University of California Pavement Research Center

In 2015, Caltrans expressed interest in studying the addition of small amounts of crumb rubber (CRM) in dense-graded asphalt mixes toincrease the total amount of recycled tire rubber used. Small amounts were defined as 5% to 10% CRM by weight of the binder orapproximately 0.25% to 0.5% CRM by weight of the aggregate. In this report, the terms “CRM binder” and “CRM mix” are used todenote the modified mixes. Four approaches for adding the rubber were proposed: 1) Addition of 5% to 10% CRM particles smaller than250 μm to the asphalt binder, not resulting in a change to the PG of the base binder, achieved by blending softer base binders and/orpolymers with the rubber at the refinery/terminal; 2) Addition of 5% to 10% CRM particles smaller than 2.36 mm to the asphalt binder,with allowable changes to the PG of the base binder, and produced using a field-blending process similar to that used for producingasphalt rubber binders with between 18% and 22% CRM; 3) Adding 0.25% to 0.5% CRM by weight of the aggregate directly into the mixusing a dry-process; and 4) Addition of 5% to 10% CRM with particles smaller than 250 μm to the asphalt binder, with allowable changesto the PG of the base binder, and produced using a field-blending process.

Laboratory test results and mechanistic-empirical performance simulations both indicate that dense-graded mixes produced withbinders containing between 5% and 10% crumb rubber modifier (CRM) by weight of the binder will generally have equal or betterperformance to dense-graded mixes produced with unmodified binders. Finer CRM gradations (i.e., smaller than 250 μm) in wet processapproaches allow binder testing with standard Superpave performance grading tests and appear to provide more consistent results.Based on literature reviews, adding between 0.25% and 0.5% CRM with particles sizes smaller than 500 μm in dry process mixes willalso provide equal or better performance to mixes that contain no CRM. If any of the approaches are adopted, more scrap tires wouldbe recycled into pavement applications. The following recommendations are proposed based on the findings from this study:

• Additional mechanistic-empirical performance simulations followed by pilot studies should be carried out to confirm the findingsdiscussed in this report, to better quantify the benefits, to expand the CalME materials library, and to identify the mostappropriate applications in pavement structures in the different California climate zones.

• Some relaxation of the solubility requirements in the PG-M specification should be considered to allow more use of Approach-4binders. Laboratory test results and performance simulations did not indicate that a relaxation in solubility requirements wouldhave a detrimental effect on performance.

• Given that dry process approaches are the simplest and cheapest method of incorporating CRM into mixes, limited additionaltesting with finer CRM particles, along with performance simulations, should be conducted to confirm that findings from researchconducted in other states and countries are applicable to California applications.