By Mike Anderson, John D’Angelo and Dwight Walker
In 1987, the U.S. Congress budgeted $150 million for the Strategic Highway Research Program and $50 million was set aside to study asphalt materials. In 1993, the Performance Graded (PG) asphalt binder specifications were introduced as part of the Superpave system. These specifications were adopted as AASHTO M320.
The PG system was primarily developed around unmodified asphalts, and G*/sin d (G-star divided by sine delta) became the high temperature specification parameter and rutting performance indicator in the specification. G*/sin d is determined by the Dynamic Shear Rheometer (DSR), one of the basic asphalt binder testing devices.
In the Superpave system, the DSR is used to characterize the stiffness and elasticity properties of asphalt binders at high and intermediate temperatures. An asphalt sample is placed between two parallel plates and a torque is applied and the response measured. The test results are used to estimate the resistance to rutting and fatigue cracking.
One of the objectives in the development of the PG asphalt binder specification was the use of performance-related criteria specific for a distress and related to climate and traffic loading. This objective implies that test measurements should be made at temperatures and loading rates consistent with conditions existing in the pavement. With this approach, the high temperature criteria stays the same for G*/sin d (1.00 kPa for unaged and 2.20 kPa for RTFO-aged binder) regardless of the location of the pavement, but the test temperature where this criteria must be met is derived from the actual pavement temperature.
While this concept worked well for conventional-speed, moderate-traffic volume pavements, research indicated that it needed some refinement for pavements that had slow-speed loading and high traffic volume. Rather than change criteria and/or test conditions to reflect a change in loading time and traffic volume, the architects of the PG system elected to simply adjust for traffic speed and volume by “grade-bumping” or testing at higher temperatures than indicated by the climate.
So, for a standard traffic asphalt pavement, the designer might use a PG 58-28 asphalt binder, but a high-volume highway pavement might require a PG 70-28 asphalt binder—even though the pavement temperature would likely never get above 58°C. This was a simple way to ensure that a stiffer asphalt binder would be used in high volume and/or slow loading conditions.
One problem with grade-bumping in the PG system was that a PG 70-28 asphalt binder would have its performance-related properties determined at a temperature that would be at least 12°C hotter than the highest pavement temperature that would be experienced. Such high specified testing temperatures in some instances have caused asphalt suppliers to manufacture binders that are very highly modified and thus difficult to use at reasonable temperatures.
Another objective of the PG system was that the performance-related properties that defined the grade of an asphalt binder be blind to modification. In other words, all asphalt binders of the same performance grade would be expected to perform the same in the same traffic/environmental conditions regardless of how they were produced. This would allow the asphalt industry to divest itself of the specification proliferation that had become increasing more common as modification of asphalt binders became more common.
While the G*/sin d parameter did capture viscous and elastic effects, it was unable to adequately capture the benefits of elastomeric modification because of the relatively small impact of phase angle (d) on the overall value of G*/sin d. As a result, additional empirical tests (commonly referred to as “PG Plus” tests) continued so that a user agency could have assurance of getting a polymer-modified asphalt binder as in the past.
These issues caused researchers to continue to look for an improvement to the high temperature parameter, G*/sin d, used in AASHTO M320. Their efforts led to the development of a new test procedure, the Multiple-Stress Creep-Recovery (MSCR) test.
MSCR test and specification
The MSCR test was developed based on creep and recovery work conducted on asphalt binders and mixtures. During SHRP, researchers at the University of California at Berkeley developed the Repeated Simple Shear Test at Constant Height (RSST-CH) for asphalt mixtures. The RSST-CH was developed to characterize the rutting performance of asphalt mixtures and was conducted using repeated cycles of 0.1-second shear load followed by a 0.6-second rest period. Based on this mixture work, the NCHRP 9-10 research used a repeated creep-recovery test to characterize the expected rutting performance of modified asphalt binders.
The MSCR test (AASHTO TP70) uses the well-established creep and recovery test concept to evaluate the asphalt binder’s potential for permanent deformation. Using the DSR, a one-second creep load is applied to the RTFO-aged asphalt binder sample. After the 1-second load is removed, the sample is allowed to recover for 9 seconds. The test is started with the application of a low stress 0.1 kPa for 10 creep/recovery cycles and then the stress is increased to 3.2 kPa and repeated for an additional 10 cycles.
The material response in the MSCR test is significantly different than the response in the existing PG tests. In the PG system, the high temperature parameter, G*/sin d, is measured by applying an oscillating load to the asphalt binder at relatively low shear strain. This is one of the reasons why the existing PG high temperature parameter does not accurately represent the ability of some polymer modified binders to resist rutting. Under the very low levels of stress and strain present in dynamic modulus testing, the polymer network is never really activated.
In the existing PG specification the polymer is really only measured as a filler that stiffens the asphalt. In the MSCR test, higher levels of stress and strain are applied to the binder, better representing what occurs in an actual pavement. By using the higher levels of stress and strain in the MSCR test, the response of the asphalt binder captures not only the stiffening effects of the polymer, but also the elastic effects.
In the MSCR test, two separate parameters can be determined—non-recoverable creep compliance (Jnr) and percentage of recovery (MSCR Recovery) during each loading cycle. Values are reported as the average of ten loading cycles at each shear stress level. This parameter has been shown through numerous field and laboratory studies to better correlate with rutting potential than G*/sin d—particularly for modified asphalt binders.
Unlike the AASHTO M320 system, the test temperature used for the MSCR test is selected based on actual high pavement temperatures with no grade bumping. Thus in the previous grade bumping discussion, the MSCR test would be performed on an asphalt binder at a high temperature of 58°C regardless of the traffic speed and loading. In the MSCR specification, AASHTO MP19, higher traffic loading is accounted for by increasing the stiffness (reducing the compliance) required for the asphalt binder at the grade temperature.
For standard traffic loading, Jnr (determined at 3.2 kPa shear stress) is required to have a maximum value of 4.0 kPa-1. As traffic increases to heavy and very heavy loading, the Jnr of the asphalt binder needs to be lower—requiring a maximum value of 2.0 and 1.0 kPa-1, respectively.
While the main requirement for Jnr is determined at 3.2 kPa shear stress, the data determined at 0.1 kPa shear stress is also important. To minimize concerns that some asphalt binders may be overly sensitive to changes in shear stress, AASHTO MP19 maintains a requirement that the difference in Jnr values between 0.1 kPa and 3.2 kPa shear stress should not exceed a ratio of 0.75.
Although it is not part of the specification, in addition to determining Jnr the MSCR test can be used to determine the amount of recovery in an asphalt binder during the creep-recovery testing. MSCR Recovery provides an indication of the delayed elastic response of the asphalt binder. A high delayed elastic response is an indication that the asphalt binder has a significant elastic component at the test temperature.
The MSCR Recovery may be used in combination with Jnr to indicate whether an asphalt binder has a sufficient elastic component. Asphalt binders that fall below the curve are considered to have low elasticity; those that are above the curve are considered to have high elasticity.
It is important to remember that the high temperature binder specification parameter from the MSCR test is Jnr. If the asphalt binder meets the appropriate Jnr specification, then it should be expected that the binder will minimize its contribution to rutting. However, if a user agency wants to validate that an asphalt binder has been polymer-modified, adding the appropriate MSCR Recovery value as a minimum requirement is an option.
If a user agency does use PG Plus tests, it makes sense that MSCR Recovery should replace those other PG Plus tests that are intended to have a similar purpose. In other words, MSCR Recovery should be expected to replace Elastic Recovery, Force Ductility, and Toughness and Tenacity tests. Other tests that have a different purpose, like the Separation test, may still be required.
Although technologists will no doubt conduct comparative testing between MSCR Recovery and other PG Plus tests, they should not necessarily expect strong correlations. Test conditions are sufficiently different between the MSCR and “PG Plus” tests that are strong relationship would be unlikely.
The MSCR test is the latest improvement to the Superpave Performance Graded (PG) Asphalt Binder specification. This new test and specification provide the user with a high temperature specification parameter that more accurately indicates the rutting performance of the asphalt binder and is blind to modification. It is an improvement in several ways:
- Jnr is better correlated with rutting potential than G*/sin d.
- The MSCR test results from just a single test can be used with modified and unmodified asphalt binders, thereby eliminating the need for additional tests to properly characterize the high temperature performance of modified asphalt binders.
- There is now criteria to eliminate binders that are overly stress sensitive, which would previously have passed the PG criteria and potentially been susceptible to rutting in the field.
- MSCR Recovery is faster/easier to determine than other “PG Plus” tests like the Elastic Recovery test and does a better job of characterizing polymer-modified asphalt binders.
- The MSCR test is conducted at the actual pavement temperature, regardless of traffic loading.