Methods and Approach

The investigations are based on a combination of daylighting simulations using Radiance [RAD] with methods used for the design of experiments (DOE) [Sch97]. The simulation allows for reproducible sky conditions.

The goal is to find the qualitative and quantitative influence of elementary architectural, i. e. structural, measures on the daylight quality in interior spaces.

The DOE methodology allows multi-dimensional factor variation at minimum experimental expense. The DOE plans containing all necessary experiments show two essential charac­teristics:

• All factors are varied in discrete steps, i. e. not continuously, with the number of steps being as small as possible. Thus in the simplest case, these are two steps defining the lower and upper border of the experimental space.

• All variables are linearly transformed so that the bounds of definition range from -1 to +1. This leads to DOE matrices with orthogonal columns, DOE plans of general validity (independent of the natural variable dimensions), and to the possibility of a direct quan­titative analysis of effects.

Base case is a room lit by one un-obstructed vertical window. Two more investigations show two windows each, in two facades adjacent to and facing each other, respectively. Two variations of an obstructed single window case and two roof lighting designs complete the research work. A few restrictions apply:

• In cases with one window only, its horizontal position is central. The results do there­fore not apply for strong asymmetric placements of the window. In cases with two win­dows, the symmetry restriction holds for one of them.

• All numbers are based on a CIE overcast sky with a sun elevation of 60° above the horizon.

• All daylight openings are rectangular.

• All room geometries are rectangular except for the roof geometries in the toplit cases.

• The interior optical description is limited to the uniform average diffuse reflection coeffi­cients of floor, walls and ceiling.

Three criteria have been looked at in [Sic03]: the average daylight factor D, the daylight factor in a room depth of 4 m, D4, and the daylight factor regularity, G, which is the ratio of minimum to mean daylight factor. The most useful criterion is D. It serves as a measure for the total amount of daylight inside under overcast conditions. The regression analysis re­sults for this criterion are presented here.