probably this is not possible without tweaking. As far as I am aware,
Design Builder uses the standard backwards algorithm of Radiance. That
one is not really made for typical solar tubes, if you want to consider
direct sun and you have no diffuser on top of the tube. If you would run
your Radiance simulations directly, you could either use a sky where the
direct sun contribution would be spread over a wider solid angle (see
Tregenza's sky subdivision scheme) or apply extensions such as the
Photon Map. I doubt that these are available in Design Builder (though I
do not really know this for sure). If you need a front end, you can have
a look at Daysim, which includes the Photon Map and also supports the
daylight coefficient method using Tregenza and higher resolution sky models.
I would recommend a look at the Radiance workshop presentation of the
recent years, which are all available at www.radiance-online.org.
I know nothing much about DesignBuilder, but if it uses Radiance for the daylighting calculations, the best way to accurately model a tubular daylight device is to obtain the photometric distribution file AKA ".ies file" for the make and size of TDD you are modeling. They are all somewhat of a hack, since they attempt to characterize the distribution and intensity of a thing that redirects an infinitely variable light source, using a single file. But there are ways to scale the intensity, and even to account for exterior obstructions at the roof aperture, by doing some evaluations of the global exterior illuminance and comparing it to the assumption that the photo metric file made, and scaling accordingly.
In any event, this is really the best way to characterize a TDD in Radiance, because you are treating the luminous aperture as a direct light source, rather than attempting to use Radiance's indirect calculation to hope enough random rays find their way up the TDD and out to the hemisphere and actually intercept the sun and other major contributors (like big puffy cumulus clouds), which is unlikely at best.
Hi Sheelam,
probably this is not possible without tweaking. As far as I am aware,
Design Builder uses the standard backwards algorithm of Radiance. That
one is not really made for typical solar tubes, if you want to consider
direct sun and you have no diffuser on top of the tube. If you would run
your Radiance simulations directly, you could either use a sky where the
direct sun contribution would be spread over a wider solid angle (see
Tregenza's sky subdivision scheme) or apply extensions such as the
Photon Map. I doubt that these are available in Design Builder (though I
do not really know this for sure). If you need a front end, you can have
a look at Daysim, which includes the Photon Map and also supports the
daylight coefficient method using Tregenza and higher resolution sky models.
I would recommend a look at the Radiance workshop presentation of the
recent years, which are all available at www.radiance-online.org.
Cheers, Lars.
I know nothing much about DesignBuilder, but if it uses Radiance for the daylighting calculations, the best way to accurately model a tubular daylight device is to obtain the photometric distribution file AKA ".ies file" for the make and size of TDD you are modeling. They are all somewhat of a hack, since they attempt to characterize the distribution and intensity of a thing that redirects an infinitely variable light source, using a single file. But there are ways to scale the intensity, and even to account for exterior obstructions at the roof aperture, by doing some evaluations of the global exterior illuminance and comparing it to the assumption that the photo metric file made, and scaling accordingly.
In any event, this is really the best way to characterize a TDD in Radiance, because you are treating the luminous aperture as a direct light source, rather than attempting to use Radiance's indirect calculation to hope enough random rays find their way up the TDD and out to the hemisphere and actually intercept the sun and other major contributors (like big puffy cumulus clouds), which is unlikely at best.
Rob Guglielmetti