Abstract
The design of a Hot Mixed Asphalt (HMA) pavement mix requires information about the Bulk Specific Gravity (Gsb) and Absorption characteristics of the fine aggregates. This data is often determined using the standard AASHTO T-84 test for fine aggregates, which usually takes 2-3 days to complete. As the test is strongly dependent on the expertise of the operator, it has encountered ongoing criticism due to the subjective nature of the test. To overcome some of the operator-dependent errors associated with the AASHTO T-84 procedures, a new method, known as the CoreLok method was developed. This method is quick, reliable, portable, and provides consistent, repeatable results for fine aggregates. The Idaho Transportation Department (ITD) has its own standard procedure for the CoreLok test, IT-144 (2008), which is based on the original ASTM standard D7370. As the CoreLok test may be completed on an aggregate sample within 30 minutes, it has become a popular replacement for the AASHTO T-84 test.
This study was conducted to develop models which could correlate the IT-144 test results with AASHTO T-84 test results. For this purpose, 22 typical aggregate samples collected from the popular quarry sites in five ITD districts used by the ITD were tested using AASHTO T-84 and Idaho IT-144 test methods. A Round-Robin experiment was carried out involving ITD (Boise), ALLWEST and STRATA to confirm that the results were comparable between the participants. A total of 68 T-84 tests and 65 IT-144 tests were run at UI for the data analysis. Regression models were developed to predict the AASHTO T-84 values using the IT-144 values which were validated using the ALLWEST values.
Simple regression analysis and multiple regression analysis were performed to develop linear and non-linear prediction models. AASHTO T-84 results were used as the dependent variable and
IT-144 test results, and other variables like particle sizes, Specific Surface Area (SSA) and Fineness Modulus (FM) of the aggregates, Coefficient of Curvature (Cc), and Coefficient of Uniformity (Cu) were used as the predictor variables. A simple linear regression model with R2 = 90.53 percent and a multiple regression model with R2 = 85.77 percent was recommended for Gsb,Dry and Absorption prediction respectively. Data validation was better for simple linear regression for Gsb,Dry and multiple regression for Absorption. It is recommended that the Idaho IT-144 test method be adopted as it is faster, easier, repeatable, and produces results which are close to the AASHTO T-84 method.
Keywords: Correlation, AASHTO T-84, IT-144, Specific Gravity, Absorption, Fine Aggregates