Variability of hot mix asphalt produced with reclaimed asphalt pavement

Abstract

Pavement recycling operations have been used in the United States from the early 1900s to date. These recycling operations have been used for a variety of reasons including solid waste disposal problems, reduction of initial cost, and conservation of natural resources (aggregate, binder and energy). Adequate techniques to assess the variability of Reclaimed Asphalt Pavement (RAP) are lacking. The objectives of this research are to determine the variabilities of RAP and Hot Mix Asphalt (HMA), determine the influence of the RAP variability on the HMA variability, and compare the variability of virgin aggregates at the quarries and the aggregates that have been handled and transported to asphalt plant stockpiles. This thesis describes the compositional data analysis including the ternary diagrams for mixes and aggregates, a multivariate beta model, variability analysis using a multivariate variance assumption, and ternary quality control charts. It also describes the hypothesis tests using a k-variate student t distribution and an F-test. The following conclusions are obtained: i) RAP is better used in producing a mix with a relatively small top size 3/8" (Type S-III mix) or 1/2'' (Type S-I mix), and not with a sieve size larger than 1" (Type S-II mix); ii) RAP percent does not significantly influence the mix variability; it is desirable to use a large percentage of RAP in the Type S-I mix, i.e., more than 40% RAP; iii) On the average, the aggregate variability increases after transportation and stockpiling process; iv) On the average, the aggregate in the quarry has less variability than the same aggregate in the asphalt plant stockpile, therefore, aggregates in quarries are better controlled than aggregates in stockpiles; v) Variabilities of aggregates in quarries are slightly larger than the variability of RAP which is slightly larger than variabilities of Type S-II and S-III mixes. Type S-II and S-III mixes have the same variability and are slightly larger than the variability of the Type S-I mix; vi) The ternary diagram and the hypothesis test can be used to judge whether a contractor did a good job; vii) Ternary quality control chart shows the dynamic trend of a mix or an aggregate during the production period.