DEVELOPMENT OF A CONCRETE MECHANICAL AND THERMAL PROPERTIES DATABASE FOR PAVEMENT MECHANISTIC-EMPIRICAL INPUT

ABSTRACTState highway agencies (SHAs) in the United States have been moving towards using the Mechanistic-Empirical Pavement Design Guide (MEPDG), which deploys mechanistic and mathematical principles to analyze the material behaviors. For the smooth transition to this updated pavement design approach for rigid pavement design, SHAs have been developing a statewide database of concrete mixture properties to select appropriate input variables and levels for rigid pavement designs. The key mechanical and thermal inputs for rigid pavement design in MEPDG are compressive strength (f’c), modulus of elasticity (Ec), modulus of rupture (MOR), coefficient of thermal expansion (CTE), portland cement concrete (PCC) heat capacity, thermal conductivity, and ultimate shrinkage. This study investigates the effects of thermal properties of concrete mixtures on the performance of rigid pavement. Thermal properties were further investigated using machine learning (ML) algorithms to understand the concrete mixture inputs’ impacts.

INDEX WORDS: Pavement ME, concrete materials properties database, thermal properties, sensitivity analysis, machine learning