Author/Authors :
Alfonso Soler، نويسنده , , Pilar Oteiza، نويسنده ,
Abstract :
The performance of a light shelf with a reflectance of about 91% after excluding the specular component (reflectance about 1%), providing for solar protection during the long, hot, dry summer season, is studied in Madrid (40.4°N, 4.4°W) using two scale models (1:10) with rectangular openings facing south, one taken as a reference and the other equipped with the light shelf. As a shading device the light shelf was designed for a vertical shade angle of 50°, and thus had both internal and external parts. The models were painted inside: walls and floor with a black matt paint, and the ceiling white matt, this being a limiting case as only the effect of a high ceiling reflectance is studied. Measurements of mean hourly illuminances were obtained in each model during one year at points at the same distances from the openings and corresponding to the working plane. In the present work we study for the points closest to the walls opposite the openings the dependence on solar elevation α and azimuth θ (both calculated at mid hour) of the ratio η = E1/Er, E1and Er respectively being the mean hourly illuminances measured in the model with the lightshelf and in the reference model. The dependence of η on θ and α is studied for days with σ = 1 or σ = 0 for all hours before and/or after the solar noon, σ being the hourly sunshine fraction. For days with σ = 1 for all hours before the solar noon (clear skies) and increasing values of α, η increases when α increases till a maximum of η (not clearly visible in winter) is obtained. For higher values of α and up to the solar noon the values of η can decrease, or η may remain approximately constant (central months of the year). Maximum values of η higher than 1 are only obtained for clear days, and approximately 110 < n < 270, where n is the day of the year, but for this range of values of n, η is not higher than 1 for all the possible values of α. For each day with clear skies maximum values of η are obtained for values of θ between 60 and 70°. For days with σ = 0 for all hours before or after the solar noon no definite dependence of η on α and σ is obtained. It is emphasized that the results presented are obtained with reflectances for the light shelf and a model that may be considered as a limiting case: practically no specular reflectance from the light shelf and walls and floor painted black in the model.