Gyratory compactor and mixture volumetric evaluation of roller-compacted concrete for pavement constructability

Standard roller-compacted concrete (RCC) test methods such as the modified proctor, Vebe, and vibratory hammer are used to determine final mixture proportions, evaluate consistency, and prepare cylindrical specimens for strength. However, these methods do not replicate the method and levels of compaction observed in the field and require multiple iterations to find a constructable and sustainable RCC mixture for pavements. Better linking RCC compaction properties with mixture volumetrics can enable more efficient RCC constituent selection and proportions. For this research, a typical RCC pavement mixture was selected from which the aggregate voids filled by paste (VFP) varied six levels to achieve mixtures ranging from underfilled (VFP=63.4 %) to overfilled (VFP=126.9 %). A gyratory compactor with a vertical and torque load cell was employed to continuously measure the compaction energy evolution in each specimen and for all mixtures. The volumetric compaction energy required to reach 90 % of the absolute density (E0-90 initial) and incremental energy to compact 90-95 % absolute density (E90-95 final ) was determined and used to evaluate RCC paveability and compactability, respectively, and relate it to mixture volumetrics. All RCC mixtures were able to achieve the initial density but at least equifilling was required to achieve final density under reasonable energy levels (<= 2.70 MJ/m3), with compaction energy decreasing exponentially as the VFP increased. The gyratory compactor more efficiently compacted the reference mixture (VFP=117.9 %) to the same density as the modified proctor test (1.88 vs 2.70 MJ/m3) or cylinders with the vibratory hammer (3.10 vs 10.0 MJ/m3). The gyratory compactor with a perforated mold also enabled quantification of paste mobility within the aggregate skeleton. The time needed for initial and final seepage through the mold perforations was better and more sensitive than the Vebe and vibratory hammer compaction times, with seepage times ranging from 5 to 166 seconds for RCC mixtures from overfilled to underfilled. The perforated mold and a wet sieve analysis also allowed for experimental determination of intergranular voids (IGV) and total paste volume (TPV) of the RCC mixtures, respectively, which had an average relative error of 3.4 % and 4.0 %, respectively.