U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA RD-03-081
Date: June 2003
Updated Minimum Retroreflectivity Levels
PDF Version (837 KB)
PDF files can be viewed with the Acrobat® Reader®
Chapter 3. Updated Factors
Several retroreflectivity-dependent factors were updated in this most recent research effort. These factors included the vehicle headlamps, the vehicle type (or size), the method used to predict retroreflective sheeting performance, the number of materials considered, and the driver's age and assumed accommodation level. This chapter summarizes the work that was completed to update these factors.
In 2001, Carlson and Hawkins completed a sensitivity analysis of the then-available headlamp isocandela profiles. (46) They used typical guide sign and street-name sign placements to compare sign illuminance values at various distances. The results showed that headlamp output directed toward overhead signs decreased by about 30 to 40 percent between the mid-1980s and mid- 1990s. For street-name signs, the drop was not as severe but still substantial at 20 to 30 percent.
Carlson and Hawkins also benchmarked seven headlamp profiles against field illuminance measurements recorded in the mid 1990s along a flat tangent section of rural interstate in Kansas.(51) The results showed that the headlamp profile representing vehicles from the mid- 1980s with halogen sealed-beam headlamps had the best correlation to the field measurements. Keeping in mind that the average age of U.S. vehicles is 9 years(52) and that the field measurements were recorded before the introduction of VOA headlamps in 1997, the results provide the earliest validation of the assumed correlation between the illuminance results of modeling headlamp isocandela data and actual field illuminance measurements. Chrysler et al. have recently provided additional confirmation of this modeling assumption.(53)
After Carlson and Hawkins completed their analyses, University of Michigan Transportation Research Institute (UMTRI) published headlamp isocandela data representing model year 2000 vehicles. (54) The UMTRI 2000 headlamp isocandela profiles were the first available that included a sample from VOA headlamp types. These headlamp types were studied by UMTRI and shown to produce even less light for nighttime sign visibility.(55) For example, compared to the conventional U.S. headlamps of the mid-1990s, the VOA headlamp (which generically describes two subclasses: VOR and VOL) reduces overhead illumination by 28 percent (VOL headlamp) and 18 percent (VOR). (The VOL headlamp is a low beam with a horizontal cutoff to the left side of the beam. The VOR has a horizontal cutoff to the right side of the beam. The VOL can reduce glare to oncoming drivers compared to conventional U.S. low beams. VOR headlamps have less ability to reduce oncoming glare but produce isocandela profiles more similar to conventional U.S. low beams.)
More recently, a newer style of headlamp has entered the U.S. market and its popularity is slowly growing. These headlamps, termed HID for High Intensity Discharge, use an arc capsule where an arc jumps between two electrodes. This arc is used as the light source, instead of the glowing filament in a conventional halogen headlamp. UMTRI's latest headlamp profile representing model-year 2000 vehicles does not include representation from HID headlamps. Therefore, as part of an earlier effort related to this project, the researchers purchased 6 HID headlamp profiles. The data from the 6 individual profiles were averaged into a composite HID profile, which was compared to various other headlamp profiles, including U.S. headlamps from the mid-1980s to 2000, and a European headlamp representing vehicles sold in Europe in model- year 2000 (see table 5 for a complete description). The researchers used three typical sign placements for the analysis: right shoulder, left shoulder, and overhead. The results were mixed. However, a consistent finding was that at distances greater than 500 ft, the composite HID headlamp profile consistently provided the least amount for traffic signs (of the five U.S. headlamp profiles). (56)
To further investigate the distance-related differences, the 2000-HID profile was compared to the 2000-UMTRI profile. Considering signs with 5-inch tall letters mounted on the left- and right-mounted shoulders and a legibility threshold defined by assuming 40 ft per inch of letter height (i.e., 200 ft), the 2000-HID profile provided illumination levels of 84 and 110 percent of 2000-UMTRI profile, respectively. For overhead signs at 650 ft, the 2000-HID profile provided an illumination level of 78 percent of the 2000-UMTRI headlamp profile.
Based on the results reported above, the 2000-UMTRI profile was selected for establishing MR levels for traffic signs. However, it is important to note that as technologies, specifications, and the vehicle fleet composition evolve, there will be a need to revisit the headlamp issues associated with MR development.
All three previous sets of recommended MR values have been based on dimensions of a vehicle that represents a large passenger car.(9,11,13) While the passenger car has traditionally been the best-selling vehicle type in the United States, for the 1999 model year, new trucks (defined as pickups, sport-utility vehicles, and minivans) outsold new cars for the first time; trucks had about 50.1 percent of the new-vehicles market versus 49.9 percent for cars.(58) This trend continued for the year 2000. Furthermore, over the past decade the number of registered passenger cars decreased by 0.1 percent, while the percent of trucks has increased over 60 percent.(52)
In November 2001, researchers measured the pertinent dimensions of the top-ten-selling light trucks, minivans, and sport utility vehicles for model year 2000. The results were averaged to develop a set of dimensions representing a typical light truck/minivan/sport utility vehicle that could be used to develop MR values (see table 6). The overall impact of this change is a larger observation angle associated with the vehicle dimensions. The larger observation angle will result in higher levels of MR values.
When the first set of recommended MR levels was published in 1993, traffic engineers did not fully understand the way retroreflectivity performed. Since then, traffic engineers have learned much about retroreflectivity and, as a result, more data are available in the public domain and many new computer tools have been developed to analyze retroreflective sheeting performance (such as Exact Roadway Geometry Output (ERGO) and TarVIP).
The first set of MR levels recommended for traffic signs used rather crude regression functions to predict the performance of retroreflectivity sheeting.(9,11) The impacts of both the orientation and rotation angles were completely neglected.
The updated retroreflectivity values use look-up functions to extract a subset of retroreflectivity values from four-dimensional matrices (observation angle, entrance angle, orientation angle, and rotation angle) that include over 250,000 data points. Retroreflectivity data come from the computer program called ERGO. (50)
Interpolation algorithms are then executed on the subset of retroreflectivity values to account for the potential nonlinearity of the data. This process results in an accurate estimate of retroreflectivity for any given geometry, as long as it is represented within the initial four-dimensional matrix.
As long as traffic signs continue to use retroreflectivity to increase nighttime visibility, the procedure described above can be used to assess the MR levels needed. As new materials are produced, the manufacturer should provide a full matrix of retroreflectivity values, similar to those in the ERGO program. A subset of the provided matrix should be validated at FHWA's photometric range.
The initial set of MR recommendations was based on the visual capabilities of a 47-year-old driver. The first field research using full-scale signs to address older drivers' needs in terms of MR levels was done by Carlson and Hawkins in 2001. (13) The effort described in this report includes further explorations of accommodating nighttime driver needs using findings from the literature and empirically derived relationships.
Using their data from an earlier effort, the researchers initially developed demand luminance curves for drivers aged 55 and older. (13) Using a subset of this data set, the researchers also considered demand luminance curves derived with the lower bound set at 65 years and older. Using these two data sets, the researchers analyzed the relative sensitivity of visual capabilities using age as a surrogate for all other visual metrics.
For either set of legibility luminance threshold curves, an accommodation level had to be established that could be used to determine the demand luminance (i.e., the percent of drivers assumed to be accommodated). Initially, sensitivity analyses were performed on various accommodation levels using just the 55-year-old driver data set. (46) When both the 55- and 65-year-old driver data sets were studied, the 50th percentile level was used. However, the researchers quickly realized that the levels under consideration did not represent the actual nighttime driving population.
Data from the National Personal Transportation Survey of 1995 were used to estimate the actual nighttime levels of driver accommodation represented by the 50th percentile levels. According to figure 6, approximately 89 percent of the nighttime drivers are under 55 years and almost 96 percent are under 65 years. If one assumes that visibility is directly correlated with age and as age increases visibility decreases, then a 50th percentile level of accommodation of drivers 55 years and older actually corresponds to nighttime accommodation levels well above 90 percent. Therefore, the 50th percentile levels were maintained for the development of the MR levels.