PhD Final Defense – Renan Santos Maia

Fatigue Cracking Mechanisms in Extended-Life Pavements: Coupled Effects of Modern Materials, Loading Configurations, and Climate Stressors

Advisor: Professor Ramez Hajj

ABSTRACT

Illinois has constructed extended-life (“perpetual”) flexible pavements for more than two decades,

with design based on limiting the tensile strain at the bottom of the asphalt concrete (AC)

layer to a fatigue endurance limit (FEL) of 70 με. Combined with conservative material and

structural inputs, this criterion can produce AC thicknesses approaching 46 cm. This dissertation

investigates the assumptions supporting this design approach and the mechanisms that control the

long-term integrity of thick AC structures through six complementary studies combining laboratory

characterization, field measurements, fatigue and fracture testing, recovery and healing analysis,

and moisture evaluation: (i) First, laboratory testing showed that modern binder-course mixtures

have higher dynamic moduli than assumed in the current design framework, with the virgin

binder low-temperature grade explaining much of the variation at design temperatures. Laboratory

measurements, structural-response calculations, and network-level Traffic Speed Deflectometer data

indicated that existing pavement sections operate under conservative tensile strain conditions. (ii)

Second, the structural integrity of nine in-service extended-life pavement sections was evaluated

using field deflections, recovered-core properties, uniaxial fatigue testing, and traffic and construction

records. The results showed no systematic evidence of traffic-induced bottom-up fatigue damage.

However, global stiffness alone was insufficient to assess damage because it does not capture

construction variability or changes in strain distribution associated with layer geometry and internal

interfaces. (iii) Third, flexural fatigue experiments showed that short-duration tests (thousands of

cycles) below approximately 150–200 με have limited sensitivity to differences in mixture fatigue

performance, whereas testing above 250 με provides a practical accelerated method for identifying

mixtures susceptible to long-term fatigue damage at low strains (millions of cycles). Base binder

properties and recycled binder content primarily influenced stiffness, while polymer modification had

a greater influence on progressive damage. (iv) Fourth, flexural and uniaxial fatigue experiments

showed that stiffness recovery after rest or thermal treatment does not necessarily indicate restored

fatigue resistance. Although stiffness could return to or exceed its initial value, subsequent fatigue

performance remained dependent on accumulated damage. This distinction reflects the combined

effects of viscoelastic recovery, healing, thermal aging, and polymer network reorganization, which

cannot be separated using stiffness measurements alone. (v) Fifth, multi-lift specimens showed that

global stiffness alone cannot characterize lift-interface integrity. Uniaxial fatigue testing identified

differences in damage evolution, including premature interface separation and accelerated integrity

loss. In contrast, single-edge notched bend (SENB) tests produced cracks that generally propagated

across the interface rather than along it, limiting their ability to characterize bond-plane fracture

resistance. (vi) Sixth, moisture exposure progressively increased the degree of saturation of densegraded

binder course AC, while severe fatigue damage increased subsequent water uptake, indicating

greater hydraulic accessibility. Moisture also influenced mechanical response: partial saturation

could increase measured modulus due to water occupying the pore structure, whereas severe freeze–

thaw conditioning reduced stiffness through moisture-induced damage. This is, therefore, a feedback

mechanism in which fatigue damage facilitates moisture ingress, and subsequent moisture damage

further reduces material integrity. Collectively, the findings indicate that the successful performance

of Illinois extended-life pavements is partly attributable to the conservatism of the current design

framework. However, structural integrity cannot be adequately characterized by a single modulus

or strain threshold. Tensile strain remains an appropriate primary design criterion, but its limiting

value should reflect mixture-specific fatigue resistance. Its interpretation should also consider the

mechanisms governing damage evolution, viscoelastic recovery and healing, lift-interface integrity,

construction variability, and moisture-induced damage.