Annual solar radiation stadium roof

Hello everyone,

I've got an issue trying to make an Annual Solar Radiation with DIVA on a stadium roof. The calculated values seems to be randomly distributed, they does not follows the expected gradient due to the orientation and inclination of the roof panels (see pictures and definition attached). I might shoot it is due to a data tree problem, but I cannot figure out a solution.

Can some help me with this problem?

Thanks so much,

Attachments:

stadium_roof_diva.gh, 37 KB

Replies to This Discussion

Permalink Reply by Jeff Niemasz on July 1, 2013 at 7:24am

Hi Matteo,

Your method of getting the analysis nodes and analysis directions is the problem (analysis vectors were all pointed up). I added a closest point component to the centroid of your panels to get the uv to analyze surface and get the normal of the surface. GH definition attached. Now it looks like this:

One more thing. Your model is not currently accounting for self-shading so it is just getting the annual solar irradiation at those points and directions but not including the panels and trusses of the model itself. If you gives these panels a material and add them to the GM input this will take self shading into account. Make sure you offset the analysis nodes from the surfaces a bit if you do this. This is discussed further here. The analysis nodes would have to be a denser grid to capture the gradient across the panel- I ran it with the panels using a generic outside facade reflectance and more nodes (gh model attached). Also, if you are using tree data you should flatten it to get the bounds of all the values for the falsecolor preview domains.

Hope that helps.

Jeff

Attachments:

stadium_roof_diva_fixed.gh, 50 KB

Permalink Reply by Jeff Niemasz on July 1, 2013 at 7:31am

Second model definition attached.

Attachments:

stadium_roof_diva_fixed_with_self_shade.gh, 51 KB

Permalink Reply by Matteo Mandrile on July 1, 2013 at 10:33am

Hello Jeff,

many thanks for your explaination, it's very clear.

I just misunedstood one thing: do I need to move a little the nodes for the analysis just when I apply a material? or I have to move these in any case?

Because in the first case here, the points were right on the surface, and in the second one (self shading) you moved the analysis point.

Thanks

Mat

Permalink Reply by Jeff Niemasz on July 1, 2013 at 12:30pm

If you are using the surface only to get the node locations and the surface normals then you do not need to move it (assumes that you are using this as an analysis surface so that you have already insured that it is not coincident with geometery with a referenced material in the model).

If you are using a surface that is "in" the model (meaning it has a material reference - it shows up in a radiance visualization), and using the same surface to derive analysis nodes and normals for analysis directions, then you should offset the nodes a bit (I usually use a fraction of a model unit or 1mm for irradiation). If the nodes are coincident with geometery in the radiance model you can get strange results. With most daylight simulations you usually run the nodes floating about 1 meter above a floor.

Jeff

Permalink Reply by Matteo Mandrile on July 1, 2013 at 4:42pm

Thanks Jeff, really appreciate your help.

One more question: there is a way to output the analysys result for visualization/presentation purposes? Including the numeric values and a" reference scale"? I see there is a component for this, but I cannot realize how to implement it!

Permalink Reply by mohammad on July 1, 2013 at 2:29pm