Citation

Fitzpatrick K, Nowlin L. Transp. Res. Rec. 1996; 1555: 42-49.

Copyright

(Copyright © 1996, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

10.3141/1555-06

PMID

unavailable

Abstract

Weaving vehicles can heavily affect the quality of service on a facility. Techniques to evaluate weaving on arterial streets, however, are limited. Similar to arterial streets, the traffic operations on frontage or access roads along freeways can also be heavily affected by weaving. Presented here is a technique for evaluating one-sided weaving areas along with recommended lengths of weaving areas. Data from six sites were used to develop a relationship between number of lane changes and weaving volume. The weaving sections at the sites ranged from three to four lanes and from approximately 137 to 335 m in length. Throughout the large range of volumes present (250 to 2,500 vph), there was a very strong relationship between weaving volume and lane changes. The field data were used to develop a computer simulation model. The relationships between different measures of effectiveness from the simulation model provided information on critical points in which the weaving speed begins to drop more rapidly. Weaving volumes (entrance-ramp plus exit-ramp volumes) of less than 1,500 vph are associated with unconstrained levels of service, volumes between 1,500 and 3,000 vph with constrained service, and volumes over 3,000 vph with undesirable service. Additional information from the models also demonstrated that it is desirable to have a weaving length greater than 300 m. If this length is not achievable, then the minimum length should be approximately 200 m.