Volumetric Requirements for Superpave Mix Design (2006)

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.

5The purpose of this report is to present the results of two closely related projects: NCHRP Project 9-25,“Requirements for Voids in Mineral Aggregate for Superpave Mixtures” and NCHRP Project 9-31, “Air Void Requirements for Superpave Mix Design.” The objectives of these projects are so closely related that the results cannot be separated in a useful way; voids in the mineral aggregate (VMA), air voids, effective binder content, voids filled with asphalt (VFA), and other fac- tors related to mixture composition are interrelated. This chapter summarizes the objectives of these projects, the scope of the research performed, and the general approach taken in performing the work. The problem statement and research objective that follow paraphrase and, in some cases, directly quote the research project statements (RPSs) for NCHRP Projects 9-25 and 9-31. Problem Statement and Research Objective Problem Statement Before the advent of the Superpave system of mixture design and analysis, 80% of the dense-graded HMA produced in the United States used aggregate gradations that passed above the maximum density gradation—that is, they were fine gradations. Under the Superpave system, most mixtures use coarse gradations that pass below the maximum density gradation. Mixture volumetric requirements developed from the 1960s through the 1980s, including VMA and air voids, were based largely on the performance of fine-graded mix- tures rather than on the typical coarse-graded Superpave mixture. Recent research at the National Center for Asphalt Technology (NCAT) and especially results of the WesTrack study have shown that some coarse-graded Superpave mix- tures can exhibit very poor rut resistance (1, 2). At the same time, durability problems have been observed in a significant number of pavements constructed using Superpave surface- course mixtures. A study in Florida has documented the rel- atively high permeability of Superpave surface-course mix- tures (3); NCHRP Project 1-42 has been initiated to evaluate the increasing occurrence of top-down cracking in hot mix asphalt (HMA) pavements since the implementation of Superpave. A closely related issue is that of design air void content. Since the early 1990s, the Marshall mix design system has allowed design air void content for HMA to vary from 3% to 5% (4). The Superpave system, as originally developed, spec- ified a single design air void content of 4%. In recent years, some agencies have modified the design air void content for Superpave mixtures in order to improve their performance. The Arizona DOT, for example, currently specifies a design air void content of 5% for Superpave mixtures. Evidence suggests that the composition of HMA—as indi- cated by VMA, air void content (total voids in mix, or VTM), effective asphalt content (VBE), VFA, and the ratio of VBE and/or VMA to aggregate specific surface (often expressed as a binder film thickness)—can affect both rut resistance and durability. Effective and efficient guidelines are needed for Superpave volumetric composition to ensure that these mate- rials exhibit adequate levels of resistance to rutting, fatigue cracking, and age hardening. An important related issue, besides how composition affects the performance of Superpave mixtures and the opti- mal ranges in composition for different applications, is how to most effectively and efficiently specify these compositions. This problem is complicated by the inter-relationship of vol- umetric factors such as VMA, VBE, VTM, and VFA and also by controversial terminology such as “binder film thickness,” which some engineers believe to be a useful concept in eval- uating HMA durability, while others strongly believe it to be misleading and potentially useless. Developing specifications involving multiple constraints on mixture compositional fac- tors, without carefully considering the full range of potential mixtures and performance, can produce overly complicated, C H A P T E R 1 Introduction and Research Approach

inconsistent, and/or redundant specifications. Any modifica- tion in the current Superpave specifications should address not only the performance of the resulting mixtures, but also the clarity and efficiency of the resulting specification. Research Objective The research objective for NCHRP Project 9-25 is stated in the RPS: The objective of this research is to develop recommended mix design criteria for VMA,VFA, or calculated binder film thickness, as appropriate, to ensure adequate HMA durability and resist- ance to permanent deformation and fatigue cracking for coarse and fine, dense-graded mixes in the context of the Superpave mix design method. The research objective for NCHRP Project 9-31 is also stated in its RPS: The objective of this research is to recommend for future field validation the range of design air void content, within the con- text of the Superpave mix design method, required for adequate durability and resistance to permanent deformation and fatigue cracking of dense-graded HMA. Scope of Study The laboratory testing for this research was limited in the RPS for NCHRP Project 9-25 to 9.5-, 12.5- and 19-mm nom- inal maximum aggregate size (NMAS) mixtures. Therefore, the laboratory work did not involve 25- and 37.5-mm NMAS mixtures, and the findings of the report tend to be focused more on the properties and performance of surface-course mixtures rather than on base course mixtures. However, it is believed that most of the findings presented in this research are applicable to all HMA, regardless of the aggregate size. The RPSs for both projects required two phases: Phase I, involving a review of literature and current practice; and Phase II, involving laboratory testing and data analysis. Nei- ther RPS contemplated using sections at test tracks, test roads, or other forms of accelerated pavement testing in performing the research. Therefore, mixture evaluations performed dur- ing this research were limited to laboratory tests. However, to verify the results of this research, significant use was made of data previously published from several test tracks/test roads, including WesTrack, the Minnesota Road Research Project (MnRoad), and the NCAT test track (2, 5, 6). Because climate, type and amount of traffic loading,and sub- grade soil types vary enormously across North America, some flexibility is desirable in HMA specifications.For this reason, the findings of this report (given in Chapter 2) are presented in gen- eral terms—equations, graphs, and summary statements describing clearly the effect of changing a particular aspect of HMA composition on rut resistance, fatigue resistance, perme- ability, and age hardening. This presentation should give pave- ment engineers the specific information they need to evaluate potential modifications in their Superpave specifications. Inter- pretation of the research findings (given in Chapter 3) is also described in general and flexible terms, for two reasons. First, many agencies have already implemented a variety of changes in Superpave, so a number of such changes are discussed in Chapter 3 of this report as an aid to agencies that are evaluating the effectiveness of these modifications. Second, some agencies may be considering changes in Superpave requirements,but the nature of such changes will no doubt vary depending on cli- mate, traffic and the nature of local materials; therefore, practi- cal application of the findings of NCHRP Projects 9-25 and 9-31 must consider a variety of scenarios. It is acknowledged that some agencies may be quite happy with the performance of HMA produced according to existing specifications and thus may feel no need for modifying their requirements for VMA,air voids, and related factors. Research Approach The initial phase of both NCHRP Projects 9-25 and 9-31 involved a review of literature and current practice. Because NCHRP Project 9-25 was initiated prior to Project 9-31, the literature review for the latter project was essentially an exten- sion and refinement of the Project 9-25 literature review. Much of the literature review focused on studies in which an attempt was made to relate volumetric properties to one or more performance-related properties. Phase I of NCHRP Project 9-31 included a survey of current practice, in which the manner that state highway agencies were specifying Superpave mixtures was reviewed and summarized. The laboratory testing performed as part of NCHRP Proj- ects 9-25 and 9-31 involved a range of procedures designed to provide information relating various aspects of HMA per- formance to mixture composition. Laboratory tests were per- formed on a variety of HMA mixtures composed of four different aggregate types, three different aggregate gradations, and four different binders. Laboratory tests performed on these mixtures included repeated shear at constant height (RSCH), uniaxial fatigue, permeability, uniaxial compressive strength, indirect tensile (IDT) strength, and dynamic mod- ulus before and after long-term oven conditioning. The results of these tests were analyzed using a variety of methods, typically including an initial graphical analysis, followed by an in-depth statistical analysis. In general, existing models for relating mixture perform- ance to volumetric composition were found to be inappro- priate for use in NCHRP Projects 9-25 and 9-31. In some cases this was because of inaccuracy of the model; in other cases, it was because the model was developed using obsolete 6

parameters, such as asphalt softening point temperature or ductility. Some models, such as the well-known Witczak model for HMA stiffness, rely heavily on parameters such as aggregate gradation data that cannot be directly related to mixture volumetric composition (7). Because of the short- comings of existing models for estimating mixture perform- ance, newer models were developed or existing models refined during the course of NCHRP Projects 9-25 and 9-31. As much as possible, these mathematical models were selected (or designed) to reflect reasonable theoretical and/or physical models for the given mode of distress. The fatigue model proposed in this research, for example, is based largely on continuum damage theory, reduced through mathematics and calibration with laboratory data to a simple formula comparable with traditional empirical equations for flexural fatigue life. To refine the models developed from analysis of laboratory data and to verify their validity in application to field data, in most cases they were applied to other data sets generated in independent research. These data sets included performance data from WesTrack, MnRoad and the NCAT test track and also permeability data collected in a study by the Florida DOT on the permeability of Superpave mixtures (2, 3, 5, 6). After refining/verifying the proposed models for estimating mixture performance from mixture composition, plots were developed showing different aspects of performance as a function of VMA, air void content, and related characteristics. These plots and the underlying analyses were further analyzed and summarized in terms of typical effects of changing VMA, VTM, VBE, and related factors on performance. These spe- cific findings are presented in Chapter 2. The final stage in the analysis involved interpreting these findings in terms of prac- tical applications to HMA mix design technology. This dis- cussion is presented in Chapter 3. The most significant sections of this chapter involve discussion of how recent changes in HMA mix design have affected pavement per- formance, discussion of how possible modifications of Super- pave requirements might affect performance, and general guidelines for modifying HMA specifications to improve fatigue resistance and durability. As discussed under Scope of Study, the findings and recommendations of this research have been presented in a format designed to provide for some flexibility in implementation so that when modifying their Superpave specification, agencies can effectively address local conditions and materials. As suggested in NCHRP report guidelines, the body of this report includes only the most important technical information and related findings, con- clusions, and recommendations. 7