U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
For 25 years, the luminance design technique has been successfully used on major motorways and tunnels in Europe. This method is based on the way the human eye sees; that is, road surfaces are made visible by light reflected from them and entering the eye of the observer.
The panel saw many examples of good lighting that resulted from the use of this design technique. Examples are shown in figures 17, 18, and 19.
Figure 17. Wevelgem Tunnel, Belgium.
Figure 18. Highway near Helsinki Airport, Finland.
Figure 19. Highway near Helsinki, Finland.
European roadways are lit to levels more than twice as high as those in the United States, and with better uniformity. Belgian experts expressed the opinion that a high degree of pavement uniformity yields good driver comfort. They are confident that driver comfort equates to driver safety. They were not, however, aware of any formal studies linking driver comfort to safety.
Based on the Belgian experience, experts suggest that roadways lit to levels between 1 and 2 cd/m2 produce good visibility, while lighting the roadway to less than 1 cd/m2 does not yield good visibility. In addition to light level, good visibility in wet conditions also depends on the locations of luminaires. For example, in Finland, the team observed lighting over the roadway.
Because the luminance design method depends on road surfaces being made visible by light reflected from roads and entering the eye of the observer, the reflection properties of pavement become an integral part of the lighting-design process. The existing pavement reflection tables, the R-tables, were published in 1976 and have been used in luminance design worldwide ever since. The R-tables refer to pavement reflection characteristics under dry road-surface conditions only.
The R-tables are based on two pavement properties: S1, the specularity or pavement shininess; and Qo, the lightness or degree of grayness, from white to black, of a road's surface.
The range of the S1 value determines the class in which pavement is assigned, R1 through R4, as shown in table 3.
|Pavement Class||Standard S1||S1 Range|
|R2||0.58||>0.42 but < 0.85|
|R3||1.11||>0.85 but <1.35|
For accuracy, the average luminance coefficient, Qo, must be determined for the particular pavement under consideration. Typically, the values for Qo are R1 = 0.1, R2 and R3 = 0.7, and R4 = 0.8. However, these typical numbers do vary.
In Belgium, the most commonly encountered road pavements were bituminous asphalts (R3, with Qo from 0.07 to 0.10 cd/m2/lux) and porous asphalts (R2, with Qo from 0.05 to 0.08 cd/ m2/lux). French experts use the real R-value of the roadway, if possible. For quick estimating purposes, however, the following luminance/illuminance conversions for roadway lighting are used in France:
1 cd/m2 is produced by 8 lux on light-colored pavement.
1 cd/m2 is produced by 18 lux on dark-colored pavement.
1 cd/m2 is produced by 14 lux on average-colored pavement.
Swiss, French, and Belgian experts mentioned that a refined analysis of pavement properties is conducted for major projects. The analysis requires that a sample of the future road pavement be measured in the laboratory and the matrix of the reduced reflection coefficient be incorporated into the specifications, along with the minimum required lighting levels and uniformities for the project.
To obtain realistic R-values when evaluating an actual road surface, the Belgians evaluate several core samples and average the results. Outliers are discarded. Furthermore, results of studies have shown that, in the case of porous asphalts, it typically takes between 6 months and a year for the pavement to stabilize in order to obtain reliable R-values.
Figure 20 illustrates why field measurements should be delayed until after the pavement has stabilized. Note that the left lane is not traveled on and is in nearly "as poured" condition, while the right lane shows the typical change in reflection properties caused by traffic. Note that the wheel-rut paths also have a much different specularity than the other pavement areas, which makes it difficult to measure the luminance of the overall pavement. All measured luminance values must be qualified as to the location on the pavement, but methods for determining the overall luminance value from collections of individual points has not been established in Europe or in the United States. This example illustrates the difficulty typically encountered when attempting to enforce luminance specifications or when verifying designs.
Figure 20. Milchbuck Tunnel, Switzerland.
Luminance measurements taken on the two lanes show the right lane, at 140 cd/m2, to have twice the luminance level as the one on the left, 70 cd/m2.
The Swiss noted problems with standard R-tables and have obtained different results initially than those designed with standard R-tables. In an additional conversation with Werner Riemenschneider, however, he clarified that after the pavement had aged for 6 to 12 months, the Swiss typically found that the measured average values were within 15 percent of the average design value, usually on the high side.
In the good cases the Belgians noted discrepancies of less than 10 percent, when comparing measured luminance levels against calculated levels for pavements, where the reflection characteristics have been determined.
As mentioned earlier, pavement types have been invented since the original R-tables were conceived. The French noted increased usage of new surfaces over the past 10 years. These surfaces include a number of wearing courses and porous asphalt, i.e., water-draining pavement.
Porous asphalt stabilizes in a unique way. It becomes more diffuse and its brightness increases, as is shown in figures 21 and 22.
Figures 21 and 22 show computer-generated images of the reflection characteristics when viewed from typical angles down the roadway. The angle Figure 21. New porous asphalt. most frequently encountered in the past is the 1-degree downward view. The French and the Swiss suggested that additional viewing angles were needed because of lower speeds and urban environments. The angles most frequently mentioned were 3 and 5 degrees.
Figure 21. New porous asphalt.
Figure 22. Porous asphalt, after 12 months.
Appendix D lists a paper by Ms. Corine Brusque that describes how to design lighting for water-draining pavements. The Dutch noted that this type of open-graded asphalt pavement seems to produce better visibility than the older, dense asphalt.
In Switzerland, experts emphasized the importance of dry roadways when conducting field measurements. There are, however, only a couple of summer months during which pavements are dry enough to be measured. In addition, the team heard warnings about dew points and pavement ages. It was noted that, during observations in cold weather (typically October through December), with a clear sky, when conditions were under the dew point, a water film could suddenly appear on the roadway. This film could provide a reflectance differential of 200 percent. Given these difficulties and variations in pavement reflection characteristics, the Swiss typically verify lighting installations with incident light measurements.
In addition to the R-tables, N-tables are applied in countries that use additional "whiteners" in pavements, which causes the pavement to become very bright. The French noted that, specifically for tunnel lighting, they are researching a special pavement that has white gravel and cream-colored bitumen. Currently, it appears as though the average luminance coefficient, Qo, changes downward, over a 3-year period. This is still under investigation.
The Scandinavian countries have developed W-tables for use on the wet roadway conditions encountered there (see figure 23).
Figure 23. Wet roadway in Finland (inset, close-up of pavement).
In Finland, the standard R- and W-tables are used, while the other countries that the team visited use only the R-tables. In Switzerland, the W-tables that were developed in Scandinavia are not used. Rather, the Swiss studied 10 typical installations and, based on experience, determined that, for their purposes, 2 cd/m2 under dry conditions was also adequate in wet conditions. Swiss experts found no operational difficulties with that approach.
|<< Previous||Contents||Next >>|