Estimation of Rideability by Analyzing Longitudinal Road Profile

A method for estimating pavement rideability by computing a statistic called ride number (RN) from measured longitudinal profiles of the pavement is presented. The computational algorithm was developed with two objectives in mind: (a) relevance, as established by correlation between RN and mean panel ratings obtained for profiled pavements, and (b) portability, as determined by the ability of different profiling systems to obtain comparable RN values for profiles taken on the same pavement. Experimental data show that the new algorithm is comparable to previously developed RN algorithms in terms of its correlation with panel ratings. The new algorithm is much more portable; comparable RN values can be obtained from different profilers and the quasi-static Dipstick device. However, measures from ultrasonic profilers are not as accurate, and in some cases these devices are not acceptable for measuring RN. Although the fundamental relationships between profile characteristics and the human perception of rideability are still not understood, the algorithm is recommended as the best way to apply the results of past research investigating rideability.