Development and validation of a Radiance model for a translucent panel

This study describes the development and validation of a Radiance model for a translucent panel. Using goniophotometer data combined with integrating sphere measurements, optical properties of the panel were derived and converted into a Radiance model using the trans and transdata material types. The Radiance model was validated in a full-scale test room with a facade featuring the translucent panel material. Over 120,000 desktop and ceiling illuminances under 24,000 sky conditions were measured and compared to simulation results using the Perez sky model and a Radiance-based daylight coefficient approach. Overall mean bias errors (MBE) below 9% and root mean square errors (RMSE) below 19% demonstrate that translucent materials can be modeled in Radiance with an even higher accuracy than was demonstrated in earlier validation studies for the plastic, metal, and glass material types. Further analysis of results suggests that the accuracy of around ±20% currently reached by dynamic Radiance/Perez/daylight coefficient calculations for many material types is sufficient for practical design considerations. A procedure is described showing how goniophotometer and integrating sphere measurements can be used to accurately model arbitrary translucent materials in Radiance using transdata function files.