This report is written to summarize the findings of a study conducted by George Mason University’s (GMU) Sustainable Geo Infrastructure (SGI) Research group to evaluate the expansion of using reclaimed asphalt pavement (RAP) and to optimize the RAP content in unbound base aggregate. The particular scope of the research presented in this report has been determined by the VTRC/VDOT team and included a laboratory and a field study.
Prior to the initiation of the research, based on the conversations with Virginia’s road building industry it was determined that the research will focus on evaluating what is referred as “fine processed” RAP (100% of the particles finer than 1-inch). Virgin aggregate (VA) used in this study complied with the VDOT’s 21A gradation and contained geologically similar aggregate pieces (diabase). During the initial phase, based on the availability, samples of RAP from 14 different asphalt plants in Virginia were collected and characterized. The goal was to assess the similarities and differences between the RAP produced throughout the Commonwealth. The results showed that all RAP samples had similar grain size distribution and primarily contained pieces of aggregate that were of diabase origin. The binder (asphalt) content of the samples ranged from 4.4 to 6.1%. Based on this characterization, samples from three different plants that represented a low, medium, and high binder content RAP were selected for detailed evaluation. The ages of the collected RAP samples were not known and most likely varied.
Laboratory evaluation focused on assessing the performance of the RAP-VA blends against the performance of the 100% VA alone. CBR, resilient modulus (Mr), and permanent deformation (PD) tests were used to evaluate performance. Up to 60% RAP (with three different binder contents) was blended with VA by weight. Results from PD tests showed a threshold where some addition of RAP into VA improved the performance but beyond a specific threshold, the overall performance started to decline. This threshold was determined to be a function of both the percentage of RAP added to a blend and the percentage of the binder content of the 100% RAP used. The optimized maximum RAP percentage in a given blend was determined in this study as up to 20% for RAP with low binder content (i.e., = 4.6%) and up 30% for RAP with high binder content(i.e., = 5.6 %). The reasons why RAP with different binder content resulted indifferent percent threshold requires further investigation as it could be due to the differences in the age of the RAP, which was not part of the scope of this study.
Field evaluation part of the study involved in constructing an actual roadway with four different base course layers consisting of 20 and 30% RAP-VA blends with low (4.5%) and high (5.7%) binder contents. Sections constructed with VA alone were used as a comparison (control sections). Field study demonstrated that Light Weight Deflectometer and modified speedy moisture content tests are suitable tools to be used for quality control of RAP-VA blends during construction. Performance evaluation in the field was monitored for a year with the embedded instruments and nondestructive tests conducted during and after the construction. Results obtained from the field were in agreement with the laboratory observations.
Based on the findings from laboratory and supported by the field observations, a relationship between the binder content of the 100% RAP and maximum allowed RAP percentage to create RAP-VA blends is created to provide guidelines for implementation.