Abstract
Fourteen samples of an acid modified bitumen for asphalt manufacture (known as M1000) were tested according to Australia's national paving-grade bitumen specification and using Multiple Stress Creep Recovery (MSCR) protocols. Testing was conducted two, three and four years after manufacture. The 14 samples represented a number of batches of bitumen of identical type and grade manufactured over a six-month period. All batches were manufactured in Australia from imported feedstock (blends of refined crude oil). Six samples were manufactured from one feedstock and the remaining eight from a different feedstock. Inter-batch/intra-feedstock variability of compliance properties at the time of release was low (4–16%). For most of the test parameters (including viscosity, penetration and creep compliance), there were significant differences for the samples manufactured from the two different feedstocks (p-values .06–<.01). The difference between samples from different feedstocks was magnified by accelerated ageing and by pretest storage time. For example, the average 17% difference in viscosity after accelerated ageing measured at the time of release had increased to an average 82% difference four years later. It is recommended that the Australian specification be modified to protect consumers from unintended variability in performance resulting from unknown changes in imported feedstocks. The tests showed that the current bitumen specification's focus on viscosity is unable to discriminate between feedstocks, resulting in significant differences in bitumen ageing, stress sensitivity and high temperature (above 64°C) response to shear stress. The MSCR test is recommended as a performance-based assessment for bitumen response to shear stress at high temperature.
Acknowledgements
The views expressed in this paper are solely those of the author. The author's employer and the bitumen supplier do not necessarily support those views. This paper examined the impact of a single change in feedstock used to manufacture M1000 bitumen. The conclusions do not suggest that M1000 properties change significantly with every change in feedstock or that M1000 is a highly variable product. It was not intended to make any conclusion about the suitability or performance of M1000 as a binder for asphalt.
Disclosure statement
No potential conflict of interest was reported by the author.