LOW TEMPERATURE PAVEMENT PERFORMANCE : AN EVALUATION USING C-SHRP TEST ROAD DATA

This research report provides a detailed compilation and comparison of field performance results on the three C-SHRP test roads constructed in 1991 and 1992 at Lamont (Alberta), Hearst (Ontario) and Sherbrooke (Quebec). The two main objectives were: 1) to assess, validate or suggest possible modifications to the Superpave system with a view to incorporating acceptable international binder and mixture specifications appropriate to Canadian climatic conditions; and 2) to validate existing fracture temperature prediction models, or to produce improved models, in order to be more useful for Canadian agencies in assessing the applicability of current and future binder and mixture specifications and pavement design practices for these environments. The performance prediction model based on fundamental mixture properties used in the Superpave system appears to yield the most favourable results, with some exceptions. The reported method of assessing the 14 binders recovered from the test roads showed that all failed the SHRP performance graded specification. Further examination of the Bending Beam Rheometer test results and conventional test methods was done using nomograph methods to estimate stiffness of the asphalt cement binder and asphalt concrete mixtures. Other fracture temperature prediction models, such as ASPELT and CAMA, were used to obtain fracture temperatures and compared with observed initial cracking temperatures. They both tended to predict cracking temperatures at close to or up to 10 degrees Celsius warmer than observed. An expanded fundamentally-based method of prediction of thermally induced cracking based on the test properties of the asphalt concrete and specific temperature regimes obtained in this project has been developed. With some exceptions, this method of prediction explains the observed field performance better than the existing prediction model.