Developing Spectrum-Based Models for International Roughness Index and Present Serviceability Index

Abstract

This paper presents a theoretical model for determining international roughness index (IRI) and present serviceability index (PSI) in term of power spectral density (PSD) roughness. Frequency response functions of quarter-car model and CHOLE-type profilometer are obtained by means of stochastic process theory. Correlation of IRI and PSD becomes possible based on their PSD-based expression. It is also shown that statistical regression may not be effective when attempting to correlate IRI with PSI due to their inherent multivalued corresponding property. Transportation agencies can use PSD-based models to precisely convert IRI to PSI given that PSD roughness of a pavement is known. Furthermore, this paper provides a correlation between typical PSI and IRI of flexible and rigid pavements whose PSD roughness is given in the form of power law at specified levels.