Flow number predictive models from volumetric and binder properties

• Key factors affecting FN identified.
• Empirical FN model developed.
• FN model minimizes need for binder extraction/recovery.

This study was conducted to develop flow number (FN) predictive models for asphalt concrete mixtures (AC) using Superpave volumetric and binder properties. The FN test is the recommended procedure for evaluating rutting resistance of AC mixtures under the AASHTO mechanistic empirical pavement design guide (MEPDG). FN tests were conducted at 54 °C on 13 plant-produced laboratory-compacted AC mixtures using repeated creep loads of magnitude 206 kPa for 10,000 cycles or until an accumulated strain of five percent. The results were compared with key Superpave mixture volumetrics and binder specification parameters. First-order multiple regression models were developed to describe the relationship among FN, mixture volumetrics, and Superpave rutting parameter (G*/sin δ). The results showed that FN was strongly correlated (R2 = 0.94) to G*/sin δ, effective binder content, air voids, and dust to effective binder ratio. The results demonstrate a clear link between FN and mixture volumetric and binder properties and suggest FN could be considered as potential rutting specification parameter for QC/QA purposes in the field. The results may also be useful for optimizing the laboratory mix design process by directly relating certain Superpave specification parameters to mixture rutting resistance.