Two-Part Power Model to Evaluate the Rutting and Moisture Susceptibility of Asphalt Mixtures Using Hamburg Wheel Tracking Test

Authors:
Publisher website
Google Scholar
Abstract
Different methods were compared for analyzing Hamburg wheel tracking test (HWTT) rutting curves to evaluate the viscoplastic and stripping-induced rutting performance of asphalt mixtures. Two common methods—the sixth-degree polynomial and logarithmic-based model—were evaluated against a new two-part power (2PP) model. The 2PP formulation provides a unified representation of the rutting curve and supports adjusted stripping inflection point (SIP) calculation based on a 12.5 mm rut depth threshold. A total of 358 mixtures from the Federal Highway Administration (FHWA) Pavement Testing Facility (PTF) project and the Texas Department of Transportation database were analyzed. An open-source software tool, AutoHWTT, was developed to automate HWTT data analysis using all evaluated methods. The results show that corrected rut depth (CRD) values at 20,000 passes from the three models were highly correlated, while the polynomial model produced unreliable stripping number (SN) estimates for mixtures. The adjusted SIP showed slight improvement, in correlation with SN (74.2 versus 68.5 percent for conventional SIP) and a reduction in misclassification against the 15,000-pass threshold (26.7 versus 45.7 percent). Results from FHWA PTF mixtures show that stripping-induced rutting dominated total rut depth (TRD) variability; this rutting measurement variability was significantly reduced when using CRD as a rutting indicator. Additionally, ideal rutting test (IDEAL-RT) results followed CRD trends, but did not correlate with those of TRD. This observation demonstrates the detrimental influence of stripping on TRD when used as a rutting metric. The 2PP model offers an efficient and integrated framework for HWTT analysis, with adjusted SIP improving stripping assessment.
Funding Information
  • Federal Highway Administration (693JJ323F00109N)

This publication has 29 references indexed in Scilit: