Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4913017 | Construction and Building Materials | 2017 | 12 Pages |
Abstract
The performance of 18 analytical models was investigated based on statistical analysis of R2, limiting deviation error, as well as their ability to predict long term rutting of AC-13 and AC-20 asphalt mixtures. An R2Â >Â 95% on fitting could not guarantee well simulated initial rutting values or accurate long term rutting predictions. However, a model showing a maximum of 200 initial cycles exceeding a deviation error of 20%, could guarantee rutting curve fittings with R2Â >Â 95%. The ability to accurately predict long term rutting was considered the best indicator of good models. Among the 18 models, only Theng - Lytton, Paute - 1, Paute - 2 and Monismith - 1 models satisfied the limiting values of the proposed performance indicators. Nonetheless, these same models required correction factors to accurately predict long term rutting when fitted to early age rutting results. Theng-Lytton model was considered the most reliable model, and a related rutting index was developed which could well be used for mixture design and optimisation.
Keywords
Related Topics
Physical Sciences and Engineering
Engineering
Civil and Structural Engineering
Authors
Barugahare Javilla, Liantong Mo, Fang Hao, Shuben An, Shaopeng Wu,