A proposed analytical technique for the design and analysis of major freeway weaving sections

Weaving occurs when merging traffic streams entering a freeway from an on-ramp cross over diverging traffic streams exiting the freeway via a nearby off-ramp. The intense lane-changing activity which typically occurs in weaving areas can create significant operational problems. Thus, weaving sections often represent bottleneck locations in urban freeway systems.;The prevalence of weaving areas on U.S. freeways warrants the need for analytical techniques which can reliably analyze and/or design these critical freeway components. However, previous research at the Institute of Transportation Studies suggested that existing analytical procedures do not accurately predict weaving operation and/or geometric design requirements.;These existing procedures typically predict performance macroscopically. Operational quality is estimated using input variables which reflect general traffic flow and geometric characteristics prevailing at a subject weaving site. No attempts are made to model the complex vehicular interactions that occur within weaving areas. As a result, these existing techniques are generally unable to predict operating conditions reliably.;The research documented in this dissertation has sought to develop and calibrate a more reliable technique for evaluating weaving performance. Specifically, a procedure is proposed for the design and analysis of major weaving areas (a subset of all weaving configurations). Such weaving sections are often used at freeway to freeway interchanges.;The proposed procedure predicts vehicle flow rates in critical regions within a weaving area as a function of prevailing traffic flow and geometric conditions. Predicted flows are then used to assess the capacity sufficiency and/or level of service of the subject weaving site. Such a methodology represents a more disaggregate approach to evaluating weaving operation than existing evaluation procedures.;The proposed model was developed using large amounts of empirical and simulation data. Much of the research effort was directed toward identifying the traffic flow and geometric factors that influence the behavior of traffic streams operating at weaving locations. Results indicate that traffic behavior in weaving areas can generally be predicted.;Validation data used in this work suggest that the proposed model reliably predicts weaving operation. Moreover, the proposed technique represents a simple and rational approach to analyzing and designing weaving sections.