Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments (2017)

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.

5 C H A P T E R 2 This chapter presents the process for developing the proposed EPG specifications. The pro- posed specifications for performance-graded emulsion used in chip seal, microsurfacing, and spray seal treatments are presented in Attachments 1 and 2, respectively. Details of the work performed in this research are presented in Appendices A through E. 2.1 Critical Distresses and Characterization of Material Properties Related to PST Performance The current specifications for the selection of emulsions for use in PSTs are based on material properties that are not related directly to the performance of PSTs (Vijaykumar et al. 2013). The proposed EPG specifications for asphalt emulsions will address this shortcoming by directly relating the binder properties to each PST type’s mixture performance. The research team conducted an extensive review of information that is relevant to asphalt binders used in PSTs, including the following: • Typical asphalt materials used in PSTs. • Key measures of PST performance. • Existing PST specifications. • Critical emulsion and residual binder properties that relate to the key performance measures of PSTs. • Test methods and equipment for measuring critical emulsion and residual binder properties. The literature review included published and unpublished reports as well as contacts with public and private agencies, industry organizations, and other domestic and foreign sources. Key findings from the literature review are presented as they relate to the identification of the critical distresses in PSTs and the performance-related mechanisms that govern the appropriate asphalt binder selection for each PST type. Test results are also provided to explain the basis for each of the test methods in the proposed EPG specifications. 2.1.1 Typical Materials Used for Surface Treatments Table 2.1 provides a summary of common materials that are known to be used in the construc- tion of PSTs. Materials were identified from a survey of maintenance engineers in the United States and Canada, a review of the literature, and the research team’s experience (Gransberg and James 2005, Caltrans 2009). The research team did not acquire and test all the materials listed in Table 2.1. Section 2.2 provides the rationale for selecting the materials used for EPG specification development. Results, Interpretation, and Applications

6 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments 2.1.2 Distress Types in Asphalt Surface Treatments That Are Affected by Binder Properties Based on the literature review as well as information gleaned from communications with pave- ment preservation professionals, a list of key binder performance characteristics (or distresses) that relate to PST mixture performance was compiled. 2.1.2.1 Chip Seal Distresses Three performance characteristics (or distresses) were found to be relevant to the assessment of chip seal surface treatment performance and the development of the EPG specification. These characteristics in order of importance are the following: 1. Raveling (aggregate loss) 2. Bleeding 3. Rutting (in multilayered seals) Table 2.1. Common materials used for surface treatments.

Results, Interpretation, and Applications 7 For this prioritization process, the research team considered only those distresses that relate directly to the binder material’s performance. For example, reflective cracking was not considered a critical distress despite its common occurrence in chip seals, because it relates to the condition and integrity of the existing pavement structure and not to the performance of the asphalt binder. The results of the survey and the literature review revealed that raveling and bleeding are the most critical performance characteristics associated with chip seal surface treatments. Raveling (or aggregate loss) is the most critical performance characteristic because the loss of aggregate particles can cause damage to vehicles and is a safety concern. Early raveling can result from the inadequate curing of an emulsion prior to traffic opening and from insufficient strength development at the time of traffic opening. Late raveling results from long-term trafficking effects at intermediate and low temperatures. At an intermediate temperature, the aggregate loss is due primarily to the loss of the bond between the asphalt and the aggregate, whereas at a low temperature, aggregate loss is due to the cohesive fracture of the brittle asphalt residue. Wet raveling is defined as raveling that occurs due to moisture damage to the asphalt binder. Moisture damage can lead to a significant reduction of the adhesive bond between the residue and the aggregate and can expedite raveling at an intermediate temperature. The proposed EPG specifications address low-temperature raveling, but the research team recommends that early, late, and wet raveling should be addressed in the mix design stage, not as part of the emulsion grading process. Loss of aggregate also contributes to pavement bleeding, which is ranked as the second most critical distress, as bare (bled) spots provide little frictional resistance. Excessive bleeding in the wheel path can lead to loss of friction and can become a safety hazard for drivers. Although the terms bleeding and flushing often are used interchangeably in the literature, the mechanism behind flushing is construction/design-related (e.g., excessive emulsion is applied or there is insufficient aggregate), whereas bleeding is related directly to the viscoplasticity of the binder at high temperatures. Therefore, only bleeding is considered as a critical distress in the development of the asphalt binder EPG specifications for chip seals. In multilayer chip seal surface treatments, such as triple seals, permanent deformation (or rutting) can occur under repeated loading in the wheel path. During rainy conditions, these ruts fill with water, which can lead to hydroplaning. Rutting resistance is related also to the viscoplastic nature of the asphalt binder. The survey results and the rationale for these rankings are provided in Appendix B. 2.1.2.2 Microsurfacing Distresses The following mixture performance characteristics (or distresses) are critical for the assessment of microsurfacing and are ranked as follows: 1. Raveling 2. Bleeding 3. Rutting 4. Thermal cracking Similar to the prioritization of chip seal mixture performance characteristics (or distresses), only distresses that relate directly to binder performance were considered for microsurfacing. Although raveling has been identified as a critical distress for microsurfacing, this performance characteristic is driven by the chemical interaction between the emulsion and the aggregate and cannot be captured by binder testing alone. This fact was confirmed through microsurfacing testing (summarized in Appendix D). Reflective cracking also is not considered a distress for microsurfacing because it relates to the structural integrity of the existing pavement and not as

8 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments much to the binder’s properties. The rutting potential in microsurfacing is related to the visco- plasticity of the residual binder at high temperatures. Thermal cracking in the mixture is related to the stiffness and thermal stress relaxation capabilities of the binder at low winter tempera- tures. The binder properties that are related to these critical mixture distresses were evaluated to determine the appropriate binder properties to include in the EPG specifications for characterizing microsurfacing mixture performance. 2.1.3 Review of Existing Specifications for Surface Treatments The existing specifications for grading binders used in surface treatments rely on the non- performance-related specifications detailed in ASTM D977 and ASTM D2397 for anionic and cationic emulsions, respectively. These emulsion standards are used to determine emul- sion properties such as viscosity, storage stability, demulsibility, coating ability, penetration, ductility, water resistivity, and solubility tests for slow-, quick-, medium-, and rapid-setting emulsions. Under the current grading system, emulsions are named based on their charge, setting behavior, viscosity, and residue properties. Cationic emulsions start with “C” followed by the setting behavior, which could be slow-setting (SS), quick-setting (QS), medium-setting (MS), or rapid-setting (RS). However, the names for anionic emulsions start with the setting behavior. The charge and setting time designations are retained in the proposed EPG specifications. The number after the setting behavior designation indicates the viscosity of the emulsion: “1” for low emulsion viscosity or “2” for high emulsion viscosity. At the end of an emulsion’s name, “h” indicates that it contains hard residue asphalt. For example, CSS-1h emulsion is cationic (C), slow-setting (SS), has low emulsion viscosity (1), and hard (h) residue asphalt. CQS-2 is a cationic quick-setting (CQS) emulsion with high emulsion viscosity (2). In addition, “P” or “L” at the end of an emulsion’s name indicates that it has been either polymer-modified or latex- modified, respectively. Emulsions that meet the existing specifications nonetheless often exhibit performance problems due to various deficiencies in the specification test requirements. For example, the current specifications (1) do not grade binders based on the material properties that are related directly to critical performance measures for each surface treatment type (e.g., aggregate loss and bleeding potential in chip seals and rutting and thermal cracking potential in microsurfacings); (2) do not fully address the fresh emulsion material properties that are related to the construc- tability of asphalt materials used in surface treatments (e.g., sprayability for chip and spray seals, drainout for chip and spray seals, mixability for microsurfacings, and curing time for spray seals); (3) do not test materials across the full temperature range to which these materials may be exposed (e.g., they do not test below 25°C or above 50°C, thereby failing to address performance at typical low- and high-temperature grades based on climate); and (4) do not grade binders based on traffic level. These shortcomings are discussed in Section 2.1.4 and addressed in the proposed EPG specifications for surface treatments. 2.1.4 Material Properties Related to Performance and Test Methods Used in Proposed EPG Specifications In developing the proposed EPG specifications for surface treatments, test methods were identified that address the fresh emulsion properties that are related to the storage and con- structability of PSTs, as well as binder properties that are related to PST mixture performance. This section details the material properties and the associated test methods recommended in the proposed EPG specifications. Figure 2.1 presents the test methods in the proposed EPG specifications.

Results, Interpretation, and Applications 9 2.1.4.1 Fresh Emulsion Testing The properties of an emulsion during its storage, transportation, and construction have a major effect on the performance of asphalt surface treatments. Table 2.2 lists the test methods that were modified to evaluate the properties of fresh asphalt emulsions that affect the overall stability, constructability, and performance of PSTs in the EPG specification. Asphalt emulsions used in the construction of PSTs are multiphase systems that are com- posed of water, asphalt binder, emulsifier, and, in some cases, modifiers. The different phases are expected to remain homogenous during storage, pumping, transportation, and construction. Storage stability is defined as the ability of an emulsion to resist significant changes in properties MSCR Fresh EmulsionResidue Min Jnr (TH) G* at Critical phase angle DSR Temperature- Frequency Sweep RV Sprayability, mixability, drain- out, storage stability (Supplier Spec) Low Temperature Aggregate Loss (Chip Seals) & Thermal Cracking (Microsurfacing) Bleeding (Chip Seals) & Rutting (Microsurfacing) Workability & Stability Figure 2.1. Recommended EPG specification tests for emulsions. Property Test Method Parameter(s) Measured Storage Stabilitya Modified ASTM D6930 –Selement and Sedimentaon A Rotaonal viscosity, η B 24-hour separaon rao (Rs) C 24-hour stability rao (Rd) Sprayabilityb Modified AASHTO TP 48 –Rotaonal Viscometer Rotaonal viscosity, η, @ high shear rate Drainoutb Modified AASHTO TP 48 –Rotaonal Viscometer Rotaonal viscosity, η, @ low shear rate Mixabilityc Modified AASHTO TP 48 –Rotaonal Viscometer Rotaonal viscosity, η, @ 5 rpm Curing Timed Modified ASTM D3121 – Rolling Ball Test Rolling distance, me to 25 cm rolling distance a Applies to all PSTs b Does not apply to microsurfacing c Applies to microsurfacing only d Applies to spray seals only Table 2.2. Test methods for fresh asphalt emulsion.

10 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments over time (Redelius and Walter 2006). In the proposed EPG specifications, storage stability is measured using a modified version of ASTM D6930 in which the conditions are consistent with ASTM D6930, but instead of comparing the asphalt content of the emulsion that is extracted from the top and bottom of the sample following storage, the viscosity of the top and bottom portions of the stored specimen is evaluated using a rotational viscometer, as viscosity is related directly to the performance of fresh emulsions. In addition, the middle portion of the sample is remixed, and the viscosity is then measured and compared to a reference, unconditioned sample to assess stability. Two parameters are reported: (1) the separation ratio (Rs), which is the ratio of the viscosity of the top and bottom portions of the specimen and is used to evaluate the resistance of the emulsion to sedimentation and creaming (Bahia et al. 2001) and (2) the stability ratio (Rd), which is the ratio of the viscosity of the remixed emulsion and origi- nal emulsion samples after storage and is used to evaluate the emulsion’s ability to be remixed following storage. If an emulsion is not storage-stable, difficulties will be encountered during construction (e.g., clogging of nozzles) and a non-uniform product will result. The viscosity of the emulsion is also critical during chip seal and spray seal construction and is characterized by the sprayability and drainout parameters. Sprayability is the ability of an emulsion to be sprayed in a uniform thickness across the surface of an existing pavement (Asphalt Institute 2008). An emulsion that is too viscous will result in streaking, spot bleeding, and partial loss of the cover aggregate in the chip seal. For spray seals, the emulsion must be fluid enough so that it can penetrate and fill the surface cracks. Drainout is defined as the ability of an emulsion to resist draining off the pavement surface via gravity after spraying (Bahia et al. 2008). High drainout leads to premature aggregate loss and reduces the amount of binder that is available for proper aggregate embedment. Sprayability and drainout were measured using a three-step shear test that employs a rota- tional viscometer. The test subjects an emulsion to three successive shear rates to quantify its thixotropic and shear thinning behavior. An initial low shear rate simulates the circulation of the emulsion in a tank; a second step, at a high shear rate, simulates spraying through a nozzle; and a third step, at a low shear rate, simulates the flow under gravity once placed. Sprayability is assessed by the viscosity value at the high shear rate. Drainout is assessed by the viscosity value in the third low-rate shear step. Sprayability and drainout are not included in the microsurfacing section of the EPG specifica- tions because emulsion is not sprayed during microsurfacing construction. For microsurfacing emulsions, workability is characterized by mixability, which is defined as the viscosity of fresh emulsion measured at 25°C at a low shear rate (i.e., 5 rpm). Because microsurfacing emul- sions are mixed with wet aggregate and water during construction at an ambient temperature, a conservative temperature of 25°C is used for this test (i.e., viscosity values would be lower at higher temperatures). The storage stability test includes measuring a reference (unconditioned) sample for its viscosity at 5 rpm. Thus, the measurement of the viscosity of the reference sample obtained during the storage stability test can be used to evaluate mixability without conducting additional testing. The curing time for spray seals is a critical aspect of the construction process. Spray seal treatments require the emulsion to cure to a certain level before the surface can be trafficked; the tackiness of the spray seal is indicative of the curing rate. If the road is reopened to traffic before sufficient curing has taken place, tracking of asphalt on tires can occur. The proposed EPG specifications for spray seals uses the rolling ball test (modified ASTM D3121) at 25°C to address this issue. The rolling ball test is used to determine the tackiness of an emulsion as a function of curing time. The test method involves rolling a steel ball down an inclined ramp at a 21° angle and then measuring the distance the ball rolls across the emulsion. As the emulsion’s tackiness decreases, the distance the ball rolls increases. The curing time that is required for the

Results, Interpretation, and Applications 11 steel ball to roll 25 cm is used for the EPG specifications and is the minimum curing time to avoid the tracking of spray seal emulsion onto tires upon traffic opening. 2.1.4.2 Existing Emulsion Specification Tests Retained in the EPG Specifications Table 2.3 lists the existing AASHTO emulsion test methods for fresh emulsion and recovered residue that are retained in the proposed EPG specifications. 2.1.4.3 Chip Seal Residual Binder Testing Table 2.4 lists the critical distresses for chip seal treatments and the test methods that have been identified for evaluating each residual asphalt binder distress for use with the EPG speci- fications. The test methods allow for the failure properties of the asphalt emulsions at high and low temperatures to be evaluated under conditions that reflect various climatic and traffic condi- tions incorporated in the EPG specifications. The test methods were identified based on previous research (Epps et al. 2001, Bahia et al. 2008, Hanz et al. 2010, King et al. 2010, Hanz et al. 2012, Schuler et al. 2011) and were found to differentiate between emulsion types and to evaluate the failure properties of asphalt emulsion residue that relate to the critical chip seal distresses. These test methods require only one piece of equipment: a DSR. Proposed specifications for the test methods listed in Table 2.4 are provided in Attachment 2. High Temperature The multiple stress creep and recovery (MSCR) test, as specified in AASHTO T 350, was employed in this study to assess the bleeding resistance of residual binders at high-temperature performance grades (PGs). This test method is used to determine the non-recoverable creep compliance (Jnr) value of the residual binder to quantify the bleeding resistance of the residue. Two stress levels are included in the procedure: 0.1 kPa and 3.2 kPa. The 3.2 kPa stress level subjects the binder to high stress conditions, similar to those experienced under field traffic loading; therefore, this stress level is incorporated in the proposed EPG specifications. For this study, the residual binders used for MSCR testing were recovered using AASHTO PP 72 Method B because this method requires only a small amount of residue for MSCR testing. Also, this method Emulsion Test Subject AASHTO Test Procedure Fresh Emulsion Tests Demulsibility T 59 Par cle charge T 59 Sieve T 59 Residue recovery PP 72 Method B Residue percentage Recovered Residue Tests Solubility T 44 Float T 50 Table 2.3. Existing emulsion tests retained in proposed EPG specifications. Table 2.4. Chip seal EPG specifications: test methods for asphalt emulsion residue.

12 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments requires a thin film that allows the residue to be recovered using a low curing temperature (60°C) and only 6 hours of curing time. Furthermore, AASHTO PP 72 Method B preserves the structure of polymer- and latex-modified emulsions during recovery and does not require significant aging of the material (Kadrmas 2010). Low Temperature The DSR frequency sweep test was selected to characterize the low-temperature fracture resis- tance of the emulsion residue in the EPG specifications following a thorough literature review and analysis of the test data. Low-temperature aggregate loss occurs as a result of the cohesive failure of the binder rather than the adhesive failure between the aggregate and the binder. To assess the low-temperature fracture risk in a binder, residual binder rheological properties were obtained through DSR frequency sweep testing at temperatures ranging from 5°C to 15°C. The parameter measured from the DSR frequency sweep test for chip seal binders is the dynamic shear modulus (G*) at the critical phase angle, which was found during testing to correlate to low-temperature aggregate loss in the chip seal mixture; this finding is detailed in Section 2.4.2.3. For this test, specimens were fabricated using residual binder recovered using AASHTO PP 72 Method B. To determine the dynamic shear modulus (G*) values at the critical phase angle of inter- est, DSR frequency sweep tests were conducted at both 5°C and 15°C regardless of PG, that is, the critical phase angle values for which the G* value was evaluated varied as a function of the low-temperature PG rather than the test temperature. However, consistent with other performance-grading specification criteria, specification limits for G* values are independent of climate. 2.1.4.4 Microsurfacing Residual Binder Testing Table 2.5 provides a summary of the critical performance characteristics for microsurfacing treatments and the test methods used in the proposed EPG specifications that have been identified for evaluating the residual asphalt binders. The selected binder test methods allow for the failure properties of asphalt emulsions at high and low temperatures to be evaluated under conditions that simulate the various climatic and traffic conditions given in the EPG specifications. High Temperature At high temperatures, rutting and bleeding are critical distresses in microsurfacing treat- ments. However, the results of experimental studies (summarized in Appendix C) show that, if the residual asphalt rate is held constant, the emulsion type does not significantly impact the skid resistance of microsurfacing mixtures. In the experimental study, all the mixtures that were fabricated using a variety of microsurfacing emulsions were tested using the British pendulum test (BPT) to measure skid resistance. The BPT number measured for each specimen was then converted to the equivalent locked wheel skid test (LWST) number. All the microsurfacing mix- tures passed the established minimum skid number threshold of 30 (Jayawickrama et al. 1996). Performance Characterisc Test Method Parameter(s) Measured Rung at High-temperature EPG Mul ple Stress Creep andRecovery Test  Non-recoverable Creep Compliance, Jnr Thermal Cracking at Low- temperature EPG Dynamic Shear Rheometer Frequency Sweep Test  Dynamic Shear Modulus (G*) at δcrical Table 2.5. Microsurfacing EPG specifications: test methods for asphalt emulsion residue.

Results, Interpretation, and Applications 13 Ultimately, rutting was identified as the critical high-temperature performance distress in microsurfacing treatments that relates to emulsion residue performance. The MSCR test for the microsurfacing binder EPG specifications follows the test used for chip seal bleeding character- ization. AASHTO PP 72 Method B is the residue recovery method recommended for residue testing at both low and high temperatures in the microsurfacing EPG specifications. Low Temperature For low temperatures, thermal cracking was identified as the most critical distress for micro- surfacing treatments. To assess the thermal cracking potential, DSR frequency sweep tests were conducted using emulsion residue in the temperature range of 5°C to 15°C, and the results were used to determine the dynamic shear modulus (G*) value at the critical phase angle. During the low-temperature testing (summarized in Section 2.4.2.5), the dynamic shear modulus (G*) at the critical phase angle was found to correlate well with thermal cracking in the micro surfacing mixture. 2.2 Test Materials for the Development of the Performance-Graded Specifications In order to develop comprehensive EPG specifications for asphalt binders used in PSTs, a set of asphalt emulsions used for chip seals, spray seals, and microsurfacing in various parts of the United States was utilized for testing and subsequently for the development of specification limits. Emulsions used in this project were obtained from seven major emulsion manufacturers that were selected to provide representative coverage of all climatic regions in the United States. Each supplier was required to meet the following requirements: • Provide standard emulsion formulations used for PST construction projects that have not been altered specifically for this research. • Own a laboratory mill for emulsion production. • Retain the original base asphalt used in the initial emulsion formulation in case more emulsion is needed for additional research. • Utilize the same emulsion formulation for each batch of emulsion produced. That is, if a CRS-2 emulsion was requested early on in the project and then needed for subsequent testing later, the supplier should produce the same emulsion formulation (base asphalt, emulsifier, etc.) as the initial CRS-2 provided. The asphalt emulsions used for this study are commercial-grade emulsions that have been uti- lized in functioning PSTs in the United States. These emulsions are expected to resist significant performance problems under the climatic and traffic conditions for which they were designed. The inclusion of such emulsions was required to ensure that the EPG specifications are not so restrictive as to preclude the use of asphalt emulsion formulations that have performed well in existing projects, that is, the research team expected that, at a minimum, most of the standard emulsions would pass the specification criteria at the lowest traffic level for certain climatic conditions. The experimental plan also included “poor-performing” emulsions. A poor-performing emul- sion is defined as a standard emulsion that has been altered by the emulsion supplier to exhibit inferior performance compared to that of the original product, while still meeting the current emulsion specification requirements. These emulsions were altered to be poor performing with regard to specific performance criteria, but not all of the performance characteristics were evaluated. The modifications of these emulsions were left up to the emulsion supplier’s dis- cretion. The individual manufacturers supplied brief details about the modification(s) they made and the specific characteristic(s) for which they expected poor performance. In cases

14 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments where the performance modification was relevant to the test results, the poor-performing emul- sion is highlighted in the results and analysis. For example, one supplier changed the emulsifier type and content of its standard emulsion such that the emulsion’s demulsibility, defined as the ability to release water or the rate at which a liquid separates from an emulsion, approached 100% in order to produce an inferior product. A high demulsibility value results in an emulsion that has a very quick rate of water release (or faster curing/breaking rate) and does not allow the bond between the aggregate and the emulsion to form properly during the chip seal fabrication process. This reduction in bond strength causes a greater aggregate loss potential in the mixture. This particular poor-performing emulsion did not meet the recommended test limits for bond strength and was deemed unsuitable for chip sealing in any climate or at any traffic level. Table 2.6 details the emulsion manufacturers’ descriptions of ways the poor-performing emulsions were developed and the effects that the supplier expected in terms of performance. It should be noted that not all of the emulsions were tested, as the modification caused specimen fabrication problems during testing for some of the emulsions. A single granite aggregate source was utilized for all the chip seal and microsurfacing specimens to ensure that any variation in the aggregate source properties (e.g., gradation, nominal maximum aggregate size, fractured faces, flakiness, etc.) did not compromise the quality of the mixture performance data and also to allow isolation of the binder effects in terms of the mixtures’ performance. The aggregate material was obtained at the onset of the experimental work and stored for use throughout the entire study. 2.2.1 Selection of Emulsion Materials Selecting a representative set of emulsions was required for developing the EPG specifica- tions. First, the total number of emulsions that could be acquired and tested was determined for each PST type. Next, survey information was utilized to ensure that a nationally representative subset of emulsion types was selected. Figure 2.2 presents example survey results for chip seals (Gransberg and James 2005). This distribution of the emulsions was used to distribute the types of emulsions acquired for this study. Figure 2.2 shows that CRS-2 emulsions constitute approximately 25% of the chip seal emulsions used by survey respondents the United States, and thus, 4 of the 15 standard emulsions used in developing the EPG specifications were CRS-2 emulsions. PST Type Emulsion Type Descrip on of Modifica on to Induce Poor Performance Expected Performance Effect Chip Seal PP-CRS-2P-E Increased chemical content and reduced demulsibility. Should reduce aggregate retenon performance slightly. PP-CRS-2-A Changed emulsifier type and emulsifier content from commercial formulaon. Increased demulsibility to 100%. Emulsion will break too fast, causing poor aggregate bonding. Microsurfacing PP-CQS-1h-E Low chemical dosage. Emulsion will break too fast. PP-CSS-1hP-B Wrong emulsifier selecon. Poor compability. PP-CSS-1h-F High emulsifier content (5%). High viscosity, increased curing me, poor aggregate bonding, and increased abrasion loss. Spray Seal PP-CSS-1-B Base asphalt is too so‹ (240–250 mm penetraon). So‹ base asphalt. Table 2.6. Asphalt emulsion manufacturer modifications for poor performance.

Results, Interpretation, and Applications 15 After identifying the emulsions required for the development of the EPG specifications, the emulsions were allocated to individual suppliers. The individual suppliers were asked to fabricate only emulsions that they typically formulate and provide for in-service projects. Thus, the research team could identify the emulsion types that each supplier typically formulates for in-service chip seal projects, establish the distribution of the emulsion types, and allocate the required emulsions to the individual suppliers. The production and subsequent delivery of the emulsions were staged to avoid the prolonged storage of any emulsion and avoid the risk of the emulsion breaking prior to testing. A maximum storage time of 2 weeks was established during which fresh emulsion testing, residue recovery, and mixture preparation were all completed. Based on an estimate of the amount of fresh emulsion that could be handled within the 2-week period, a schedule for emulsion delivery was prepared. Emulsions were received in 5-gallon batches and were stored in calibrated forced-draft ovens until testing was complete. 2.2.2 Emulsion Acquisition The emulsions tested during this research were acquired in two phases. In Phase 1, an initial set of emulsions, listed in Table 2.7, was acquired for performance testing. These emulsions were used for all the fresh emulsion, binder, and mixture initial testing. However, analysis of the test results revealed that changes to the low-temperature test procedures were necessary. However, by that time, these emulsions were no longer useable because they had been stored much longer than 2 weeks (the maximum allowed duration of emulsion storage according to ASTM D977). Although each emulsion supplier had agreed to retain the base asphalt used in the initial formu- lations in case additional amounts of any emulsion needed to be produced during the project, many suppliers were unable to provide the same original formulations upon request. Therefore, a smaller second subset of emulsions, listed in Table 2.8 and referred to as Phase 2 emulsions, was acquired to develop the low-temperature chip seal and microsurfacing EPG specifications. Note that in Table 2.7 and Table 2.8, the letter in parentheses following the emulsion name identifies the emulsion company that supplied the material for testing in this project. The “PP” designation before the emulsion name in Table 2.7 indicates that the emulsion was produced to be poor performing. 0 10 20 30 40 % o f S ur ve y R es po nd en ts Asphalt Emulsion Grade AC CR S-2 CR S-1 RS - 1 RS - 2 HF RS - 2 HF RS - 2P HF MS - 2P CR S-2 h CR S-1 hP CR S-2 P/L Figure 2.2. Asphalt binder grades typically used in chip seals.

16 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments 2.3 Performance-Graded Specifications Framework This section describes the conceptual framework of the EPG specifications in terms of the intended use of the EPG specifications, temperature grading concepts, traffic designations, binder grade designations, and overall conceptual development. 2.3.1 Intended Use of the EPG Specifications The EPG specifications provide specifications for performance-graded emulsions for use in chip seals, microsurfacing, and spray seals. The EPG specifications were designed to determine whether an asphalt emulsion is suitable for use in constructing a PST that is capable of effectively resisting critical distresses under specific climatic and traffic conditions. The EPG specifications incorporate performance test methods to guide the selection of asphalt emulsions for use in these applications. The EPG specifications assume that no pre-existing structural distresses are present within the pavement structure on which the PST is to be applied. Prior to the application of a PST, any pre- existing problems in the pavement should be remedied in order to achieve acceptable structural performance, as PSTs do not increase the structural capacity of the pavement structure. Because PSTs do not contribute to the structural capacity of pavements, the EPG specifications do not Chip Seal Microsurfacing Spray Seal CRS-2 (A) CSS-1h (A) CRS-2 (A) CRS-2P (A) CSS-1h (B) SS-1 (B) HFRS-2P (A) CSS-1h (C) CSS-1 (B) CRS-1 (B) CSS-1hP (C) Revive (E) CRS-1h (B) CSS-1hP (D) SS-1h (E) RS-1 (B) CQS-1h (E) CSS-1h (E) CRS-2P/L (C) CQS-1hP (E) PP-CSS-1 (B) HFRS-2 (C) CSS-1hL (F) PP-CSS-1h (E) CRS-2 (E) CSS-1h (F) CRS-2P (E) CSS-1hL (NC) CRS-2P-hP (E) PP-CSS-1h (F) CRS-2 (F) PP-CSS-1hP (D) CRS-2L (F) PP-CSS-1h (B) CRS-2 (NC) PP-CQS-1hP (E) CRS-2L (NC) PP-CRS-2 (A) PP-HFRS-2 (C) PP-CRS-2P (E) Table 2.7. Phase 1 emulsions acquired for EPG specification development by PST type. Chip Seal Microsurfacing CRS-2 (NC) CSS-1h (C) CRS-2L (NC) CSS-1hP (C) CRS-2 (F) CQS-1h (E) CRS-2L (F) CQS-1hP (E) CRS-2P (A) CSS-1hP (D) CRS-2L (C) CSS-1hL (F) Table 2.8. Phase 2 emulsions tested for EPG specification development by PST type.

Results, Interpretation, and Applications 17 include material property testing for critical distresses, such as fatigue and reflective cracking, which are related to the strength and integrity of the pavement structure. The only distresses addressed within these EPG specifications are those distresses that are not related directly to pavement structural integrity, are not the result of pre-existing pavement issues (e.g., rutting, cracking, etc.), or are not caused by poor construction practices. In addition, the EPG specifications do not provide recommendations for the design and construction of the PST. The end-user of the EPG specifications is responsible for determining the appropriate design and construction practices for the PST type and location. 2.3.2 EPG Specification Design Temperature Grade Concepts 2.3.2.1 High Temperature High-temperature climatic HMA PG requirements for a given location are determined using a model based on a depth of 20 mm, yearly degree days above 10°C, and assumed threshold for allowable rut depth of 12.7 mm (Mohseni et al. 2005). Climatic grades are typically determined based on 98% statistical reliability using climatic data from a representative set of years. Reliability is defined as the probability that in a single year the actual pavement surface temperature will exceed the design pavement surface temperature. Because the proposed EPG specifications use surface temperature instead of temperature measured at a 20-mm depth, as recommended in the existing HMA PG specifications, the research team conducted a study to determine the appropriate temperature grades for emulsions used at the surface. The rutting damage model approach used to determine HMA high-temperature PG requirements does not include provisions for depth correction. However, the Long-Term Pavement Performance (LTPP) High Pavement Temperature Model (Mohseni 1998) indicates that the difference between the temperature of a pavement at its surface and that at a depth of 20 mm is 3.9°C. Therefore, the research team determined that in the EPG specifications, surface temperature grade increments should be shifted 3°C from the existing HMA PG specification grade increments. While a shift of 4°C would adhere more closely to expected temperature gradients, it is deemed more practical to use a 3°C shift since existing HMA PG specifications are based on 6°C increments. For example, instead of keeping the existing HMA performance high-temperature grades of PG 58, PG 64, PG 70, etc., the EPG specifications recommend the use of high-temperature grades of EPG 61, EPG 67, EPG 73, etc. The main purpose of this 3°C shift is to keep the EPG map for emulsions similar to the PG map for HMA binders, except for the 3°C difference. Figure 2.3 presents a map of the United States that shows the theoretical climate-based PGs measured at a depth of 20 mm, as defined in the current HMA performance specifications. EPG requirements will be 3°C higher than the grade requirements indicated in Figure 2.3. It is important to note that this change in the specification grade does not mean that the base asphalt PG used to fabricate the emulsion would necessarily be higher than the PG of asphalt binders currently used successfully in the field. In fact, all of the emulsions tested in this research were made from an original base asphalt with a high-temperature PG of either 58°C or 64°C, yet every emulsion tested had a higher temperature grade in the EPG specifications than the original base asphalt PG used to fabricate the emulsion (Table 2.9 provides a summary of this finding). Section 2.4.2.1 provides discussion of the methodology used to determine the high-temperature PGs in the EPG specifications. 2.3.2.2 Low Temperature The low-temperature grade in the EPG specifications is defined as the annual 1-day minimum pavement surface temperature (similar to the PG specifications for asphalt binder used in HMA).

18 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments High PG at 20 mm Depth Figure 2.3. Theoretical high-temperature performance grade map based on current HMA specifications. Table 2.9. Original base asphalt versus EPGs (EPG specifications) at high temperature. Therefore, the depth correction required at high temperature does not apply to low-temperature grade determination. However, it would make sense to shift both the high- and low-temperature grades from the HMA PG specifications to the EPG specifications for simplicity. Therefore, temperature data from the LTPP database were analyzed to determine what the grade requirements for a given location would be if the grade thresholds were increased by 3°C. When determining the grade requirement for a given location, the continuous grade is rounded (conservatively) to the nearest available grade. For example, if the low-temperature continuous grade is determined

Results, Interpretation, and Applications 19 to be -18°C, the HMA low-temperature grade requirement is -22°C. The corresponding EPG low-temperature grade requirement if available low-temperature PGs were increased 3°C is -19°C. Results of the LTPP database analysis are presented in Figure 2.4. Figure 2.4 (a) presents the nationwide theoretical HMA low-temperature grades based on climate data at 98% reliabil- ity, and Figure (b) presents the low-temperature EPG grade with thresholds shifted 3°C from HMA grades (e.g., -25°C as opposed to -28°C). Figure 2.4 (a) and (b) appear very similar, indicating in most cases that shifting current HMA low-temperature grades by 3°C for the EPG Low EPG Thresholds Low PG Thresholds (a) (b) Figure 2.4. Maps of LTPP program climate data showing (a) theoretical low-temperature performance grades for current HMA specifications and (b) low-temperature performance grades for EPG specifications.

20 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments specification is acceptable, and using the HMA low-temperature grade requirement plus 3°C as the EPG low-temperature grade requirement is appropriate. 2.3.3 Traffic Designations for the EPG Specifications Because the performance of PSTs such as chip seals and microsurfacings depends upon the traffic conditions that the PST undergoes, the developed EPG specifications provide material test limits for grading binders according to the expected traffic. A survey of emulsion professionals, a literature review, and the research team’s experience have led to categorizing traffic classes that are proposed in the EPG specifications into three traffic classes, according to average annual daily traffic (AADT) volume for vehicles, as follows: • Low Traffic: 0–500 AADT • Medium Traffic: 501–2,500 AADT • High Traffic: 2,501–20,000 AADT The research team recommends 20,000 vehicles as the upper AADT limit for high traffic based on a study of high traffic chip seal practices across the United States (Gransberg and James 2005) as well as the research team’s experience. According to this study, chip seals are commonly con- structed at AADT counts that can exceed 20,000 vehicles in California, Colorado, and Montana. Also, the research team has constructed chip seals at an AADT exceeding 15,000 vehicles with no reported performance issues (Im 2013). However, the performance of chip seals constructed at high traffic volumes is heavily dependent on local factors such as climate, traffic speed, aggre- gate quality, contractor’s experience, equipment, etc. Therefore, the high traffic upper limit is conservatively set at 20,000 vehicles. One main reason that conservative traffic volumes are adopted is that, although the binder properties in the developed EPG specifications are impor- tant for the performance of a chip seal, other factors need to be considered carefully in chip seal construction, especially in high traffic situations. It is recommended that aggregate with a highly uniform gradation, at least two crushed faces, high durability (as measured by the Los Angeles Abrasion Test), and low dust proportion (< 2% passing the No. 200 sieve) should be used. Also, a performance-based mix design should be used to ensure an appropriate initial aggregate embedment depth. Lastly, optimal construction practices (e.g., optimized roller types/patterns and proper traffic opening time and sweeping protocol, etc.) should be utilized to ensure construction quality. The selected high traffic upper limit also applies to micro- surfacings, which are known to be utilized at traffic volumes up to 20,000 vehicles in California (Caltrans 2009). 2.3.4 Binder Grade Naming Designations for EPG Specifications The developed EPG specifications grade binders as part of a performance-grading system that consists of high- and low-temperature grade designations that relate to climate. The grad- ing system uses a single letter traffic designation to denote the AADT range at which the binder is graded to perform acceptably. The EPG specifications retain the currently used designations for the emulsifier charge and set rate. In the chip seal EPG specifications, for example, a sample emulsion grade may be CRS-EPG 67-19M, which is defined as follows: CRS = Cationic rapid-setting; the binder type designation denotes the charge, specialty float type if any (i.e., high float), and setting rate for the binder. EPG = Emulsion performance grade. 67 = Average annual 7-day maximum pavement surface temperature (°C). –19 = Minimum pavement surface temperature (°C). M = Medium traffic volume; traffic volume categories are low (L), medium (M), or high (H) in the EPG specifications.

Results, Interpretation, and Applications 21 2.3.5 Underlying Concept and Outline for the Development of the EPG Specifications 2.3.5.1 Fresh Emulsion Concept Statistical analysis was conducted to develop the specification test limits for fresh emulsion properties under the assumption that the standard emulsions tested in this study represent a broad range of materials that perform acceptably in service. Both standard and poor-performing emulsions, as defined previously, were tested in developing these EPG specifications. However, because the poor-performing emulsions were fabricated by the emulsion suppliers intentionally to exhibit inferior performance, test data for the poor-performing emulsions were not considered in determining the fresh emulsion specification limits using statistical analysis, but were used to validate the EPG specifications (the emulsions that were designed to have poor stability failed the storage stability test). Note that the poor-performing emulsions would not necessarily fail all the EPG specification tests, but rather these materials are expected to perform poorly when tested for the specific material property that directly relates to the cause of the poor performance. A box-and-whisker statistical analysis procedure that displays the distribution of data and identifies outliers statistically within a normally distributed dataset (Sim et al. 2005) was utilized for deriving limits for the fresh emulsion properties. The test results for the poor-performing emulsions were removed from the dataset. The outliers among the test data were then deter- mined based on the box-and-whisker plot method for normally distributed data, as shown in Figure 2.5 and as described below: • The mean, median, and standard deviation were calculated. • The sample minimum and the sample maximum were determined. • The first quartile, median, and third quartile were calculated based on the dataset. • The inter-quartile range (IQR), which is the difference between the top of the first quartile and the bottom of the fourth quartile, was calculated. • The outliers among the dataset were considered when the results were greater than the third quartile + 1.5*IQR or less than the first quartile – 1.5*IQR. • The outliers were removed from the dataset and the remaining data were considered in order to determine the specification limit. Once the outliers were removed from the dataset, the specification limit was determined based on the 98% reliability concept coupled with engineering judgment. 0 25 50 75 100 125 150 Te st P ar am et er Third Quartile First Quartile Median Outlier Outlier Maximum Minimum Figure 2.5. Example of box-and-whisker plot method used to remove outliers.

22 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments 2.3.5.2 Residual Binder Concept For developing the specifications for residual binders, the relationship between each binder property measured for the proposed residual binder test method was compared against the mixture performance to which the binder property relates. Based on this relationship, the cor- responding binder test limit could be determined to ensure that the target mixture performance was achieved. Figure 2.6 shows an example of the underlying concept used in the EPG specifica- tions for residual binders. The MSCR test results are plotted against mixture bleeding for both modified and unmodified binders. A temperature-independent relationship was found between the binder property, i.e., the non-recoverable creep compliance, Jnr, and bleeding performance in the chip seal mixtures that were fabricated using a consistent performance-based mix design (Kim and Adams 2011). Here, the upper limit for bleeding performance was used to define the maximum Jnr value for the MSCR test. The underlying concept is that if a binder meets this MSCR Jnr limit, the bleeding of the mixture should not exceed the maximum bleeding threshold due to inadequate binder performance. 2.3.5.3 Process for the Development of the EPG Specifications The overall process for developing the EPG specifications for surface treatments is summarized as follows: • Identify the test methods required for the proposed EPG specifications. • Obtain test materials. • Perform asphalt material tests for the assessment of emulsion and binder performance properties at the critical temperature range associated with each related PST distress. • Perform PST mixture performance tests at the critical temperature range associated with each PST distress. • Develop relationships between PST binder properties and mixture performance. • Establish threshold values for fresh emulsion and residual binder properties. y = 11.161ln(x) + 67.126 0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 % B le ed in g Jnr (in kPa–1) Unmodified 46C Modified 46C Unmodified 52C Modified 52C Unmodified 58C Modified 58C Binder Property M ix tu re P er fo rm an ce Target Mix Performance Corresponding Binder Property Figure 2.6. Example of underlying concept used in EPG specifications for residual binders.

Results, Interpretation, and Applications 23 • Develop proposed EPG specifications. • Develop draft specifications for test methods used in the proposed EPG specifications. 2.4 Performance-Graded Specification Limits This section describes the validation of the test methods in the EPG specifications and the development of preliminary specification limits. The results presented herein were derived using data obtained from the set of emulsions listed in Section 2.2.1. This EPG specification frame- work is similar to the Superpave PG framework for hot asphalt binder, which prescribes a set of performance-related test methods with specification limits to cover critical distresses and constructability. Grades that dictate the climatic and traffic conditions for which the emulsion can be used were determined by the test temperatures at which the specification limits pass or fail. To develop the specification criteria that relate to constructability for fresh emulsions, information from the existing literature, statistical analysis, and engineering judgment was used to identify appropriate limits. To develop the specification criteria that relate to critical distresses for residual binder properties, relationships were developed between the binder properties and mixture performance over representative temperature ranges associated with each distress. Binder speci- fication test limits were derived based upon established mixture performance thresholds. All fresh emulsion, binder, and mixture test methods discussed are fully detailed in the attachments and appendices of this report. 2.4.1 Fresh Asphalt Emulsion EPG Specification Limits This section presents the methodology used for deriving the specification limits that relate to the stability and constructability of fresh emulsions used in chip seal, microsurfacing, and spray seal treatments. In addition, the procedures for establishing specification limits for each fresh emulsion material property addressed in these EPG specifications are described in this section. The fresh emulsion is representative of the material’s state during storage, transport, and appli- cation prior to completion of the breaking and curing processes. 2.4.1.1 Fresh Emulsion Limit Derivation Figures 2.7 through 2.12 present the statistical procedures that were used to determine the preliminary specification limits for each of the fresh emulsion properties included in the EPG specification. Figures 2.7 through 2.9 show examples of how the fresh emulsion specification test limits were determined using the test data obtained from the storage stability and separation ratio tests conducted using chip seal, microsurfacing, and spray seal emulsions, respectively. For the data shown in Figure 2.7 (a) and (c), outliers were determined and removed based upon the box-and- whisker plot method. The specification limits were then determined based on the 98% reliability concept and engineering judgment [see Figure 2.7 (b) and (d)]. Engineering judgment was used in determining the stability ratio specification limits for the chip seal emulsions, as only a maximum limit was necessary because a stability ratio close to 1 is desirable, and all the test measurements exceeded 1. A maximum stability ratio limit of 2 was determined for chip seals based on the data presented in Figure 2.7. From this limit derivation approach, the acceptable range of separation ratios for chip seal emulsions was determined to be 0.5 to 1.5. The same approach was used to derive limits for microsurfacing and spray seals, as shown in Figure 2.8 and Figure 2.9, respectively.

24 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Se pa ra tio n Ra tio Emulsion Number Unmodified Modified Outlier (a) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 S ep ar at io n R at io Emulsion Number Unmodified Modified 98% Reliability (b) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 St ab ili ty R at io Emulsion Number Unmodified Modified Outlier (c) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 St ab ili ty R at io Emulsion Number Unmodified Modified 98% Reliability (d) Figure 2.7. Examples of (a) outlier identification for separation ratio data, (b) separation ratio specification limit determination, (c) outlier identification for stability ratio data, and (d) stability ratio specification limit determination for chip seal fresh emulsions. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 7 8 9 Se pa ra tio n Ra tio Emulsion Number Unmodified Modified (a) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 7 8 9 Se pa ra tio n Ra tio Emulsion Number Unmodified Modified 98% Reliability (b) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 7 8 9 St ab ili ty R at io Emulsion Number Unmodified Modified Outlier (c) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 7 8 9 St ab ili ty R at io Emulsion Number Unmodified Modified 98% Reliability (d) Figure 2.8. Examples of (a) outlier identification for separation ratio data, (b) separation ratio specification limit determination, (c) outlier identification for stability ratio data, and (d) stability ratio specification limit determination for microsurfacing fresh emulsions.

Results, Interpretation, and Applications 25 Figure 2.10 and Figure 2.11 show examples of the specification limit determination for the sprayability and drainout constructability parameters that are applicable to chip seals and spray seals in the EPG specifications, respectively. A maximum limit for sprayability viscosity in chip/ spray seals in the EPG specifications ensures that the binder has a low enough viscosity value to ensure consistent output when sprayed through a nozzle during the emulsion application process. Failure to meet the sprayability threshold values would result in streaking in a chip seal or spray seal treatment. A minimum drainout limit also is needed in the EPG specifications to ensure that after application the emulsion has a high enough viscosity level to resist flow due to the slope of the road. Failure to meet the drainout threshold would result in uneven material application rates along the roadway as well as performance problems. Figure 2.12 shows the specification limit determination for mixability, which is the parameter that is applicable to micro- surfacing. Because microsurfacing emulsions are mixed with the aggregate prior to application, defining the maximum viscosity allowable to ensure good mixability during microsurfacing construction is required (see Figure 2.12). 2.4.1.2 Fresh Emulsion Limit Summary Table 2.10 presents a summary of the specification test limits that were derived for the fresh emulsions used in chip seal, microsurfacing, and spray seal treatments in the EPG specifications. 2.4.2 Residual Binder Test Limits This section describes the methodology used to derive the binder specification criteria that relate to the performance of chip seals and microsurfacing treatments with respect to critical 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 Se pa ra tio n Ra tio Emulsion Number Unmodified (a) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 Se pa ra tio n Ra tio Emulsion Number Unmodified 98% Reliability (b) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 St ab ili ty R at io Emulsion Number Unmodified Outlier (c) 0.0 0.5 1.0 1.5 2.0 2.5 3.0 1 2 3 4 5 6 St ab ili ty R at io Emulsion Number Unmodified 98% Reliability (d) Figure 2.9. Examples of (a) outlier identification for separation ratio data, (b) separation ratio specification limit determination, (c) outlier identification for stability ratio data, and (d) stability ratio specification limit determination for spray seal fresh emulsions.

-200 0 200 400 600 800 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Vi sc os ity (c P) Emulsion Number Sprayability Unmodified Modified (a) -200 0 200 400 600 800 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Vi sc os ity (c P) Emulsion Number Sprayability Unmodified Modified 98% Reliability (b) -400 0 400 800 1200 1600 2000 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Vi sc os ity (c P) Emulsion Number Drainout Unmodified Modified Outlier (c) -400 0 400 800 1200 1600 2000 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Vi sc os ity (c P) Emulsion Number Drainout Unmodified Modified 98% Reliability (d) Figure 2.10. Examples of (a) outlier identification for sprayability data, (b) sprayability specification limit determination, (c) outlier identification for drainout data, and (d) drainout specification limit determination for chip seal fresh emulsions. 0 50 100 150 200 250 1 2 3 4 5 6 Vi sc os ity (cP ) Emulsion Number Sprayability Unmodified Outlier (a) 0 50 100 150 200 250 1 2 3 4 5 6 Vi sc os ity (cP ) Emulsion Number Sprayability Unmodified 98% Reliability (b) -200 0 200 400 600 800 1000 1 2 3 4 5 6 Vi sc os ity (c P) Emulsion Number Drainout Unmodified (c) -200 0 200 400 600 800 1000 1 2 3 4 5 6 Vi sc os ity (cP ) Emulsion Number Drainout Unmodified 98% Reliability (d) Figure 2.11. Examples of (a) outlier identification for sprayability data, (b) sprayability specification limit determination, (c) outlier identification for drainout data, and (d) drainout specification limit determination for spray seal fresh emulsions.

Results, Interpretation, and Applications 27 distresses. The residual binder is the asphalt binder material that remains after all the water has evaporated from the asphalt material. Relationships between the binder properties and the PSTs’ mixture performance were used to establish the specification criteria. The chip seal mixture specimens were fabricated using a performance-based mix design and a sample fabrication procedure that closely simulates the field construction process (Kim and Adams 2011). Because the residual asphalt content of emulsions can vary, each chip seal specimen was designed such that the residual asphalt content was kept consistent (at approximately 67%) after curing for all fabricated mixture specimens. This process required varying the emulsion application rate slightly prior to curing such that the embedment depth and residual asphalt content for each specimen were consistent after curing. Granite 78M aggregate was utilized for all the fabricated specimens, and the aggregate application rate that yielded a single stone coverage of 15.5 lb/yd2 was kept consistent for all specimens. The microsurfacing mixture specimens were fabricated according to ASTM D3910 and ASTM D6372. The mix design was conducted in accordance with the International Slurry Surfacing Association (ISSA) A143 specifications using emulsion, mineral filler, aggregate, and water. 0 200 400 600 800 1000 1 2 3 4 5 6 7 8 9 Vi sc os ity (cP ) Emulsion Number Unmodified Modified (a) 0 200 400 600 800 1000 1 2 3 4 5 6 7 8 9 Vi sc os ity (c P) Emulsion Number Unmodified Modified 98% Reliability (b) Figure 2.12. Examples of (a) outlier identification for mixability data and (b) mixability specification limit determination for microsurfacing fresh emulsions. Treatment Type Test Temperature (°C) EPG Specificaon Test EPG Specificaon Parameter Specified Limit Chip Seal 60 Storage Stability Test Separaon Rao 0.5 to 1.5 Stability Rao Maximum 2 Three-Step Shear Test Sprayability Maximum 400 cP Drainout Minimum 50 cP Spray Seal 25 Storage Stability Test Separaon Rao 0.5 to 1.5 Stability Rao Maximum 1.5 Three-Step Shear Test Sprayability Maximum 100 cP Drainout Minimum 100 cP Microsurfacing 25 Storage Stability Test Separaon Rao 0.2 to 1.5 Stability Rao Maximum 1.5 Mixability Maximum 600 cP Table 2.10. EPG specification limits for fresh emulsion properties.

28 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments Similar to the chip seal specimen fabrication process, the residual asphalt content and aggregate rate were held constant for all specimens (9% asphalt content was used in this case). For all specimens, as a mineral filler, 1% cement was used, and the water content was kept between 10% and 12% of the dry aggregate weight to ensure enough workability to fabricate specimens using different emulsions. 2.4.2.1 High-Temperature Residual Binder Test and Limits At high temperatures, bleeding and rutting are the major distress types for chip seals and microsurfacing treatments, respectively. It is common in chip seal construction to use a thick triple seal on road surfaces that experience substantial traffic loading. Likewise, microsurfacing treatments often are applied in thick applications. These thick seals are susceptible to both bleeding and rutting at high temperatures, whereas single-seal chip seals are susceptible only to bleeding. Binders with superior resistance to permanent strain will result in PST mixtures that are resistant to both bleeding and rutting. Before discussing the recommended high-temperature tests and limits for the EPG specifica- tions, it is important to evaluate the existing high-temperature parameters for emulsions. Evaluation of Existing High-Temperature Parameters in HMA Performance Grade Specifications for Emulsions Used in Surface Treatments For high temperatures, the current HMA PG specifications originally included only the elastic component of the complex shear modulus (G*/sin d) obtained from the DSR test to characterize rutting resistance, although recent evidence suggests a widespread shift to the use of the MSCR test. The research team conducted a study to determine whether the G*/sin d value for the binder obtained from DSR testing correlates well with the percentage of bleeding for chip seals and with rutting in microsurfacing mixtures at test temperatures of 46°C, 52°C, and 58°C. The results showed a relatively good correlation between the G*/sin d value and performance for chip seals, but a weak correlation for microsurfacings, as shown in Figures 2.13 and 2.14, respectively. Fig- ure 2.14 illustrates one of the flaws of the G*/sin d approach, which is its inability to differentiate between modified binders whose performance diverges outside the linear viscoelastic range. In addition to the lack of correlation found for the microsurfacing emulsions, the current prevailing thought is that the G*/sin d value is not capable of capturing the effects of elastomeric modification due to the relatively small impact of d on G*/sin d (D’Angelo et al. 2016). The research team found a strong correlation between MSCR Jnr values and high-temperature mixture performance for both chip seals and microsurfacings, as detailed in Section 2.4.2. The MSCR test better captures the elastic response of the polymer network because the MSCR test subjects the 0 10 20 30 40 50 60 70 80 90 100 0 10000 20000 30000 40000 50000 60000 % B le ed in g G*/sin delta 46C 52C 58C Figure 2.13. G*/sin c versus percentage of bleeding for chip seal emulsions.

Results, Interpretation, and Applications 29 material to higher levels of stress and strain than the DSR test used to measure G*/sin d. Also, the MSCR test better simulates the stress levels that the binder experiences under field traffic loading in surface treatments. Furthermore, using the same high-temperature parameter and test method for both chip seal and microsurfacing emulsions has significant practical merits to be considered when emulsion specifications are developed. Based on these observations, the research team recommends the MSCR test for characterizing high-temperature binder performance in the EPG specifications. The research team also investigated the use of the percentage of recovery (% recovery) param- eter obtained from the MSCR test for its ability to identify modification in residual binders. Figure 2.15 shows that % recovery can effectively distinguish between polymer-modified and unmodified binders, but it does not effectively identify the presence of latex modification (latex-modified and unmodified emulsions show similar % recovery values on average). Because of the inability of the % recovery parameter to identify the presence of latex modi- fication, a threshold value for detecting the presence of a modifier could not be determined. Instead, a review of the existing literature revealed that a maximum phase angle (d) of 80°, as 0 1 2 3 4 5 0 500000 1000000 1500000 R ut D ep th (m m) G*/sin delta 46C 52C 58C Figure 2.14. G*/sin c versus percentage of bleeding for microsurfacing emulsions. 0 5 10 15 20 25 30 35 40 45 50 % R ec ov er y 61C 67C Unmodified Latex- Modified Polymer-Modified CR S-2 -NC CR S-2 -A CR S-2 L-N C CR S-2 L-C CR S-2 L-F CR S-2 P-H P-E PP -CR S-2 P-E CR S-2 P-E CR S-2 P-A CR S-2 -F CR S-1 -B CR S-2 -E Figure 2.15. Percentage of recovery results for chip seal emulsions.

30 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments obtained from DSR testing, has been recommended for binders used in chip seal treatments (Vijaykumar et al. 2013). Figure 2.16 shows that all the modified emulsions have a phase angle value that is below the maximum threshold of 80° whereas all of the unmodified emulsions have a measured phase angle value above 80°, indicating that this test method could effectively identify the presence of modification. The presence of a modifier is recommended for emulsions used in chip sealing a high traffic roadway (Gransberg and James 2005, Im 2013); however, no maximum phase angle is recommended in the EPG specifications to determine whether modi- fiers are present in emulsions used at the high traffic level. Recommended High-Temperature Parameter for EPG Specifications: Non-Recoverable Creep Compliance (Jnr) In order to develop criteria for the EPG specifications that address high-temperature perfor- mance, the non-recoverable creep compliance (Jnr) of binders at a stress level of 3.2 kPa, which is an indicator of permanent strain resistance, was evaluated against the critical mixture perfor- mance measures, bleeding and rutting, for chip seal and microsurfacing mixtures, respectively. MSCR binder testing was conducted at temperatures ranging from 46°C to 70°C at 6°C incre- ments that were selected originally in accordance with the existing HMA high-temperature PG specifications. To verify that the binder’s non-recoverable creep compliance (Jnr) values relate to bleeding and rutting performance and that these Jnr values could be used subsequently to establish speci- fication limits, the binder Jnr results were correlated with corresponding chip seal performance that was measured using the one-third scale model mobile load simulator (MMLS3), which simulates traffic loading at an accelerated rate of 990 wheel applications every 10 minutes inside a temperature-controlled chamber (Lee et al. 2006, Lee and Kim 2010) For direct comparison to the high-temperature binder properties measured in the MSCR test, the mixture specimens were fabricated and subjected to high temperatures of 46°C, 52°C, and 58°C. The test temperature of 58°C is the highest temperature that the MMLS3 chamber was capable of maintaining during mix- ture testing. Extrapolation of the relationships between the binder and mixture results allowed for consideration of the binder results at typical EPG temperatures of 61°C and 67°C. The percentage of bleeding for a chip seal specimen was obtained by dividing the area of bleeding on the chip seal specimen due to MMLS3 loading by the total area of the specimen. 65 70 75 80 85 90 Ph as e A ng le (° ) Emulsion Type 61C 67C 73C C-C RS -2- AE C-C RS -2L -AE C-C RS -2L -C C-C RS -2L -F C-C RS -2P -A C-C RS -2- F Figure 2.16. Phase angle as a function of high EPG grade for chip seal emulsions.

Results, Interpretation, and Applications 31 The rut depth in microsurfacings was determined by measuring the difference in height between the wheel path and non-wheel path areas of the microsurfacing specimen that was subjected to MMLS3 loading (the processes for measuring bleeding and rutting performance are fully detailed in Appendix C). In order to establish traffic-dependent specification criteria, MMLS3 traffic loads were applied to represent low, medium, and high traffic conditions by converting MMLS3 wheel loads to equivalent field wheel loads (detailed in Appendix D). Specification limits were derived based on the Jnr value that corresponded to the maximum allowable PST mixture performance threshold at each traffic level; these limits were determined to be 80% bleeding for chip seals and a 5-mm rut depth for microsurfacing. The 80% maximum bleeding limit for chip seals defines the amount of bleeding under MMLS3 loading that, when exceeded, is unacceptable for chip seal performance (Lee 2007). The bleed- ing test results shown in Figure 2.17 were utilized to determine the 80% bleeding threshold for MMLS3-loaded chip seals. Lee’s work (Lee 2007) showed that, as key conditions (e.g., the emulsion and aggregate application rates) were varied to increase the bleeding potential, the bleeding performance under MMLS3 loading stayed consistently below 50% and then suddenly exhibited a sharp increase to 80% bleeding and higher. 0 20 40 60 80 100 6 8 10 12 14 Aggregate Application Rate (lb/yd2) Bl ee di ng (% ) 0.20 EAR 0.25 EAR 0.30 EAR 0.35 EAR 0.40 EAR 0 20 40 60 80 100 6 8 10 12 14 Residual Aggregte Rate (lb/yd2) Bl ee di ng (% ) 0.20 EAR 0.25 EAR 0.30 EAR 0.35 EAR 0.40 EAR 0 20 40 60 80 100 10 12 14 16 18 Aggregate Application Rate (lb/yd2) Bl ee di ng (% ) 0.15 EAR 0.20 EAR 0.25 EAR 0.30 EAR 0 20 40 60 80 100 10 12 14 16 18 Residual Aggregte Rate (lb/yd2) Bl ee di ng (% ) 0.15 EAR 0.20 EAR 0.25 EAR 0.30 EAR Li gh tw ei gh t Ag gr eg at e Gr an ite Ag gr eg at e Figure 2.17. Bleeding versus material application rates for granite and lightweight chip seals (Lee 2007).

32 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments The MMLS3 test for multilayered seals represents harsher loading conditions than field traffic due to the accelerated loading rate used in MMLS3 tests that causes increased bleeding potential compared to field traffic (Lee 2007 and Adams and Kim 2013), i.e., 80% bleeding under MMLS3 traffic loading does not equate to 80% bleeding under field traffic loading. The reason for this difference is that the MMLS3 applies 990 wheel loads every 10 minutes (i.e., a rate of 1.65 wheel loads applied every second), which is higher than the loading rate that a chip seal would experience in the field, and thus allows the binder less rest and recovery time between MMLS3 loadings. Under field traffic loading at high temperatures, asphalt binder can recover during the rest period between traffic loads. During this recovery period, the elasticity of the binder allows it to recover from some of the non-permanent deformation that it experi- ences due to the loading stress. In the MMLS3 test, the binder is not allowed this recovery time and the maximum MMLS3 bleeding threshold should take into account this higher magnitude of “% bleeding” that occurs in MMLS3 testing (further explanation is provided in Appendix D). MSCR Test Limits for Bleeding in Chip Seals at High Temperatures Figure 2.18 presents the correlations between the binder Jnr results and the amount of bleed- ing measured from the MMLS3 testing of chip seal mixtures at low, medium, and high traffic 0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 % B le ed in g Jnr (in kPa-1) Jnr (in kPa-1) Unmodified 46C Modified 46C Unmodified 52C Modified 52C Unmodified 58C Modified 58C 0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 % B le ed in g Unmodified 46C Modified 46C Unmodified 52C Modified 52C Unmodified 58C Modified 58C 0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 % B le ed in g Unmodified 46C Modified 46C Unmodified 52C Modified 52C Unmodified 58C Modified 58C High Traffic Medium TrafficLow Traffic Jnr (in kPa-1) Figure 2.18. MSCR results versus percentage of bleeding relationship at high temperatures.

Results, Interpretation, and Applications 33 volumes. Note that the binder Jnr values do not change with traffic volume. Each data point represents the average of three MMLS3 bleeding test replicates and four MSCR test replicates. The results show a strong temperature-independent relationship between the binder Jnr values and the mixtures’ bleeding performance. The results indicate that the binder Jnr captures the binder’s contribution to the bleeding in the chip seal and matches expected trends, as a lower Jnr value indicates more permanent strain resistance and thus less susceptibility to bleeding. Figure 2.19 shows the relationship between the Jnr and “% bleeding” plotted as a function of traffic level. As expected, the results shown in Figure 2.19 indicate that the severity of the bleeding increases with increased MMLS3 traffic loading. Curves were fitted to the bleeding versus Jnr relationships shown in Figure 2.18 to allow for the extrapolation of Jnr values for 61°C and 67°C that could not be measured due to the limita- tion of the MMLS3 temperature control system. The results are presented in Figure 2.20 for low, medium, and high traffic volumes, which were used to establish preliminary specification limits for the EPG specifications. In order to establish limits for the binder Jnr value at each traffic level, a maximum allowable bleeding threshold of 80% was utilized for the laboratory-tested chip seal mixture specimens. The maximum limit for the Jnr value for each traffic volume is defined as the corresponding Jnr value at the point where the Jnr versus bleeding curve intersects the 80% bleeding severity threshold, as depicted in Figure 2.20. Table 2.11 presents the corresponding MSCR test limits that were derived for the high- temperature grading of binders used in chip seal treatments in the EPG specifications. Table 2.12 shows how the binders tested in developing these specifications would grade using the preliminary MSCR Jnr limits shown in Table 2.11. The values in the cells labeled “high,” “medium,” and “low” meet the EPG criteria for high, medium, and low traffic volumes, respectively, and the values in the cells labeled “fail” failed to meet the criteria for low traffic volume. These results appear reasonable because they necessitate the use of modified emulsions at high traffic volumes at 61°C and 67°C (most of the unmodified binders failed to meet the required high-temperature performance standard at 67°C). Most of the United States have grades of either EPG 61 or EPG 67 at high temperatures (see Figure 2.3). The emulsions obtained for this research were formulated for and have been used successfully in climates with EPG high-temperature grades of 61 and 67. 0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 5 6 7 8 9 10 11 12 % B le ed in g Jnr (in kPa-1) High Traffic Medium Traffic Low Traffic Figure 2.19. MSCR versus percentage of bleeding relationship as a function of traffic level for chip seals.

34 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments y = 13.126ln(x) + 53.066 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 5 6 7 8 9 10 % B le ed in g 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 5 6 7 8 9 10 % B le ed in g 0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 5 6 7 8 9 10 % B le ed in g (a) (b) (c) Pass and grade Fail High Medium FailLow Medium FailLow Jnr (in kPa-1) Jnr (in kPa-1) Jnr (in kPa-1) Figure 2.20. Developed maximum MSCR Jnr limits established for chip seal high-temperature grading in the EPG specifications based on (a) low traffic, (b) medium traffic, and (c) high traffic volumes. EPG Specificaon Traffic Grade Maximum Jnr @ 3.2 kPa Low 8 kPa-1 Medium 5.5 kPa-1 High 3.5 kPa-1 Table 2.11. Chip seal high-temperature EPG specification limits. MCSR Test Limits for Rutting in Microsurfacing at High Temperatures Rutting was identified as a critical high-temperature distress for microsurfacing emulsions. Similar to the bleeding specifications for chip seal residue, the resistance to rutting of micro- surfacing residue was characterized using the Jnr value at the 3.2 kPa stress level in the MSCR tests. The binder Jnr results were correlated with the mixture rutting measurements that were taken using the MMLS3 at low, medium, and high traffic volumes in order to verify that the Jnr value is a good indicator of microsurfacing rutting resistance and subsequently to establish specification limits. A maximum allowable rutting threshold of 5 mm for laboratory-tested mix- ture specimens was utilized in the development of the preliminary specification limits. Because

Results, Interpretation, and Applications 35 no microsurfacing rut depth threshold value existed prior to this research, the 5-mm maximum rutting limit was derived from the fact that each applied MMLS3 wheel load is one-third of a typical dual wheel single-axle load (Lee 2003). Therefore, the rutting limit for microsurfacing was determined to be approximately one-third of the acceptable rut depth limit for HMA pavements of 12.7 mm or about 5 mm for microsurfacings subjected to MMLS3 traffic (Mohseni et al. 2005). Curves were fitted to the rutting versus Jnr relationships over the temperature range of 46°C to 58°C and used to extrapolate the rut depths that correspond to the Jnr values measured at the high-temperature EPGs of 61°C and 67°C (i.e., the temperatures at which the corresponding mixture performance could not be measured due to MMLS3 temperature chamber limitations). Figure 2.21 presents the relationships between the measured MSCR test parameters and the mixture rutting performance at high temperatures for low, medium, and high traffic volumes. The results demonstrate a fairly strong relationship between the binder Jnr and the mixture rut depth, which follows expected trends, indicating that a lower Jnr value will lead to more rutting resistance. These results indicate also that the Jnr value is a suitable high-temperature specifica- tion parameter for microsurfacing emulsion residue. Using the maximum rut depth of 5 mm as well as the correlation between the Jnr value and rut depth at each traffic volume shown in Figure 2.21, the specification limits for the maximum allowable Jnr value were determined for low, medium, and high traffic volumes. Ultimately, the long-term validation of these EPG specifications will determine whether the maximum Jnr limits that were determined using this maximum rut depth threshold of 5 mm are appropriate for specifying high-temperature microsurfacing performance or whether these limits need to be adjusted. The preliminary specification limits derived for the medium and high traffic volumes shown in Figure 2.21 (c) were found to be similar (i.e., 1.75 kPa-1 for medium traffic and 1.25 kPa-1 for high traffic) because the microsurfacing rut depths are similar at the medium and high traffic volumes, as shown in Figure 2.22. The limited increase in rut depth value from medium to high traffic volumes is due to limitations regarding rut depth. Initially upon trafficking, densification occurs in uncompacted microsurfacings as traffic repetitions increase. However, after a large volume of traffic repetitions, the thin microsurfacing mixture can no longer continue to densify, Emulsion High-Temperature EPG 61°C 67°C CRS-2-NC 3 (High) 7 (Low) CRS-2-F 7.5 (Low) 15 (Fail) CRS-2-E 6 (Low) 11 (Fail) CRS-2-A 5.25 (Med) 21.5 (Fail) PP-CRS-2-A 5.1 (Med) 21.5 (Fail) CRS-2L-C 5.5 (Med) 10 (Fail) CRS-2L-F 3.2 (High) 6.75 (Low) CRS-2L-NC 2.4 (High) 4.75 (Med) CRS-2P-A 1.1 (High) 2.5 (High) CRS-2P-E 0.8 (High) 2 (High) CRS-2P-hP-E 1.2 (High) 3 (High) PP-CRS-2P-E 2 (High) 5 (Med) CRS-1-B 9.5 (Fail) 35 (Fail) Table 2.12. Summary of chip seal high-temperature emulsion performance grades.

36 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 R ut D ep th (m m) Jnr (kPa-1) Jnr (kPa-1) Jnr (kPa-1) 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 R ut D ep th (m m) 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 R ut D ep th (m m) (a) (b) (c) Pass and grade Fail @ test temp H M Fail @ test tempLow Med Fail @ test tempLow Figure 2.21. Developed maximum MSCR Jnr limits derived for microsurfacing high-temperature grading in the EPG specifications based on performance at (a) low traffic, (b) medium traffic, and (c) high traffic volumes. 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 R ut D ep th (m m) Jnr (kPa-1) Low Traffic Medium Traffic High Traffic Figure 2.22. MSCR versus rut depth relationship as a function of traffic level for microsurfacing.

Results, Interpretation, and Applications 37 and the rut depth approaches an asymptotic depth, which was achieved for most mixtures at the medium traffic level. Therefore, one maximum Jnr limit will define acceptable performance for both high and medium traffic volumes for the microsurfacing EPG specifications. The maximum allowable Jnr limit for both medium and high traffic volumes was selected to be 1.5 kPa -1, as this value is the average of the two limits. Tables 2.13 and 2.14 list the proposed MSCR test limits for the high-temperature grades of binders used in microsurfacing treatments and the high- temperature EPGs for the microsurfacing emulsions used in this study. 2.4.2.2 Intermediate Temperature Testing This section explains the rationale for not including intermediate temperature testing in the proposed EPG specifications to assess raveling potential. Instead, intermediate temperature test- ing, which is driven by chemistry and compatibility between aggregate and emulsion, should be assessed during the mix design stage for both chip seals and microsurfacings. Chip Seal Raveling, or aggregate loss, is the primary distress in chip seals at intermediate temperatures. The bitumen bond strength (BBS) test, a modification of AASHTO TP 91, was evaluated initially during EPG specification development for its ability to assess raveling potential at an intermediate temperature. The BBS test measures the stress that is required to detach a binder specimen that is adhered to an aggregate substrate (i.e., bond strength) and is used for quantifying the emulsion residue’s resistance to aggregate loss. Because the emulsions used in chip seals are cured to their residual binder state while they are in contact with the aggregate surface, BBS testing requires the emulsion samples to be cured on the actual aggregate substrate (referred to as the curing on rock, or COR, method) to allow a natural adhesive bond to occur between the emulsion and the aggregate. This requirement presents a problem when attempting to grade emulsions in a purchase specification because the specific aggregate source to be utilized in the design chip seal mixture is unknown at the time of grading. The remaining portion of this section highlights the significant impact that the curing EPG Specificaon Traffic Grade Maximum Jnr @ 3.2 kPa Low 5.0 kPa-1 Medium/High 1.5 kPa-1 Table 2.13. Microsurfacing high-temperature EPG specification limits. Table 2.14. Summary of microsurfacing high-temperature emulsion performance grades.

38 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments method and natural bond formation have on the correlation between bond strength and aggregate loss for chip seals. Initially in this study, the BBS test was conducted using the so-called residue on rock (ROR) curing method. In this method, residual binder from the chip seal emulsion is recovered and then applied to the aggregate substrate using a heated stub and conditioned for 24 hours at the test temperature for BBS test specimen fabrication. The BBS test results using the ROR method are plotted against the percentage of aggregate loss (“% aggregate loss”) measured from the Vialit test as shown in Figures 2.23 and 2.24 for the intermediate temperatures of 15°C and 25°C, respectively. The results presented in Figures 2.23 and 2.24 show that the Vialit mixture tests were able to capture the aggregate loss difference between the modified and unmodified emulsions, as the modified binders show a clear reduction in aggregate loss. However, the BBS test results for the ROR-cured specimens showed a minimal ability to differentiate between the unmodified and modified emulsions and poor correlation with the aggregate loss of the chip seal mixture 0 5 10 15 20 25 0 500 1000 1500 2000 2500 3000 % A gg re ga te L os s BBS (kPa) Unmodified Emulsions Modified Emulsions Figure 2.23. Aggregate loss versus BBS at 15°C using the recovered ROR curing method for BBS test specimen fabrication. 0 5 10 15 20 25 0 500 1000 1500 2000 % A gg re ga te L os s BBS (kPa) Unmodified Emulsions Modified Emulsions Figure 2.24. Aggregate loss versus BBS at 25°C using the recovered ROR curing method for BBS test specimen fabrication.

Results, Interpretation, and Applications 39 at either 15°C or 25°C. The problem with the ROR curing method is that it does not accurately simulate the initial bond development and chemical interaction that occur between the fresh emulsion and aggregate when they are introduced during chip seal construction. Due to this shortcoming, a curing method that could accurately simulate the initial bond development and chemical interaction between the aggregate and asphalt emulsion for BBS testing was needed. The COR method involves placing fresh emulsion on the aggregate substrate at the start of the curing process, which simulates the natural chemical interaction between the emulsion and aggregate and allows a more realistic adhesive bond to form between the materials. This method simulates the bond development that occurs in the chip seal mixture when the aggregate is spread onto the fresh emulsion during construction. Through use of a more natural curing condition, the aggregate-to-residual binder interface that is formed for the BBS test is more representative of the interface that is formed within the chip seal mixture, thus leading to a better correlation between the binder and mixture test results. Compared to the ROR method, the COR curing method better simulates the bond strength development that occurs between the emulsion and aggregate during the first 21 hours of interaction in a chip seal mixture, as evidenced by the results shown in Figure 2.25 where the BBS test data are plotted against aggregate loss mixture performance for those same materials. Figure 2.25 shows a reasonable correlation between the BBS of the residue binder and the “% aggregate loss” in the mixture when the COR method was used to fabricate the specimens, as opposed to the poor correlations observed using the ROR method, as shown in Figures 2.23 and 2.24. In addition, a relatively unique relationship was found between “% aggregate loss” and the BBS regardless of the binder modification. This finding indicates that for both unmodified and modified emulsions assessing the compatibility via bond strength is important to constructing chip seals that resist raveling. The research team does not believe that compatibility can be addressed effectively in the proposed EPG specifications, which are developed as emulsion purchase specifications. Instead, BBS testing should be conducted during the mix design phase in order to ensure that sufficient compatibility exists between the emulsion and specific aggregate materials to be used in con- structing the chip seal mixture. y = -0.0111x + 24.641 R² = 0.7493 0 5 10 15 20 25 0 500 1000 1500 % A gg re ga te L os s BBS (kPa) Unmodified at 15°C Modified at 15°C Unmodified at 25°C Modified at 25°C Poor Performing at 15°C Figure 2.25. Aggregate loss versus BBS using the COR method for BBS test specimen fabrication.

40 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments Microsurfacing For intermediate temperatures, raveling and reflective cracking were identified as the most critical distress types in microsurfacing treatments. However, raveling of microsurfacing mixtures was determined to be driven by the chemical interaction between the emulsion and other mixture constituents (i.e., aggregate, cement, lime) as exhibited by the results presented in Appendix D. Thus, raveling resistance cannot be linked solely to the quality of the binder. In fact, microsurfacing emulsion manufacturers formulate emulsions specifically for the aggregate with which the emulsion is to be blended. Reflective cracking is related to the pavement structure and is driven largely by the condition of the underlying pavement surface and thus is also not related directly to binder quality. The EPG specifications recommend conducting the Wet Track Abrasion Test (in accordance with ASTM D3910) during the mix design stage, after combining the emulsion with the aggre- gate, to ensure adequate raveling resistance at intermediate temperatures. 2.4.2.3 Low-Temperature Residual Binder Test and Limits Aggregate loss is the primary distress in chip seals at low temperatures. Aggregate loss in chip seals at low temperatures was found to occur in a cohesive fracture pattern within the binder during low-temperature mixture performance testing. Therefore, it was speculated that chip seal aggregate loss at low temperatures is driven largely by the properties of the residual binder rather than the compatibility of the aggregate and binder. Thermal cracking is the primary distress in microsurfacings at low temperatures. Thermal cracking in microsurfacings is related to the ability of the residual binder to withstand thermal contraction upon cooling without fracture. The critical low-temperature distresses in both chip seals and microsurfacings typically occur during the first winter following the PST construction. Because the low winter temperatures occur only a few months after the initial construction, it was assumed that the low-temperature distresses were the most critical distresses before the residue aged significantly. Therefore, unaged residue was used for the low-temperature grading. The evaluation of existing low-temperature HMA PG specification parameters and the rec- ommended test methods in the EPG specifications and the derivation of performance-based limits are discussed in this section. Evaluation of the Existing Low-Temperature Specification Parameters on Emulsions Used in Surface Treatments The existing low-temperature HMA Superpave PG specification parameters, S(60) and m(60), were evaluated as well as another emulsion specification parameter, S(8), proposed by Vijaykumar et al. (2013). The current asphalt binder PG specification parameters of creep stiff- ness at 60 seconds, S(60), and the m-value at 60 seconds, m(60), were investigated as potential parameters to address the low-temperature performance of surface treatments in the EPG speci- fications. Standard Bending Beam Rheometer (BBR) tests were conducted to determine S(60) and m(60) for chip seal emulsion residues (AASHTO T 313). For microsurfacing emulsion residues, the DSR was used to obtain S(60) and m(60) using the interconversion of frequency sweep mastercurves. The DSR frequency sweep test was conducted using emulsion residue at 5°C, 10°C, and 15°C. Then, interconversion methods developed by Ferry (1980) and Anderson et al. (1994) were applied to obtain the low-temperature creep properties (i.e., the stiffness values and m-values) from the shear properties measured at intermediate temperatures. The creep stiffness at 8 seconds, S(8), also was evaluated in terms of mixture performance. The S(8) parameter is a proposed low-temperature parameter for chip seals in surface performance-graded (SPG) specifications (Walubita et al. 2005, Vijaykumar et al. 2013).

Results, Interpretation, and Applications 41 Figures 2.26 and 2.27 present the correlations between these binder parameters and mixture performance where binder S(60) is compared to the low-temperature aggregate loss measured from Vialit tests of chip seal mixtures and fracture energy is measured from single-edge notched beam (SENB) tests of microsurfacing mixtures, respectively (these tests are described in Appen- dix C). Figures 2.28 and 2.29 show the relationships observed between m(60) and the critical mixture performance measures for chip seals and microsurfacings, respectively. Figures 2.30 and 2.31 show the relationships observed between S(8) and the critical mixture performance measures for chip seals and microsurfacings, respectively. The results show that for S(60), m(60), and S(8), the relationships with the mixture perfor- mance are binder-dependent, and thus, the correlations between these binder parameters and mixture performance are inadequate for use in the proposed EPG specifications. Therefore, a recommended low-temperature parameter for the EPG specifications was developed; details are provided below. 0 5 10 15 20 25 30 35 40 45 0 200 400 600 800 1000 A gg re ga te L os s (% ) S(60) (MPa) C-CRS-2-AE C-CRS-2-F C-CRS-2L-C C-CRS-2P-A C-CRS-2L-AE C-CRS-2L-F Figure 2.26. Binder S(60) versus low-temperature Vialit aggregate loss for chip seal mixtures. 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035 0 200 400 600 800 1000 1200 Fr ac tu re E ne rg y (J) S(60) (MPa) M-CSS-1h-C M-CSS-1hL-F M-CQS-1hP-E M-CQS-1h-E M-CSS-1hP-D Figure 2.27. Binder S(60) versus fracture energy for microsurfacing mixtures.

42 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments 0 5 10 15 20 25 30 35 40 45 0.00 0.10 0.20 0.30 0.40 0.50 A gg re ga te L os s (% ) m(60) C-CRS-2-AE C-CRS-2-F C-CRS-2L-C C-CRS-2P-A C-CRS-2L-AE C-CRS-2L-F Figure 2.28. Binder m(60) versus low-temperature Vialit aggregate loss for chip seal emulsions. 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Fr ac tu re E ne rg y (J) m(60) M-CSS-1h-C M-CSS-1hL-F M-CQS-1hP-E M-CQS-1h-E M-CSS-1hP-D Figure 2.29. Binder m(60) versus fracture energy for microsurfacing mixtures. 0 5 10 15 20 25 30 35 40 45 0 200 400 600 800 1000 1200 1400 A gg re ga te L os s (% ) S(8) (MPa) C-CRS-2-AE C-CRS-2-F C-CRS-2L-C C-CRS-2P-A C-CRS-2L-AE C-CRS-2L-F Figure 2.30. Binder S(8) versus low-temperature Vialit aggregate loss for chip seal mixtures.

Results, Interpretation, and Applications 43 Recommended Low-Temperature Parameter for EPG Specifications Fracture-mechanics-based and rheology-based residual binder properties were evaluated for both chip seal and microsurfacing low-temperature specifications. The strength of the correlation between mixture performance and binder properties was used to evaluate the appropriateness of the parameter for low-temperature specifications. Chip seal mixture low-temperature aggregate loss was measured using Vialit tests conducted at temperatures ranging from 0°C to -28°C. Microsurfacing mixture resistance to thermal crack- ing was quantified using fracture energy measured via SENB testing at -16°C, -22°C, and -28°C. Fracture properties of the residue binders were obtained via SENB testing at -16°C, -22°C, and -28°C. Residual binder rheological properties were obtained via DSR frequency sweep testing at temperatures ranging from 5°C to 15°C. Although both rheology-based and fracture-mechanics- based properties were found to relate to chip seal and microsurfacing mixture performance at low temperatures, the rheological properties, which demonstrated the strongest relationship to mixture performance, were ultimately selected for the specifications on the basis of efficiency and cost-effectiveness. DSR testing also is required for high-temperature EPG determination, and hence, the use of the DSR as opposed to the SENB test set-up minimizes equipment requirements. In addition, DSR tests require less residual binder, involve simpler specimen preparation, and take less time than SENB tests. (Details are provided in Appendix D.) The rheological residual binder property that demonstrated the strongest relationship to both chip seal aggregate loss and microsurfacing fracture energy is the dynamic shear modulus (G*) at a critical phase angle (dc). The critical phase angle values varied as a function of the low-temperature PG of interest. The use of G* at critical phase angle values for low-temperature specifications was motivated by the discovery of a strong relationship between the crossover modulus (Gc*) and low-temperature mixture performance. The crossover modulus is defined as the G* value that corresponds to the reduced frequency where the phase angle equals 45°. At reduced frequencies lower than the crossover frequency (i.e., higher temperatures or slower rates), the loss (viscous) component of the G* exceeds the storage (elastic) component. In the study of polymers and other viscoelastic materials, it has been postulated that the crossover modulus is an indicator of intermolecular forces that comprise a material’s microstructure (Winter 1987). At temperatures below or frequencies above the crossover point, the excitation applied to the material is insufficient to overcome the forces of molecular interaction contained within the material’s microstructure. However, once the crossover point is exceeded (either by 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035 0 500 1000 1500 Fr ac tu re E ne rg y (J) S(8) (MPa) M-CSS-1h-C M-CSS-1hL-F M-CQS-1hP-E M-CQS-1h-E M-CSS-1hP-D Figure 2.31. Binder S(8) versus fracture energy for microsurfacing mixtures.

44 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments increasing the temperature or decreasing the rate), the excitation overcomes the forces of the material’s microstructure, and hence, a tendency toward fluid behavior ensues. Thus, a lower crossover modulus value in asphalt binders theoretically implies a higher degree of compatibility (i.e., less structure). Materials that are structured are inherently more brittle and susceptible to cracking than more compatible materials. Thus, it is intuitive that a lower crossover modulus value relates to better fracture resistance, which is confirmed by the results obtained herein. The finding that G* at a critical phase angle is related to fracture resistance is corroborated by findings reported in the literature that demonstrate the existence of a relationship between asphalt binder R-value and asphalt mixture fracture resistance (Rowe et al. 2016). The R-value is the difference between the G* value at the glassy state and the G* value at the reduced frequency where phase angle equals 45°. For the majority of binders, the glassy modulus is approximately 109 Pa (Anderson et al. 1994). Thus, the R-value is essentially an indicator of asphalt binder crossover modulus (i.e., the G* value at a critical phase angle). However, crossover modulus and R-value are temperature-independent parameters. Because the crossover modulus is a temperature-independent parameter whereas the mixture performance changes as a function of temperature, the crossover modulus cannot be used in the framework of the EPG development. Therefore, to adapt the concept of the crossover modulus to allow its incorporation into the EPG specification framework, critical phase angle values were determined as a function of mixture test temperature to produce a temperature-independent relationship between the mixture low-temperature performance and corresponding G* values. To determine the G* values at the critical phase angle values of interest, DSR frequency sweep tests were con- ducted at both 5°C and 15°C regardless of PG, that is, the critical phase angle values for which the G* value is evaluated vary as a function of low-temperature PG rather than test temperature. However, consistent with other performance-grading specification criteria, the specification limits for G* values are independent of climate. To determine the specification limits, the residual binder property values that correspond to critical mixture performance thresholds were determined. Mixture performance thresholds were determined based on the results herein. Chip seal aggregate loss at low temperatures is dependent on traffic level; therefore, specification limits were defined for high, medium, and low traffic levels. Microsurfacing thermal cracking is induced by thermal rather than vehicular loading. Therefore, the microsurfacing residual specifications contain a single specification limit that is independent of traffic level. 2.4.2.4 Critical Phase Angle Values and Dynamic Shear Modulus Test Limits for Raveling in Chip Seals at Low Temperatures The research team utilized trends in the test results combined with a priori knowledge of the emulsions’ performance to establish critical performance thresholds for chip seal mixture Vialit aggregate loss at low temperatures, as shown in Figure 2.32. To establish thresholds based on temperatures of interest, aggregate loss data were first translated to the three most common low- temperature climatic EPGs, -19°C, -25°C, and -31°C, via interpolation and extrapolation of measured test data. Figure 2.32 presents the resultant relationship between chip seal mixture aggregate loss and temperature. The aggregate loss threshold for low traffic was set at 35% based on the known performance of CRS-2-NC emulsion on low traffic roadways in a -19°C region. To determine the high traffic threshold, trends in aggregate loss with temperature were used. As shown in Figure 2.32, the CRS-2L-NC emulsion experiences a significant increase in aggregate loss when the temperature is decreased from -19°C to -25°C. Similarly, CRS-2L-C experiences an abrupt increase in the sensitivity of aggregate loss performance to temperature when the temperature drops below -25°C. In both instances, the aggregate loss before the observed abrupt loss of raveling resistance with the decrease in temperature is 25%. In addition, the CRS-2P-A

Results, Interpretation, and Applications 45 emulsion has been used successfully in -25°C regions on medium traffic roadways and meets this criterion, further validating its reasonableness. The medium traffic aggregate loss threshold was selected simply as the average of the high and low traffic thresholds. To determine the critical phase angle values as a function of the low-temperature EPG, chip seal mixture aggregate loss was correlated with binder G* values that correspond to varying phase angle values. Phase angle values used for each mixture performance test temperature were adjusted until a temperature-independent relationship between the mixture aggregate loss and binder G* was obtained. Figure 2.33 presents the results. Figure 2.33 (a) shows that the relationship between aggregate loss and the crossover modulus depends on temperature because the mixture performance changes as a function of temperature, but the crossover modulus value does not. Figure 2.33 (b) shows the relationship between aggregate loss and G* when the critical phase angle values for which G* is determined are adjusted as a function of temperature. The dc value that corresponds to -25°C remains 45°. However, the dc values that correspond to -19°C and -31°C are 48° and 42°, respectively. These results demonstrate 10 15 20 25 30 35 40 45 -35 -30 -25 -20 -15 -10 A gg re ga te L os s (% ) Temperature (°C) CRS-2-NC CRS-2-F CRS-2L-C CRS-2P-A CRS-2L-NC CRS-2L-F High Traffic Threshold Medium Traffic Threshold Low Traffic Threshold Figure 2.32. Low-temperature aggregate loss for chip seal mixtures as a function of temperature. 0 5 10 15 20 25 30 35 40 45 0.E+00 1.E+07 2.E+07 3.E+07 4.E+07 5.E+07 A gg re ga te L os s (% ) G* @ δ = 45° (Pa) G* @ δc (Pa) -19 -25 -31 CRS-2L (a) (b) -F R2 = 0.63 0 5 10 15 20 25 30 35 40 45 0.E+00 1.E+07 2.E+07 3.E+07 4.E+07 5.E+07 A gg re ga te L os s (% ) -19 -25 -31 CRS-2L-F R2 = 0.86 Figure 2.33. Chip seal mixture aggregate loss versus residual binder G*at critical c corresponding to (a) c = 45° and (b) adjusted c as a function of temperature.

46 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments good agreement between the aggregate loss and G* values irrespective of the mixture test tem- perature (R2 = 0.86), with the exception of CRS-2L-F. Aggregate loss in the CRS-2L-F mixture does not vary appreciably with test temperature, unlike the other emulsions, but its rheology does change with temperature, which explains its outlier behavior. The relationships between low-temperature EPGs and dc values shown in Figure 2.33 (b) were extended to determine dc values that correspond to all low-temperature EPGs. Table 2.15 presents the results. Residual specification limits were defined based on the relationships between aggregate loss and G* values that correspond to the dc values and the established allowable aggregate loss thresholds for low, medium, and high traffic volumes. Figure 2.34 presents the results. The corresponding maximum G* limits for low, medium, and high traffic volumes are 30, 20, and 12 MPa, respectively, and are summarized in Table 2.16. Table 2.17 provides a summary of the corresponding EPGs of the emulsions evaluated. CRS-2-NC can be used only at low traffic levels in -19°C and -25°C regions. The results match intuition because this binder is currently used successfully on low traffic roadways in a -19°C region. With the exception of CRS-2-F at -19°C, medium and high traffic volumes require the use of a modified emulsion. Based on the known benefits of modification, these results seem to be reasonable. Note that only one of the emulsions evaluated (CRS-2P-A) meets the high traffic criterion at both -25°C and -31°C. CRS-2P-A is Low-Temperature EPG (°C) δc (°) −7 54 −13 51 −19 48 −25 45 −31 42 −37 39 Table 2.15. Critical phase angle values for chip seal residual binders. 0 5 10 15 20 25 30 35 40 45 0.E+00 1.E+07 2.E+07 3.E+07 4.E+07 5.E+07 A gg re ga te L os s (% ) CRS-2-NC CRS-2-F CRS-2L-C CRS-2P-A CRS-2L-NC CRS-2L-F Low Traffic Threshold Medium Traffic Threshold High Traffic Threshold G* @ δc (Pa) Figure 2.34. Maximum critical G* limits for chip seal residual binders.

Results, Interpretation, and Applications 47 the only emulsion evaluated that is currently used in a -25°C or colder region, and hence, the performance-grading results appear reasonable. 2.4.2.5 Critical Phase Angle Values and Dynamic Shear Modulus Test Limits for Thermal Cracking in Microsurfacing at Low Temperatures Trends in the test results were used to establish critical performance thresholds for micro- surfacing fracture energy at low temperatures; Figure 2.35 presents the SENB test results. These results show that the fracture energy for each microsurfacing mixture decreases as the temperature decreases. However, for microsurfacing mixtures CSS-1hP-E, CQS-1h-E, and CQS-1hP-E, little decrease in fracture energy occurs when the temperature is decreased below a EPG Specificaon Test Performance Parameter Traffic Level Temperature EPG Specificaon Limit DSR Frequency Sweep Maximum G* @ δc Low 5°C and 15°C <30 MPa Medium <20 MPa High <12 MPa Table 2.16. Low-temperature chip seal EPG limits. Emulsion Low-Temperature EPG −19°C −25°C −31°C CRS-2-NC L L FAIL CRS-2-F M L FAIL CRS-2L-C H M M CRS-2P-A H H H CRS-2L-NC M L FAIL CRS-2L-F M L L L = low traffic volume, M = medium traffic volume, and H = high traffic volume Table 2.17. Summary of low-temperature EPGs of chip seal emulsions evaluated. 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035 -30 -25 -20 -15 -10 Fr ac tu re E ne rg y (J) Temperature (°C) CSS-1h-C CSS-1hL-F CQS-1hP-E CQS-1h-E CSS-1hP-D Figure 2.35. Microsurfacing mixture fracture energy results.

48 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments certain critical temperature. (Note that for CQS-1h-E, this critical temperature exceeds the range of temperatures considered in testing.) The lack of sensitivity of fracture energy to temperature implies that the test temperature is below the glass transition temperature. Glassy materials are inherently brittle and susceptible to cracking. Therefore, a mixture should not be placed in a climate where it will be exposed to temperatures below its glass transition. Based on the micro- surfacing fracture energy results, it is apparent that the value of fracture energy at or below the glass transition temperature is less than 0.0010 J. Therefore, the critical microsurfacing mixture performance threshold was set to 0.0010 J. The critical phase angle values for microsurfacing residual binders were established in a pro- cedure analogous to that used for chip seal residual binders. Microsurfacing mixture fracture energy values were correlated with binder G* values that correspond to phase angle values that varied as a function of mixture test temperature. The phase angle values used for each mix- ture performance test temperature were adjusted until a temperature-independent relationship between mixture fracture energy and binder G* was obtained. Figure 2.36 presents the results. Figure 2.36 (a) shows that the relationship between fracture and the crossover modulus depends on temperature because the mixture performance changes as a function of temperature, but the crossover modulus does not. Figure 2.36 (b) shows the relationship between fracture energy and G* when the critical phase angle values for the determined G* values are adjusted as a func- tion of temperature. The dc value that corresponds to -22°C remains 45°, whereas the dc values that correspond to -16°C and -28°C are 47° and 43°, respectively. These results demonstrate good agreement between fracture energy and G* values irrespective of mixture test temperature (R2 = 0.85). The relationships between low-temperature EPGs and dc values shown in Figure 2.36 (b) were used to determine the dc values that correspond to all low-temperature EPGs. Table 2.18 presents the results. Residual specification limits were defined based on the relationships between fracture energy and G* values that correspond to dc values and the established allowable fracture energy thresholds. Figure 2.37 presents the results. The corresponding maximum G* limit is 16 MPa and is sum- marized in Table 2.19. Table 2.20 provides a summary of the results of low-temperature EPG determination for the binders evaluated at the three most common low-temperature EPGs. The results demonstrate that CSS-1h-C is the only emulsion that meets specifications at -31°C. R² = 0.85 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035 0.E+00 1.E+07 2.E+07 3.E+07 4.E+07 Fr ac tu re E ne rg y (J) -16 (Delta = 47) -22 (Delta = 45) -28 (Delta = 43) (b)R² = 0.65 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035 0.E+00 1.E+07 2.E+07 3.E+07 Fr ac tu re E ne rg y (J) -16 -22 -28 (a) G* @ δ = 45° (Pa) G* @ δc (Pa) Figure 2.36. Microsurfacing mixture fracture energy versus residual binder G* at critical c corresponding to (a) c = 45° and (b) adjusted c as a function of temperature.

Results, Interpretation, and Applications 49 Low-temperature EPG (°C) δc (°) −7 50 −13 48 −19 46 −25 44 −31 42 −37 40 Table 2.18. Critical phase angle values for microsurfacing residual binders. 0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 0.0035 0.E+00 1.E+07 2.E+07 3.E+07 4.E+07 Fr ac tu re E ne rg y (J) CSS-1h-C CSS-1h-L CQS-1hP-E CQS-1h-E CSS-1hP-D G* @ δc (Pa) Figure 2.37. Maximum critical G* limit for microsurfacing residual binders. EPG Specificaon Test Performance Parameter Traffic Level Temperature EPG Specificaon Limit DSR Frequency Sweep Maximum G* @ δc Low Medium High 5°C and 15°C 16 MPa Table 2.19. Low-temperature microsurfacing EPG limits. Emulsion Low-temperature EPG −19°C −25°C −31°C CSS-1h-C PASS PASS PASS CSS-1h-L PASS FAIL FAIL CQS-1hP-E PASS PASS FAIL CQS-1h-E FAIL FAIL FAIL CSS-1hP-D PASS PASS FAIL Table 2.20. Summary of low-temperature EPGs of microsurfacing emulsions evaluated.

50 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments Both CQS-1hP-E and CSS-1hP-D have a low-temperature EPG of -25°C. CSS-1h-L has a low- temperature EPG of -19°C, and CQS-1h-E fails the specification at -19°C. 2.5 Performance-Graded Specifications in Practice This section describes how the proposed EPG specifications can be used in practice. The aver- age testing time required for each test method in the EPG specifications is provided for reference. Also included is a demonstration example of how the chip seal, microsurfacing, and spray seal emulsions are graded in the respective frameworks within the overall EPG specifications. 2.5.1 Estimated EPG Specification Testing Time Table 2.21 provides an estimate of the time required to complete each test (for the recom- mended number of replicate specimens) for a single emulsion type tested at a single temperature and the number of replicates recommended for each test method in the EPG specifications. Although the times needed for specimen conditioning and curing are exact, the specimen preparation and testing times are estimates, as these times would vary based on the experience of the user conducting each test and do not account for any potential coordination of the curing/ conditioning of the specimens to shorten the test preparation times. The residual binder testing portion of the EPG specifications applies only to chip seal and microsurfacing emulsions. The residual binders were recovered from emulsions using AASHTO PP 72 Method B for all test methods for residual binders in the EPG specifications. 2.5.2 Equipment Required for Test Methods Used in EPG Specifications For chip seal and microsurfacing emulsions, only the DSR and rotational viscometer are required to complete all the tests in the EPG specifications. For spray seal emulsions, only the rotational viscometer and rolling ball test device are required. 2.5.3 Implications of Three Levels of Traffic in the EPG Specifications Although the EPG specifications include three different traffic levels at each temperature grade, there is no need to be concerned that they must carry three emulsion storage tanks per Table 2.21. Estimated testing time for proposed EPG test methods.

Results, Interpretation, and Applications 51 grade. The EPG specifications have flexibility in that an EPG 67 (low) also can be used in an EPG 61 climate at high, medium, or low traffic levels, for example. Likewise, an EPG 67 (high) could be used in an EPG 67 climate for low and medium traffic levels as well. Therefore, in an EPG 67 location, an emulsion supplier would not necessarily need to store 67 (low), 67 (medium), and 67 (high) emulsions. The flexibility of the EPG specifications allows suppliers to make decisions about how best to store emulsions cost-effectively, while promoting better emulsions that meet performance-based emulsion specification standards. 2.5.4 Emulsion Grading Example for EPG Specifications This section presents an example that uses measured data from the emulsions tested during the development of the EPG specifications to show how the EPG specifications grade typical asphalt materials that are used for chip seal construction. The emulsions were graded for use in an EPG 67-19 climate and at low, medium, and high traffic volumes. Table 2.22 presents the results. Table 2.22 identifies both the fresh emulsion and residual binder test methods included in the EPG specifications. For each test method, the binder property measured and its corresponding specification limit are provided. The first column provides the existing emulsion name under the naming convention used in this study, with the letter in parentheses anonymously denoting the supplier of that particular emulsion material. For each emulsion listed, the measured test values are provided along with a pass/fail designation based on each test limit. The far right column denotes either that the emulsion has failed for the traffic level at the test temperature being considered, or denotes that the emulsion has passed and grades the asphalt material accordingly. It is important to note that any material that fails the EPG specification criteria in these examples could very well pass the EPG specification criteria in a different climate. 2.6 Short-Term Validation of Performance-Graded Specifications 2.6.1 Chip Seal Short-Term Validation For the short-term validation of the proposed EPG specifications for chip seals, 305-meter field sections were constructed in Knightdale, North Carolina, on a road section with an AADT of approximately 2,000 vehicles (i.e., a medium traffic roadway in the EPG specifications). Two single-seal and two triple-seal sections were constructed using the CRS-2L-NC emulsion. The emulsions that were used to construct the field sections were the same materials tested using both the mixture and binder test protocols in the proposed EPG specifications. The field emul- sions were sampled directly from an asphalt spray tanker on the day of construction. The field performance was then observed during the first year in service to provide short-term validation of the chip seal specifications. Asphalt binders with a high-temperature grade of 67°C would be used in North Carolina where these validation sections were constructed. Therefore, the emulsion used was tested at the high temperature of 67°C. The results of the MSCR test that measured the high-temperature binder performance for the EPG specifications yielded the Jnr value of 4.8 kPa -1. The chip seal EPG specifications define the acceptable medium traffic performance of a binder as having a non-recoverable creep compliance (Jnr) measurement below 5.5 kPa -1 at a stress level of 3.2 kPa. Therefore, this binder meets the high-temperature EPG specification limit for medium traffic at 67°C and would be graded as an EPG 67 residual binder for the high-temperature specifi- cations. Based on these test results, the field sections constructed using these materials should not

Existing Emulsion Name Proposed EPG Grade Fresh Emulsion Tests Residual Binder Tests Pass & Grade or Fail at Test Temp/ Traffic Level Emulsion Type Climate Traffic Level Storage Stability Rd < 2 Separation 0.5 <Rs < 2 Sprayability < 400 cP Drainout > 50 cP High Temp. Grade Test at 67°C Low Temp. Grade Test at −19°C Bleeding MSCR Max Jnr @Low <8 kPa-1 @Med <5.5 kPa-1 @High <3.5 kPa-1 Low Temp. Raveling DSR Max G* at δcrit @Low < 30 MPa @Med < 20 MPa @High < 12 MPa CRS-2 (NC) CRS 67-19 Low 1.4 Pass 1.4 Pass 190 Pass 400 Pass 7 Pass at Low 20 Pass at Low CRS-EPG67- 19L Med 7 Fail at Med 20 Pass at Med Fail High 7 Fail at High 20 Fail at High Fail CRS-2L (F) Low 1.1 Pass 1.0 Pass 180 Pass 350 Pass 6.75 Pass at Low 14 Pass at Low CRS-EPG67- 19L Med 6.75 Fail at Med 14 Pass at Med Fail High 6.75 Fail at High 14 Fail at High Fail CRS-2P (A) Low 0.3 Pass 1.0 Pass 80 Pass 450 Pass 2.5 Pass at Low 4 Pass at Low CRS-EPG67- 19L Med 2.5 Pass at Med 4 Pass at Med CRS-EPG67- 19M High 2.5 Pass at High 4 Pass at High CRS-EPG67- 19H Table 2.22. Chip seal EPG specification grading examples.

Results, Interpretation, and Applications 53 exhibit bleeding in the chip seal; indeed, none of the single-seal or triple-seal sections exhibited any signs of bleeding after over 1 year in service. At low temperatures, raveling due to binder fracture is the critical distress addressed in the EPG specifications. The results of the DSR frequency sweep tests yielded 18 MPa as the dynamic shear modulus (G*) value at the critical phase angle (dc). The chip seal EPG specifications define the acceptable medium traffic performance of a binder as having a measured G* value at dc below the 20 MPa threshold. Therefore, this binder meets the low-temperature EPG specification limit for medium traffic at 67°C and would be graded as EPG 67-19 according to the proposed EPG specifications. Based on binder test results, the chip seal field validation sections constructed using these materials should not exhibit low-temperature aggregate loss; indeed, none of the single-seal or triple-seal sections exhibited any signs of low-temperature aggregate loss after a full winter in service. Figure 2.38 shows a photograph of one of the short-term field validation sections that was constructed using CRS-2L-NC emulsion after 15 months in service. None of the four sections constructed using the CRS-2L emulsion exhibited bleeding under medium volume traffic load- ing and no excessive aggregate loss was observed after the first winter. Also, these sections did not exhibit any shelling from the effect of aging during this year in service. This finding sup- ports the recommendation not to include tests of aged binders in the EPG specifications because low-temperature aggregate loss can occur during the first winter prior to any significant aging. Therefore, these short-term field observations, although limited in terms of emulsion types included, are in agreement with the expected chip seal performance based on the EPG specifica- tion test results measured for these emulsion materials under the climatic and traffic loading conditions experienced by the sections in North Carolina. Figure 2.38. Pavement conditions after 15 months in service for the CRS-2L validation sections.

54 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments 2.6.2 Microsurfacing Short-Term Validation Microsurfacing field validation sections were constructed by the North Carolina Department of Transportation (NCDOT) in Forsyth County, North Carolina, in June 2013. Initially, the short-term validation plan was to include different types of emulsion used in microsurfacing field construction. However, only projects using CSS-1hL emulsion were available during the summer of the planned construction; it was the only microsurfacing emulsion type used in the short-term validation sections. For these sections, the CSS-1hL emulsion was used with cement as the mineral filler. Type III aggregate gradation specified by the ISSA was used. The mix design was 1% mineral filler and 6.5% residue content based on the dry weight of aggregate with the required amount of water to obtain the proper consistency. The mix design was developed by the NCDOT staff based on previous experience and local design calibrations. On the day of construction, the emulsion and aggregate materials were collected for laboratory EPG specifica- tion testing. However, pre-existing cracks on the pavement surface were not properly remedied prior to construction, and the existing surface displayed a wide variety of cracking types, includ- ing transverse, longitudinal, and block cracking. The dimensions of the block cracks were about 300 mm × 300 mm on average. The contractor applied only a tack coat before applying microsurfacing on the pavement, but the cracks were not sealed. The cracks can be observed in Figure 2.39. The first field visit was made in November 2013, 5 months after construction. On the day of the field visit, a visual inspection was performed to compare the pavement condition before and after the microsurfacing application. The visual inspection revealed that cracks had developed in the pavement through the microsurfacing layer. In addition, a representative of the NCDOT reported that the cracks had propagated through the microsurfacing within weeks of construction; these cracks can be observed in Figure 2.40. The research team conducted a study to determine the locations of the crack initiation and the direction of the propagation. For this study, field cores were extracted at three different locations within the sections where cracking had occurred. Cores were extracted using a 150-mm diameter coring machine at the full depth of the pavement layer above the base layer. Figure 2.41 shows the field cores, with cracking observed throughout the pavement layer in addition to the surface. a b Figure 2.39. Images taken prior to field construction of (a) patched areas and (b) cracks.

Results, Interpretation, and Applications 55 b ca Figure 2.40. Microsurfacing cracks visible prior to the first winter following construction. b ca Core-2 Core-3Core-1 Figure 2.41. Field cores showing cracks throughout the pavement layer and at the surface.

56 Performance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments Based on the cores, the existing pavement layer was determined to be 300-mm deep and was found to be composed of several HMA layers, an asphalt binder layer, and a concrete layer at the bottom. Each core was covered with an approximately 6.25-mm thick layer of microsurfacing. This microsurfacing layer was observed to be compacted from its original thickness of 12.5 mm during construction. For Core-1 and Core-2 (see Figure 2.41), cracking was observed to run through all of the HMA layers and even into the concrete layer at the bottom of the core. In Core-3, the crack ran through all of the HMA layers, but was not present in the concrete layer. Additionally, debonding was observed between the concrete and bottom HMA layers in Core-2 and Core-3. These observations suggest that the cracks that were visible in the microsurfacing layer were existing cracks that had propagated up through the microsurfacing (i.e., reflective cracks). Because existing cracks propagated through the microsurfacing shortly after construction, an assessment of binder performance and an evaluation of the microsurfacing field sections were not appropriate.