Atterberg limits and their relationship to longitudinal cracking in granular pavements

Abstract

[Abstract]: Australia has approximately 300,000km of sealed roads and maintenance of these roads imposes a significant financial burden on road agencies. Premature pavement failure exacerbates this burden. Longitudinal pavement cracking often occurs independent of traffic loading and may be attributed in many instances to moisture changes in expansive subgrade soils. This project investigates a possible relationship between the Atterberg limits of low strength subgrade materials (CBR < 3) and the incidence of longitudinal cracking in unbound granular (flexible) pavements supported by them. Unbound granular pavements are the most common form of pavement construction in Australia. Design of these pavements is undertaken in accordance with individual authorities’ empirical design charts. These design charts are usually presented as a series of curves whereby the depth of pavement is set by the relationship between subgrade strength, expressed in terms of a four day soaked California Bearing Ratio (CBR) and traffic loading over the pavement’s design life, expressed in Equivalent Standard Axles (ESAs). These charts typically provide pavement depths for subgrade CBR values of 3 and above. Where the subgrade CBR is less than 3, the guidelines recommend an additional depth of pavement gravel depending on the CBR. The primary objective of this research is to determine if any of the Atterberg limits can be used as a predictor of longitudinal cracking in unbound granular pavements designed in accordance with authority guidelines. This would enable consultants and authorities to determine if alternate methods of pavement construction should be considered (using tensile reinforcement for example) in lieu of unbound granular construction. Analysis of results indicated that a relationship exists between two of the Atterberg limits of a subgrade material and longitudinal cracking in unbound granular pavements designed in accordance with existing authority empirical design charts. Due to the small sample size, recommendation on specific values of these limits to determine when alternate pavement designs should be considered would be premature.