Proposal for Standard U.S. Headlamp Beam Pattern for Evaluation of Retroreflection

Transcription

1 TRASPORTATO RESEARCH RECORD Proposal for Standard U.S. Headlamp Beam Pattern for Evaluation of Retrorefletion THEODORE J. SZCZECH AD SUSA T. CHRYSLER Confusion arises among researhers and designers hen different headlamp output data files are used to produe luminane profiles of retrorefletive sheeting. t is proposed that a standard U.S. headlamp data file be established to redue onfusion and improve ommuniation among researhers in the area of retrorefletive materials. A omposite 15th perentile headlamp data file for alulating luminane values for hih 85 perent of sampled headlamps produe equal or greater luminane is offered. The alulation of luminane depends on headlamp performane and retrorefletor harateristis. Use of this 15th perentile headlamp is justified by omparison of luminane values alulated for enlosed lens sheeting for various headlamp data files. This referene headlamp is intended as a standard for retrorefletive material researh. t is not intended to be a design target for the development or testing of ne headlamps. The headlamp data file is available on request. Retrorefletive materials in traffi ontrol devies, pavement markings, and personal safety lothing provide drivers ith information about the roaday environment. These materials return light from a vehile's headlamps to the driver's eyes in a speifi diretion. Several researhers have used luminane as a tool to assess sign performane and pavement marking and pedestrian visibility (1-3). Luminane measurements may also be used to ompare headlamp performane aross a set of lamps by using a onstant retrorefleti ve target. Luminane profiles allo a glimpse of the dynami performane of retrorefletive materials by displaying the amount of light reahing a driver at different distanes. The alulation of luminane depends on retrorefletor harateristis and headlamp performane. When different headlamp data files are used in luminane alulations by different laboratories, it is diffiult for researhers to ompare luminane profiles of various retrorefletive materials. We propose that a standard data file for U.S. headlamps be established to redue this onfusion and improve ommuniation among researhers in this area. We offer a omposite 15th perentile headlamp data file for alulating luminane values for hih 85 perent of sampled headlamps produe equal or greater luminane. n their reommendations for minimum sign retrorefletivity, Paniati and ae (3) use a omposite headlamp. This omposite is based on the median value of a sample of ne headlamps. As suh only 5 perent of the sampled headlamps ill give alulated luminane values of equal or greater amplitude. f the luminane distribution as a funtion of headlamp type is suffiiently narro, this hoie is valid. f the luminane spread is greater than 2 to 1 hoever, Jenkins and Gennaoui (2) propose a better engineering hoie. They suggest a headlamp for hih more than 8 perent of sampled vehiles ould produe equal or greater luminanes. n addition to this problem of variability in headlamps, many vehiles on Optis Researh Department, Traffi Control aterials Division, 3 Company, 3 Center , St. Paul, inn the road are older and may have dirty, misaligned, or missing headlamps. These anomalous situations are not onsidered here. The data given here are further idealized in that no atmospheri or indshield loss as inluded in the analysis and all headlamps analyzed ere ne. The use of a 15th perentile instead of a 5th perentile headlamp gives a minimum realisti safety fator in attempting to aommodate all vehiles on the road. ETHODOLOGY We used a ombination of omputer modeling and atual photometri measurements to evaluate different individual and omposite headlamps. Our omputer model ( 4) alulates geometry for left and right headlights for a speified roaday, sign position, and range of distanes. The model takes as input headlamp information in the form of andela values at 1, 71 measurement points. Data onerning headlamp plaement on the vehile, driver eye position ithin the vehile, sign material type, and sign position on the road are also inluded. The model outputs the presentation, observation, entrane, and orientation angles for eah headlamp for a speified set of distanes. The amount of illumination (in lux) reahing the sign from eah headlamp at eah distane is also provided. nput Variables Signing aterial To simplify the analysis a single sheeting material as studied. The material, hite enlosed lens retrorefletive sheeting, had a oeffiient of retrorefletion at a.2-degree observation angle and a -4-degree entrane angle of 112 d/lx/m 2 As suh this material is representative of a typial ne sheeting sample. This Type sheeting is ommonly used internationally for all types of signing. Roaday, Vehile Type, and Position of Signs Roaday length is defined as the enterline distane from the vieer to the target sign, here the vieer's eye position is alulated as a single point at the bridge of the nose. A straight roaday of 457 m (1,5 ft) long ith a 3.7-m (12-ft) lane idth as hosen. This represents a ommon situation and provides for analysis over a ide range of vieing angles. A Ford Taurus ith a driver vieer approahes the sign positions don the enter of this lane. The headlamp and vieer positions are speified in the x, y, and z oordinates ith a point of origin on the ground belo the front enter of the ar. The headlamps are.61 m to the left or right of

2 92 TRASPORTATO RESEARCH RECORD 1456 CETERED OVER LAE ST PAUL 1 T 2' (6.1m) T m 8' (2.4m) 8' (2.4m) 1.1m 3.7m 3.7m 3.m 1.1m (3.5') (12') (12') (1') (3.5') LEFT RGHT PARKG enter and.68 m off the ground. The pith and roll points of the vehile are 2.2 m bak and.68 m up, respetively. The vieer's eye is positioned 2.8 m behind the front,.41 m to the left, and 1.16 m above the ground. Luminane values ere alulated for the geometri enter of right shoulder, overhead, and left shoulder sign positions. Figure 1 speifies the positions of the three signs. The exat loations of the signs are not ritial to the analysis. These three sign positions give a very ide luminane range. Headlamps Reently, the University of ihigan Transportation Researh nstitute (UTR) released headlamp output data for a sample onsisting of 43 ne U.S. headlamps (5). Tenty-six of these headlamps ere measured for an earlier HTSA study. Eah file ontains 1,71 test points that result from a 21 X 81 matrix of andela values ranging from - 2 to + 2 horizontal degrees and - 5 to + 5 vertial degrees at.5-degree intervals. For the UTR data, andela values varied as muh as 53 to 1 at some test points. Also, at most test points the distribution of values aross headlamps as not Gaussian. We analyzed 42 of these headlamp files. These headlamps are identified by an arbitrary number; all manufaturer-speifi information remains anonymous. n addition, three omposite headlamps ere inluded in the analysis: the median (5th perentile) of the UTR data, the 15th perentile of the UTR data, and the median from the HTSA data set of 26 headlamps. This latter omposite is alled CARTS5 beause it is used in the CARTS model of sign visibility (3). The 15th perentile omposite headlamp files ere derived by using the statistially derived perentile for eah of the available 1,71 test points. The natural logarithm of the andela values as taken, and test points orresponding to those in 1579 DEC84 of SAE (6) ere tested for normality. The onversion to logarithmi values suessfully produed Gaussian distributions at these test points. The 15th perentile headlamp file as statistially derived from the log(andela) distributions by taking the mean value at eah test point and adding 1.4 standard deviation units (7). These log values ere then onverted bak to ra andela values by taking the antilog of the statistially derived values. This omposite headlamp meets the 1579 DEC84 test point speifiations of SAE (6) as shon in Table 1. TABLE 1 SAE J579 Photometri Test Points from HS-34 and (Proposed) 15th Perentile Headlamp Value Degrees Vertial 1. up.5 up.5 up 1.5 up.5 up.5 don 1. don 1.5 don 1.5 don 1.5 don 2. don 2. don 4. don Horizontal 1.5 left 1.5 left 1.5 left 1. right 1. right 1.5 right 6. left 2. right 9. left 9. right 15. left 15. right 4. right Candelas aximum inimum Proposed Headlamp ote: A tolerane of + or.25 deg is alloed at any test point. Luminane Calulation On the basis of these input variables, intrinsi photometri angle sets ere alulated for the left and right headlamps for vehile to sign distanes ranging from 3- to 457-m (1- to 1,5-ft) intervals. The omputed photometri angles ere dupliated on a photometri range here the enlosed lens sheeting as measured at eah angle set to determine the retroreftetane (RA) for eah vieing distane. By using the headlamp data (d), illumination (x) at the sign enter at eah distane as alulated separately for the left L u A 1 E D s Q DSTACE (ETERS) FGURE 2 Luminane profiles for 42 UTR headlamps for right shoulder signs.

3 L u 1 A E D s a ---- UTR EDA z a: a.. > TH PERCETLE -.- HEADLAP >-- HEADLAP LUACE VALUES FOR LAPS OF TEREST: HEADLAP 18 = TH PERCETLE = 4.6 UTR EDA = 9.6 CARTS5 = DSTACE (ETERS) LUACE (CD/SQ) FGURE 3 Luminane profiles for loest, highest, and omposite headlamps for right shoulder signs. Headlamp 18 mathes losely the performane of the 15th perentile omposite. 1 FGURE 5 Cumulative perentile luminane distribution for 42 headlamps for right shoulder signs at a vieing distane of 183 m (6 ft). Luminane values for ritial lamps are provided for referene z a: a.. > LUACE VALUES FOR LAPS OF TEREST: HEADLAP 18 = TH PERCETLE = 5.3 UTR EDA = 8.7 CARTS5 = 9.1 z a: a.. > LUACE VALUES FOR LAPS OF TEREST: HEADLAP 18 = 2. 15TH PERCETLE = 2.2 UTR EDA = 4.3 CARTS5 = l-'-+---l l---l---l---l---l---l LUACE (CD/SQ) FGURE 4 Cumulative perentile luminane distribution for 42 headlamps for right shoulder signs at a vieing distane of 61 m (2 ft). Luminane values for ritial lamps are provided for referene. ote the 15th perentile value of 5.3 d/m 2 falls on the 15 perent value LUACE (CD/SOl FGURE 6 Cumulative perentile luminane distribution for 42 headlamps for right shoulder signs at a vieing distane of 365 m (1,2 ft). Luminane values for ritial lamps are provided for referene.

4 94 TRASPORTATO RESEARCH RECORD 1456 and right headlamps. The measured RA (d/lx/m 2 ) values ere then multiplied by the alulated illumination (lx) to give luminane values ( d/m 2 ) for the left and right headlamps. Finally, the luminane values for the left and right headlamps ere summed to give the total luminane at the sign enter. RESULTS AD AALYSS Luminane Curves Figure 2 shos the luminane urves for all 42 UTR headlamps for the right shoulder sign. The spread in performane is surprising. Figure 3 shos the luminane urves for six seleted headlamps. These six headlamps ere seleted for omparison beause they represented the minimum (Headlamp 34), maximum (Headlamp 11), the three omposite beams, and Headlamp 18, hih produed luminane similar to that of the 15th perentile headlamp. ote that the CARTS5 and UTR median headlamps are very lose to eah other in performane. Figures 4, 5, and 6 sho the luminane distributions plotted as umulative perentiles at 61, 183, and 365 m (2, 6, and 1,2 ft), respetively. These figures sho that at these important vieing distanes the 15th perentile headlamp gives a luminane value for hih approximately 85 perent of the headlamps have an equal or greater value. Data for the overhead and left shoulder signs are presented in an abbreviated form to save spae. The overall spread of the 42 headlamps and their umulative luminane distributions for overhead and left shoulder signs ere similar to those for the right shoulder sign. Figure 7 shos the luminane urves of the six seleted headlamps for an overhead sign. Comparing the plot in Figure 7 ith that in Figure 3 shos that the overhead sign appears approximately five to six times dimmer to the driver than the right shoulder sign. Figure 8 shos the plot for the six seleted headlamps for a left shoulder sign. From Figure 8 the brightness of the left shoulder sign is seen to be omparable to that of the overhead sign, although the distribution idth of these urves appears to be slightly narroer. Headlamp soandela Plots Figures 9 to 14 sho isoandela plots for the six seleted headlamps for hih luminane urves ere given in Figures 3, 7, and 8. n these figures the horizontal sale differs from the vertial sale by a fator of about 2 to make the figures more easily readable. Also on these plots the lines sho the angular loation for the right (R) and left (L) headlamps for the three sign loations. The overhead sign loators terminate at 76 m (25 ft) sine these lines ould extend beyond the figure's border. The peak andela value of the data set is given along the right side of the plot. The CARTS5 and UTR (Figures 9 and 1, respetively) isoandela plots are nearly identi HEADLAP UTR EDA --- CARTS5 15TH PERCETLE L..._ HEADLAP 18 u 1 -->--- HEADLAP 34..,,...-- A ==e-... E D s.j>... - """"" """- - Q _ '»--->--- L u 1 A E D s Q HEADLAP UTR EDA --- CARTS5 15TH PERCETLE..._ HEADLAP 18 -->--- HEADLAP ' ' DSTACE (ETERS) FGURE 7 Overhead sign luminane profiles for loest, highest, and omposite headlamps. Headlamp 18 mathes losely the performane of the 15th perentile omposite DSTACE (ETERS) FGURE 8 Left shoulder sign luminane profiles for loest, highest, and omposite headlamps. Headlamp 18 mathes losely the performane of the 15th perentile omposite.

5 Szzeh and Chrysler (/) 1. Q._,. E::: -1. P:: > ti) g -1. > -2. p LL...::::L::r::::t,,_i L:l::::,,J.---L..LJD:::::r:::t:::::1::D.:b...L..L-L...L::::::l ::::t::=:L:...bdt:::::t:::C:L!_i..:::t::JC -5. HORZOTAL (DEGREES) FGURE 9 Headlamp isoandela plot for CARTSSO headlamp. al, hih is not surprising onsidering that the CARTS5 data are a subset of the neer UTR data. Headlamp 11 (Figure 11) provides high sign luminane beause it has a large peak value and the light distribution is quite broad. Headlamp 34 (Figure 12) on the other hand has a loer peak value. ore important, this headlamp has an extremely sharp utoff leading to lo sign illumination. The 15th perentile headlamp and Headlamp 18 (Figures 13 and 14, respetively) have isoandela plots that are learly different. Yet the luminane urves alulated ith eah headlamp are quite similar. This disrepany highlights the importane of examining luminane urves to evaluate headlamps instead of relying solely on isoandela plots. Different intensity patterns an deliver very similar luminanes to the driver. COCLUSOS t is diffiult to kno preisely ho the headlamp population examined in the present study relates to the headlamp population of vehiles on U.S. highays. All ne U.S. headlamps presumably meet SAE speifiations. Furthermore, it is reasonable to assume that manufaturers of headlamps ill not manufature a ertain headlamp type unless the sales of that headlamp are large enough to be profitable. This means that one ould expet any of the headlamps investigated in the present study to represent those used on a signifiant population of vehiles on the road. An important aspet of this investigation is the finding that the performane range of U.S. headlamps is surprisingly broad, so that (/) P:: 1. ; Q._,....l -1. P:: > l -1. > -2. p HORZOTAL (DEGREES) FGURE 1 Headlamp isoandela plot for UTR median headlamp.

6 96 TRASPORTATO RESEARCH RECORD Ci) 2. r- 1. t'l 1. e > > p HORZOTAL (DEGREES) FGURE 11 luminane. Headlamp isoandela plot for Headlamp 11, hih gave the highest sign Ci) 2. r- t'l e > > p HORZOTAL (DEGREES) FGURE 12 luminane. Headlamp isoandela plot for Headlamp 34, hih gave the loest sign luminane values alulated ith a median headlamp file an be 5 to 6 times greater than those of the poorest performing headlamps. This is a ause for onern. The 15th perentile headlamp on the other hand gives luminane values roughly tie those of the poorest-performing headlamp. Finally, hen assessing the performanes of signs and other retrorefletive materials from a moving vehile, it is important to kno the harateristis of the headlamps used. deally, one ould measure the isoandela harateristis of the headlamp and rank the performane of the hosen headlamp by alulating luminane and omparing it ith the plots given in this paper. A headlamp set meeting the 15th perentile riterion as defined here ould be our preferene for most researh needs. Headlamps at either extreme of the performane range should be avoided. Furthermore, e believe that use of a 5th perentile headlamp signifiantly overestimates the luminane reeived by the majority of drivers. We propose that researhers interested in visibility issues adopt a standard headlamp data file. Those ishing to verify our results

7 Szzeh and Chrysler Ul D o.jo aj t!> e.. :...:i > ::.: > p HORZOTAL (DEGREES) FGURE 13 Headlamp isoandela plot for the 15th perentile headlamp. ill find the photometri angles defined iri Figure 1 of Federal Test ethod Standard 37 (8). Table 2 provides the headlamp oordinates for a right shoulder sign. The data presented here sho that the 15th perentile omposite performs nearly identially to an atual headlamp. The omposite as reated from a broad range of headlamps on the road today. For this reason the shape of the isoandela plot is smoother and overall more representative of all headlamp types than a single, atual headlamp. The 15th perentile headlamp provides an adequate safety fator above th median omposite typially used. For these reasons e suggest that this 15th perentile omposite be onsidered a standard for retrorefletive material researh. t is not, hoever, implied to be a target speifiation for ne headlamp designers. ACKOWLEDGET We thank Doug ae and ihael Sivak for their assistane in providing headlamp data. We also thank Efren Dizon and Trent Kay for providing assistane ith data aquisition and preparation of this paper Ul 1. t!> i:::i....;i -1. > D D 1.. :...:i -1. > ::.: l'zl -2. p HORZOTAL (DEGREES) FGURE 14 Headlamp isoandela plot for Headlamp 18, hih gave a luminane performane similar to that of the 15th perentile headlamp.

8 98 TRASPORTATO RESEARCH RECORD 1456 TABLE2 Positions and Angles for Vieer and Right Shoulder Sign Vieer Distane Lgfl Qr Right Degrees eters Feet Lo beam Observation Entrane Headlamp An9le Angle left right left right left right left right left right left right left right left right left right left right left right left right left right left right left right Presentation Orientation Headlamp Headlamp AnQle AnQle Horizontal Vertial : REFERECES 1. Woltman, H. L., and T. J. Szzeh. Sign Luminane as a ethodology for athing Driver eeds, Roaday Variables, and Signing aterials. n Transportation Researh Reord 1213, TRB, ational Researh Counil, Washington, D.C., 1989, pp Jenkins, S. E., and F. R. Gennaoui. Terminal Values of Road Traffi Signs. Speial Report o. 49. Australian Road Researh Board Ltd., Vermount South, Vitoria, Australia, Paniati, J. F., and D. J. ae. inimum Retrorejf.etivity Requirements for Traffi Signs. Report FHWA-RD FHWA, U.S. Department of Transportation, Egan, J.C. User Referene anual for ightsee: An nterative Program for the Simulation and Analysis of ighttime Driving Report. UTR Transportation Researh nstitute, University of ihigan, Ann Arbor, Sivak,.,. J. Flanagan, and T. Sato. Light Output of U.S., European, and Japanese Lo-Beam Headlamps. Report UTR Transportation Researh nstitute, University of ihigan, Ann Arbor, SAE Ground Vehile Lighting anual. HS-34. J579 DEC84. SAE, Warrendale, Pa., p Hays, W. L. Statistis, 4th ed. Holt, Rinehart, and Winston, n., e York,.Y., Federal Test ethod Standard 37, arh 1, nstrumental Photometri easurements of Retroreftetive aterials and Retrorefletive Devies. Federal Supply Servie, General Servies Administration, Washington, D.C., Publiation of this paper sponsored by Committee on Visibility.