Effect of Windshield Veiling Glare on Driver Visibility

Abstract

This paper presents results of a veiling glare prediction model that was developed by modeling veiling glare luminance measurements and by using Blackwell's visual contrast threshold data. A critical visibility condition for a driver approaching a tunnel with the sunlight falling on the windshield and attempting to detect a target inside the tunnel was modeled. The luminance of windshield reflections of 38 instrument panel samples was measured in a specially designed veiling glare simulator by varying the sun angle, the windshield angle, and the instrument panel angle. The measurements were used to develop a regression model to predict the veiling glare coefficient as a function of windshield angle, sun angle, instrument panel angle, and instrument panel material gloss value. The developed visibility–distance prediction model was exercised by varying eight factors: (a) windshield angle, (b) instrument panel angle, (c) sun angle, (d) sun illumination incident on the windshield, (e) gloss level of the instrument panel material, (f) driver's age, (g) target size, and (h) illumination incident on the target in the tunnel. For example, under a worst-case situation of 45,000-lux sunlight illumination incident on the windshield, a 65-year-old driver will not be able to see a 2-ft2 10% reflectance target while approaching a dark tunnel with 100 lux of tunnel lighting. In contrast, if the tunnel lighting is raised to 5,000 lux, the 65-year-old driver can see the target from 650 ft, and a 25-year-old can see it from 4,900 ft.