Existing annual daylight simulation software fall short\nwith respect to variable fenestration optics that change\ninterior daylight distribution with sun position and/or\noperating schedule, thus limiting the ability to compare\nthe performance of advanced fenestration systems.\nMany of these window or skylight systems can be\ndescribed efficiently as a bidirectional scattering\ndistribution function (BSDF), which characterizes their\nflux output as a function of input for a particular\nconfiguration. In this paper, we describe a new method\nthat employs measured or simulated BSDFs to permit\nfast, matrix-based annual daylighting calculations. The\nmatrices themselves are precomputed by Monte Carlo\nray-tracing in a modified daylight coefficient approach\nwe call Dynamic Radiance. The inner time-step loop\nthen consists of multiplying the desired sky luminance\nvectors against three matrices in the general case, where\na separate BSDF matrix permits dynamic fenestration\ncontrol strategies. In this paper, the authors describe\ntheir implementation of the Dynamic Radiance method\nand demonstrate its application to a set of 61 real spaces\nmodeled for a research project to determine new\ndaylight metrics. We present results from these\nsimulations and discuss advantages and limitations of\nthe new approach.
[1]
Jacob C. Jonsson,et al.
Simulating Complex Window Systems using BSDF Data
,
2009
.
[2]
J. Mardaljevic.
Daylight simulation : validation, sky models and daylight coefficients.
,
1999
.
[3]
Lisa Heschong,et al.
61 FLAVORS OF DAYLIGHT
,
2010
.
[4]
Christoph F. Reinhart,et al.
Standard daylight coefficient model for dynamic daylighting simulations
,
2008
.
[5]
Lisa Heschong,et al.
Improving Prediction of Daylighting Performance
,
2010
.