Evaluation of Indirect Tensile Test (IDT) Procedures for Low-Temperature Performance of Hot Mix Asphalt (2004)

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.

25 GENERAL RECOMMENDATIONS FOR LOW-TEMPERATURE CREEP AND STRENGTH TESTING OF ASPHALT CONCRETE The IDT geometry was originally selected during SHRP for use in low-temperature characterization of asphalt concrete mixtures primarily because the specimen preparation methods available at that time did not include ways of making speci- mens suited for uniaxial measurement of creep compliance, relaxation modulus, or strength. The simple performance tests developed as part of NCHRP Project 9-19 and the dynamic modulus master curve characterization methods for struc- tural design recommended in NCHRP Project 1-37A require specimens 100 mm in diameter and 150 mm high to be used in uniaxial testing. Therefore, this obstacle to uniaxial testing no longer exists. Uniaxial testing would also potentially allow the use of relaxation modulus tests, rather than creep tests, which would eliminate the need to calculate the relaxation modulus from the creep compliance. However, relaxation tests have not been widely performed on asphalt concrete mixtures; and, for practical purposes, the creep test should probably be retained regardless of test geometry. A review of the equip- ment required to perform dynamic modulus master curve test- ing indicated that, with only minor modifications, it could be used to perform low-temperature uniaxial creep tests. This would have several advantages: • Cost savings on purchase of test equipment; • Cost savings on purchase of specimen preparation equip- ment and test accessories; • Cost savings on training engineers and technicians to prepare specimens and perform tests; • Greater reliability of data due to greater experience with a single test geometry and test device; and • Greater flexibility in scheduling testing, if more than one device is needed in a lab. For these reasons, significant effort was expended in the labo- ratory testing of Phase III of NCHRP Project 9-29 to evaluate uniaxial tensile creep testing as the standard low-temperature test for asphalt concrete. Unfortunately, the laboratory testing and analysis indicated that compliance values determined in uniaxial tension were significantly higher than those determined using the IDT test. Furthermore, the correlation between the two sets of data was not extremely strong. In fact, the compliance values deter- mined using the IDT appear to agree more closely with com- pliance values determined in uniaxial compression. Because of the extensive work done on the IDT test and analysis— especially the calibration of the Superpave thermal cracking model to field studies—it is recommended at this time that the IDT creep and strength test be retained as the primary method of evaluating the low-temperature properties of asphalt concrete mixtures. Compliance Measurements Although determining compliance in uniaxial compression is potentially simpler, quicker, and more economical than using the IDT test, these procedures do not provide inter- changeable data. The compliance determined using the IDT test is generally somewhat lower than that determined in uni- axial compression and much lower than that determined in uniaxial tension. This is most likely the result of anisotropy in asphalt concrete specimens prepared using the gyratory compactor. The compliance in the diametral plane appears to be significantly lower than that in perpendicular planes (e.g., along the length of the specimen). Although the gyratory compactor may not always replicate the conditions of field compaction, it seems likely that similar anisotropy exists in situ and that the IDT creep and strength test is probably the best approach to providing estimates of the properties of asphalt concrete in place. Uniaxial compression is suitable for determining creep compliance for research purposes, but it must be realized that the resulting data may not accurately reflect in situ properties or the results of the IDT or other pro- cedures. In general, pavement engineers and researchers should recognize the anisotropic nature of asphalt concrete and make certain that the properties they are using for spec- ification and design purposes are determined using appropri- ate and uniform methods. Strength Measurements The IDT strength procedure as currently described in AASHTO T322 involves using LVDTs to determine the true point of failure and associated tensile strength. This procedure often results in damaged or destroyed LVDTs and is not prac- tical. Phase III of Project 9-29 found that a reasonably good relationship exists between uncorrected IDT strength and IDT strength determined using the more accurate, instrumented CHAPTER 3 INTERPRETATION, APPRAISAL, AND APPLICATIONS

26 procedure of AASHTO T322. It is recommended that the IDT strength test be performed without LVDTs and that the un- corrected strength determined using the maximum load then be adjusted to estimate the corrected IDT strength using the empirical relationship presented in this report as Equation 8. Proposed Changes to AASHTO T322 The requirements of AASHTO T322 have been reviewed, and a number of relatively minor changes have been recom- mended. Specific changes in transducer specifications and most other requirements included in this procedure were pre- sented previously in Table 2. Recommended requirements for specimen dimensions and uniformity were listed in Table 3. A revised loading protocol is given in Table 4. Another important recommendation made in this study is that the temperatures used for low-temperature creep and strength tests should vary according to the stiffness of the mix- ture. For asphalt concrete mixtures made using PG XX-22 and PG XX-28 binders, the current test temperatures of −20, −10 and 0°C should be retained. For mixtures made using PG XX-16 binders or harder, these temperatures should all be increased by 10°C. Similarly, for mixtures made using PG XX-34 binders or softer, test temperatures should be decreased by 10°C. Highly aged mixtures should also be tested at the higher test temperatures. Tensile strength tests should always be performed at the middle creep test temperature. This pro- tocol will help ensure good test precision and will also help avoid problems that occur when the maximum relaxation time in the Prony series is exceeded during analysis of creep data. Precision of the IDT Creep And Strength Tests Anderson and McGennis (3) evaluated the precision of the IDT strength test and reported a standard error for n = 3 repli- cates of about 7 percent. A precision study of the IDT creep tests was performed as part of NCHRP Project 9-29 Phase III, which included numerous mixtures from six different labora- tories. Evaluation of these data resulted in estimated standard errors for compliance for n = 3 replicates of 8 to 11 percent expressed as a percentage of the mean (coefficient of variation, or C.V.). This corresponds to a d2s precision of 22 to 32 per- cent. The laboratory testing executed in this project gave nearly identical results, with an estimated C.V. of 9 percent. The precision for the IDT strength test appears to be accept- able. The precision for the IDT compliance procedure, on the other hand, needs to be improved as part of the implemen- tation process. Further standardization of the procedure and equipment should help achieve improvements in precision. COEFFICIENT OF THERMAL CONTRACTION The equation developed during SHRP for estimating mix- ture coefficient of thermal contraction is not accurate and should be abandoned. Methods for the laboratory measurement of thermal contraction of asphalt concrete mixtures have not been well developed or widely used and are not highly accurate. A simple improved procedure for estimating the co- efficient of thermal contraction was developed in this project and provides reasonably accurate results. ESTIMATING CREEP COMPLIANCE AND STRENGTH VALUES NCHRP Project 1-37A recommended a three-level hierar- chical system for determining inputs for flexible pavement design and analysis. For thermal cracking, IDT creep and strength measurements in accordance with AASHTO T322 are needed for the most reliable, Level 1 determination. Level 2 uses reduced IDT testing at a single temperature; Level 3 is based on typical compliance and strength values for mixtures. In Project 1-37A, predicted thermal cracking based on Level 3 input data did not correlate well with measured thermal cracking for 36 Long Term Pavement Performance (LTPP) sections used to calibrate the Level 3 analysis. Work performed during Phase III of NCHRP Project 9-29 suggests that better Level 2 and 3 thermal cracking input data might be obtained by determining compliance values using the Hirsch model (4) and estimating tensile strength from VFA using Equations 8 and 9. In evaluating the effects of differ- ences in air void content on creep compliance, compliance estimates were made for the mixtures tested in this study using the Hirsch model; binder compliance values were estimated from bending beam rheometer (BBR) test data. Mixtures made with the modified asphalt (PG 76-22) were not included in this analysis, because only one set of BBR data was avail- able, rather than the two needed to develop reasonable creep stiffness estimates over a range of temperatures and loading times. BBR data were empirically adjusted from the Pres- sure Aging Vessel (PAV) to the Rolling Thin Film Oven Test (RFOT) condition, based upon typical test data as reported by Christensen and Anderson in their study of the SHRP asphalt binders (14). The resulting estimated compliance values were in excellent agreement with those measured with the IDT test, as shown in Figure 23. This figure demon- y = 1.04x R2 = 88 % 1.E-07 1.E-06 1.E-05 1.E-04 1.E-07 1.E-06 1.E-05 1.E-04 Compliance, Hirsch, 1/psi Co m pl ia n c e , I D T, 1/ ps i MD DBASE PA GRVL VA GRNT VA LMSTN Equality Figure 23. Compliance values as estimated using the Hirsch model and as measured using the IDT test.

27 strates the feasibility of using estimated compliance values to evaluate the low-temperature properties of asphalt con- crete. Additional effort is needed to determine if critical cracking temperatures estimated in this way agree reason- ably well with those determined using the IDT creep and strength test. If positive results are obtained, the approach should be further developed and documented for possible use in future revisions of the pavement design guide developed in NCHRP Project 1-37A. IMPLEMENTATION Based on the findings from Phase III of NCHRP Proj- ect 9-29, additional efforts to implement AASHTO T322 and the compliance and strength predictive equations developed in Project 9-29 are warranted. Initial plans for these future implementation efforts are presented in this section. AASHTO T322 Three activities associated with AASHTO T322 should be considered. The first involves the incorporation of the changes recommended in Tables 2, 3, and 4 into AASHTO T322. These recommendations as well as Appendix A, which docu- ments them in detail, have already been forwarded to the task force responsible for recommending revisions to this test method to AASHTO. The next logical step in the implementation of AASHTO T322 is the completion of ruggedness testing for the creep testing procedure in AASHTO T322. As outlined below, this is a substantial effort requiring a significant commitment of equipment and resources. Unfortunately, the IDT equipment originally purchased for the Superpave Centers cannot be used in the ruggedness testing because of its documented poor per- formance and the lack of technical support for the equipment. The ruggedness testing should be performed using properly calibrated servo-hydraulic equipment meeting the revised AASHTO T322 requirements. There are two options for gain- ing access to such equipment. The first is to procure second generation IDT devices specifically for the ruggedness test- ing. The second is to contract with laboratories who currently have the equipment meeting the requirements. Guidance on the statistical design of a ruggedness testing program is presented in ASTM C 1067 “Standard Practice for Conducting a Ruggedness Screening Program for Test Methods for Construction Materials.” The standard design tests seven factors that are anticipated to significantly affect the results at two levels. Eight measurements are made using predetermined combinations of the seven factors, and the entire experiment is replicated within a given laboratory. This results in a total of 16 measurements within each labora- tory. Ruggedness testing of the creep procedure in AASHTO T322 is complicated somewhat by the trimmed mean analy- sis approach used in this procedure. In the trimmed mean approach, data from two sides of three specimens are needed to develop a single creep compliance curve. Thus, complete replication of the creep testing procedure at a specific tem- perature requires the collection and analysis of data from three specimens. Table 19 presents one possible scenario for the factors and levels to be used in ruggedness testing for the AASHTO T322 creep procedure. The factors included in Table 19 are based in the research team’s experience with AASHTO T322 and may require modification as additional data on factors affecting IDT creep tests are published by other researchers and practitioners. In addition to the factors and their levels, the ruggedness testing should be conducted over a range of compliances and include mixtures with a range of nominal maximum aggre- gate sizes. Table 20 presents possible mixture combinations and testing temperatures that may be included in the rugged- ness testing. This design includes four mixture/temperature combinations. Ruggedness testing involves a significant level of effort from the participating laboratories. For the design outline above, each participating laboratory would perform 192 creep tests. Assuming that four laboratories participate in the AASHTO T322 creep procedure ruggedness testing experi- ment and that all specimens are fabricated at a single location, the specimen fabrication laboratory will prepare 384 test specimens. Rules of thumb for estimating levels of effort are 1.5 hours for each creep test and 2.5 hours per test spec- imen for fabrication. Thus a ruggedness testing experiment involving 4 laboratories, 2 mixtures, and 2 temperatures will require approximately 2,112 person-hours of testing effort. An additional 400 hours professional time should be bud- geted for initial planning, coordination, data compilation, data analysis, and reporting. The third implementation item associated with AASHTO T322 is future research to better characterize the relationship between uncorrected IDT strength and corrected IDT strength as determined using the procedure given in AASHTO T322. This will provide an improved equation for estimating the corrected IDT strength from the uncorrected strength calcu- lated using the maximum load. An additional 16 mixtures combined with the 16 mixtures tested in this project should Factor Low Level High Level Equilibrium temperature X – 1 °C X + 1 °C Strain level < 0.025 < 0.05 Specimen air voids 5 % 8 % Specimen thickness 40 mm 60 mm Loading strips With neoprene Without neoprene Load application First load Second load End parallelism < 1.0 ° < 2.0 ° TABLE 19 Example ruggedness testing factors for AASHTO T322 creep testing Number Mixture Type Binder Temperature, °C 1 Coarse 9.5 mm PG 76-16 10 and –10 C 2 Fine 25 mm PG 58-28 0 and –20 TABLE 20 Example mixture and temperature combinations for AASHTO T322 creep procedure ruggedness testing

28 provide a very robust data set for the development of an improved predictive model. The data collected in this effort can also be used for the development of improved empirical models for estimating tensile strength from volumetric prop- erties as discussed below. The level of effort for this testing is estimated to be approximately 300 person-hours of testing effort and 160 person-hours of professional effort. Compliance and Strength Predictive Methods In addition to work associated with AASHTO T322, future research is needed to further develop and evaluate procedures for estimating resistance to low-temperature cracking using binder test data and mixture composition through application of the Hirsch model to determine mixture creep compliance and application of empirical methods to estimate strength. Such approaches would be very useful for general mixture selec- tion, mixture design guidance, quality control applications, and as possible replacements for the current Level 2 and 3 thermal cracking data input for the pavement design guide developed in NCHRP Project 1-37A. The major effort for esti- mating creep compliance is the development of methods to predict the binder master curve from limited AASHTO M320 test data. Approximately 240 person-hours of profes- sional effort should be budgeted for this task. The 16 addi- tional mixtures described above when combined with the 16 tested in this project should provide a very robust data set for comparing estimated and measured creep compliance and for developing an improved model for estimating ten- sile strength from mixture volumetric properties. Approxi- mately 220 person-hours of testing effort should be included for conducting creep tests prior to the strength testing described in the preceding section. Finally, approximately 240 person-hours of professional effort should be budgeted for analyzing this data and the strength data and preparing a report documenting the work.