New Approach to Roadway Performance Indices

Abstract

In 1964, Carey and Irick defined the present serviceability index (PSI) by assuming that the PSI can be expressed on a scale of 0 to 5 as the sum of a logarithmic function of slope variance, a quadratic function of mean rut depth, a square root function of cracks and patches, and a random error term using the American Association of State Highway Officials (AASHO) road test data. These functional forms have been used widely although they yield intercept values for PSI over 5 and some other nonphysical results for both rigid and flexible pavements; similar difficulties appear in the work of others performed over the years since then. To resolve these difficulties, we propose a new methodology dealing with the test data by applying Fechner's psychophysical law. It is found that the AASHO road test data, road test data obtained in Texas, road test data from Canada, and data from the international road roughness experiment are organized well by the simple summation of logarithmic terms of roadway characteristic variables. The methodology can be further applied to handle similar types of road test data.