Another look at delineation of uniform pavement sections based on falling weight deflectometer deflections data

Abstract

The large amount of data commonly used to characterize the pavement surface and structural conditions offer a challenge to practitioners making decisions about the representative value of a particular parameter for design. While a large number of observations along the length of a road allow a better quantification of the expected value and variance of a parameter, basing a design on an average parameter along the project length will typically be uneconomical and less reliable. Therefore, pavement surface and structural condition data along a project length needs to be delineated into uniform sections. The design can be performed individually for each of these uniform sections to achieve economy without compromising reliability level. This paper documents delineation methods that explicitly address the problem of segmentation of measurement series obtained from Falling weight deflectometer deflections. Modifications in the existing American Association of State Highway and Transportation Officials (AASHTO) delineation procedure were incorporated to address the mean differences and the local variability. The results of delineation show that the AASHTO methodology ignores the local variations along the project length which may not be valid from a practical standpoint while designing rehabilitation or preservation strategies. The inclusion of restrictions on mean difference and section length resulted in better delineation than the AASHTO method but it could be sensitive to local variations of the deflections within a section. The delineation approach can handle the local deflection variations within a section if appropriate constraints on the local variations are imposed. The results from the delineation of field deflections showed that the restrictions on mean difference, minimum section length, and location variability are vital to delineate the project length into appropriate homogenous sections which can be different from each other from both statistical and practical viewpoints.