Lighting simulations take various forms. Daylight simulations may be for a single moment in time with an overcast sky, perhaps to establish daylight factors. Simulations with sunny skies often need longer simulation periods to show the effect of the sun path, a shading or shadow analysis. Applications include feasability of energy saving devices such as light shelves, light penetration in atria, perimeter daylighting and alternative glazing specifications, shadow casting and glare. Radiance is capable of producing extremely detailed images with multi-colour. However, in construction most simulations are done in grayscale. This is because colour schemes have often not been prepared at the time of simulation or because to do so would require significantly more effort and unnecessary expense. The goal of daylight simulations is most often satisfied by numerical results producing contours or spot values. A visual impression from a design point of view can be produced in grayscale, taking care with exposure values. Simulations are not restricted to daylight, artificial lighting can be included where necessary.
A proposed roof design for the National Railway Museum in Shildon incorporated a bespoke pattern of glass reinforced plastic strips to allow daylight penetration over the centre of the engine shed as well as via the glazed facades. Daylight simulation calculated daylight factors for a standard CIE Overcast sky and results were compared to real recorded daylight data for the area. This allowed a reasonable estimate of the number of annual hours that useful natural light would be available.
A redevelopment at Manchester airport included the addition of new office space. Light shelves were proposed and a section of office space was chosen for analysis. Radiance software was used to predict Daylight factors, using a CIE overcast sky and and a maintenance factor (cleaning). Results were compared to recorded local daylight levels.Contours and graphs were produced to visualise the results and tabulated data identified the required external illuminance to achieve internal design targets.
On a domestic building daylight simulation was used to study potential improvement in a poorly daylit kitchen by making open plan to a lounge with a south-easterly aspect.Radiance allowed comparison under overcast sky but also showed the difference on a sunny morning. The comparison is shown using Lux levels at the floor, but could equally be irradiance.
Building Performance uses Radiance specialist software developed at Lawrence Berkeley National Laboratories.
Radiance uses backward ray tracing calculations to produce accurate values of radiance (W/m2 per steradian). Rays are ‘spawned’ from
the model surfaces at an density defined by the user and reflected around the 3D model. They are modified at each surface they hit by
appropriate reflection, absorption, transmission properties until a light source is found. There are a number of different surface models
in order to account for the different nature of altenative materials. Radiance allows basic material descriptions to be modified repeatedly
adding properties, textures and colour to produce an infinite variety of surface finishes. Backward ray tracing is computationally
significantly more efficient than forward ray tracing from light sources. The deterministic (consistant) and stochastic (random) elements
are explained in the book referenced below.Both electric and daylight sources can be included. Electric lighting can include manufacturers
data while several sky models (uniform, overcast, sunny) can be used to generate natural daylight for any global location and time.
The visual part of irradiance, illuminance, is detemined by luminous efficacy, defined as the ratio of illumance to irradiance as a
function of wavelength. Many other functions are available to calculate and present derived values, for example daylight factors and glare.
Tutorials, applications and calculation methods are provided in the book ‘RENDERING WITH RADIANCE The art and Science of Lighting Visualisation’
by Greg Ward Larson & Rob Shakespeare with contributions from Peter Apian-Bennewitz, Charles Ehrlich, John Mardaljevic, and Erich Phillips.
©Morgan Kaufmann Publsihers 1998, ©2003, 2011, 2021 Greg Ward Larson and Rob A Shakespeare.
For Radiance software and further reading visit the Radsite at Lawrence Berkeley National Laboratories