Effect of Mineral Filler Type and Amount on Design and Performance of Asphalt Concrete Mixtures

Abstract

Three methods for classifying aggregate particle shape and texture—AASHTO TP33, ASTM D3389, and the flow rate method—were evaluated. These methods were used to rank four natural river sands and a crushed granite from good to poor performance based on the criteria established by each method. Test results indicate that all methods easily distinguished the crushed aggregate from the natural river sands. The AASHTO TP33 and the flow rate method were found to be somewhat less sensitive to slight differences in particle shape and texture than was ASTM D3398. All the test methods were found to be repeatable, each having low coefficients of variation for all the aggregates tested. To evaluate the effect of particle shape and texture and mineral filler content on mix performance in permanent deformation, one natural sand ranked as average performing was selected and blended with the crushed granite in the proportion of 20 percent natural sand and 80 percent crushed granite. Asphalt-aggregate mixtures containing 4, 6, 8, and 12 percent mineral filler were designed by using the Marshall procedure and were tested in repeated shear test at constant height. Within the range of mineral filler type and contents used, results indicate that mixtures containing 100 percent crushed granite show lower accumulation of permanent strain than does an 80/20 blend of crushed granite and natural sand, and an increase in the mineral filler content of a mixture was found to decrease its accumulated permanent strain while increasing the mixture shear resilient modulus.