What if BMD is a Failure? There Are Many Different Measures of Success

Trenton M. Clark, PE, President, Virginia Asphalt Association

For the better part of two decades, researchers and departments of transportation have invested millions of dollars into laboratory asphalt mixture testing in order to improve design and performance. Much effort has been spent on directly measuring the physical engineering properties of asphalt mixtures through devices such as the AMPT (Asphalt Mixture Performance Tester) in an effort to improve the predictive modeling for flexible pavement design. While this approach has benefits, it also has several shortcomings, such as the time required during mix design and usefulness during actual asphalt production. Therefore, researchers and agencies have turned their attention to less complicated testing procedures that can be used during mix design and production. While these procedures do not produce fundamental engineering properties such as modulus and flow number, they do provide a method to assess material performance against a set of minimum or maximum thresholds. Essentially, agencies are hoping to improve cracking resistance while not resulting in rutting/deformation. In other words, balance distress modes to maximize performance. In the end, agencies such as VDOT have stated that to be successful, a Balanced Mix Design (BMD) process must result in new asphalt mixes that perform better than the “recipe” mixes resulting from conventional Superpave design.

But if BMD mixes are no better, will BMD be a failure?


Most design procedures evolve, change and adapt over time. Asphalt mix design is no different. Just look through decades of state highway agency specification books and this will be obvious. What started out as a trial and error approach more than 125 years ago has evolved into a very complicated process. However, the basic ingredients in asphalt mix have not changed—fine aggregates, coarse aggregates and asphalt binder tossed in with some recycled materials and at times polymer modified binder. The evolution over the decades has been the recognition of how these basic ingredients and other additives in various proportions impact performance.

During the Marshall Mix Design era (1980s and 1990s), much emphasis was placed on mixture volumetrics—specifically air voids or voids in total mix (VTM), voids filled with asphalt (VFA) and voids in mineral aggregate (VMA). Understanding the role of these mixture properties contributed to significant strides in mixture design. However, it became clear in the late 1970s and early 1980s that properties other than volumetrics must be considered to combat the asphalt rutting and bleeding experienced across the US. These concerns led to the asphalt mix design research efforts under the Strategic Highway Research Program (SHRP) in the mid-1980s. Ultimately, the outcomes from SHRP led to more stringent aggregate properties, a new asphalt binder grading system and design mixture compaction levels as a function of pavement loading. While envisioned as a three-level mixture design process (Level 1 recipe design, Level 2 lab performance design and Level 3 engineering properties design), limitations in lab testing equipment kept state highway agencies at Level 1. However, the lack of robust testing equipment and procedures did not deter researchers and agencies from pursuing Level 2 and Level 3 design processes.

Asphalt mix testing.

The Asphalt Mixture Performance Tester or AMPT device was developed through National Cooperative Highway Research Program (NCHRP) projects 9-19 (Superpave Support and Performance Models Management) and 9-26 (Simple Performance Tester for Superpave Mix Design)1 to achieve Level 3 designs for mixes where the results can be used in mechanistic/mechanistic-empirical flexible pavement design methodologies. While the AMPT has progressed since its initial inception in the early 2000s, there has not been widespread adoption by highway agencies or the contracting industry. One drawback to the AMPT is the time required to achieve results. During the mix design process, the turnaround time from specimen preparation to final results is less critical (typically 4 to 5 days). However, if adjustments to the design are needed then the time to final approval can be extensive. Likewise, during the production of asphalt, test results are needed in “real-time” to make adjustments. If the mix design approval is based on fundamental engineering properties and not standard properties such as asphalt content, aggregate gradations and volumetrics, then a contractor will not be able to adjust their mix during production in order to ensure conformance with the mix design. Many days will pass from the time the sample is retrieved to final test results.

An asphalt sample in IDT test equipment.

Recognizing other challenges exist with the production use of AMPT, research moved to focus on Level 2 mix design using lab performance tests, generally referred to as BMD. Many extensive efforts have been underway for the past decade at numerous universities to develop lab performance tests that correlate to in-service performance of asphalt mixes. The most notable early tests developed have been the Illinois I-FIT, the Texas Asphalt Overlay Cracking Test, the Disc Shaped Compact Tension and the Semi-Circular Bend test. Since the development of these procedures and use by state agencies, the Indirect Tensile (IDT) has gained broader acceptance for assessing cracking potential; the Cantabro test for evaluating asphalt mix and durability, and either the Asphalt Pavement Analyzer (APA) or Hamburg wheel tracking device to assess mixture rutting potential. Based on a recent survey of state agencies, 10 states are implementing while others are in the process of either implementing or evaluating BMD tests.2

The Basic Premise

As stated earlier, state agencies such as VDOT are investigating or implementing BMD in the hopes that new asphalt mixes perform better than current asphalt mixes. Better performing mixes, while potentially more expensive initially, will have lower life cycle costs due to the better performance and longer service life.

Recently, the National Center for Asphalt Technology (NCAT) completed a report for NCHRP (NCHRP Project 20-07 Task 406) outlining a model for BMD implementation. A few important points were made in that report. First, an agency needs to evaluate the numerous lab performance tests available that fit the state. For example, the predominant cracking experienced in Montana will be different than the cracking in Florida. While fatigue cracking may occur in Montana, temperature-induced thermal cracking may be more relevant to the Montana Department of Transportation (MDT) in selecting a cracking test protocol. Florida DOT may be more concerned with fatigue cracking and rutting due to high pavement temperatures.

Once a test or set of tests are selected, a second important step is to baseline the current or standard mixes used in the state. As the old adage goes, if you do not know where you are, how can you know where you are going? Through extensive lab testing, contractors and the agency can determine a benchmark or baseline for existing materials by nominal maximum aggregate size and binder type. By cross-referencing the standard mixes with performance data from a pavement management system, an agency can begin to determine what “better” means by using BMD.

BMD experimental mix being laid at VTTI.

But What if Better Isn’t Achieved?

While so much time, effort and funding has been invested in BMD, what if BMD does not produce better mixes? What if the only outcome is a different mix design process? Would that make BMD is a failure?

According to Merriam-Webster’s Dictionary, failure is defined as “omission of occurrence or performance, state of inability to perform a normal function…a lack of success.” In those terms, when compared to the impetus behind BMD implementation, an asphalt mix of equal performance will be deemed a failure. However, is that really the case?

BMD can be a success even if the overall performance of asphalt mixes is not improved. There are three main successes agencies can achieve as well as the contracting industry. The first success comes in the area of knowledge. For too long asphalt mixes have been designed to meet a set of metrics—asphalt content, aggregate gradation bands and volumetric properties. In general, this approach has worked. Like any normal distribution of performance, some mixes have performed better than expected, many have performed as expected and others have not performed well at all. Without immediate failures, the bad performers are not identified without tying the job mix formula to measured field data many years after the mix was produced. By understanding the overall lab performance of mixes and establishing actual baselines, each existing or new mix design can be assessed. The contractor and the agency can evaluate ways to improve lab performance outcomes that may improve performance in the field that would benefit the agency.

The second success can be seen in reduced costs to the taxpayer. These reductions may be in terms of initial cost, i.e., producing the asphalt, or long-term in lower life cycle cost. In a performance-based BMD approach, the actual performance of the mixture is paramount. Except for certain criteria such as design and production VTM as well as top-sized aggregate (function of lift thickness and pavement design), contractors are free to design mixes that meet or exceed the performance criteria. For some contractors, that may change the amount of recycled asphalt pavement (RAP) or asphalt shingles (RAS) used in a mix. This could lead to cost savings for the contractor and agency. The contractor may be able to use more local aggregates and not have to truck in special materials to meet gradation bands and volumetric properties. Again, this will be a cost savings. A combination of recycled materials, specialty binders and mix additives could result in lower costing or better performing materials. Whether it is lower costs or longer performance by eliminating poorer performing mixes, the cost to the taxpayer should be reduced.

Finally, success can be seen in terms of sustainability. Performance-based BMD mixes will be more sustainable, especially in urban areas where RAP and RAS are abundant. Incorporation of RAP and RAS beyond the current limits in most state agency specifications will reduce the demands for virgin aggregates and virgin asphalt binders. In terms of aggregates alone, this will extend the life of quarries and reduce the amount of energy needed to blast, haul, crush and truck materials to asphalt plants. Binder additives such as softeners and rejuvenators will allow aged asphalt binder in RAP and RAS to be better utilized in new mixes. Mix additives such as ground tire rubber, crumb rubber and various fibers can be part of performance designed mixes. Many of the additives are products of other waste streams or by-products. And while the research is still being conducted and consensus conclusions still not obtained, plastics may be in future asphalt mixes. To understand the impacts on performance, BMD will be an assessment tool.

Don’t Miss the Forest for the Trees

Will BMD result in better mixes? The hopeful answer to that question is “yes.” However, it will be many years before an agency will be able to definitively answer that. Even if the new mixes meet certain criteria during design and production, will the change in actual performance be measurable? Given the number of factors that contribute to the actual performance of a mix, improvement may be a nebulous term. In the meantime, agencies should not miss the forest for the trees. Proper implementation of a performance-based BMD approach will have long term benefits through knowledge gained, reduced costs, and improved sustainability.


  1. “TechBrief: Asphalt Mixture Performance Tester,” FHWA 2013, https://www.fhwa.dot.gov/pavement/asphalt/pubs/hif13005.pdf
  2. “Facilitating Balanced Mix Design Implementation,” NCHRP 20-44(27) Project Description, https://apps.trb.org/cmsfeed/TRBNetProjectDisplay.asp?ProjectID=4920