Aggregate Properties and the Performance of Superpave-Designed Hot-Mix Asphalt (2005)

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.

60 CHAPTER 3 SURVEY OF CURRENT STATE AGENCY SPECIFICATIONS 3.1 INTRODUCTION A survey of U.S. state agencies was conducted to assess the adoption of the Superpave consensus and source aggre- gate properties, gradation bands, and volumetric properties. In addition, the survey included questions regarding problematic aggregates, importation of aggregates, and future research needs. A copy of the survey is included in the appendix. In total, 48 agencies—47 U.S. states and 1 Canadian province— responded to the survey. 3.2 SUPERPAVE CONSENSUS AGGREGATE PROPERTIES 3.2.1 Sand Equivalent Test The sand equivalent test (AASHTO T176) is used to iden- tify the quantity of clay-like fines in a sample of fine aggre- gate. Clay-like fines may coat the aggregate such that the asphalt coating the aggregate particles will adhere to the clay instead of to the aggregate particles. In the presence of mois- ture, the asphalt can then separate from the clay-coated aggre- gate, leading to moisture damage. AASHTO T76 is specified by 92% of the states that responded to the survey. Two of those states, California and Texas, specify agency test methods that are essentially the same as AASHTO T176. Nevada specifies AASHTO T90, “Plastic Limit and Plasticity Index.” Mis- sissippi specifies AASHTO T88, “Particle Size Analysis of Soils.” Alaska does not specify a test to address plastic fines. For the states specifying AASHTO T176, 56% specify the same minimum criteria as in AASHTO M323. Four agen- cies specify less restrictive criteria, the lowest being Geor- gia DOT’s specification minimum of 28 for limestone aggre- gates. Minnesota eliminated the sand equivalent requirement for less than 3 million ESALs. The majority of the remaining 10 agencies specify a minimum of 45 for all traffic levels or for up to 30 million ESALs, above which a minimum of 50 is specified in accordance with AASHTO M323. Hawaii requires a minimum of 50 and Arizona a minimum of 55 for all traffic levels. Louisiana only requires the sand equivalent test to be run on natural sands. The frequency distribution of sand equiv- alent specifications by traffic level is shown in Figure 25. Some aggregate producers have noted that crushed fines from some aggregate types can produce clay-size fines, which lower the sand equivalent value. They do not believe these fines to be detrimental. 3.2.2 Fine Aggregate Angularity Superpave specifies the uncompacted voids in fine aggre- gate test (AASHTO T304 or ASTM C1252, Method A) to ensure that the blend of fine aggregates in an HMA mixture has sufficient angularity, texture, or both to provide resis- tance to rutting. Prior to the adoption of the Superpave method, many states had limited the amount of natural sand in mixes. In 1997, 23% of states responding to the NCHRP Project 4-19 survey on aggregate properties specified limits on natural sand (170); however, not all natural sands are rounded. Therefore, it was felt that a test such as AASHTO T304 could better qualify fine aggregate. Of the states responding to the current survey, 85% stated that they specify AASHTO T304 or ASTM C1252 Method A to determine FAA. Arizona uses its own modified version of the test. Oregon ran AASHTO T304 for a period of 2 years and determined that only two sources produced fine aggre- gate with FAA values below 45, so the test was discontinued. Five other state agencies (10%) do not specify any test to measure FAA. All of these states limit the amount of natural sand that may be used in HMA from 0% to 15%. Texas spec- ifies the Hamburg test to measure the rutting propensity of the HMA mixture in lieu of a test to measure FAA. Califor- nia has its own test procedure, CTM 205, which measures the number of crushed particles retained on the 4.75-mm and 2.36-mm sieves. California requires 70% crushed particles in the fine aggregate. Only 51% of the agencies specifying AASHTO T304 spec- ify the requirements outlined in AASHTO M323. The remain- ing agencies’ criteria are summarized in Table 13. Six of the 17 states report relaxed criteria; the remaining states have more stringent criteria. Oklahoma is considering reducing the minimum to 43 for blends that do not include natural sand or gravel. Ontario allows fine aggregate with a minimum uncompacted void content of 43% if the mixture meets all of the volumetric requirements. One reported concern is that states were being forced to import fine aggregate to meet FAA requirements. Of the states responding, 28% reported that they imported some aggregate.

010 20 30 40 50 60 70 80 90 28 35 37 40 45 50 55 Sand Equivalent Value Fr eq ue nc y, % < 3 million 3-30 million > 30 million Figure 25. Frequency distribution of sand equivalent specifications by traffic level. U.S. State Agency/ Canadian Province Minimum FAA Specifications Arizona 42 < 3 M. ESALs, 45 >=3 M. ESALs Colorado 45 for all traffic levels Iowa Use original 7 traffic levels: 40 < 3 M. ESALs, 43 for 3 to 10 M. ESALs, 45 >= 10 M. ESALs Kansas Surface Mixes (< 100 mm from surface): 42 < 3 M. ESALs, 45 > 3 million ESALs, 40 Shoulder Base Mixes (> 100 mm from surface): 42 < 30 M. ESALs, 45 > 30 M. ESALs, 40 Shoulder Kentucky Surface Mixes: Surface Mixes: 40 < 3 M. ESALs, 45 >= 3 M. ESALs Base Mixes: 40 < 30 M. ESALs, 45 >= 30 M. ESALs Louisiana 40 < 3 M. ESALs, 45 > 3 M. ESALs Michigan Same except 43 for 1 to 3 M. ESALs when gradation enters restricted zone Minnesota 40 > 1 M. ESALs, 42 for 1 to 3 M. ESALs, 44 for 10-30 M. ESALs, 45 >= 30 M. ESALS Non Wear Mixes: 40 for all traffic levels Mississippi 4.75 mm NMAS mixes: 45 for all traffic levels 9.5 mm NMAS mixes: 40 < 3 M. ESALs, 44 3 to 10 M. ESALs, No pavements > 10 M. ESALs All other mixes must be coarse graded unless FAA > 44 Missouri Eliminated reduced requirements for > 100 mm from surface in AASHTO M323 Nebraska 40 for low traffic, 43 for medium traffic, 45 for Interstate pavements New Mexico Eliminated reduced requirements for > 100 mm from surface I M323 Ohio 44 for single source or blend, all traffic levels Ontario Allow 43 if mixture volumetric properties are satisfied. Utah 45 for all traffic levels Virginia 45 for all traffic levels, 40 for 9.5 mm NMAS Subdivision mix Washington 45 for all traffic levels Wisconsin 40 < 1 M. ESALs, 43 for 1 to 3 M. ESALs, 45 > 3 M. ESALs TABLE 13 Summary of AASHTO T304 Specifications differing from AASHTO M323

62 Some of these states only imported aggregate close to their borders where it was actually cheaper to bring in material from out of state. Most state agencies stated that they had done so prior to the use of Superpave. Three states reported importing aggregate to meet frictional requirements. Only one state, New Hampshire, cited importing aggregate to meet FAA values. Two other states, Mississippi and Oklahoma, reported that meeting FAA values could be difficult. Three states reported that minimum VMA requirements were difficult to meet— this may be related to the angularity of the locally available fine aggregate. FAA requirements and the restricted zone were designed to limit the amount of rounded natural sand allowed in HMA based on “performance” criteria. However, 46% of the responding states continue to limit natural sand by specifica- tion; 79% of these also have FAA requirements. As shown in Figure 26, the limits on natural sand ranged from 0% to 50% with most falling between 10% and 15%. Some states had more than one criterion, depending on expected traffic, mix type, or frictional properties. Prior to the adoption of the Superpave method, FHWA recommended limiting natural sand to less than 15%. 3.2.3 Coarse Aggregate Angularity The coarse aggregate angularity test is used to measure the number of fractured faces on a coarse aggregate particle according to ASTM D5821 or AASHTO TP61-02. A frac- tured face is defined as a fractured area having sharp edges whose area is equal to at least 25% of the greatest projection (2-D) of the particle. The original ASTM D5821 test method included a provision for a “questionable” pile. The technician could place an aggregate particle in the questionable pile if he or she were unsure that the fractured face was at least 25% of the projection or if the fractured face had been weathered since the fracture occurred. The mass of particles in the ques- tionable pile could not be more than 15% of the mass of the total sample. This provision is still included in AASHTO TP61. AASHTO M323 specifies a percentage of both one and two fractured faces, by mass, to help provide resistance to rutting. The coarse aggregate angularity test is used by 83% of the responding agencies who specify ASTM D5821, AASHTO TP61, or an agency version of the test reported to be similar to ASTM D5821 or AASHTO TP61. Twelve of those agen- cies (25%) specify their own test methods, which are similar to ASTM D5821 or AASHTO TP61. An additional three states (6%) have a test method but did not indicate whether this method was similar to ASTM D5821, and copies of the method have not been obtained. Five states (11%) specify a crushed percentage by definition. Only 14 agencies (39% of those using ASTM D5821 or a similar procedure) reported that their specified criteria matched AASHTO M323. Four states (11%)—Missouri, North Car- olina, Nebraska, and Wyoming—do not use AASHTO’s reduced fractured face requirements for pavement layers deeper than 100 mm (4 in.) in the pavement structure. Altered criteria for the remaining states are shown in Table 14. The altered criteria for one state were not obtainable. Missis- sippi’s specification addresses previously expressed concerns 0 5 10 15 20 25 30 35 0 10 15 20 25 30 50 Maximum Allowable Percentage of Natural Sand Fr eq ue nc y, P er ce nt Figure 26. Frequency distribution of natural sand specifications.

63 regarding the influence of the parent gravel cobble size on the ability to produced high percentages of fractured faces. Quarried aggregates are generally expected to have two or more fractured faces. However, for the states that specify a fractured face test, 81% stated that the test was performed on all aggregates, while 19% stated that it was only performed on gravel sources. 3.2.4 Flat and Elongated Particles Superpave specifies that the percentage of flat and elon- gated particles with a maximum-to-minimum (length-to- thickness) ratio greater than 5 be determined according to ASTM D4791. Flat particles are defined as those particles whose width to thickness exceeds some ratio, typically 51 or 31. Elongated particles are defined as those particles whose length to width exceeds some ratio, similar to flat par- ticles. AASHTO M323 considers both flat particle and elon- gated particle shapes to be susceptible to breakdown during production, placement, and compaction and, therefore, spec- ifies the ratio of the maximum-to-minimum dimension of the particle. ASTM D4791 is specified by 79% of the responding agen- cies. An additional three agencies (6%) use an agency pro- cedure similar to ASTM D4791. Georgia DOT also uses a method similar to ASTM D4791, but instead of defining the minimum dimension of the particle as the maximum thick- ness, it instead defines the minimum dimension as the aver- age thickness. This method is a more restrictive test than is ASTM D4791. West Virginia also has its own test proce- dure. Four states (8%) do not specify any requirements for F&E. One of these states, Colorado, tested all of its sources and found only two sources that exceeded the 51 ratio by more than 2%. The current criterion specified by AASHTO M323 for pavements with more than 1 million ESALs is 10%. Of the state agencies that specify ASTM D4791 or a simi- lar agency test method, 63% specify criteria that match those outlined in AASHTO M323. Kentucky specifies less than 10% of 51 particles for all traffic levels. Connecticut, Hawaii, New Mexico, Utah, and Wyoming specify a maximum of 20% particles exceeding the 31 ratio. In Utah, these criteria only apply to the +9.5-mm material. Texas specifies a maximum of 10% of 31 particles. Minnesota specifies flat or elongated (width to thickness or length to width, respectively) based on a 31 ratio. Only Idaho specifies less restrictive criteria, allow- ing 15% by weight of particles to exceed the 51 ratio. In addition to the 10% 51 requirement, Ontario specifies a maximum of 15% or 20% of particles exceeding the 41 ratio. Six agencies report having more restrictive require- ments for aggregate used in stone mastic asphalt, which are typically a maximum of 20% of 31 and 5% of 51 particles. 3.3 SOURCE PROPERTIES 3.3.1 Introduction Source property levels are not specified in the AASHTO Superpave Specifications. Source property tests are generally related to the durability of the aggregate during construction, wetting and drying, freezing and thawing, and resistance to abrasion under traffic as well as to contaminants such as dele- terious materials. The need for a different level of source prop- Agency Criteria Arizona -/851 for all traffic levels and depths Arkansas 98/80 for all traffic levels/depths Idaho 90/60 for all traffic levels and depths Kansas Added requirement of 50/- for < 0.3 million ESALs at a depth greater than 100 mm. 50/- specified for shoulder mixes regardless of depth Kentucky 75/- < 3 million ESALs, 95/90 3-30 million ESALs, 100/100 > 30 million ESALs Louisianna 75/- low traffic, 95/- medium traffic, 98/- high traffic Minnesota 30/30 < 1 million ESALs, 55/55 1-3 million ESALs; > 3 million ESALs match AASHTO M323 Mississippi -/90 12.5 mm NMAS or smaller, -/80 for 19.0 mm NMAS, -/70 for 25.0 mm NMAS Montana 75/60 0.3-3 million ESALs; > 3 million ESALs match AASHTO M323 New Jersey Surface and Intermediate lifts: 95/90 for low, medium and high traffic levels, 100/100 for very high traffic levels Base lifts: 80/75 for low, medium and high traffic levels, 100/100 for very high traffic levels Utah 95/90 for Category 1 traffic (National Highway System and truck routes); 85/80 for 19.0 and 25.0 mm, 90/90 for 12.5 and 19.0 mm NMAS Category 2 pavements (all others) 1 Percentage of particles with one or more and two or more fractured faces, respectively. TABLE 14 Coarse aggregate angularity criteria for state agencies with altered criteria

64 erties is effected by climate and the ability of various geolo- gies to meet the criteria. 3.3.2 LA Abrasion Test Information on aggregate hardness and its resulting resis- tance to degradation during handling and construction is almost universally measured using AASHTO T96, “Los Ange- les Abrasion Test.” AASHTO T96 may also be related to the expected polish resistance of the aggregate under traffic. The survey indicated that 96% of the responding agencies spec- ify AASHTO T96. California DOT (CalTrans) and Illinois DOT specify their own version of the test. Only two agencies that responded to the survey, Maine and Ontario, do not spec- ify AASHTO T96. Instead, they specify the micro-deval test. In addition, Ontario specifies the British Standard for Polish Stone Value and Aggregate Abrasion Value. The aggregate abrasion value test was developed by Ontario and produces results similar to the LA abrasion test, with more portable equipment (115). Figure 27 shows a frequency distribution of the AASHTO T96 criteria specified by state agencies based on 43 responses. In some cases, the agencies specify varying levels depending on traffic or aggregate type. The frequency distribution in Figure 27 represents the agencies’ most stringent require- ments. The states with multiple levels are shown in Table 15. 3.3.3 Sulfate Soundness Aggregates can deteriorate from wetting and drying or freezing and thawing cycles. The sulfate soundness test simulates the effects of the expansion of water in the aggre- gate pores during freezing. An aggregate sample is satu- rated with either a magnesium or sodium sulfate solution, placed in a drying oven, and dried to a constant mass, which causes the sulfate to crystallize in the aggregate pores. Upon reintroducing the sample into the sulfate solution, the sul- fate crystals expand when they are rehydrated. This expansion is similar to the expansion of water freezing in the aggregate pores. Of the responding agencies, 73% specify AASHTO T104 for aggregate durability in HMA. Sodium sulfate is specified by 64% and magnesium sulfate by 30% of the agen- cies specifying AASHTO T104. Two agencies (6%) allow either sodium or magnesium sulfate. Three agencies specify 0 10 20 30 40 50 60 30 35 40 45 50 55 60 Maximum Loss, Percent Pe rc en t o f A ge nc ie s Sp ec ify in g Figure 27. Frequency distribution of AASHTO T96 specifications. Agency Maximum Loss (%) by AASHTO Kentucky 40 for most, 50 for sandstone and 60 for slag Rhode Island 40 for friction course, 50 for others South Dakota 45 < 0.3 M. ESALs, 40 for 0.3 to 3 M. ESALS, 35 > 3 M. ESALs Utah 35 for Category 1 and 40 for Category 2 Routes Wyoming 35 or 40 depending on class T96 TABLE 15 AASHTO T96 specifications for states with multiple levels

65 freeze-thaw testing. Equipment currently available to easily conduct freeze-thaw testing in the laboratory was not avail- able when the soundness test was developed (115). The distribution of soundness specifications for coarse aggregate is shown in Figure 28. A maximum allowable loss of less than 12% is specified by 53% of the agencies specify- ing sodium sulfate soundness. There is little consensus on the criteria for magnesium sulfate soundness, with values rang- ing from 12% to 30% loss. 3.4 MIX DESIGN PROPERTIES 3.4.1 Gradation Superpave gradation control consists of control points on four sieve sizes: the maximum aggregate size, NMAS, 2.36-mm (No. 8) sieve, and 0.075-mm (No. 200) sieve. Of the responding agencies, 33% have altered the Superpave grada- tion bands. In some cases, these changes are as simple as adding additional control points between the sieves specified by the Superpave method or altering the range for the percent passing the 0.075-mm (No. 200) sieve. In other cases, agen- cies have tightened the Superpave gradation bands to pro- duce mixes that more closely resemble dense-graded mixes used prior to Superpave. One agency has modified the Super- pave gradation bands to include the 16.0-mm sieve, used prior to the introduction of the Superpave method. Of the responding agencies, 25% differentiate between coarse- and fine-graded Superpave mixes. Pavement perme- ability has been a concern with some coarse-graded Superpave mixes. However, only two agencies specify different in-place pavement densities for coarse- and fine-graded Superpave mixes. In addition, Florida DOT includes permeability speci- fications for coarse- and fine-graded mixes. Two other states have permeability specifications for use during design, and four states are considering permeability specifications. 3.4.2 Aggregate Specific Gravity The Superpave method specifies the use of the dry bulk aggregate specific gravity for the calculation of VMA. Of the responding agencies, 89% use dry bulk specific gravity to calculate VMA. Four agencies (9%) use the aggregate effec- tive specific gravity to calculate VMA. The effective gravity is determined using the HMA maximum specific gravity or rice value, asphalt content, and binder specific gravity. The effective specific gravity is always larger than the bulk spe- cific gravity and, therefore, results in a larger calculated VMA. The use of the effective aggregate specific gravity to calcu- late VMA includes the volume of absorbed asphalt as part of the void volume between particles. One agency uses the appar- ent specific gravity to calculate VMA. This would result in a larger calculated VMA then would be determined using either the bulk or effective aggregate gravities. One state does not calculate VMA. 0 10 20 30 40 50 60 9 10 12 15 18 20 30 Loss, Percent Fr eq ue nc y, P er ce nt Magnesium Sodium Figure 28. Frequency distribution of sulfate soundness specifications.

66 3.5 SUMMARY OF AGENCY SPECIFICATION SURVEY One of the goals during the implementation of the Super- pave method was the establishment of consistent test meth- ods and specifications for HMA across the Untied States. Based on the agencies responses, it appears that the first goal has been relatively successful. The majority of the respond- ing agencies specify the Superpave consensus aggregate prop- erties and source property test methods; however, lower per- centages follow the specification values for the consensus properties. Agencies provided little indication as to the rea- soning behind their changes. Past experience was cited in several instances. The aggregate specific gravity is generally determined by the contractor or its representative (65% of respondents). Of the agencies responding, 17% determine the aggregate spe- cific gravity. The dry bulk aggregate specific gravity deter- mined during the design process is used to calculate VMA during production by 52% of the responding agencies. Seven agencies (17%) use the effective specific gravity to calculate VMA during production. Three of these agencies apply a cor- rection factor to the effective gravity to estimate the bulk spe- cific gravity. Four agencies (9%) do not calculate VMA dur- ing production. Other agencies publish aggregate specific gravities to be used for design and production. Only two agen- cies reported measuring the aggregate bulk specific gravity during production.