Date of Award

2017

Document Type

Open Access Dissertation

Department

Civil and Environmental Engineering

Sub-Department

College of Engineering and Computing

First Advisor

Sarah Gassman

Abstract

In 2008, the American Association of State Highway and Transportation Officials (AASHTO) released a modified pavement design method (i.e., the Mechanistic Empirical Pavement Design Guide (MEPDG)) based on Long Term Pavement Performance (LTPP) data from all over the United States. The MEPDG default design parameters developed from the LTPP database are expected to be significantly different than those for South Carolina material, traffic and weather conditions, thus the default design parameters may not be accurate for South Carolina. Therefore, the new pavement design method should be calibrated for South Carolina conditions by performing MEPDG local calibration.

Different input variables should be studied to run the pavement design program to minimize the difference between the measured and predicted distresses of pavements. Rutting is one of the most important asphalt pavement distresses because it is responsible for both the functional and structural condition degradation of the flexible pavement. There are limited studies on the effect of resilient modulus (𝑀𝑅) of subgrade on pavement rutting in the MEPDG. Therefore, the purpose of this study was to characterize the subgrade 𝑀𝑅 and study the effects of subgrade 𝑀𝑅 on pavement rutting.

Firstly, pavement performance evaluation models were developed in this study using data from primary and interstate highway systems in the state of South Carolina, USA. Twenty pavement sections were selected from across the state and historical pavement performance data for those sections was collected. A total of 9 models were developed based on regression techniques, which include 5 for Asphalt Concrete (AC) pavements and 4 for Jointed Plain Concrete Pavements (JPCP). Five different performance indicators were considered as response variables in the statistical analysis: Present Serviceability Index (PSI), Pavement Distress Index (PDI), Pavement Quality Index (PQI), International Roughness Index (IRI), and AC pavement rutting. Annual Average Daily Traffic (AADT), Free Flow Speed (FFS), precipitation, temperature, and soil type (soil Type A from Blue Ridge and Piedmont Region, and soil Type B from Coastal Plain and Sediment Region) were considered as predictor variables. Results showed that Type A soil produced statistically higher PDI, PQI (p < 0.01), and rutting (p < 0.001) compared to Type B soil on AC pavements; whereas, Type A soil produced statistically higher IRI and lower PSI (p < 0.001) compared to Type B soil on JPCP pavements. Using the developed models, local transportation agencies could estimate future corrective actions, such as maintenance and rehabilitation, as well as future pavement performances.

Next, resilient modulus (𝑀𝑅) of subgrade soils for different geographic regions in South Carolina was characterized in this study. Shelby tube samples of subgrade soils were collected from existing pavements in different regions: SC-93 in Pickens county (Upstate Area), US-521 in Georgetown county (Coastal Plain), and US-321 in Orangeburg county (Coastal Plain, near the fall line). Statistical analysis was performed to develop 𝑀𝑅 estimation models for undisturbed soils using soil index properties. A correlation between laboratory measured 𝑀𝑅 with the modulus from Falling Weight Deflectometer tests was also developed. Finally, the effects of 𝑀𝑅 on subgrade rutting were studied using MEPDG. Results showed that the developed models offer higher reliability than the universal Long-Term Pavement Performance models in estimating the resilient modulus of undisturbed soils and predicting subgrade rutting for South Carolina.

Pavement rutting depends largely on subgrade soil stiffness, which is a function of the in-situ moisture content and soil index properties. The subgrade soil moisture content may vary from the specified condition due to variations in the compaction procedure during construction and fluctuations in the ground water table from seasonal changes. The resilient modulus (MR) is used to define the subgrade soil stiffness, and is one of the most important material inputs for the Mechanistic-Empirical (M-E) pavement design method. In this study, California Bearing Ratio (CBR) tests and laboratory MR tests were performed on remolded samples of soils collected from different regions in South Carolina. The samples were prepared at moisture contents above and below the optimum moisture content (wopt). Correlations between the results from the two tests were developed as a function of moisture content and statistical models were developed to correlate generalized constitutive MR model parameters with soil index properties. Furthermore, pavement rutting was studied using the resilient modulus determined for the subgrade soils compacted at wopt and Β±2%wopt. Statistical analysis showed that a slight change in moisture content during compaction has a significant effect on pavement rutting. The peak value of both CBR and MR was found on the dry side of optimum and at a dry density less than the maximum. It is also found that the subgrade soil moisture condition has a significant influence on subgrade rutting if graded aggregate base is used. However, if a higher strength base layer is used (i.e., cement stabilized base or asphalt treated aggregate base), the effect of moisture content is less significant.

Share

COinS