Date of Award

Fall 2026

Abstract

Balanced Mix Design (BMD) has emerged as a promising approach for improving asphalt mixture performance by balancing rutting and cracking resistance through performance tests rather than relying solely on volumetric criteria. Therefore, its application is increasingly being explored by transportation agencies for both highway and airfield domains. Despite its advantages, several challenges still limit the widespread implementation of BMD. These challenges include uncertainties associated with the compatibility of asphalt mixture components, the limited practicality of existing cracking performance tests, and the lack of performance prediction tools suitable for process control during mixture production. This dissertation addresses these critical challenges by developing practical methodologies for material selection, simplifying an existing test protocol, and developing a machine learning (ML)-based predictive tool for asphalt mixtures used in transportation infrastructure.This doctoral research first proposes a performance-oriented framework for identifying compatible asphalt mixture components, such as virgin binders, reclaimed asphalt pavement (RAP), and recycling agents, using binder rheological and thermal properties. These binder properties are also validated through mixture-level performance tests commonly employed in BMD practice. The dissertation also investigates the characterization of intermediate-temperature cracking and proposes a methodology to incorporate the climatic conditions at the project location while reducing test duration. By applying the time-temperature superposition principle, the study introduces a procedure to adjust loading rates so that the test can be performed at room temperature while still representing field-related intermediate temperatures. This approach enables faster testing while maintaining representative cracking characterization for BMD applications. To further improve cracking characterization within the BMD framework, a simplified low-temperature semi-circular bend (SLT-SCB) test is developed as a practical, fundamental method for evaluating low-temperature cracking resistance of asphalt mixtures. A novel low-temperature cracking index was introduced, while the proposed method reduces testing complexity and test time. Finite element simulations using cohesive zone modeling are conducted to validate the proposed test approaches and to examine their sensitivity to loading rates and temperature variations, supporting the experimental findings. Finally, an ML-based model is developed to predict rutting performance using mixture design parameters, binder characteristics, and aggregate gradation properties. The resulting model demonstrates strong predictive capability for asphalt pavement analyzer rut depth and is implemented in a Microsoft Excel-based tool that allows practitioners to rapidly estimate mixture rutting performance for asphalt mixtures used in airfield applications. Importance and interaction analyses further show that binder properties govern rutting resistance, with volumetric and gradation parameters modulating their influence, thereby guiding mixture adjustments for process control. As a result, the findings of this dissertation provide transportation agencies and practitioners with methodologies for developing practical tools to improve the reliability, efficiency, and broader implementation of BMD.

Document Type

Dissertation

First Advisor

Eshan V Dave

Second Advisor

Navneet Garg

Third Advisor

Hassan Tabatabaee

Department or Program

Civil and Environmental Engineering

Degree Name

Doctor of Philosophy

Download

Share

COinS