A Roadware ARAN inertial profiler was used to calculate the International Roughness Index (IRI) for the Auburn University National Center for Asphalt Technology (NCAT) test track tangent sections as well as three paving projects in Michigan. IRI values were calculated for short intervals [4.6 to 7.6 meters (15 to 25 feet)] rather than for the traditional 160 meters (0.1 mile). These short intervals were able to identify easily localized anomalies in the pavement smoothness due to changes in the construction activities. They were also useful in evaluating relationships between IRI and mix variables in the short NCAT test track test sections [60 meters (200 feet)].

IRI repeatability (standard deviation for three replicates) immediately after onstruction was found to be 0.063 m/km (3.99 inches/mile) in the right wheel path for mixes at the test track. The left wheel path tended to have higher and more variable IRI values. This was attributed to the left laser path being close to potentially segregated mix at the edges of the screed extension. The IRI variability increased with traffic. Using this nitial estimate of variability, some statistically significant differences due to mix variables such as aggregate source and/or binder grade can be seen.

Results from the test track indicate that the finer mixes tended to have IRI values that ncrease slightly as density increases. This suggests that working the fine mixes to obtain higher densities results in a higher initial roughness.This observation was also generally seen in the data from the Michigan field projects. IRI values decreased significantly within 6 months of trafficking for the majority of the mixes at the test track. This suggests that traffic can have a smoothing affect on the pavement for up to 6 months after the start of trafficking.

All results from this research were based on the construction of test track sections. These results should be verified on actual paving projects.