opinions, findings, and conclusions expressed

Transcription

1 OF CHEVRON PATTERNS FOR USE EVALUATION DEVICES IN STREET AND HIGHWAY CONTROL H. Cottrell, Jr. Benjamin Research Scientist Highway TRAFFIC 0N ZONES WORK opinions, findings, and conclusions expressed ( are those of author and not necessarily report Highway & Transportation Research Virginia Organization Sponsored Jointly by Cooperative with U. S. Cooperation Highway Federal February 80-R32 VHTRC Council Virginia by this in of those sponsoring agencies.) (A and of Highways & Transportation Department University of Virginia) of Transportation Department Administration Charlottesville, Virginia

2 H. E. PATTERSON, Senior Traffic Engineer, Norfolk Department MR. Public Works of TRAFFIC RESEARCH ADVISORY COMMITTEE MR. L. C. TAYLOR II, Chairman, District Traffic Engineer, VDHgT MR. L. H. DAWSON, JR., Assist. Traffic g Safety Engineer, VDHST MR. J. B. DIAMOND, District Traffic Engineer, VDHgT MR. J. E. GALLOWAY, JR., Assist. State Materials Engineer, VDHgT DR. JAMIE HURLEY, Assist. Professor of Civil Engineering, VPI g SU MR. C. 0. LEIGH, Maintenance Engineer, VDHST MR. R. F. MCCARTY, Safety Coordinator, FHWA MR. J. P. MILLS, JR., Traffic $ Safety Engineer, VDH$T MR. W. C. NELSON, JR., Assist. Traffic $ Safety Engineer, VDHST MR. R. L. PERRY, Assist. Transp. Planning Engineer, VDHgT MR. F. D. SHEPARD, Highway Research Scientist, VHgTRC ii

3 chevron pattern consists of alternate orange and stripes that form an arrow pointing in direction in white traffic is being diverted. objectives of this which were (I) to select most effective design for research pattern, and (2) to evaluate effectiveness of chevron chevron designs under road conditions as compared to selected most effective chevron pattern was selected by a rating of groups of patterns used on channelizing subjective ABSTRACT presently used designs. In general, selected chevron designs were preferred devices. presently used patterns. A black stripe separating over orange and white stripes proved effective in reducing haloation. measure of performance used in field tests was position of lane changing relative to transition taper. was found that driver response was not strongly dependent on It channelizing device employed in taper. subjective revealed chevron patterns to be preferred over evaluation presently used patterns because of ir clear directional message. iii

4

5 OF CHEVRON PATTERNS F0R USE ON TRAFFIC EVALUATION DEVICES IN STREET AND HIGHWAY WORK ZONES CONTROL H. Cottrell, Jr. Benjamin Research Scientist Highway total system of traffic control devices is installed A assure smooth, safe vehicular movement in vicinity o to and highway work zones. Barricades and channelizing street essential elements in total system, are employed devices, guide and direct drivers safely past hazards. ''( marking on standard barricade rail consists of orange and white stripes of equal width. stripes alternate downward at an angle of 45 in direction traffic slope to pass. This directional information does not provide is since many drivers do not understand message guidance is meant to impart. Moreover, field crews often are design as to which diagonal sign (slope right vs. slope left). confused be installed. In general, pattern of stripes on should standard barricade rail promotes confusion in directional guidance. Virginia Department of Highways and Transportation extensive use of orange cones for channelization on recommends only work. Since cones are not reflectorized, simulated daytime panels, vertical panels with orange and white horizontal drum of reflective material, are used at sites where stripes is extended into evening. se vertical panels have work research efforts have investigated appli- Recent of chevron pattern to barricades and channelizing cation chevron pattern consists of alternate orange and devices. stripes that form an arrow pointing in direction in white by INTRODUCTION AND PROBLEM warn and alert drivers of hazards created by construction "to maintenance activities in or near traveled w, and to or replaced drums because drums are scarce in Virginia.

6 traffic is being diverted. Studies have documented which of signs th chevron pattern in providing bearing(2, )but effectiveness optimal design for chevron pattern has significant potential as a design on barricades and channelizing devices. standard effectiveness of selected chevron designs under evaluate conditions. A traffic sign's performance measures its road traffic control devices bearing chevron design to that of barricades and channelizing devices bearing presently of scope of research was limited to use of and channelizing devices to provide directional barricades achieve two-objectives, study comprised four To tasks as listed below. major Review of literature covering completed and ongoing A. on traffic control devices used in work zones. research Selection of chevron designs by subjective evaluation for B. in field tests. examination Field tests on selected chevron patterns and presently C. patterns to obtain data on average driver's used guidance information, positive sign has not been determined. this OBJECTIVES AND SCOPE objectives of this research were (I) to select effective design of chevron pattern and (2) to most to command attention and convey a clear, meaningful ability to driver. evaluation compared performance message used stripings. This restriction was important because re are a guidance. of situations in which barricades and channelizing variety devices are applicable but directional guidance is not warranted. response to se devices. D. Comparative evaluation using results of field testing.

7 search of available literature was conducted A facilities of Transportation Research Information through Reports selected from abstracts received through Service. literature search were obtained. Additional reports were by transportation professionals and through a less identified literature search. Information derived from literature formal of objectives was to select most effective One for chevron pattern. approach used to perform design barricades are in use in France, Canada, and state chevron Utah. se devices are usually large and are used to of effectiveness of chevron pattern has been in research efforts by CalifQ ia Department of addressed parameters must be specified in order to define Several chevron designs, including length and width of alternative sign face, ratio of orange stripes to white stripes, width of stripes. Or considerations are and target value of sign depends on size of face and width of stripes. Both chevron research sign cited above concluded that diagonal pattern was efforts at a greater distance than chevron pattern of recognizable LITERATURE REVIEW review is documented throughout report. CHEVRON DESIGN SELECTION this task and results are presented in this section..exper i.e ce with Chevron Designs Previqus concept of chevron patterns is not new. Standard supplement a taper or to close a roadway. and BioTechnology, Inc.. and conclusions Transportation(2) from reports on those efforts influenced selection of drawn chevron patterns in present study. Chevron Pattern Design of thin black stripes between white and orange placement and use of borders (black or orange). Encapsulated stripes lens sheeting is used on all channelizing devices except cones.

8 size and stripe width. reason for this appears similar be haloation, a phenomenon by which reflection of to stripe dominates and distorts orange. Haloation white point of chevrons to appear less distinct. causes haloation. objective of borders is to outline eliminating sign. comparison of striping patterns, size In signs was varied little. On or hand, since of point of legibility. At point of detection, and ft. (154 m) from devices, patterns are visible 500 message pattern imparts is unclear. message but clear at point of legibility (300 ft. [93 m] from is se two points are important and are discussed devices). subsequent sections. demonstrations were conducted in both day and night conditions. At night,, groups of under were observed under both high and low beam headlights. designs various groups are displayed in Figure l.and of groups are given in Appendix A. Table I photographs each group in Figure i with a specific channelizing identifies devices. aonly one rail displayed. ratio of orange to white and placing thin black Increasing between white and orange stripes are measures for stripes haloation is important, stripe ratio was varied overcoming widely and black stripe was employed. quite Chevr_on D.e..s..ig n D..e.m.o.ns..t.r.atio.n of chevron designs were subjectively rated by Groups in vehicles at two points -- point of detection observers Tab le i Sign Groups and Device Types GROUP DEVICE Type II barricades 1,2,3 Type I barricades 4,4A 5,6,7 Vertical panels

9 (I) x 24" 8".:,..i!,. o nge ½" Figure i. Chevron pattern groups. :::_.:!::.:..::::.'..::: :: :':':,;::':.%:: ':::.9...:.:-:.::.':.:..:::.-.:.:.:.'.:. ½" black border I" : 2.54 cm 4"0 4"W (½"B) (2) x 24" 12" 4" 0 4"W (½"B) (3) x 24" 8" 12" x 24" (½"B) 6"W 6"0 ":" i:g'::%:.. ".$-:$ : :.': "::!@:.:::::::h-. ":':::::':'$'.': { ":i:::i h{b. "::::::::':::".K:'.:. ":::'-":::' "':::": :':'-';-':' ": : :-" S"W 6"0 ½"B 3"W 6"0 ½"B (4A) S"W 6"0 ½"B Legend, black ----.'.':.!:..: (5) W white, 6.#! x ]..2" 24" x 18" 24" 30" x 18". :'k-:. "::::, :.:.;$. ; ::.: ::::.:::.';, ::!: ::: l'tb border i 314"B -i B

10 observers rated pattern groups at two in terms of parameters given in Table 2. points rating form is shown in Figure 2. to command attention (a) Ability to convey clear, dis- Ability messages (d) tinct Ability to warn and alert (b) Ability to guide and direct (e) rating scale was varied depending on number of in a group. parameters were summed to obtain one patterns at each point for each pattern by an observer (i); in measure x. (a+b+c)i; Yi = (d+e+f).. se measures were symbols, formall observers' ratings 1 of each pattern at each summed in symbols, se two point; X = i x. Y = v.. all i - all measures were compared with those for or cumulative in group. mean and standard deviation were patterns and Wilcoxon Ranked Sign Test was used to calculat.ed rank patterns with a 0.05 level of significance statistically a two-tailed test level of significance was selected for it is customary to consider a test significant if null since hyposis is rejected at this level. ratings, three demonstrations were conducted For minor differences in sign groups in first with as compared to those in last two. results demonstration first demonstration prompted Omission f groud 4A of which one pattern was obviously superior and addition of in 7 to identify threshold width for effectiveness group black stripe. dates of demonstrations and of are listed in Table 3. participants Rating Procedure Table 2 Parameters for Rating Sign Groups At Point of Detection At Point of Legibility Overall appearance (c) Overall appearance (f) l E

11 Inadequate Fair Adequa t e signs are identified from left to right by following order alphabet. of Examp I.e" Following Rate with Respect Patterns Figure 2. Chevron rating form. Title Your Rating Scale" A B C Observe all of patterns before rating m. Group No. Poi.nt of Detection Point of Legibility To ir" Sign Description Comments

12 April 4-5, 1979 April 19-20, 1979 Traffic Research Advisory Committee A and district traffic engineers members assistants at Research Student Council average numbers of observers in groups were 14 for A, for B, and 8 for C. visual acuity of participants i0 assumed to be representative of that of general population. was majority of persons in groups B and C were not involved in cases where it was not possible to conclude which In was preferred, results for Groups B and C were pattern to increase data available for Wilcoxon Ranked combined Test. This combination is reasonable because grouping Sign procedure was followed on pattern groups for This day and night observations. Where re were inconsistencies both day and night observations or between demonstration between a subjective decision was made to select one of groups, results of ratings are given in Table 4. and night rankings are displayed for each pattern group day along with sign selected. results of ratings of pattern groups, From following conclusions and recommendations were developed. general, observers were not aware of In message imparted by diagonal pattern. directional some variations in wear, Despite were consistent. ratings Table 3. Demonstration Schedule Group Participants Date May 2-3, 1979 members at Research Staff Council traffic oriented research. arrangements, location, and wear conditions were similar. patterns.

13 Design Selection Results Chevron to Figure 1 for a sketch of chevron patterns) (Refer # Day Night Grou.p. a chevron striped a chevron striped 1 chevron a chevron b diagon.al c diagonal b a chevron striped a chevron striped 2 chevron b chevron a a 12" chevron striped a 12" chevron striped 3 12" chevron non-striped a 12" chevron non-striped a 8" chevron striped c 8" chevron striped c 8" chevron d 8" chevron d "W, 6"0, I"B a 6"W, 6"0 a 5"W, 6"0, I/2"B a W"W, 6"0 d 3"W, 6"0, I/2"B d 6"W, 3"0 f 4"W, 6"0, I"B a 5"W, 6"0, I/2"B b 1 3/ "B a ii 2"B c Legen.d indicates rank alphabet I" 2.54 cm %"W, 6"0, I"B a 3"W, 6"0, I/2"B b 5"W, 6"0, I/2"B b "W, 6"0 b 6"W, 6"0 e 6"W, 3"0 f "W, 6"0, I"B a 5"W, 6"0, I/2"B b 1 3/ "B a I/2"B c 4"W, 6"0, I"B Table 4 Note" designs are listed in order of decreasing preference Overall chevron striped chevron striped c diagonal c diagonal 12" chevron striped "B 4A 4"W, 6"0, with border with border a drum chevron drum; a drum a a chevron chevron c diagonal c diagonal I"B I"B striped chevron a I"B a l/2"b b i/2"b b c standard c standard d orange border d orange border I"B l"b striped chevron a I"B a

14 24 in. (61 cm) long barricade rails, For was little difference between three re at point of detection, but both patterns chevron patterns were preferred over of use of black borders on barricade rails 4a) was not favored and is not recommended. (group patterns (more orange than white) Nonsymmetric preferred for night use for 12 in. X 36 in. were cm X 91.4 era) barricade rails (group 4). (30.5 to haloation, nonsymmetric patterns would Due expected to be more suitable for night use. be pattern with 4 in. (10.2 cm) white, 6 in. cm) orange, and! in. (2.54 cm) black (15.2 was highly rated both day and night. stripes black stripe separating orange and white was effective in improving effectiveness stripes 30 in. X 18 in. (76.2 cm X 45.7 cm) and 24 in. 18 in. (16.0 cm X 45.7 cm) drum panels were X rated at detection, but chevron was equally at legibility (g oup 5). For vertical preferred black stripes larger than i in. (2.5 cm) panels, is 1-3/4 in. [4.4 cm]) were equal to (that in. (2.54 cm) stripe in ratings (group 7). i i in. (2.54 cm) black stripe is recommended. A general, selected chevron designs were In preferred over presently used designs eir as opposed to moving observation adversely influenced observation demonstration. first two situations were imposed by on availability of appropriate sites. However, limitations is realistic to assume that channelizing devices will be it in several different environments where visual distractions used as construction and maintenance equipment) may exist, (such only one observer comjnented on group layout. In regard and mode of observation., it was reasoned that ratings by to observers would increase number of variables moving and thus introduce complications. involved, pattern at point of legibility. diagonal that diagonal pattern was rated slightly Note at detection. Even with 1/2 in. (1.3 era) higher stripe, chevron point is distorted. black 12-in. (30.5 cm) wide rails were preferred 8 in. (20.3 cm) wide rails (group 3). over of patterns (groups 4, 7) or rated equal to m. observers noted that visual distractions in background, Some layout (spacing of patterns), and stationary group i0

15 subjective ratings and statistical analysis, Through with some judgement, following patterns were combined 6 in. (15.2 cm) orange, and white, in. (2.54 cm) black stripes. I II barricades 12 in. X 24 in. (30.5 cm X 61.0 cm) Type with 4 in. (!0.2 cm) white, chevron in. (10.2 cm) orange, and 1/2 in. 4 era) black stripes. (1.3 panels 24 in. X 18 in. (61.0 cm X 45.7 cm) Vertical panel with a i in. (2.54 cm) chevron evaluating effectiveness of chevron designs, input n motorist is desired. objective of field tests from to obtain data on motorists' response to a given was device in a taper arrangement. Preliminary work, channelizing as development of a test procedure and conduct of such testing, and field tests are discussed below. pilot scope of research was limited to Since in Which directional guidance is necessary, a single- situations closure arrangement was selected. measure of effec- lane deemed most appropriate was position of tiveness lane change. To minimize number of lane changes motorists' a four-lane divided highway was specified. A right- occurring, closure was desired because most motorists drive in lane lane and for m a lane change in work zone would, right refore, be necessitated. zonal system was devised to facilitate collection A data on a driver's lane change as a response to a specific of device. basic premise for using lane changing channelizing measure of effectiveness is that earlier a driver as selected for field testing. I barricades 12 in. X 36 in. (30.5 cm X 91 4 cm) Type with 4 in. (i0.2 cm) chevron black stripe. FIELD TESTS Test Procedure ii

16 3. Two zones, each 350 ft. (107.7 m) long, were set up Figure to transition taper zone. zone length is based prior i included point of detection (500 ft. [154 m]) Zone zone 2 point of legibility (300 ft. [93 m]). point and legibility is critical since this is point at which of age conveyed by pattern becomes clear. It is important mess note that legibility distance (300 ft. [93 m]) does not to estimated distance for negotiating provide lane change a zones were used in data collection, with a Four being prior to zone i. Lane changes occurring prior to zone I were mostly due to supplemental devices, such as zone arrow panel and warning signs, rar than to flashing on channelizing device. For this reason, only 3 pattern were considered in data analysis. zones counters were placed at boundaries of Traffic with rubber tubes extending across right lane zones traffic. By determining difference in volume count of traffic recorders bounding a zone, number of vehicles on lanes in that zone was obtained. Note that changing 3 was zone critical zone because forced mergers occurred re. transition taper is single most important element system of traffic control devi in work within where zones reduction in pavement width is desired. Much care was used a arranging taper in accordance with MUTCD and Virginia in field tests were conducted to compare presently Four and proposed barricades and channelizing devices. used lanes, more effective channelizing device is changes providing guidance. zone system is displayed in in est.m, gted time required to change lanes, which is 4 to on seconds 5 (350 ft. [107.7 m]). Department of Highways and Transportation guidelines. of se tests are summarized in Table 5 and devices features displayed in Figure 4. are 12

17 Zone system at test site. 13

18 Design of Field Tests stripe Chevron vertical panel Day $ night Horizontal panel vertical were collected at each site for an average of 21 Data for each channelizing device except cone, for which hours each channelizing device, and percentages of lane changes for zones were obtained based on total of all lane changes by in {one system. n, percentages for occurring used and proposed barricades and channelizing devices presently an effort to establish a single parameter for in in a test and to relate zone of lane change to comparison position relative to work area, zonal lane changes were its amount of weighting depended on distance weighted. zone to work area behind taper. percentage from lane changes within a zone and weighted factors were of weighted lane changes 3 x zone i + 2 x zone 2 + i x zone 3 more effective channelizing device is indicated refore, higher weighted lane changes. se measures were by Type Test No. Presently Used Proposed Time of Day i Cone Chevron Type I barricade Day Type II barricade Chevron Type II barricade Day g night stripe Chevron vertical panel Day g night Diagonal panel vertical data were collected for 8 hours in daytime only. of AnalY..S,is Method distribution of lane changes by zones was determined in each test were compared. mu!tiplied and summed. Thus, where zone i percentage of lane changes in i. calculated for day, night, and total (day and night) time periods. 14

19 Test No. I Test No.. 2 Test No. 3 Test No. 4 PRESENTLY USED PROPOSED 4 Presently used and proposed devices rated in field Figure tests. 15

20 pilot test was conducted to examine adequacy of A test procedure and method of analysis. In pilot test, X cm] black with orange background)mounted on cm left corner of each barricade and oversized chevrons top in. X 48 in. [91.4 cm X cm] with a I-I/2 ft. [0.46 m] (36 white arrow on an orange background). Each zone was 533 ft. wide m) in length since se channelizing devices were larger (164 devices to be tested later in main study and than visible from a greater distance. warning sign refore is in Appendix B and channelizing devices are shown layout results of pilot test are given in Table 6. weighted totals it appears that Type iii barricade From slightly more effective than oversized chevron. is III barricade and arrow sign is significantly larger than Type in Essex County, approximately 2 miles (3.2 kin) north of South southbound traffic was channeled across Tappahannock. second site was on Interstate Route 81 South in County between Edinburgh and Woodstock. A right- Shenandoah closure channeled traffic into left lane and through lane work area on Narrow Passage Creek Bridge. average traffic volume was 13,120. Since bridge work was daily before field tests were finished, a third site, completed at norrn end of Shenandoah County on Interstate 81 located was studied with a similar work zone arrangement and South, sign layouts for three sites are illustrated in B. Appendix Pilot Test was conducted on U. S. Route 460 East in Giles County, which devices compared were Type Iii barricades with channelizing and white diagonal striped rails (12 in. X 48 in. orange cm X cm]) with an arrow sign (24 in. X 48 in. [61.0 [30.5 in Appendix C, Figure C-I. chevron panel. Also note that chevron did not have black stripe separating orange and white (recall that this a was quite effective in chevron design selection). basis of pilot test, minor adjustments were On in field test and analysis procedure. made Site Selection sites, all on four-lane divided highways, were Three for testing. first site was on U. S. Route 17 selected to inside northbound lane due to construction on median southbound Mt. Landing Creek Bridge. average daily traffic volume was 6,695. traffic volume. 16

21

22 field data collected in three tests are in Tables 7-9. shown U,S..Route 17 Mt. Landing Creek test site on U.S. Route 17--Mt. Landing Creek was in order to observe a work zone situation or than a selected right-lane closure. From Table 7, it is noted that a typical percentage of lane changes for all channelizing devices high forced. An early lane change is not encouraged at were crossover setup because (i) lane width ;s narrowed median to addroximately 16 ft. (4.9 m), and (2) driver can only channelizing devices extended across entire roadway. see range of total weighted lane changes is to with Type I chevron barricade and cone as respectively. A 40-ft. (12.3 m) taper spacing was endpoints, used. test site at Cedar Creek consisted of a right-lane due to bridge resurfacing. cone (184.3) and closure drum panel (193.2) represent low and high simulated respectively (Table 8). A linear taper with endpoints, test site at Narrow Passage used an 80-ft. (24.6 m) of channelizing devices in taper. Type I spacing (187.8) was ranked highest and cone (169.3) ranked chevron (Table 9). Note from test #I that for day only 40-ft. lowest m) and 80-ft (24.6 m) spacing of devices are equal for (12.3 I chevron. This result prompted use of 80-ft. (24..8 m) Type at this test site. re were four incidents of spacing colliding with channelizing devices in taper. vehicles devices involved were simulated drum, chevron, and panels. All of incidents occurred during diagonal week. In previous week, re were no collisions with same with similar taper arrangements. area of contact barricades middle of taper. No reasons were identified involved collisions. However, holding of one-week for of channelizing devices are ranked by total All lane changes for each site in Table i0. weighted Field Test Results U.S. Interstate 81 Cedar Creek devices spaced 40 ft. (12.3 m) apart was used. U.S,_!nter_s..tate 81 --Narrow Passage county fair during that week may be related to incidents. 18

23 0 CO C) 0 0

24 0 c3

25

26 (126.7) Cone panel (124.4) Diagonal panel (123.9 Drum panel (123.8) Chevron II diagonal (123.7) Type I chevron (120.6) Type panel (193.2) Drum I chevron (189.0) Type panel (186.9) Diagonal II diagonal (186.1) Type II chevron (!85.8) Type panel (184.5) Chevron panel (185.5) Chevron pane! (182.2) Diagonal II chevron (175.7) Type panel (174.2) Drum II diagonal (173.2) Type (169.3) Cone weighted total for cone includes day only data whereas those Note" all or channelizing devices incorporated both day and night for general, day weighted lane changes were greater In night weighted lane changes, probably because of than results of four tests are discussed below with to each field testsite. reference Type I chevron barricade ranked over cone in cases except for Route 17 --Mt. Landing Creek site. all explanation for this exception is evident in data. No are more mobile than barricades, and this advantage is Cones about same. Type II chevron barricade performed ranked at night than Type II diagonal pattern at Narrow better Table i0 Rank of Total Weighted Lane Changes Rte. 17--Mt. Landing Creek Cedar Creek Narrow Passage I chevron (187.8) Type Type II chevron (-) Cone (184.3) N J barrier (I 5.6) data. For direct comparison for day only data, refer to Tables 7-9. C0.mpar, is.od.s,, o f F.!e,,l d Test..Re,su,, 1.tS decreased visibility at night. T.e.s..t.,#i,. Type. I Chevron Barr!c de. vs.,. Cone important. #2 Type,! I Barricad,es_: Chevron vs. Diagonal Test Type II barricade chevron and diagonal patterns Passage site. 22

27 simulated drum panel ranked over chevron at Creek, equal to chevron at Mt. Landing Creek, and Cedar at Narrow Passage. low ranking of simulated drum lower at Narrow Passage was not expected, since it has panel chevron and diagonal vertical panels consistently close enough to be considered equal. Under night ranked are also noted. Except for results of test #3, equal Creek and Narrow Passage test sites have similar results. Cedar Channelizing Device Designs Selected from Field Test Results Test #3 -- Chevron Panel vs. Simulated Drum Panel largest surface area of all vertical panels. Test #4- Chevron Panel vs. Diagonal. Panel diagonal appeared mostly white and chevron conditions, orange. predominantly channelizing devices favored for each of four are shown in Table II. Channelizing devices which are tests Table!! Test # Rt Mt. La_nding Creek Ceder Creek 1-81 Narrow Passagf! I Cone Type I chevron Type I chevron Not available II chevron; Type II chevron; Type II diagonal Type II diagonal Type Drum panel; chevron panel Drum panel Chevron panel Chevron Diagonal panel; Diagonal panel; panel; panel panel diagonal panel chevron chevron Note" listing of two channe!izing devices indicates equal ratings. 23

28 chevron design selection, legibility distance In defined as 00 ft. (9 m). refome, message imparted was pattemn was not clear until drivem was in zone 2. by this eason, zone 2 was examined alone to determine if Fore was a diffemence in pemcentage of lane changes in this me fop channelizing devices in a given test. Table 12 zone percentages of lane changes in zone 2 (based on shows which entered zone 2 in might lane) fore Ceda vehicles chevron devices had a lower percentage of lane than presently used devices at a 40-ft. (12.3 m) changes except in test #I. On or hand, chevron spacing, had a higher percentage of lane changes than presently devices devices at an 80-ft. (24.6 m) spacing. refore, it is used that chevron devices are more effective at an concluded #I Test I chevron Type # 2 Test II chevron Type #3 & 4 Tests Chevron Drum Diagonal Percentages of Lane Changes for Zone Passage Creek Narrow (24.6 m) 80-ft. Day N i, ght, 0 t.a Zone 2,.a_nd Legibi!i.tY..ni._stanc e Cmeek and Narrow Passage Cmeek. increased spacing, based on zone 2 results. Table 12 Percentage Lane Changes Cedar Creek 40-ft. (12.3 m) Day To tal_ Cone Type II diagonal S

29 s _Te s t o.,mpari s,q.ns CrOS results may be drawn by making comparisons Additional channelizing devices at different test sites. Because between type of closure used on Route 17 --Mt. Landing Creek, of Ced.ar,,Creek.,vs.,,., a,,r0 w,,s, g e primary differences between two sites on 81 are road geometrics, taper alignment, and interstate of channelizing devices. zone system at Cedar Creek spacing [12.3 m] spacing) was set up on a curve and downgrade. (40-ft. at Cedar Creek was linear and not dependent on alignment roadway as at Narrow Passage (80-ft. [24.6 m] spacing), curving taper alignment was parallel to road alignment. where linear taper is more gradual in slope than a curved taper. A total weighted lane changes for each channelizing except chevron panel and Type I barricade were device and chevron panel did not vary much with respect to barricade change in spacing (refer to Tables 8 and 9). positions lane changes are more evenly distributed between zones at of 40-ft. (12.3 m) spacing than at 80-ft. (24.6 m) spacing. seems that chevron patterns would be more It if entire pattern were visible to approachin effective rar than if it was obscured by preceding motorists A 40-ft. (12.3 m) spacing presents image of a pattern. with taper devices running toger (Figure 5). wall is beneficial for nondirectional patterns. However, This patterns are more effective when spaced far directional apart to appear as separate devices. On or hand, enough nondirectional patterns do not convey a message when separately. viewed it is not possible to compare this site with or two. Narrow Passage Creek zone system was on a slight curve slight upgrade (refer to Appendix B). Also, taper and at a 40-ft. (12.3 m) spacing than at an 80-ft. (24.6 m) greater total weighted lane changes for Type I chevron spacing. 25

30 5. Views of Figure fore spacings Typ_e! Barricade vs. Type II Barricades (12.3 m) and 80-ft. ( ft. Type I chevron barricade. Type I chevron barricade rated over smaller II barricades. Barricade rails 2 ft. (0.62 m) long with Type (10.2 cm) stripes are less effective than 3 ft. (0.92 m) 4-in. rails with wider stripes. In some cases, barricades less long 3 ft. (0.92 m) long are necessary due to limited lateral than Moreover, 2-ft. (0.62 m) barricade rails are clearance. New compare for devices work addressing reconstruction additional test was into study to incorporated with channelizing barrier Based on reports $te. work and pavement bridge with a greater increase in ociated or construction activities. e work setu p for se zone activities typical in Virginia oncrete barrier and flashing arrow panel. c Jersey New a employs m) used by contractors. use of wider stripes.on commonly barricade rails, especially in reference to chevron, smaller would increase effectiveness of devices. New..Jersey Concrete Barrier An Jersey concrete Narrow Pas S a e. zone safety, [b, are as s ar accident rates than channelizing Since a supplement to as were being tested while serving devices Jersey barrier, a supplemental taper New seemed like an obvious alternative. 26

31 New Jersey barrier was rated equal to cone for only and lower than all or devices based on weighted day changes at Narrow Passage (Table 9). Steady-burn beacons lane 9eflectors about 6 in. (15.2 cm) long were mounted on and Jersey barrier, which had a slope of 16"1 for 192-ft. New m) taper. From visual observations, all channelizing (52.1 appeared more effective than New Jersey barrier in devices early shifts of traffic. New Jersey barrier influencing not visible from as great a distance as channelizing was However, flashing arrow board panel is probably devices. dominant channelizing element in New Jersey barrier setud. Subjective,.E..va_!uatiq. n brief and limited subjective evaluation was obtained A Virginia Department of Highways and Transportation from Narrow Passage. Two responses were obtained at each site. and evaluation form and results are shown in Appendix D. included a district traffic engineer, district respondents officers, and an on-site inspector. nong comparisons safety 4 tests, chevron pattern devices were preferred for all except that of cone vs. Type I barricade. in cone is preferred for daytime use because of its portability effectiveness. chevron patterns are selected for use and lateral movements are required. channelizing device where among all chevrons is chevron panel. recommended arrow chevron appears to excell n providing a clear single of required movement. picture costs of sheetin_, labor, and Daint for devices are given in Table 13. channelizing cost estimates are based on a processing rate se 100/hour and do not include costs of metal backing of chevron pattern is more expensive than Although presently used design except simulated drum, a cost may be achieved by a reduction of inventory. Since savings may be used to guide traffic in eir direction, an chevrons responsible for traffic control in work areas personnel This evaluation was performed at Mt. Landing Creek tested. Cost Analys i_s and heat pressure treatment. of both left-and right-channelizing devices is not inventory This savings is applicable only to comparison of necessary. chevrons with diagonal patterns. 27

32 backing, and heat pressure treatment) are not included. I" 2.54 cm drum panel simulated panel with i" black chevron $ direct cost comparison between chevrons and cones A not possible because all costs (including stands, metal is Table 13 Cost Estimates Test # I orange cone 36" X 36" chevron with i" stripe rail 12" X 24" 12" X 24" 12" barricade rail diagonal with I/2" black stripe chevron Test # rail 4.28 X 30" 18" X 24" 18': 7.78 Tests #3 & 4 stripe X 24" 12" diagonal panel inch for 24" orange high $0.33/linear sheeting intensity inch for 30" orange high $0.42/linear sheeting intensity Source- Salem District Sign Shop, B Cockram 28

33 conclusions dmawn from this research indicate effectiveness of vamious channelizing devices in melative directional guidance to motorists where latemal pmoviding ame requimed. Conclusions on melated factors (such movements taper spacing) identified in this meseamch ame discussed. as. conclusions listed below have been drawn from esults on black stripe separating orange and white i. improved effectiveness of chevron stripes For barricade rails, 12 im (30.5 cm) wide 2. was recommended over 8 in. (20.3 cm) rail rail. wider rail was clearer at equal wide A nonsymmetric chevron pattern with distances. orange than white partially eliminated more effect. Note that a black stripe was haloation with nonsymmetric striping. included rails 24 in. (61..0 cm) in length and Barricade a stripe width of 4 in. (10.2 cm) received with ratings and were less visible than barricade low Since taper arrangement at Cedar Creek 3. standard 40-ft. (12.3 m) spacing, represents conclusions of field test have been basic from this site. channelizing devices drawn Type I chevron barricade and Type I chevron barricade were panel effective. diagonal and chevron panels most equal and Type Ii diagonal and chevron were were also equal. barricades CONCLUSIONS chevmon sign selection and field tests. pattern rails with a greater stripe width. listed, below in decreasing order of, re weighted lane changes. total Drum panel Diagonal panel; chevron panel ii chevron barricade; type II diagonal Type barricade Cone devices large in surface area appear Channelizing be more effective than small ones. drum to 29

34 chevron panel was selected in subjective as recommended channelizing device evaluation 80-ft. (2 two pat those in device 80-ft. (2 th Th (2 th Type I barricade and chevron chevron 80-ft. (12.3 m) and 40-ft. (24.6 m) same results. refore, ded m) spacing is preferred for 4.6 terns. taper 80-ft. (24.6 m) e question to be e m) spacing be ft. (12.3 m) e pattern on more visible. is acceptable replacement for an number of spacing? required would be halved. is recommended over a curved taper r road geometrics. Since se two e spacing and alignment, were per it is not possible to taneously, of influence that each exerts degree channelizing. Note that road to changed; refore, se also be viewed with caution. should re are no distinct differences attributable Since difference in patterns used on a specific to of device (panel or barricade) in general, type may be concluded that effectiveness of a it device is not based primarily on channelizing used. Due to short legibility pattern [93 m])and detection (500-ft. [154-m]) (300-ft. used, it is difficult to measure differences distances driver's responses with respect to patterns. in legibility distance for warning signs used in re are present, zones than on work chevron alignment sign m).(l yellow background) should be with for at least 500 ft. ( i) ctable) legibility distance would be much e hazards exist at work areas protected e However, at distances. legibility distances at lower all seven reviewed. single chevron among appeared to.convey clearest directional arrow message. For panel, spacings yiel each channelizing Although than lower m) 4.6 m) 2.3 is rated spacing for spacing performs spacing or 40-ft. (i patterns, satisfactorily. 80-ft. addressed is will ch devices anneli zing linear tape A th reflecting ta variables, changed simul indicate with respect geometrics conclusions zones is between 400 and 500 ft. (124 and 154 m) work on a legibility distanc97 f ; 50 ft. per inch based (38.7 m/era) of letter height. legibility distance for guide s o}gns is 600 to 800 ft. m to 246 (185 stripe (black visible (dete refore, th less. Becaus by channelizi ng devices, re is a need for greater legibility normal roadway sections. 3O

35 study mecommends a minimum visibility distance One 900 ft. (275 m) for channelizing devices at of (b) decision and mesponse, and (c) me.cognition, changing by drivem. lane re is a need to supplement New Jersey barrier 7. a taper of channelizing devices. Although with should warrant use of supplemental taper. barrier field test results do support warrant for a supplemental taper. As an additional using supplemental taper of diagonal panels note, results of this study do not support a reco mmendation adoption of use of chevron patterns on all for!i zing devices. Except for those rela-ting to Type I channe barricade, conclusions do not clearly and consistently chevron chevron patterns. In chevron design selection, favor chevron patterns were generally slightly favored or rated. to related presently used patterns. responses of equal as measured by position of lane changing were similar drivers two types of patterns. chevron patterns, especially for chevron vertical panel, were preferred in subjective for clear directional message y convey. evaluation although directional information provided by However, devices is desirable, it does not appear necessarf channelizing on reaction of drivers. That presently used patterns based not effective is evidenced by considerable amount of are being conducted on work zone safety and driver infor- research use of a supplemental taper of channelizing devices i. to New Jersey concrete barrier is recommended. prior supplemental taper provides advance warning to transition and reduces number of vehicular lane wheme speed is 55 mph (96.5 kin/h). ( ) locations distance permits (a) detection and This arrow panel, primary channelizing flashing used with barrier, has proven to be device (8), severity of accidents involving effective collision of a vehicle with New Jersey a for regular use at Cedar Creek proved to employed effective in reducing vehicle contact with be New Jersey barrier. RECOMMENDATIONS mation. recommendations of this study are discussed below. collisions with New Jersey barrier. 31

36 use of stripes wider than 4 in. (10.2 cm) on 2. rails less than 3 ft. (0.92 m) long is barricade pending furr study comparing different recommended widths. In addition, a Type I diagonal barricade stripe ft. (0.92 m) long with 6 in. (15.2 cm) stripes should 3 tested i,n field for direct comparisons with be Type I chevron. This preliminary work indicates that 80-ft. (24.6 m) 3. has potential for effectively channelizing spacing traffic with less devices than used in 40-ft. m) spacing. Type I chevron barricade and (12.4 panel are especially promising if an 80-ft. chevron m) spacing is acceptable. A double savings is (24.6 since half as many devices are used in taper achieved a smaller inventory is possible with reversible and patterns in place of diagonal patterns. chevron.tests which compare different taper arrangements Field as spacing of devices (40 ft. [12.3 m] and such ft. [24.6 m]) and taper alignment should be conducted 80 same site with one variable being altered at at time. a legibility and detection distances of all channelizing 4. employed in a taper deserve furr consideration. devices is a problem because legibility distance is re than estimated distance required to perform a less imparted. Moreover, since legibility and being distances are dependent on speed, locations detection This research focused on observation of drivers' 5. Future related research should consider reactions. perception, understanding, and preferences on drivers' equal basis with drivers' reactions. an This research effort focused on one element of 6. traffic control system in work zones. impact total or traffic control elements should be of because drivers' reactions are influenced considered, change maneuver. Consequently, motorists are making lane without discerning directional message maneuvers with lower speed limits should be studied. by total system. 32

37 author expresses sincere appreciation to J. D. Shelor his efforts in collecting field data and to R. N. Robertson for study, especially T. Talley, R. Harmon, J B. Diamond, S. Sheets, R. Hiner, L. C. Taylor II, and B. Cockram. L. particular, thanks go to Harry Craft for report editing; In Jan Kennedy and Susan Kane for typing draft manuscript; to Barbara Turner for typing reproduction masters; to to Baker for graphics; and to Edward Deasy for photographs. Allen research was financed from Highway Planning and funds administered through Federal Highway Research AC KNO W LE DGE ME NT S for his advice given during study. is due all those personnel of Virginia Appreciation of Highways and Transportation who participated in Department services provided by staff of Virginia Finally, and Transportation Research Council are acknowledged. Highway Administration. 33

38

39 on Uniform Traffic Control Devices, Federal Highway Manual U. S. Department of Transportation, 1978 Administration, Inc. prepared f0r"natmonal Cooperative BioTechnOlogy Research Program, Transportation Research Board, Highway D. C., 1979 Washington, S. N., and Nail, R. E., "Evaluation of Barricade Bailey, --Chevron versus Diagonal", State of California Patterns Department of Transportation, March 1977 Work Zone Traffic Control_ Devices", BioTechnology, Inc., of for FHWA, Washington, D. C., July 1978 prepared H. L., "Work Zone Safety" Transportation Research Anderson, :No.: 693, Washington, D. C., 1978 ecord R. J., Harwood, D. W., Graham, J. C., and Glenn, J. Paulsen, of Traffic Safety in Highway Construction Zones" "Status CITED REFERENCES i H. W., Pain, R. F., and Knapp, B. G., Evaluation of McGee,.Co.n_.trols..fo_r.Street and Highway Work one's Traffic Requirements McGee, H. W., and Knapp, B. G., "Visibility Research Record No. 693, Washington, D. C., Transportation 1978 R. N., "Evaluation of High Intensity Sheeting Robertson, Overhead Highway Signs", Virginia Highway and Transportation for Research Council, Charlottesville, Va., December 1974 F. D., "Highway Signing for Safety", Virginia Shepard, Research Council, Charlottesville, Va., June 1971 Highway 35

40

41 Sung C. Cho, Prentice-Hall, Course in Construction and Maintenance Zone Safety "Short Notebook" prepared by Virginia Highway and Course SELECTED REFERENCES Applied Statistics in Science, Introductory Englewood Cliffs, New Jersey, 1978 Inc. T. E., Robertson, H. D., Price, H. E., Alexander, G. J., Post, Lunenfield, H. A User's Guide to Positive Guidance, and Federal Highway Administration, Washington, D. C., June 1977 Research Council, Charlottesville, Va., Transportation 1978 December 37

42

43 APPENDIX A PHOTOGRAPHS OF CHEVRON PATTERN GROUPS

44 Figure A-I. Group #I. Figure A-2. Group # 2. Figure A-3. Group #3. Figure A-4. Group #4.

45 (I) x 24" 8" 4"0, 4"W x 24" 8" x 24" 12" ½" black border Legend black, i" : 2.54 cm (½"B) (2) x 2 " 12" " 0, "W (½"B) (½"B) (½"B) ( A) orange, white, W

46 Figure A-5. Gro up #4A. Figure A-6. Group #5. Figure A-7. Group #6. FigUre A-8. Group # 7.

47 ½vv B Legend black, l" : 2.5 cm. ". orange, white, W 310" X 18" 24" x 12" 24" x 18" orange ½" border I"B (7) x 18" 2 " i 3 iq-" B- I"B

48

49 APPENDIX B TRAFFIC CONTROL SCHEMES FOR FIELD TESTS

50 .jp----i,,.,i 533' UNDER ROAD CONSTRUCTION B-I. Pilot test site U. Figure County. Channelizing devices Taper length I000 ft. Road signs are Note: on both used Guardrail LIMIT 533' i000' lo00' SPEED LIMIT LEGEND AND NOTES 45 Traffic counter tube I foot spacing 1 foot.308 meter SPEED i AHE_AO lo00' sides of road. S. Rte. 460, Giles B-2

51 UNDER ROAD CONSTRUCTION Taper length 560 ft.! foot.308 meter Road signs are Note on both used Figure B-2. U.S. Rte. 17 Mt. Landing Creek site. KEEP LEFT KEEP LEFT 350' 350' LEGEND AND NOTES Channelizing devices 3 0' Traffic counter tube 40 foot spacing 350' sides of road. 250'

52 m m m m REDUCED SPEED flashing or Illuminated amber arrow sequential Taper length 560 ft. Degree of curvature 1 Road signs are Note on both used New Jersey barrier 350' 1200' LEGEND AND NOTES m Channelizing devices 2000' = Traffic counter tube 40 foot spacing 1 foot..308 meter Grade -4.0% sides of road. 1550' _1.,,,,I Figure B-3. U.S. Interstate I-8! South Cedar Creek site.

53 SPEED LIMIT flashing or Illuminated amber arrow sequential Taper length 560 ft. Degree of curvature.417 Road signs are Note: on both used m m New Jersey barrier m 3SO' 1200' AHEAD LEGEND AND NOTES m Channe lizing devices 2000' = Traffic counter tube 80 foot spacing 1 foot.308 meter Grade +0.5% sides of road. 1550' B-4. U.S. Interstate 1-81 South Narrow Passage Figure site. Creek

54

55 APPENDIX C FIELD TEST PHOTOGRAPHS

56 C i. Pilot test. Figure Type III a) barricade. diagonal vertical panel. chevron b) Oversized

57 C-2. Test Figure a) Type I chevron barricade 2115 #I. Cone b)

58 C-3. Test #2. Figure Type II diagonal barricade. a) b) Type II chevron barricade.

59 C-4. Test #3. Figure Drum vertical panel. a) b) Chevron vertical panel.

60 C-5. Test #4. Figure Diagonal vertical panel. a) 2118 b) Chevron vertical panel.

61 FIELD TEST SUBJECTIVE EVALUATION RESULTS APPENDIX D

62 is a comparative evaluation between presen and ch evron This for channei izing evices used to Form tapers. channe.izing pa : rns under consideration ar. displayed in Figure.. An evaluation Form devices available for each st. Some general questions i=o.low comparative is to amount of time r quir d at test sit. to.valuate Due of channelizing devic. s, emphasis will be placed on observing th, e all patterns. However, observation of present patterns is also chevron If you are unable to observe he present pattern but you do desirable. additional.space is necessary for answering questions, use If of page. You may wish to make a note of your observations rar back rely on your memory since a day or mor may pass between set up than present and proposed patterns. Two blank pages at end are of OF CHEVRON PATTERNS EVALUATION USE ON CHANNELIZING DEVICES FOR Instructions eva ua i on. t,he chevron pattern l or a given test, complete questions based observe your familiarity with present pattern. If you are unable o observe on he chevron pattern, n ignore questions i or that est. available for this. Please return to Ben Cottr ll. Name Summary of responses.

63 If he 2. reason of motorists and protect:ion workers 3,, day night of (day time nigh:) observation changes vs. your preferences, iscuss s. hi for chevron!- cone in daylight I- 3 observed 1 not observed ObServed i _no : observed IchecX ao,oropriat space.s) questions!. Which of se pa ::erns do you prefer with r s eci: to" Day Nigh: Chevron Pr?sen Presen: KShevron i-n guiding motorist effectiveness of design wi:h or uniformity i 2 control evices traffic decision sigh di s:ance adequate i i which pa:: =rn in :. s: do you,or, er and why?

64 i _nigh: 3_day =he appropria=a spac.=s) (check U EST O, IS of =,he mo:oris=s and pro=ec:ion workers time of observa=ion If for =his. r=.as n i 2 I 2 i 3 observed 1 no= observed o served no observed day 2 ni ght:. ' hich of se pa== :rns do you prefer with r=.spec: Oay Chevron Both Presen= Presen= Chevrun C in guiding.he,too:otis= effec=iveness of design wi h o=her uniformity 2 I 2 con=rol devices :r:.ffic decision sigh= distanc.= acequat.= 2 (day vs. changes your pr=.fer nc s, ; Overal R, rn,n :his 3- chevron