A comparative analysis of the thermal conductivity measurement of asphalt concrete under steady-state and transient heat transfer conditions


Creative Commons License

DEMİRTÜRK D., ÖZTÜRK H. I., GÜLER A. M.

Materials and Structures/Materiaux et Constructions, cilt.59, sa.4, 2026 (SCI-Expanded, Scopus) identifier identifier

Özet

The long-term performance of pavement structures is strongly governed by temperature-dependent distress mechanisms, including rutting and thermally induced stresses. As heat transfer properties regulate internal temperature gradients within asphalt layers, the accurate determination of thermal conductivity is essential for reliable service-life prediction. Yet, accurately assessing the thermal properties of such a complex and heterogeneous material remains challenging due to variations in mixture composition, inconsistent heat transfer mechanisms, and experimental limitations. To overcome these challenges, this work employs both steady-state and transient-state testing methodologies to determine the thermal conductivity of asphalt concrete samples. Following this, the study conducts a detailed comparative evaluation of the two approaches to better understand their strengths and limitations. Accordingly, a new, well-insulated apparatus equipped with a heat flux sensor was designed for the steady-state testing approach, significantly enhancing the thermal insulation process necessary for achieving one-dimensional heat transfer, a capability lacking in current studies. In addition, transient heat transfer experiments were performed using the Hot Disk TPS 2500S instrument. In this study, twelve different mixtures were prepared with two aggregate sources, three distinct air void levels, and two different maximum aggregate sizes, and their thermal conductivities were compared under two heat transfer conditions. The TPS method, which is quick and easy, consistently showed higher values due to aggregate particles in the sensor’s contact zone. Conversely, the improved steady-state approach yielded relatively lower but more reliable values, effectively reflecting the sample's thermal conductivity by considering the distribution of mixture components.