The Influence of Aggregate Gradation Curve on Predicting the Indirect Tensile Strength of HMA Mixtures Using Soft Computing Approach

Indirect Tensile Strength (ITS) of HMA mixtures is a fundamental parameter in pavement design and evaluating the pavement distresses. Researchers have proposed some predictive models to estimate the ITS of asphalt mixtures. However, the impact of aggregate size distribution has not been directly considered in those models. In this research, attempts were undertaken to develop a new predictive model for ITS of HMA mixtures considering the effect of aggregate gradation curve. Artificial Neural Networks (ANN) was used as the computational tool using 259 ITS test results from LTPP database in which Indirect Tensile Tests (IDT) have been conducted according to AASHTO T-322 standard. The new ANN model could successfully predict the ITS of HMA with the high accuracy of R² =0.97. Moreover, Principal Component Analysis (PCA) was performed that transfer the initial variables to orthogonal groups to divide the similar function inputs into the separated groups. Finally, sensitivity analysis was carried out to examine the level of influence of input parameters on the model output. The outcomes of this study indicate that the aggregate gradation has a significant effect on ITS of asphalt mixtures and needs to be directly considered in predictive models.

• Record URL:
  https://annualmeeting.mytrb.org/OnlineProgram/Details/15748
• Supplemental Notes:
  • This paper was sponsored by TRB committee AKM30 Standing Committee on Asphalt Materials Selection and Mix Design.
• Corporate Authors:

  Transportation Research Board

  ,    
• Authors:
  • Heidaripanah, Ali
  • Hassani, Abolfazl
  • Moaveni, Maziar
• Conference:
  • Transportation Research Board 100th Annual Meeting
  • Location: Washington DC, United States
  • Date: 2021-1-5 to 2021-1-29
• Date: 2021

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 16p

Subject/Index Terms

• TRT Terms: Aggregate gradation; Asphalt mixtures; Hot mix asphalt; Neural networks; Predictive models; Sensitivity analysis; Tensile strength; Tension tests
• Identifier Terms: Long Term Pavement Performance (LTPP) Program; Principal Component Analysis
• Subject Areas: Highways; Materials; Pavements;

Filing Info

• Accession Number: 01763612
• Record Type: Publication
• Report/Paper Numbers: TRBAM-21-04158
• Files: TRIS, TRB, ATRI
• Created Date: Feb 4 2021 10:57AM