Multivariate statistical techniques were used to determine the feasibility of quantifying segregation in hot-mix asphalt pavements. Data obtained by using visual ratings of segregation, nuclear density gauge readings, and laboratory tests at nine field sites as reported by the National Center for Asphalt Technology were analyzed. Data were screened, before multi variate statistical analyses were performed, to develop relationships among segregation level, change in key mixture properties (air voids, asphalt content, and percent passing various sieves), and change in mixture density. The results revealed that segregation level could not be accurately identified by visual ratings and that multivariate statistics based on actual measured properties were appropriate for evaluating segregation level. Laboratory-prepared samples were made to assess the accuracy of the model developed in the study. The verification results indicated that the model could predict segregation level of these samples with accuracy. From the analysis of the models, it was concluded that a change in mixture density could predict segregation. This allows screening of segregated pavement sections on the basis of numerical field measurements instead of reliance on subjective visual observations. Given the promising results from this work, further studies on segregation models based on density measurements are suggested.
Analysis of Multivariate Models for Evaluating Segregation in Hot-Mix Asphalt Pavements
Transportation Research Record: Journal of the Transportation Research Board
Transportation Research Record: Journal of the Transportation Research Board ; 1900 , 1 ; 33-40
2004-01-01
Article (Journal)
Electronic Resource
English
Segregation in hot-mix asphalt pavements
TIBKAT | 2000
|Engineering Index Backfile | 1895
Methods for Evaluating Longitudinal Joint Quality in Asphalt Pavements
Online Contents | 2009
|Methods for Evaluating Longitudinal Joint Quality in Asphalt Pavements
Transportation Research Record | 2009
|