Abstract
To the extent that daylight can be effectively admitted and distributed in large buildings, daylighting can reduce dependence on artificial light and conserve energy. Effective in all climates, an atrium can serve more building area than other usual methods of daylighting large buildings. The objective of this research is to analyze the daylighting performance in actual building atriums including 4-sided, 3-sided, and linear atriums, and also to develop and validate algorithms for the prediction of daylight levels in these atriums under various sky conditions. Daylight patterns of three generic forms of building atriums are tested through extensive actual building measurements in combination with scale model tests and first principle theories. Also, actual measured data are compared with scale model data by using statistical analysis tests to show the degree of correlation. Algorithms are developed for the prediction of daylight levels at the wall and on the ground floor of the atrium, and then conventional techniques can be used to calculate daylight levels in perimeter offices adjoining the atrium. Final algorithms which are based on the scale model tests are validated by comparing with actual building measurement data for selected buildings in the Texas region. Office lighting performance and energy performance are not dealt with in this research, but rather constitute a logical extension of this work. Excerpts from this work have previously been presented at the 2nd International Daylighting Conference in Long Beach, CA. and 3rd Annual Symposium Improving Building Energy Efficiency in Hot and Humid Climates, Arlington, TX during the fall of 1986. All of the above work is part of a daylighting studies project supported by the National Science Foundation under Grant No. MSM 8504104.
Kim, Kang-Soo (1987). Development of daylighting prediction algorithms for atrium design. Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -25382.