ANALYSIS OF FREEWAY WEAVING SECTIONS
Project No. 3-75Copy No. 1ANALYSIS OF FREEWAY WEAVING SECTIONSFINAL REPORTPrepared forNational Cooperative Highway Research ProgramTransportation Research BoardNational Research CouncilTransportation Research InstitutePolytechnic UniversitySix Metrotech CenterBrooklyn, NY 11201andKittelson & Associates, Inc.610 SW Alder Street, Suite 700Portland, Oregon 97205January 2008
ACKNOWLEDGMENT OF SPONSORSHIPThis work was sponsored by the American Association of State Highway and TransportationOfficials, in cooperation with the Federal Highway Administration, and was conducted in theNational Cooperative Highway Research Program, which is administered by the TransportationResearch Board of the National Research Council.DISCLAIMERThis is an uncorrected draft as submitted by the research agency. The opinions and conclusionsexpressed or implied in the report are those of the research agency. They are not necessarilythose of the Transportation Research Board, the National Research Council, the FederalHighway Administration, the American Association of State Highway and TransportationOfficials, or the individual states participating in the National Cooperative Highway ResearchProgram.
TABLE OF CONTENTSCHAPTER 1 - OVERVIEW, HISTORY, AND CONCEPTS . 1PROJECT OBJECTIVES AND PRODUCTS.1HISTORY AND BACKGROUND .2THE HCM2000 WEAVING ANALYSIS MODEL.7AN OVERVIEW OF THE RECOMMENDED MODEL .10CONCEPTS, TERMINOLOGY, AND VARIABLES.11THE DATA BASE .19CLOSING COMMENTS .21REFERENCES FOR CHAPTER 1 .22CHAPTER 2 - PREDICTION OF LANE-CHANGE PARAMETERS . 25PREDICTING THE RATE OF WEAVING LANE CHANGES IN A WEAVING SECTION .25PREDICTING THE RATE OF NON-WEAVING LANE CHANGES IN A WEAVING SECTION .31CLOSING COMMENTS .39CHAPTER 3 - PREDICTION OF SPEED PARAMETERS . 41THE HCM2000 MODEL .41PREDICTING THE AVERAGE SPEED OF WEAVING VEHICLES .43PREDICTING THE AVERAGE SPEED OF NON-WEAVING VEHICLES.46THE ISSUE OF FREE-FLOW SPEED.51SENSITIVITY OF SPEED.53CLOSING COMMENTS .59CHAPTER 4 - CAPACITY OF FREEWAY WEAVING SECTIONS . 61THE HCM2000 APPROACH .61CAPACITY OF WEAVING SECTIONS IN THE DATA BASE .62ESTIMATING WEAVING SECTION CAPACITY .64OTHER LIMITATIONS ON CAPACITY.67MAXIMUM LENGTH OF WEAVING SECTIONS.70SENSITIVITY OF CAPACITY.71CLOSING COMMENTS .73CHAPTER 5 - CONCLUSIONS AND RECOMMENDATIONS. 77A RECOMMENDED METHODOLOGY .77TWO-SIDED WEAVING SECTIONS .83OTHER RECOMMENDATIONS .84APPENDIX I - AERIALS & SCHEMATIC OF SITES . 87APPENDIX II – DRAFT CHAPTER FOR THE HCM 2010 . 103APPENDIX III – RESPONSES TO REVIEWER COMMENTS. 159i
ii
National Cooperative Highway Research Program Project 3-75ANALYSIS OF FREEWAY WEAVING SECTIONSFinal ReportCHAPTER 1 - OVERVIEW, HISTORY, AND CONCEPTSThis document represents the Final Report of the Project Team for National CooperativeHighway Research Program Project 3-75, Analysis of Freeway Weaving Sections. The ProjectTeam includes individuals from the prime contractor, the Transportation Research Institute ofPolytechnic University, and a major subcontractor, Kittelson and Associates, Inc. When theproject was initiated, a second subcontractor was included, Catalina Engineering, Inc. Catalinasubsequently merged into Kittelson and Associates. The Project Team consists of:Polytechnic UniversityRoger P. Roess, Professor of Transportation Engineering (PI)Jose M. Ulerio, Industry Associate Professor of Transportation EngineeringElena S. Prassas, Associate Professor of Transportation EngineeringKittelson and Associates Inc.Jim Schoen, Vice-President (Co-PI)Mark Vandehey, Principal EngineerWilliam Reilly, Principal EngineerWayne Kittelson, President and Principal EngineerIn the early stages of the project, Dr. Alexander Skabardonis of the University ofCalifornia at Berkeley participated as a consultant.PROJECT OBJECTIVES AND PRODUCTSNCHRP Project 3-75, Analysis of Freeway Weaving Sections has had a clear and focusedprimary objective: calibrate new and/or updated models for prediction of performance infreeway weaving sections, and draft a replacement chapter for the Highway Capacity Manual.The Project Team, working with the Project Panel and the Highway Capacity and Quality ofService Committee of the Transportation Research Board (HCQSC), established the followingdesirable characteristics that the new models should embody:1. The new model(s) should be conceptually logical.2. The new model(s) should be based on a significant modern data base covering abroad range of weaving designs, configurations, and demand flow rates.3. The new model(s) should attempt to eliminate the need for separate algorithms basedupon weaving configuration type and constrained vs. unconstrained operation.4. The new model(s) should attempt to incorporate parametric measures that directlydescribe the impact of configuration, and (if possible) constrained vs. unconstrainedoperation.5. The new model(s) should provide demonstrably improved predictions of performanceparameters when compared to the current models of the HCM2000.1
Once such models were developed, the following products were to be prepared:1. A draft replacement Chapter 24, Freeway Weaving, for the Highway CapacityManual.2. Draft material to replace freeway weaving portions of HCM Chapter 13, FreewayConcepts.3. Recommendations for changes in other related chapters of the HCM, including, butnot limited to Chapter 22, Freeway Facilities and Chapter 25, Ramps and RampJunctions.4. A spreadsheet-based computational engine that replicates the methodology developedfor freeway weaving sections.5. A Final Report documenting the research efforts leading to the new model(s) andmethodology.This report provides a detailed explanation of how and why models were developed, andhow they fit together to provide a cohesive methodology for analysis of freeway weavingsections.HISTORY AND BACKGROUNDWith the possible exception of signalized intersections, it is doubtful that anymethodology of the Highway Capacity Manual has been so frequently studied, and so frequentlya subject of technical controversy. It is, therefore, beneficial to place this current effort into thehistorical context of research in the general area of freeway weaving section capacity andperformance.Table 1-1 provides a capsule summary of various approaches that have been used tomodel freeway weaving areas, beginning with the work leading to the 1965 HCM model – thefirst to specifically address weaving behavior.The 1965 HCM actually contained two approaches to freeway weaving areas, as part of amodel that attempted to address weaving on all types of facilities. The primary model,developed by Leisch and Normann [1], was based upon a set of curves that plotted total weavingvolume vs. weaving length. Each curve represented a “k-factor” that varied from 1.0 to 3.0, andwas used as a multiplier to develop an “equivalent non-weaving volume” according to thefollowing equation:vEQ v ( k 1) vw 2where:vEQvkvw2 total equivalent non-weaving flow rate, veh/h.total flow rate, veh/h (unadjusted)weaving intensity factorsmaller weaving flow rate, veh/h2
Table 1-1Summary of Alternative Modeling Approaches to Freeway Weaving ivalent NonWeavingVehiclesNot directly.Macroscopic,Lane DistributionYesFreewayCapacityControlsMoscowitz &Newman1963Microscopic,Lane Distributionand LaneChanging by CellYesFreewayCapacityControlsRoess & McShane1973-1980Macroscopic,RegressionBased,Speed PredictionLeisch1983Macroscopic,Equivalent NonWeavingVehiclesLeisch, Normann1965Hess1963AddressLOS?MOECommentsBased on very sparse data. Quality of Flowused to map into LOS. Approach notsuccessfully calibrated in later studies.3YesApprox.SpeedYesMerge, Diverge,and FreewayVolumeRegression-based model focuses on lane 1 ofthe freeway and the ramp, general LOS criteriabased upon flow rates loosely tied to verbaldescription of operating characteristics.YesMerge, Diverge,Weaving, andFreewayVolumeFocus on high-volume cell among freewaylane 1 and auxiliary lane, general LOS criteriabased upon flow rates loosely tied to verbaldescription of operating characteristics.Not directly.YesAverage Speedof Weaving andNon-WeavingVehiclesAppeared in several forms, with final formappearing in Circular 212, iterative process,introduced configuration and type of operationinto the analysis process.Not directly.YesAverageSpeedA re-calibration of the 1965 Leisch/Normannwork. Nomographs used.
Table 1-1 (Continued)Summary of Alternative Modeling Approaches to Freeway Weaving AreasModelReilly et al19841985 HCMRoess et alFazio 19854Cassidy,Skabardonis,May, Ostrom1988-1995BasicTypeMacroscopic,Theoretical andRegressionBased,Speed PredictionAddressCapacity?Not directly.Not directly.AddressLOS?MOECommentsYesAverage Speedof Weaving andNon-WeavingVehiclesIntroduced a different “density” concept tied toweaving intensity, introduced basic modelform still used in HCM 2000.YesAverage Speedof Weaving andNon-WeavingVehiclesDeveloped as a merger of the earlierRoess/McShane and Reilly models. The Reillymodel form was stratified to considerconfiguration and type of operation.YesAverage Speedof Weaving andNon-WeavingVehiclesAdded lane-changing parameter to Reilly-typemodel, eliminating the need for differentconfiguration types to be considered.MacroscopicMacroscopic,Theoretical andRegressionBasedMicroscopic,Lane-Distributionand LaneChanging by CellHCM 2000Roess et alMacroscopicLertworawanich& Elefteriadou2001-2002Microscopic,Gap Acceptanceand LinearProgrammingNot directly.Yes, Based onMax Cell FlowRates and MaxLane-Changingper CellYesDensityA modern look at the Moscowitz/Newmanmodel form, with far greater precision. Lanedistribution modeled for each component flowof the weaving section.YesYesDensityAddition of density model and capacitypredictions to 1985 HCM methodology.YesNoN/ACapacity model based upon gap acceptanceand linear programming optimization treatsweaving capacity as function of basic freewaycapacity.
The 1965 HCM model did not refer to the term “equivalent non-weaving volume” orflow rate. The algorithm was, however, used to determine the number of lanes needed in theweaving section, essentially dividing vEQ by an appropriate capacity or service volume per lane.NCHRP 3-15, Weaving Area Operations Study, conducted at Polytechnic University [2]in the early 1970’s made extensive attempts to calibrate the weaving curves of the 1965 HCM.Calibrated k-factors, however, could not be systematically related to the length of the weavingsection and weaving volume or flow rate, even when different constructs of the “equivalent nonweaving volume” concept were attempted.Another unique aspect of the 1965 HCM model was the clear definition of “out of therealm of weaving.” The weaving curve for a k-factor of 1.0 essentially identified the limit ofweaving length that resulted in weaving movements. Beyond these lengths, which dependedupon weaving volume or flow rate, the section was believed to operate as a basic freewaysection, with merging at one end and diverging at the other. The curve depicted lengths of up to8,000 ft, based largely on data from a single long weaving site. Subsequent weaving studieshave focused on lengths no longer than 3,000 – 3,500 ft, due to the cost of data collection and thelikelihood that longer sections do not operate as weaving sections. The issue of maximum lengthof weaving sections is addressed as part of the current research.The 1965 HCM contained another model that could be applied to ramp-weaveconfigurations. The model, developed by Moskowitz and Newman [3], was actually presentedas a merging and diverging model for ramp junctions operating at levels of service D and E. Itdefined lane-changing distributions between lane 1 (right freeway lane) and the auxiliary lane,and identified the 500-ft segment that had the most intense lane-changing activity. This modelprovides the theoretic basis for subsequent algorithms based upon microscopic lane-changing ofother characteristics. The primary weakness of the model was that the lane-changing distributionwas based solely on the length of the section, and did not vary with other factors, such as volumeor flow rate, or the split between weaving movements.The first significant post-1965 HCM study of weaving sections was NCHRP Project 315. It was also the first in a string of NCHRP and FHWA-sponsored efforts directed specificallytowards the development of the 1985 HCM (which was originally supposed to be the 1983HCM). The results of NCHRP 3-15, Weaving Area Operations Study, were published in anNCHRP Report [4]. The model introduced the issue of configuration, and involved complexiterations. As part of an FHWA-sponsored study of Freeway Capacity Analysis in the late1970’s, the model was re-formatted by Roess and McShane and published in TRB Circular 212[5], Interim Materials on Highway Capacity. This model continued to be complex and iterative,but broke the original model into discrete steps that were more easily explained andimplemented. It also introduced the concept of constrained vs. unconstrained operation, even tothe point of defining the degree of constraint that might exist.While some of the concepts of this model were interesting, and survive in current models,the algorithms were difficult to implement, and their subdivision into various components madecalibration an issue, given the limited size of the data bases available at the time.5
While the NCHRP and FHWA studies progressed, Leisch [6] independently developed amodel similar to the 1965 HCM in form and concept. FHWA later funded the documentation ofthe method. In the meantime, the model was also published as part of TRB Circular 212. Thus,from 1980 through the publication of the 1985 HCM, several different weaving area analysismethodologies were in active use: the two models from the 1965 HCM, the Roess/McShanemethod of Circular 212, and the Leisch method of Circular 212. The Leisch model continued todepict weaving lengths for which no data existed, and produced results that differed substantiallyfrom the Roess/McShane model, even though both were calibrated with the same data.In 1981, another weaving research effort was launched to answer the question of whetherthe Roess/McShane model or the Leisch model should be chosen for the forthcoming 1985HCM. Conducted by JHK and Associates, the study included additional data collection, andrecommended a third model for inclusion in the HCM. This model, developed by Reilly et al[7], resulted in the algorithm form that is currently used in the HCM2000. The model did not,however, address configuration or type of operation.The 1985 HCM model was based upon the Reilly algorithm, modified by Roess (at thebehest of the HCQSC) to incorporate the impact of configuration and type of operation. Themodel has been updated twice since 1985, based upon a single data base from the Reilly studyconsisting of 10 sites with 1 hour of data each. Both revisions were made to constants ofcalibration in the primary algorithm, and were published in the 1994 update to the manual andHCM2000. Other changes in the HCM2000 included the elimination of multiple weaving areaanalysis, the development of a complex capacity estimation procedure, and conversion to adensity-based level of service definition.Since 1985, a number of additional weaving area studies have taken place. All werehandicapped by small data bases, but a number of interesting concepts resulted.Fazio [8] developed a model around the Reilly algorithm, but added a lane-changingparameter that eliminated the need to pre-categorize weaving areas by configuration. This isessentially the approach recommended herein, with more attention paid to the development ofthe lane-changing parameter(s). Fazio, due to a small data base, was forced to assume entrylane-distribution behavior of weaving vehicles to estimate lane-changing.CALDOT and the University of California at Berkeley conducted a number of weavingstudies through the 1980’s and early 1990’s that focused on recalibration of a model similar tothe Moskowitz/Newman approach in the 1965 HCM [9, 10, 11, 12].Over the years, several different calibrations were researched by CALDOT/Berkeley.The methodology(ies) have both strengths and weaknesses. All of the configurations studied inthe California work would be classified (in the original terms of the 1965 HCM) as one-sidedweaving sections in which weaving activity is focused on the right-most lanes of the section.Two-sided weaving sections were not included in the studies. A major issue is the calibration oflane distribution models.6
Given that separate distributions are needed for each of several lanes, for three or fourcomponent flows (including ramp-to-ramp), and for various lengths and configurations, thenumber of such models nee
realm of weaving.” The weaving curve for a k-factor of 1.0 essentially identified the limit of weaving length that resulted in weaving movements. Beyond these lengths, which depended upon weaving volume or flow rate, the section was believed to operate as a basic freeway section, with merging at one end and diverging at the other.
freeway merge and diverge segments to address limitations of the Chapter 14 methodology. VOLUME 4: APPLICATIONS GUIDE 25. Freeway Facilities: Supplemental 26. Freeway and Highway Segments: Supplemental 27. Freeway Weaving: Supplemental 28. Freeway Merges and Diverges: Supplemental 29. Urban Street Facilities:
"The Gulf Freeway Observes its 500 Millionth Mile of Use," press release prepared by W. J. Van London, head of the Houston Urban Expressway System, 18 November 1954. 39. "Highway Week," Texas Parade, January 1952. 40. "Freeway bond drive is launched," Houston Chronicle, 27 August 1956; "Freeway bond rally Thursday," Houston .
ments to and along U.S. 59 (Southwest Freeway) between Beltway 8 and I.H. 610 (West Loop) (see Fig 1). The im provements provide for: adding lanes to the freeway, con struction of a transitway and transitway interchanges along the freeway, widening and overlaying the existing freeway pavement, frontage road improvements, improving the free 1
Dark Age Tablet Weaving for Viking and Anglo-Saxon re-enactors 1 Introduction Tablet weaving, also known as card weaving, is a method of using square tablets with holes in the corners to weave narrow decorative bands made of wool, linen or silk threads. Tablet weaving was widespread in E
Free Weaving Patterns from Eplore Techniques for nlay Plain Weave and uck Lace The Draft: How to Read Weaving Patterns P atterns for weaving are written in a form called a “draft.” The draft is a standardized short-hand way
Jul 01, 1995 · ROUTE CONCEPT REPORT . STATE ROUTE 91 . ARTESIA FREEWAY (I-710 to I-5) RIVERSIDE FREEWAY (I-5 to I-215) 12-ORA-91 . 0.00/R18.90 . SUMMARY State Route 91 (SR-91) in Orange County, also known as the Riverside Freeway and the Artesia Freeway is an access controlled, grade
56,351 Freeway Ford Bloomington, MN 952-888-9481 freeway ford.net 4.7 / 5.0 465 Veried Re views Vehicle ser viced 02/25/2010 Minnesota Motor Vehicle Dept. Inver Grove Heights, MN Title #H3110W793 Registration issued or renewed - Titled or registered as commercial vehicle 03/12/2010 56,615 Freeway Ford Bloomington, MN 952-888-9481 freeway ford .
Japanese Language School 1 This guide for university study abroad students, advisors, and staff is intended to help you determine which of your courses is equivalent to each of the levels (1–8) of the KCP Intensive Japanese Language course. Total immersion KCP International teaches Japanese in the Direct Method —learning entirely in Japanese without a vehicular language such as English, at .
