Date of Award

Spring 2022

Document Type

Open Access Thesis

Department

Civil and Environmental Engineering

First Advisor

Sarah Gassman

Abstract

This thesis presents the effect of concrete material inputs on rigid pavement design in the state of South Carolina. The impact of concrete strength input parameters in the historical AASHTO pavement design method and MEPDG method is investigated using three newly constructed concrete sections, I-85 in Cherokee County, I-20 in Lexington County, and I-85 in Spartanburg County. Strength data obtained from the sections were the result of QA/QC testing provided by the SCDOT during construction, as well as from UofSC laboratory testing. The compressive strength data from these tests was correlated to the elastic modulus and modulus of rupture (which were also directly obtained from the UofSC testing). Different types of strength data were used as input to understand the effects of the input strength parameters on the resulting optimized pavement thickness.

The optimized pavement thickness obtained from the AASHTO method was 1.5 to 3 in. larger than the optimized pavement thickness found from the MEPDG method. It was also found that the use of MEPDG Level 1 inputs resulted in a design pavement thickness that was 1 in. larger than when using the MEPDG Level 2 inputs. The effect of using local calibration coefficients instead of the default global calibration coefficients was also studied. The local calibration coefficients predicted an IRI at 20 years of 158 in./mile, while the global calibration coefficients predicted an IRI at 20 years of 72.7 in./mile.

A sensitivity analysis was performed using the Hall and Beam (2005) method and the Schwartz et al. (2011) method. From the Hall and Beam (2005) method, the smoothness and faulting distress models were found to be sensitive to f’c for all segments, and the cracking model was insensitive to f’c for segments of Cherokee/I-85 and Lexington/I-20 but sensitive for segments of Spartanburg/I-85. For the segments of Spartanburg/I-85, based on the Schwartz et al. (2011) method, the smoothness and faulting models were either sensitive or very sensitive to f’c (depending on the individual segment), while the cracking model was very sensitive to f’c for all segments.

Lastly, the study implemented a new method of segmenting pavement sections based on concrete compressive strength and the cumulative difference method. For this study, the number of f’c measurements in a final pavement group prior to segmentation ranged from 17 to 246; and the typical spacing between the measurements was 500 ft. The segmenting procedure resulted in 16 unique pavement segments being identified.

Share

COinS