UC Davis Institute of Transportation Studies

This report presents the results of the study, "Pilot Project for Fixed Segmentation of the Pavement Network."

The goal of this pilot project was to study a small sample of the California Department of Transportation (Caltrans) network to determine the feasibility of expanding the pilot approach to the entire pavement network. The project's work included evaluating the effectiveness of ground penetrating radar (GPR) and limited coring for measuring pavement layer thicknesses and types, application of an algorithm for determining "fixed" segmentation of the pilot network, population of a database for the pilot network, then assessing costs of these activities.

Fixed segmentation for use in the Pavement Management System (PMS) is required to develop the capability to do pavement performance modeling, which is essential for the following pavement management tasks:

* Predicting future performance of segments of the network, and * Identifying the most cost-effective maintenance and rehabilitation (M&R) strategies based on life-cycle costs.

Pavement layer-type and thickness data are also needed to develop effective pavement performance models and to conduct effective condition surveys of composite pavements (asphalt overlays of PCC pavement). The data are also useful for project-level engineering.

Background information summarizing the experiences of several other states in using GPR for pavement work is also presented.

The pilot network consisted of a total of eight roadways: three interstate highways (I-5, I-505, and I-80), four state routes (SR-16, SR-45, SR-99, and SR-113), and one U.S. highway (US-50). The roadways chosen are mostly in District 3, except for the I-80 section and part of the I-505 section, which are both in District 4. GPR data was collected on 681 lane-miles of the pilot network and analyzed for 305 lane-miles. Traffic data was obtained from Caltrans. Climate regions were determined from a recent map developed by Caltrans and the University of California Pavement Research Center (UCPRC).

The UCPRC collected coring data for some of the locations on the pilot network. Some of the cores were provided to Infrasense, Inc., for GPR calibration and some were retained by the UCPRC for checking the accuracy of the layer thicknesses and types that Infrasense determined from the GPR data. The UCPRC also collected available as-built information and maintenance records, and the most recent Pavement Condition Survey (PCS) data from Caltrans.

The UCPRC then used the data collected to develop fixed segmentation for the pilot network, resulting in 236 segments for the 305 lane-miles analyzed, with an average segment length of 1.27 miles.

Comparison of the cores retained by the UCPRC with the layer types and thicknesses identified by the GPR showed that the GPR data was reliable, especially for the top two layers of the pavement.

Extrapolation of the costs on the pilot network for data collection and analysis and segmentation results in an estimate of approximately $7.0 million of contracted field work consisting of GPR use and coring (including collection and analysis), plus 12.3 person-years of additional analysis work to complete the segmentation for the entire Caltrans 49,000 lane-mile network.

If Caltrans moves ahead with collection of pavement structure data and fixed segmentation, it will be important to document as-built information in the structural database as future maintenance, rehabilitation, and reconstruction work occurs, in order to keep the database accurate.

Work beyond this pilot study is underway to determine:

* Whether PMS performance data can be used with the fixed segments to develop reasonable performance histories for the segment, and * Whether the performance models developed by the UCPRC from Washington State Department of Transportation (WSDOT) PMS data can be verified with Caltrans PMS performance histories using the fixed network segments and other necessary data developed in this pilot project.