Abstract
Fatigue and rutting equations are essentially used in design of asphalt pavements by mechanistic-empirical method and they are developed from the laboratory and/or the field performance data. Conventionally, ordinary least square regression analysis is employed for developing these equations. The present paper discusses the underlying assumptions of ordinary least square regression analysis and shows that these assumptions are violated while developing fatigue/rutting equations from the data obtained from field or laboratory measurements. This paper proposes the use of a measurement error based estimator in order to alleviate the inaccuracy associated with use of least square regression analysis while developing fatigue and rutting equations. A step-by-step methodology is presented on how to develop the fatigue or rutting equation from a given data set using the proposed method. The confidence intervals of the estimators are obtained using bootstraping technique. Finally, a design example is presented to show the differences in design thickness values obtained by the conventional and the proposed method.
Acknowledgements
The authors wish to acknowledge useful discussions with Dr Shalabh, Department of Mathematics and Statistics, Indian Institute of Technology Kanpur.
Notes
1. Current affiliation: Institute of Numerical Mathematics, Dresden University of Technology, Dresden, Germany.