|Abstract:||A field performance study of 19 bonded concrete overlays on asphalt pavements (BCOA) projects in 9 states were completed in this research study. BCOA pavements ranged in age from 7 to 26-years old and average thickness per project of 3.8 to 9.3 inches. The panel size varied from 4 ft x 4ft to 12 ft x 12 ft. The goal of the study was to determine the main factors influencing favorable and unfavorable field performance of BCOA pavements. For each project site, 8 cores, 4 on joints and 4 in center of slabs were extracted which included concrete, asphalt layer, aggregate subbase and subgrade samples from each project. FWD testing, distress surveys, IRI, and ultrasonic MIRA testing was completed at each project location.
First, Falling weight deflectometer (FWD) and MIRA data was collected to determine slab thickness, effective thickness of structural sections, and joint load transfer efficiency across the panels. The extracted field cores and material sample were tested in the CEE materials laboratory. Concrete samples were tested for compressive strength, split-tension strength, and coefficient of thermal expansion. Asphalt samples were characterized by their bulk specific gravity, Hamburg wheel testing device, and indirect tensile strength. The shear strength and bond failure mechanism were measured on composite cores (bonded concrete-asphalt cores). The fine material fraction of the aggregate subbase and subgrade were classified according to AASHTO soil classification. Additionally, gradation and Atterberg limit testing was performed on the unbound samples.
The field performance of the sections (longitudinal cracking) was compared to existing AASHTO Pavement ME (PaveME). PaveME runs predicted the longitudinal cracking of the pavement section given over 100 input variable including loading, material conditions, environmental conditions, geotechnical information, and geometry of sections. New BCOA pavement thickness was designed for each section also with the Pitt BCOA ME, Colorado DOT, and Illinois DOT methodologies and a comparison made with the as-built thickness. The different methods used traffic, climatic, existing asphalt properties, sublayer properties, concrete material properties, and slab geometry to design thickness sections.
From the non-destructive testing data, backcalculated effective thickness did not significantly vary within sections classified by visual distress as good or poor. Smaller slab sizes resulted in excessively large modulus of subgrade reactions, which may been from contraction joints not activating. MIRA correlated well to extracted cores. The average difference between cores and MIRA was 0.1-inches. Composite cores extracted from the majority of project locations were not bonded between concrete and asphalt layers. Bonded composite cores that were shear tested had a mixture of failure mechanisms of cohesive failures within the asphalt layer and adhesive failure at the asphalt-concrete interface. The interface shear strength ranged from 283 kPa to 1224 kPa but did not correlate to visual distress classification of good and poor sections.
Improvements to PaveME are needed to update failure mechanisms of software to include corner breaks, transverse cracking, and bond condition of overlay. Longitudinal cracking predictions from PaveME were reasonable for 82% of the locations. BCOA-ME designs compared to field constructed section resulted in 1-inch thicker for approximately 71% of the sections. For the Illinois DOT BCOA method, section designs were within 1-inch of constructed sections only 55% of the time. The Colorado DOT PCC overlay design resulted in 50% of the sections predicted to be under designed by more than 1-inch.