Development of Performance-Related Specifications for Asphalt Mixtures in Ontario
MetadataShow full item record
Superpave mix design relying on volumetric specifications is not completely able to predict the long-term performance of asphalt pavements. Therefore, implementation of suitable performance specifications is crucial to maintain sustainability in highway infrastructure. Overall, performance specifications can be divided into two categories: performance-based specifications (PBS) and performance-related specifications (PRS). PBS are Quality Assurance (QA) specifications that describe the desired level of fundamental engineering properties and can be used in performance prediction, while PRS are QA specifications that describe the desired levels of key materials and construction quality characteristics that have been found to correlate with fundamental engineering properties. The implementation of efficient and practical performance-related specifications can be used in asphalt mix design process and Quality Control/Quality Assurance (QC/QA) specifications. In this research, tests such as Disc-Shaped Compact Tension (DC(T)) test, Semi-Circular Bend (SCB) test, Illinois Flexibility Index Test (I-FIT), IDEAL-CT test, complex modulus and cyclic tests, and Hamburg Wheel Tracking (HWT) test, which have shown promise to be considered as performance-related tests, were investigated through laboratory research on five plant-produced asphalt mixtures. The characterization of these asphalt mixtures by DC(T) and SCB tests revealed that no statistically significant difference could be found between fracture energies of DC(T) and SCB tests at three testing temperatures, namely -18℃, -24℃ and -30℃ for asphalt mixtures investigated. However, SCB test was not able to distinguish and rank low temperature cracking resistance of asphalt mixtures at -18ºC and -30ºC. This can be attributed to the behavior of asphalt mixtures when they are too ductile and too brittle, respectively, and the discrepancies existing with the geometry of DC(T) and SCB specimens, and CMOD rate. Furthermore, DC(T) test was not sensitive to the two methods of long-term aging (forced-draft oven aging at 85°C for 120 hours and forced-draft oven aging at 95°C for 72 hours) employed in this research. Moreover, the characterization of these asphalt mixtures by the I-FIT test revealed that forced-draft oven aging of I-FIT specimens at 95°C for 72 hours produced statistically similar Flexibility Index (FI) values as the specimens aged at 85°C for 120 hours. The results of FI values showed that asphalt mixtures containing hard PG asphalt binder (PG64-28 and PG70-28) were more sensitive to testing temperature variability compared to asphalt mixtures having softer PG asphalt binder. Additionally, a survey was distributed to asphalt mixture laboratories in Ontario to investigate the capability of laboratories for carrying out the aforementioned performance tests. Overall, responses from forty-six laboratories revealed that several of them are capable of conducting I-FIT, IDEAL-CT, DC(T) and HWT tests. Based on the analysis of the tests results from five plant-produced asphalt mixtures and the survey data, three performance tests, namely I-FIT, DC(T) and HWT tests were selected for further research. For this purpose, sixteen plant-produced asphalt mixtures with their corresponding field cores were investigated by conducting the tests selected. The statistical analysis conducted by t-test on sixteen mixtures determined that there was a significant difference between the average FI of PPLC (Plant-Produced Laboratory Compacted specimens) and their corresponding field cores with some exception. In addition, the CV of FI results for most of PPLC specimens and pavement field cores was less than 20% showing the low variability for I-FIT test for post-production mixtures and field cores. The statistical analysis conducted by t-test on sixteen mixes determined that there is not a significant difference between the average fracture energy of PPLC specimens and their corresponding field cores except for one mix in DC(T) test. Furthermore, the CV of fracture energy results for most of PPLC specimens and pavement field cores is less than 20% showing the low variability of DC(T) test. Moreover, the analysis of the tests results provided preliminary specifications for I-FIT, DC(T) and HWT tests. According to the data analysis, a preliminary threshold FI value of 10 can be used for all mixtures except for SMA mixtures. A preliminary threshold FI value of 15 is suggested for SMA mixtures. Moreover, based on the data analysis in this study, a premilitary threshold DC(T) fracture energy value of 700 J/m2 can be considered for traffic category E mixtures, and a preliminary threshold DC(T) fracture energy value of 600 J/m2 can be considered for all other traffic categories. With regard to HWT test results, as a preliminary threshold criterion, mixes containing PG70-XX, must not reach a rut depth of more than 6.0 mm at 50°C after 20000 passes. In addition, mixes containing PG64-XX must not exhibit a rut depth of more than 12.5 mm at 50°C after 20000 passes. For mixes containing PG58-XX and PG52-XX, the rut depth should not exceed 12.5 mm at 44°C after 20000 passes. These results provide, besides of a comprehensive evaluation of performance-related tests in Canadian asphalt mixtures, a basis for implementation of performance specifications for I-FIT, DC(T) and HWT tests in Ontario.
Cite this version of the work
Saeid Salehiashani (2021). Development of Performance-Related Specifications for Asphalt Mixtures in Ontario. UWSpace. http://hdl.handle.net/10012/17258