«FIELD MEASUREMENTS OF DIRECT A N D REARVIEW-MIRROR GLARE FROM LOW-BEAM HEADLAMPS Michael Sivak Michael J. Flannagan Brandon Schoettle Yoshihiro ...»
OF DIRECT A N D
FROM LOW-BEAM HEADLAMPS
Michael J. Flannagan
FIELD MEASUREMENTS OF DIRECT AND REARVIEW-MIRROR GLARE
FROM LOW-BEAM HEADLAMPSMichael Sivak Michael J. Flannagan Brandon Schoettle Yoshihiro Nakata The University of Michigan Transportation Research Institute Ann Arbor, Michigan 48109-2150 U.S.A.
Report No. UMTRI-2001-3 January 2001 Technical Report Documentation Page
1. Report No. 2. Government Accession No. 3. Recipient’s Catalog No.
4. Title and Subtitle 5. Report Date Field Measurements of Direct and Rearview-Mirror Glare January 2001 from Low-Beam Headlamps 6. Performing Organization Code 302753
7. Author(s) 8. Performing Organization Report No.
Sivak, M., Flannagan, M.J., Schoettle, B., and Nakata, Y. UMTRI-2001-3
9. Performing Organization Name and Address 10. Work Unit no. (TRAIS) The University of Michigan 11. Contract or Grant No.
Transportation Research Institute 2901 Baxter Road Ann Arbor, Michigan 48109-2150 U.S.A.
12. Sponsoring Agency Name and Address 13. Type of Report and Period Covered The University of Michigan 14. Sponsoring Agency Code Industry Affiliation Program for Human Factors in Transportation Safety
15. Supplementary Notes The Affiliation Program currently includes Adac Plastics, AGC America, Automotive Lighting, BMW, Corning, DaimlerChrysler, Denso, Donnelly, Federal-Mogul Lighting Products, Fiat, Ford, GE, Gentex, GM NAO Safety Center, Guardian Industries, Guide Corporation, Hella, Ichikoh Industries, Koito Manufacturing, Libbey-Owens-Ford, LumiLeds, Magna International, Meridian Automotive Systems, North American Lighting, OSRAM Sylvania, Pennzoil-Quaker State, Philips Lighting, PPG Industries, Reflexite, Renault, Schefenacker International, Stanley Electric, Stimsonite, TEXTRON Automotive, Valeo, Vidrio Plano, Visteon, Yorka, 3M Personal Safety Products, and 3M Traffic Control Materials.
Information about the Affiliation Program is available at: http://www.umich.edu/~industry
16. Abstract This study measured direct and rearview-mirror glare illuminances produced by low-beam headlamps in a sample of 22 passenger vehicles. The glare illuminances were measured for 12 common glare situations that were defined by a full factorial combination of three scenarios (oncoming driver, center rearview mirror of a preceding driver, or driver-side mirror of a preceding driver one lane to the right), two longitudinal distances (25 m or 50 m), and two vertical locations (glared vehicle being either a car or a light truck/van/SUV). The measurements were made outdoors at night on asphalt pavement.
The median illuminances ranged from 0.5 lux for an oncoming driver of a light truck/van/SUV at a distance of 50 m, to 3.4 lux at the driver-side mirror of a preceding car at 25 m one lane to the right. (These values do not take into account window transmittance or mirror reflectance.) The ratios of the maxima and the minima measured for each of the 12 glare situations were large, ranging from about 5:1 to 36:1.
The median actual illuminances were compared to the median expected illuminances based on a recent, laboratory-measured, representative sample of U.S. low-beam patterns, taking into account the possible effects of dirt, voltage, misaim, and pavement reflectance.
This analysis indicates that the actual illuminances could be very well modeled using the laboratory-measured beam patterns and assuming a linear relationship between the light output of clean and dirty headlamps. Additional analyses evaluated the relationships between headlamp mounting height and glare illuminance.
17. Key Words 18. Distribution Statement glare, oncoming traffic, rearview mirrors, cars, light trucks, vans, SUVs, Unlimited low beams, passing beams, lamp mounting height, field measurements
19. Security Classification (of this report) 20. Security Classification (of this page) 21. No. of Pages 22. Price None None 20
Two recent studies provide market-weighted information about low-beam headlighting patterns in the U.S. and Europe (Sivak, Flannagan, Kojima, and Traube, 1997; Sivak, Flannagan, and Schoettle, 2000). The photometric data in these two reports are in the form of detailed candela matrixes. From this information, it is possible to calculate the amount of light that would be expected to be directed to a particular point in space, such as the eyes of an oncoming driver or the eyes of a preceding driver via a rearview mirror. To achieve this would first require calculating, for the particular point of interest, the horizontal and vertical angles with respect to each headlamp location. Next, luminous intensity would be looked up in the respective candela matrix for the calculated angles. Finally, the sum of the two candela values (one for each headlamp), divided by the square of the distance, would provide an estimate of the illuminance reaching the point of interest.
However, the available photometric data are based on laboratory measurements for new and clean headlamps that are correctly aimed and energized at a controlled voltage level. Consequently, the calculations described above would not take into account several important factors that influence headlamp illumination, such as lamp voltage (Ammerlaan and Vellekoop, 1996; Silva, 1998; Sivak, Flannagan, Traube, and Miyokawa, 1998), lens dirt (Cox, 1968; Rumar, 1974; Padmos and Alferdinck, 1988), misaim (Padmos and Alferdinck, 1988; Sivak, Flannagan, and Miyokawa, 1999a), and pavement reflectance (Jackett and Fisher, 1974; Sabey, 1972). It is possible to correct the originally calculated values by using estimated effects of the intervening factors. Another approach would involve obtaining measurements under actual field conditions, and that is the approach taken in this study.
Specifically, this study was designed to obtain a set of field glare illuminance readings (representing the glare experienced by oncoming drivers and the glare experienced by preceding drivers via rearview mirrors), and to compare these values with expected illuminances based on laboratory photometric data.
Experimental setup Measurements were made in an asphalt-paved parking area near the UMTRI building. The experimental setup was designed to represent 12 common glare situations that were defined by a full factorial combination of 3 lateral locations, 2 longitudinal locations, and 2 vertical locations (see Figure 1).
Lateral locations. There were 3 lateral locations, representing vehicles in 3
different lanes of traffic:
(1) Direct glare for an oncoming driver in the left adjacent lane.
(2) Indirect glare via inside, center mirror for a preceding driver in the same lane.
(3) Indirect glare via outside, driver-side mirror for a preceding driver in the right adjacent lane.
Longitudinal locations. Two distances were used, representing vehicles separated by 25 m and 50 m.
Vertical locations. There were two heights above the pavement, representing two types of glared vehicles (passenger cars and light trucks/vans/SUVs).
Table 1 lists the spatial coordinates of all 12 test locations. These coordinates are based on the data from Sivak, Flannagan, Budnik, Flannagan, and Kojima (1996) for the locations of driver eyes; Reed, Lehto, and Flannagan (2000) for the locations of rearview mirrors on cars; and Reed, Ebert, and Flannagan (2001) for the locations of rearview mirrors on light trucks, vans, and SUVs.
The lateral coordinates differ slightly between the two classes of vehicles for both driver eye positions (a difference of 0.07 m) and driver-side mirrors (a difference of
0.12 m). Because small changes in horizontal angles have only minor effects on the light output, these differences were disregarded and in each case were averaged to derive the common lateral coordinates for both types of vehicles.
Figure 1. A schematic diagram of the experimental setup.
For each of the six positions shown, measurements were taken at two different heights above the pavement, for a total of 12 measurements. (See text for details.)
The photometric measurements were taken at least 1 hour after sunset. It took about 30 minutes to take the 12 measurements for each vehicle.
Each vehicle was positioned by the volunteer subject, with the assistance of two experimenters outside of the vehicle. The vehicle was centered within a lane 3.66 m wide, with the headlamps at the baseline longitudinal distance (0 m).
The driver was instructed to turn on the low-beam headlamps, leave the engine running for the duration of the measurements, and remain in the vehicle. At the time of recruitment, the drivers were asked not to make any adjustments to their headlamps (such as cleaning, aiming, or bulb replacement) just because they were participating in this study.
Before the photometric measurements were taken, the headlamp type and mounting locations were recorded.
The photometric measurements were then recorded using a tripod-mounted illuminance meter (Minolta T-1). The tripod was calibrated to allow for vertical height adjustments as needed. Because the illuminance meter was not inside a vehicle, the measured illuminance values do not take into account window transmittance or mirror reflectance.
The existing fixed lighting in the vicinity of the experimental setup was turned off during the measurements. Ambient light levels were recorded several times during each session. They averaged 0.14 lux. The average ambient light levels for each experimental session were subtracted from the recorded measurements in that session to obtain the actual illuminance values.
The sample for this study consisted of 22 vehicles owned by UMTRI employees or UMTRI. The sample included 16 passenger cars (73%) and 6 light trucks, vans, and SUVs (27%). The model years of the vehicles ranged from 1989 to 2000 (see Table 2).
The sample included 15 vehicles with two-lamp systems (68%) and 7 vehicles with fourlamp systems (32%). A breakdown by the optical design of the lamps is shown in Table 3, and a breakdown by bulb type is shown in Table 4. The median headlamp mounting height (center to ground) was 0.64 m, and the median headlamp separation (center to center) was
The photometric readings are summarized in Tables 5 and 6. Table 5 lists the median illuminances for the 12 conditions of interest, while Table 6 provides the ratios between the maximum and minimum illuminances.
For each of the 22 test vehicles, we calculated the expected illuminances at each of the 12 points. These calculations took into account the actual mounting positions of the two lamps on each individual vehicle, and the corresponding laboratory photometric data for the respective vehicle class in Sivak et al. (1997). The median expected illuminances are shown in Table 7. The median actual illuminances (from Table 5) as percentages of the expected illuminances (from Table 7) are listed in Table 8. We will return to the patterns in Table 8 after we discuss the effects of dirt, voltage, misaim, and pavement reflectance.
Dirt deposits on headlamp lenses have two major effects: a reduction in the total amount of emitted light and an increase in scattered light. Sivak, Flannagan, Traube, Kojima, and Aoki (1996) have shown that the relation between “dirty” and “clean” luminous intensities is well described by a linear function y = ax + b, where y is the “dirty” luminous intensity, x is the “clean” luminous intensity, a, the slope ( 1) specific to the dirt accumulation in question, is an estimate of the proportional reduction in luminous intensity throughout the beam pattern caused by both absorption and scattering, and b, the intercept specific to the dirt accumulation in question, is an estimate of the amount of the superimposed luminous intensity caused by scattering.
The net result of these effects is to increase intensities at points in the beam pattern that have relatively low intensity when the lamp is clean, and to decrease intensities at points that have relatively high intensity when the lamp is clean (Sivak, Flannagan, Traube, Kojima, and Aoki, 1996).
Applying those findings to the present data leads to a prediction that the expected glare illuminances (based on measurements with clean headlamps) involving points in the beam pattern that are relatively weak should underestimate the actual illuminances.
Conversely, the expected illuminances involving points in the beam pattern that are relatively strong should overestimate the actual illuminances.
To test this prediction, we calculated the luminous intensities that the lamps needed to emit to produce the median actual glare illuminances in Table 5 and the median expected illuminances in Table 7. (These calculations assumed equal contributions from the two lamps.) These two sets of luminous intensities are shown in Table 9. Consistent with the prediction, the expected luminous intensities that were less than 1,200 cd tended to underestimate the actual intensities, while the expected luminous intensities that were more than 1,200 cd tended to overestimate the actual intensities. (There was only one exception to this general pattern.) To describe this relation formally, we regressed the actual intensities on the expected intensities (both from Table 9). The results (see Figure 2) are, again, consistent with the findings of Sivak, Flannagan, Traube, Kojima, and Aoki (1996).