|
|
|
Santiago L. Torres |
|
Sakamoto Laboratory |
|
The University of Tokyo |
|
Graduate School of Engineering |
|
Department of Architecture |
|
|
|
|
|
|
Sections in the main window and position of
daylight sensors |
|
|
|
|
|
|
|
|
Light shelf |
|
Louvers |
|
Ceiling shape |
|
Orientation |
|
|
|
|
|
Sequence in the simulation process (repeated for
each hour) |
|
|
|
|
|
Diagram of the modules used |
|
|
|
|
|
|
|
|
The use of only one observer to assess glare
conditions |
|
|
|
The use of UGR, not specific for daylight |
|
|
|
Low quality renderings (computing time) |
|
|
|
Impossibility to determine the sky condition for
low solar altitudes |
|
|
|
Differences between the modeled building and the
real building |
|
All blinds in each room open or close together |
|
Daylight sensors are placed every three or four
sets of lamps |
|
The reflectance of the louvers is much lower |
|
|
|
Differences between simulated behavior of
occupants and real occupants (unknown) |
|
|
|
|
|
|
The variation of the glare rates was always
related to a variation of the energy consumption. |
|
|
|
The light shelf showed a protective effect
without reducing the daylight levels inside the room. |
|
|
|
The horizontal ceiling redirected the light from
the light shelf further into the rear part of the room. |
|
|
|
Changing the orientation of the main facade
proved to be more effective than other measures. |
|
|
|
Some results were counterintuitive, indicating
that glare rates should be considered in energy consumption assessments. |
|
|
|
Further research should include the comparison
of this methodology with real operating conditions, especially regarding
the behavior of real occupants. |
|