On the impact of the filler on the complex modulus of asphalt mixtures

Abstract

Mineral fillers have been used in asphalt mixtures from their very beginning. Rapidly, formulators observed that the filler could have a strong influence on the final mixture. Nowadays, mineral fillers are essentially characterised by their ability to stiffen the bitumen, leading to the development of the celebrated “Rigden air voids” RV (EN 1097-4), now the corner stone of the European specifications for fillers in asphalt mixtures. If the way the filler modifies the rheological properties of mastics is now well described, the relationship between mastic and mixture properties remains poorly understood. Therefore, this paper describes a systematic study on the impact of increasing filler contents on asphalt mixtures. Experimental results are presented for two very different asphalt mixture formulas, a regular semi-dense Asphalt Concrete (AC) and an open-graded AC for very thin layers (BBTM). The complex modulus of each formula was measured as a function of filler content with two materials having very different stiffening behaviour, that is, limestone filler (LF) and hydrated lime (HL). For both formulas and regardless of the type of filler, the modulus versus filler content curves exhibited a well-defined maximum. The filler content at which the maximum was found was strongly filler dependent, with HL formulas peaking at 4 wt.% and LF formulas peaking at 10 wt.%. An original interpretation based on reduced variables was proposed, showing that the maximum in modulus occurred for reduced volume fraction of filler close to 66%. This suggests that, before the peak, modulus increased with filler content as a consequence of a stiffer mastic, and after the peak, modulus started to decrease due to poor aggregate coating as a consequence of too-viscous a mastic. As a practical consequence, the studied mix formulas could accept higher filler loadings than currently specified, which would potentially stiffen the mixtures by ∼20%. Therefore, one way to benefit from these findings without changing the specifications would be to use blends of HL and LF. Of course, these findings must be validated with other mixture formulas and raw materials, and similar systematic study on the filler impact on the resistance to fatigue and moisture damage would be needed before implementing the conclusions based on modulus only.