Abstract

This study presents the development of an improved temperature correction methodology for falling weight deflectometer (FWD) deflections in full-depth asphalt pavements. The objective was to address the limitations of existing correction methods by integrating viscoelastic pavement behavior, structural properties, and realistic thermal gradients. Field temperature data from instrumented sites in Indiana were used to train a machine learning model capable of predicting pavement temperature profiles in flexible pavements. These predictions informed a series of finite element simulations that captured pavement responses to FWD loading under a range of thermal and structural conditions. Temperature correction factors were then derived and modeled using a unified sigmoidal function, whose parameters depend solely on asphalt thickness, subgrade stiffness, and effective pavement temperature. For full-depth asphalt pavements, the temperatures at the surface–base interface and quarter-depth were found to be the most accurate representations of the effective pavement temperature for use in the proposed deflection correction methodology. Field validation demonstrated that the proposed methodology consistently outperformed the traditional AASHTO 1993 correction approach, offering a more accurate and practical solution for temperature correction in pavement structural evaluation.

Keywords

pavement evaluation, falling weight deflectometer, temperature correction, full-depth asphalt pavements, viscoelastic analysis, finite element modeling, machine learning, deflection basin

Report Number

FHWA/IN/JTRP-2026/01

SPR Number

4717

Performing Organization

Joint Transportation Research Program

Publisher Place

West Lafayette, Indiana

Date of Version

2026

DOI

10.5703/1288284318609

SPR-4717 Technical Summary.pdf (5031 kB)
SPR-4717 Technical Summary

Share

COinS