Different types of artificial skies have been realised in the past: mirrorskies, dome skies, spotlight sky simulators or scanning skies, each characterised by different advantages and disadvantages7,8,9,10,11,12.

Comparing the potentialities and limits of each type of artificial sky, at the Daylighting Laboratory ofthe Politecnico di Torino it was decided to design and achieve a scanning artificial sky, able to reproduce the diffuse skylight component, supplemented by an artificial sun, able to reproduce direct sun-light component13,14. The facility was conceived not only for research purpose, but also and especially as a tool for designers (architects, engineers, lighting designers) to predict which way daylight characterises outdoor and indoorenvironments, since it allows both determining daylight levels (illuminance and daylight factor values, spatial distribution of daylight over an indoor room) and reproducing how a daylighted environment appears as well as what is the dynamic behaviour of sun penetration.

The “sky” reproduces one sixth of the vault, consisting of 25 individually dimmable luminaires (figure 1), based on the model ofsubdivision ofthe sky hemisphere proposed by Tregenza for sky luminance measurements and assumed by the CIE in the IDMP (International Daylighting Measurement Program)15,16. In orderto reproduce the entire sky dome, the model’s stand produces a six-step scan rotation, modifying foreach scan the luminance distribution. Global photometric quantities and pictures are therefore obtained adding the partial values and images. Different sky conditions are reproducible according to both standard models and real luminance values recorded at IDMP measuring stations. The “Sun” is in a fixed position, so the model’s stand is rotated an tilted to suitably reproduce the relative Sun-Earth position, according to solar geometry equations.

In short, main advantages as tool for daylighting design may be summarised as follows17:

• good adherence with real situations

• possibility of simulating different sky conditions, referring to both standardised daylighting models and real skies, experimentally measured

• possibility of comparing performances of different daylighting systems

• possibility to carry out an objective measurement of photometric data (quantitative evaluation) and a perceptive assessment of the daylighted environment (qualitative evaluation), by taking photographs of the indoorsimulated environment

• possibility of carrying out studies with different aims and on different scales, from site planning to indoor environment to daylighting systems.

Apart from the advantages already listed a scanning artificial sky presents some limits, linked to the finite distance between the model’s stand and the portion of dome: especially when dealing

with large models, an horizon line error and a parallax error (different parts of the considered model receive different quantities ofdaylight and sun-light18) may occur.