A Laboratory and Statistical Evaluation of Factors Affecting Rutting

Abstract

This research identifies the most significant factors from those factors evaluated which affect rut potential of Hot-Mix Asphalt (HMA). Mixture rut potential is determined using an Asphalt Pavement Analyzer (APA) in the laboratory. The experimental program employed in this study consists of three sets of tests and each set represents a matrix whose elements are rut factors. In Set A, seven factors, each at two levels, are examined using a mixture of limestone aggregates designed in accordance with Superpave method. The test results are analyzed statistically. The analysis results presented using Set A show that binder's performance grade (PG), specimen type, test temperature and moisture in test specimen affect a mixture's rutting performance significantly. Wheel load, asphalt content and hose pressure at the selected levels in Set A are shown to be less significant. In Set B six factors, aggregate gradation, temperature, moisture, asphalt content, load and hose pressure, are investigated using one Hveem designed mixture with gravel aggregates. Gradation, temperature, asphalt content and moisture are found to be significant. The levels of asphalt content selected for Set B to be one at optimum and the other at one percent more than the optimum asphalt content. Similar results are found for the test Set C with five factors, temperature, gradation, moisture, load and hose pressure. A predicted rut depth and its range under the influence of the significant factors studied in Set A are also determined and verified by additional experiments. The paper developed and described a detail statistical procedure to analyze a designed experimental program to interpret test results without the need for a full factorial approach.

Keywords:

• Rutting
• Hot mix asphalt
• Rut factor
• Asphalt pavement analyzer
• Performance grade
• Statistical analysis

Acknowledgements

The authors express their deep appreciation to Oklahoma Department of Transportation for funding the project. We acknowledge T.J. Campbell Construction Company for providing the aggregates and the Oklahoma binder suppliers whose products were used in this study. The authors wish to thank undergraduate students, Brent Caswell and Matt Gearhart for their assistance in testing. The authors also appreciate the assistance of Reynolds H. Toney, P.E., Materials Engineer for ODOT, for his support during this work.

Notes

^†[email protected]

^‡[email protected]