Asphalt Mixture Variability and Its Effect on Balanced Mix Design Testing from Design to Production

About the report:

VDOT’s roads have been experiencing increased surface cracking for several years, and BMD is seen as providing a means to address cracking concerns while also balancing the mixture’s susceptibility to other distresses. Another benefit of BMD implementation that is attractive to VDOT is the ability to consider innovations in mix designs.

However, one of the gaps in available information regarding BMD is the influence of production variability on laboratory BMD test results. As VDOT moves forward with implementation, an effective quality assurance plan is needed to ensure that the performance evaluated in the laboratory is replicated in the field. However, to develop this plan, the influence of production variability on performance, as defined by the laboratory performance testing criteria, must first be considered.

The objective of this study was to compare mixture and test variability determined under controlled laboratory conditions with that found in actual production conditions. This comparison is important because it indicates the ability of volumetric quality control measures to keep mixture performance under control during production. The controlled laboratory conditions simulated the designed job mix formula and variations in gradation and binder content based on the maximum acceptable production limits. Production conditions were considered to include the original job mix formula volumetrics and BMD test data as generated by the producer for mix design submittal as well as subsequent production sampling and the resulting volumetric and test data.

In general, the results indicate that the test variability is similar to or wider than the variation induced from the production tolerance limits on coarse and fine aggregates, suggesting that the current material tolerance limits on aggregates are appropriate for the BMD concept from the perspective of the Indirect Tensile Cracking Test. However, the variations induced due to the production tolerance limits on asphalt content, in this case ±0.3%, were significantly wider than the test variability. Overall, it was found that asphalt mixtures that have been designed with an average performance result within the single-operator precision estimates of CT index performance thresholds may have a risk of failures during production due to the inherent material and test variability. Similar assessments of production variation and test variability with respect to the Cantabro mass loss and APA rut depth were unable to be pursued, as precision estimates are not available for these tests.

The study recommends that VDOT’s Materials Division should continue the use of current production tolerance limits on coarse and fine aggregates for acceptance of BMD performance plus volumetric optimized mixtures. In addition, the Virginia Transportation Research Council, in collaboration with the Materials Division, should evaluate alternative methods to support and encourage controlling materials variability, such as binder content, aggregate gradation, and reclaimed asphalt pavement, during production. Finally, the Virginia Transportation Research Council should develop precision estimates for the Cantabro mass loss and the asphalt pavement analyzer rut tests.

Authors