A methodology for obtaining traffic data input to the NCHRP 1-37A PDG

Jingjuan Li

Traffic loading is an essential input to the pavement analysis and design process because it significantly affects pavement performance. Therefore, it is important to predict it accurately over the life of pavements. This is specially challenging, given the limited information available at design time. The NCHRP 1-37A Pavement Design Guide (PDG) uses mechanistic-empirical relationships to predict pavement performance. Traffic input is in terms of axle loading, axle configuration, and number of axle passes. Traffic loading induced pavement damage accumulation depends on the mechanical properties of the layers, which are affected by environmental conditions (e.g. temperature, humidity), which vary with time. Therefore, the temporal variation in traffic loading parameters needs to be specified. This thesis addresses two objectives, which are addressed by extracting and analyzing data from Long Term Pavement Performance database (LTPP). The first objective is to develop a methodology for computing the traffic data input necessary to the new PDG. User-friendly software TI-PG is developed to generate traffic input to the PDG. TI-PG uses daily traffic volume or axle passes as data sources to compute the traffic input elements. The daily data can be continuous over extended periods of time or discontinuous for short time spans. Site-specific or Regional data sets are combined for these computations. The general data storage file in Microsoft Access Table format is used as input. Site-specific or Regional traffic information can be computed for different purposes. The second objective is to document the extent of variation in traffic input as a function of the traffic data collection scenario. Seventeen traffic data collection scenarios are simulated using daily WIM (Weigh In Motion) data from the LTPP database. For each scenario, 30 sites are used for the simulation. Statistical analyses are performed for the main traffic input elements for the PDG. Results show that for traffic volume estimation, one month per season and one week per season of site-specific truck class data show similar accuracy in predicting AADTT (Annual Average Daily Truck Traffic) and MAF (Monthly Adjustment Factors). Site-specific truck class data collected periodically (monthly or seasonally) is very important for AADTT estimation. For axle loading information, one month per season site-specific data has much better accuracy than one week per season site-specific data. It is concluded that the length of data coverage can improve the quality of the axle load distribution estimation.

A methodology for obtaining traffic data input to the NCHRP 1-37A PDG

Files and links (1)

Abstract

Metrics

Details