.⁽³⁷⁾ When the resistance of the paste is of concern, as in the majority of cases, the air-void system can be evaluated instead of testing for freezing and thawing. ASTM C 457 describes the determination of the air-void parameters and can be used to predict the performance of concrete exposed to a severe environment.

The most commonly used test for resistance to freezing and thawing is AASHTO T 161 (ASTM C 666). There are two procedures: Procedure A involves rapid freezing and thawing in water; procedure B requires rapid freezing in air and thawing in water. Procedure A is a severe test and concretes performing well in this test have done well in field applications. However, concretes failing the test may also have satisfactory field performance; however, such performance must be proven in the field. In procedure A, the specimen is subjected to cycles of freezing and thawing much faster than expected in the field. Specimens are also continuously kept moist and are moist cured for two weeks and tested without allowing any drying. Critics raise concerns because of the harsh environment of the test procedure. HPC designed for longevity is less likely to become saturated because of its low permeability. In fact, the ACI 318 Building Code permits the reduction of the required air contents by 1 percent for concretes with specified strengths exceeding 34 MPa (5000 psi).⁽¹⁸⁾ The ACI commentary states that HSC has low water-cementitious materials ratios and porosity and, therefore, improved frost resistance. However, it should be emphasized that strength alone does not ensure freeze-thaw durability and a proper air-void system is required. Also, in procedure A, the specimens are moist cured for 2 weeks without a drying period. In the field, concretes are generally dry before exposure to cycles of freezing and thawing. Therefore, in procedure A, a more realistic approach to curing all specimens (not just HPC specimens) would be to extend the required 2 weeks of curing to a longer time period of 1 or 2 months, and then to let the concrete dry for at least a week before starting the test.

Grade 1 of the current definition has a lower limit for the relative dynamic modulus of 60 percent after 300 cycles. This limit corresponds with the end point given in the test procedure, which was developed for conventional concretes. For HPC, a higher standard needs to be set. Consequently, a new set of values, as shown in table 19, is proposed since HPC structures are expected to have longer service lives.

Scaling Resistance

Scaling is a freezing- and thawing-related deterioration starting at the surface. It progresses into the concrete with additional freezing and thawing. Hydraulic and osmotic pressures cause great stresses that lead to cracking of the concrete. Osmotic pressure is enhanced by the presence of deicing salts. The deterioration is greater for intermediate concentrations of chemicals (3 to 4 percent salt solution by weight) than for lower or higher concentrations.⁽³⁸⁾ The resistance to scaling is improved by providing an adequate air-void system, proper finishing and curing, a drying period before the salt application, and reduced permeability and water-cementitious materials ratios. Air-void analysis (ASTM C 457) can be used to indicate the level of resistance to scaling of concretes properly proportioned, placed, finished, and cured.

The standard test for scaling resistance is ASTM C 672. In ASTM C 672, specimens are subjected to a 4 percent solution of calcium chloride and are rated from 0 to 5 (0 represents sound concrete with no scaling and 5 represents severe scaling with coarse aggregate visible over the entire surface). However, it is also possible to modify the ASTM C 666 test with the addition of sodium chloride to water surrounding the specimens to determine the effects of salt scaling. Virginia DOT uses 2 percent sodium chloride in the test water of ASTM C 666, procedure A.⁽³⁹⁾

Table 19. Proposed grades of performance characteristics
for high-performance structural concrete.¹

¹ This table does not represent a comprehensive list of all characteristics that good concrete should exhibit. It does list characteristics that can quantifiably be divided into different performance groups. Other characteristics should be checked. Only one characteristic is sufficient for HPC.

² For non-heat-cured products, all tests are to be performed on concrete samples moist, submersion, or match cured for 56 days or until test age. For heat-cured products, all tests are to be performed on concrete samples cured with the member or match cured until test age. See table 13 for additional information and exceptions.

³ A given HPC mix design is specified by a grade for each desired performance characteristic. A higher grade indicates a higher level of performance. Performance characteristics and grades should be selected for the particular project. For example, a concrete may perform at grade 3 in strength and elasticity, grade 2 in shrinkage and scaling resistance, and grade 2 in all other categories.

⁴ Based on SHRP C/FR-91-103, p. 3.52.

⁵ Based on SHRP S-360.

⁶ Based on SHRP C/FR-91-103.

⁷ Based on PCA Engineering Properties of Commercially Available High-Strength Concretes, RD104.

⁸ Based on SHRP C/FR-91-103, p. 3.17.

⁹ Based on SHRP C/FR-91-103, p. 3.25.

¹⁰ Based on SHRP C/FR-91-103, p. 3.30.

The lower limit for grade 1 in the current definition is a rating of 5.0. Concrete that has deteriorated to the extent that coarse aggregate is visible over the entire surface should not be considered as HPC. Consequently, a revision to the grades, as shown in table 19, is proposed. A range of ratings is proposed rather than single values since the rating should be the average of at least two specimens.

Sometimes, the use of pozzolans or slag in air-entrained HPC exposed to a severe environment has been attributed to causing increased scaling. Such scaling has been limited to only a very thin surface layer. When pozzolans or slag are substituted for a portion of the portland cement, concrete strength and maturity develop at a slower rate, especially in cold weather. Concretes with lower strengths and less maturity are expected to scale more. Consequently, it is important to ensure that the concretes with pozzolans or slag have the same strength as the control concretes when this test is used for comparison purposes.⁽⁴⁰⁾

Abrasion Resistance

Abrasion resistance is defined by ACI 116R as the ability of a surface to resist wear from rubbing and friction.⁽⁴¹⁾ The abrasion resistance is improved by increasing the compressive strength, using hard and dense aggregates, and proper finishing and curing methods.^(42-44) This property is of great importance in transportation facilities. The traveled surfaces must have adequate skid resistance for proper vehicular control. Skid resistance is affected by both the microtexture provided by the aggregate particles and the macrotexture mainly provided by the grooves formed on freshly mixed concrete or the grooves cut in the hardened concrete (ACI 325.6R).⁽⁴⁵⁾ Deep texture also enables the drainage of water, preventing hydroplaning as the tire loses contact with the pavement surface.

In some locations on the roadway, such as the acceleration and deceleration lanes, tollbooth areas, where chains or studded tires are used, and for elements in water exposed to abrasive material, abrasion resistance is required. Studded tires can cause considerable wear and an NCHRP report addresses pavement wear in the presence of studded tires.⁽⁴⁶⁾ Abrasive materials such as sand are used widely with deicing salts and have caused little damage to quality concrete surfaces. Thus, in bridge structures, the need for an abrasion test is limited. ASTM C 944 is the standard test procedure used for HPC. This test method is similar to procedure B of ASTM C 779. A rotating cutter abrades the surface of the concrete under load. It has been successfully used in the quality control of highway and bridge concrete subjected to traffic. The equipment is not common and many laboratories do not test for abrasion. Also, the test results have a large variability of 12.6 percent for the single-operator coefficient of variation.

It is recommended that this test be specified only for locations where abrasion resistance is a critical issue. No change to FHWA’s abrasion resistance values is recommended. However, clarification is needed in the FHWA definition concerning the force to be applied to the test specimen.⁽³¹⁾ In table 13, a force of 98 N is listed. However, in the text of the original article, a force of 196 N is listed. Both force levels are permitted by the test procedure.

Chloride Penetration

The durability of concrete exposed outdoors depends largely on its ability to resist the penetration of water and aggressive solutions. There are four major types of environmental distress in reinforced concrete: (1) corrosion of the reinforcement, (2) alkali-aggregate reactivity, (3) freezing and thawing deterioration, and (4) attack by sulfates.⁽⁴⁷⁾ Corrosion of the reinforcing steel is the most extensive of these. In each case, water or solutions penetrating into the concrete initiates or accelerates the distress, making costly repairs necessary. Air-entrained concretes that have low penetrability are necessary to resist infiltration of aggressive liquids and provide the necessary resistance to freezing and thawing when exposed to the environment.

The recommended chloride penetrability test for HPC is AASHTO T 277 (ASTM C 1202). In this test, the charge passed in coulombs through a saturated specimen 50-mm (2-inches) thick and 100 mm (4 inches) in diameter, and subjected to 60 volts (V) direct current (dc) in a 6-h period is determined.⁽⁴⁸⁾ Low values indicate high resistance to penetration by solutions. This test gives a good indication of permeability with proper testing, the absence of interferences, and proper interpretation.

All concretes exposed outdoors or cured in a moist room exhibit a reduction in coulomb values with time, and different concretes have different rates of reduction.⁽²³⁾ However, specimens air-dried in the laboratory do not exhibit the expected reduction.⁽⁴⁹⁾ Thus, the method and duration of curing of the concretes subjected to the AASHTO T 277 test are important. Virginia DOT moist cures specimens 1 week at 23 °C (73 °F) and 3 weeks in water at 38 °C (100 °F) to obtain coulomb values similar to those obtained after 6 months of curing at 23 °C (73 °F). The test age of 56 days in the original HPC definition paper is too early for concretes with pozzolans and slag. The classification should allow testing at later ages when the construction schedule permits.

Interferences, such as the presence of calcium nitrite, can produce misleading results from the test method. One option is to test concrete with and without the interfering ingredient and make corrections to the required values. Another option is the new test developed for FHWA. In this new test, the actual penetration of chlorides is measured.⁽⁵⁰⁾ The chlorides are driven into the concrete using a direct voltage similar to AASHTO T 277. Then the sample is split open and treated with a silver nitrate solution. Reaction of the silver nitrate with chlorides forms white silver chloride, which can be visually detected. If there is doubt about the effect of ingredients on the results of the electrical tests, a correlation with the ponding test (AASHTO T 259) is recommended.

The limits for grade 1 in the current definition are 3000 and 2000 C. Data from the FHWA showcase bridges show that most States specified 2500 C or less for both girders and deck. The measured results showed many values less than 1000 C. Consequently, it is recommended that the upper limit be lowered to 2500 C and the ranges for each grade be adjusted accordingly. A new set of values, shown in table 19, is suggested. These are more conservative than the current values since HPC is expected to have a longer service life. Coulomb values should be as low as possible as long as it is practical, economical, and does not have adverse effects on other properties. A possible revision to AASHTO T 277 would allow testing at a concrete age of 28 days, provided that accelerated curing of concrete is used prior to the test. This accelerated test will make it easier to meet the requirements in less time since it produces similar values to tests at later ages.

Compressive Strength

Concrete compressive strength is the only performance characteristic that is always specified for both conventional concrete and HPC. With time, the upper strength level of HSC has increased. However, in the United States, the lower limit has remained at 41 MPa (6000 psi) for many years.⁽¹⁵⁾ This is reflected in the lower limit for grade 1 compressive strength in the current definition.

The primary application of HSC in bridge structures is in precast, prestressed concrete girders. With today’s technology, most precast concrete producers can achieve compressive strengths of 41 MPa (6000 psi) relatively easily. Therefore, it is recommended that the lower limit for the strength characteristic be raised to 55 MPa (8000 psi) and that the following performance grades be used:

Grade 1: 55 ≤   < 69 MPa  (8000 ≤   < 10,000 psi)
Grade 2: 69 ≤   < 97 MPa (10,000≤   < 14,000 psi)
Grade 3:   ≥ 97 MPa (≥ 14,000 psi)

Modulus of Elasticity

The modulus of elasticity of concrete is the ratio of stress to strain in the elastic range of a stress-strain curve. According to ASTM C 469, it is calculated as the slope of a straight line between two points on the stress-strain curve. The upper point corresponds to a stress equal to 40 percent of the measured compressive strength. The lower point corresponds to a strain of 50 millionths. It has long been accepted that the modulus of elasticity is approximately proportional to the square root of the concrete compressive strength, i.e., E_c = 57,000 .

The current lower limit for the modulus of elasticity in grade 1 is 28 GPa (4000 ksi). Using the current conventional equation relating modulus of elasticity to compressive strength, this corresponds to a compressive strength of about 34 MPa (5000 psi). Based on the data shown in table 17, and the proposed changes in the grades for strength, it is recommended that the lower limit for the modulus of elasticity be increased to 34 GPa (5000 ksi) and the following three grades be used:

Grade 1: 34 ≤ E_c < 41 GPa (5000 ≤ E_c < 6000 ksi)
Grade 2: 41 ≤ E_c < 48 GPa (6000 ≤ E_c < 7000 ksi)
Grade 3: E_c ³ 48 GPa (E_c ³ 7000 ksi)

It is also possible to have situations such as bridge decks or seismic zones where a maximum value for modulus may be appropriate.

Shrinkage

Drying shrinkage is a shortening that results from loss of moisture from the concrete. The magnitude and rate of shrinkage depend on many factors, including concrete constituent materials, size of the member, amount of nonprestressed reinforcement, and ambient environment. Consequently, in a test procedure for shrinkage, it is important to specify the conditions for the test. The test procedure, AASHTO T 160 (ASTM C 157), involves measuring the length change of a concrete prism made of concrete similar to that to be used in the actual structure. The length change is generally measured using a comparator. For this test, it is important that the concrete be stored at a constant temperature and humidity.

The curing procedure for the shrinkage specimens states that measurements are to start at 28 days after moist curing. To be consistent with the curing procedures of ASTM C 157, measurements should begin at a concrete age of 28 days. Generally, this includes 27 days of moist curing.

The stated curing procedure in the current definition provides a means to compare the shrinkage of concretes subjected to the same curing procedures. It does not provide data that represent shrinkage in a heat-cured product since heat curing affects the amount of shrinkage. If the intent of the test is to provide a measure of the shrinkage in the product, the specimen curing should be as close as possible to that of the product. This is true for all concretes and needs to be clarified in the procedure.

It should be noted that shrinkage measured in the ASTM C 157 test is that of a small specimen stored at 50 percent relative humidity. Shrinkage in a bridge member is less because the member is thicker and the outdoor relative humidity is generally higher than 50 percent.

Creep

Creep is the change in length of a concrete member when subjected to a sustained load. The amount and rate of creep depend on concrete constituent materials, age and strength of concrete at time of load application, length of time under load, size of the member, amount of nonprestressed reinforcement, and ambient environment. In the FHWA definition, creep is defined as specific creep, which is the change in length divided by the applied sustained stress. It does not include the initial length change that occurs when the load is first applied.

Creep tests in accordance with ASTM C 512 are made using special rigs so that the stress on the specimen remains reasonably constant over the duration of the test. These special rigs are expensive and only a few testing laboratories in the United States are equipped to perform the test. This may explain some of the reluctance to specify the creep characteristic. It is recommended that t