Validated Finite-Element Model for Predicting Asphalt Concrete Overlay Reflective Cracking Potential

Aravind Ramakrishnan, Imad L. Al-Qadi

Research output: Contribution to journalArticlepeer-review

Abstract

The reflective cracking potential of overlay systems is often studied through large-scale testing, which is time-consuming and expensive. Therefore, a generalized 3D finite-element model was developed to simulate a large-scale test setup to predict the reflective cracking potential. Crack propagation was modeled, and fracture parameters, such as stress intensity factor (SIF) and J-integral, were obtained. Mode I cracking dominated the model’s crack propagation, while in-plane shear cracking was insignificant. The model was validated with large-scale testing results; the average Mode I SIF was found to be a valid measure to assess the overlays’ reflective cracking potential. The model could be used qualitatively to compare different overlay configurations. A database of 128 cases was generated to compute fracture parameters for extreme scenarios or a combination of inputs. A data-driven surrogate model that predicts reflective cracking potential was developed to allow easy application by agencies. Shapley Additive Explanations (SHAP) analysis confirmed the model’s robustness. The surrogate model may be used by engineers to select asphalt concrete mixtures and design overlay thicknesses to control reflective cracking potential.

Original languageEnglish (US)
Pages (from-to)1409-1422
Number of pages14
JournalTransportation Research Record
Volume2678
Issue number11
Early online dateApr 29 2024
DOIs
StatePublished - Nov 2024

Keywords

  • asphalt overlay
  • finite element
  • large-scale testing
  • overlay modeling
  • pavement distress
  • reflective cracking

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Mechanical Engineering

Fingerprint

Dive into the research topics of 'Validated Finite-Element Model for Predicting Asphalt Concrete Overlay Reflective Cracking Potential'. Together they form a unique fingerprint.

Cite this