Surface Problem

Replies to This Discussion

Permalink Reply by uğur özgür on November 18, 2013 at 5:39pm

EPW File

Attachments:

• TUR_Izmir.172180_IWEC.epw, 1.5 MB

Permalink Reply by Mostapha Sadeghipour Roudsari on November 18, 2013 at 5:57pm

Hi Ugur,

I can't check your file right now but from the image I guess the normal direction of the surfaces are reversed. Check the normal directions and make sure they are outward. In Rhino you can check the DIRection of the normal using dir. Type f and press enter to Flip the direction of the normal.

Hope it helps,

Mostapha

Permalink Reply by uğur özgür on November 18, 2013 at 6:54pm

Thank you for response. I did changes what you said and a bit changed something but I think main problem is about so different results on surfaces considering sun direction. When I choose any specific time, results are different between horizantal and vertical surfaces in same levels. I mean, results should be close. 

Attachments:

• Adsız.jpg, 359 KB

Permalink Reply by Abraham Yezioro on November 19, 2013 at 8:09am

Hi Ugur,

I played a bit with your files. First of all, as Mostapha said, there are a LOT of surfaces with their normals pointing inwards. Those are not small changes in the results. But saying that, i don't understand why you think the results are not making sense. I find them pretty logic. I recommend you to increase the resolution of the analysis grid (you have 4, i set 2, but recommend even smaller). In some cases may be confusing seeing one color supposedly wrong just because the calculation was done on the center of the face. In your case this is clear specially on those "terraces" on the north (they are north?) side.

Also i recommend to show the legend and even more than that to set a fixed scale for the legend so you can have all visuals comparable with the same scale.

Hope this helps,

-A.

PS. Mostapha, question for you: For the RealTimeAnalysis component to work well there is need to execute the GenCumulativeSkyMtx for the whole year (analysis period)? In Ugur's case he has only one day, one hour. I'm not sure about the relation between the results, so i'm wondering.

Permalink Reply by uğur özgür on November 19, 2013 at 8:45am

I tried on another surface with considering flipping. You will see the differences..

Attachments:

• ANOTHER SURFACES.jpg, 222 KB
• ANOTHER SURFACES.3dm, 52 KB
• ANOTHER SURFACES.gh, 79 KB

Permalink Reply by Abraham Yezioro on November 19, 2013 at 2:07pm

Ugur,

I've playing a lot with the last file you uploaded. The results completely make sense to me. First of all i suggest to show always the legend so you can see what the colors represent.If you show a summer day (say 21.6) you'll see big differences between the vertical and the inclined planes. The radiation in vertical is much lower than in horizontal or inclined. As for winter (say 21.12) the vertical or slightly inclined (as in your example) the radiation is pretty similar. I hope i understood your points and that this helps.

-A.

PS Mostapha: Another one for you. See the attached. I tried to show the results of the RealTimeRadiation in the surfaces but didn't succeed. The results list is different than the one from the RadiationAnalysis component. Don't know how to change the List format. Can you suggest?

Attachments:

• ANOTHERSURFACES_A.gh, 83 KB

Permalink Reply by Mostapha Sadeghipour Roudsari on November 20, 2013 at 7:52pm

Hi Abraham and Ugur,

Abraham, I think all you needed was flattening the test points (Image below), however, It reminded me that I should unFlatten the test points from the realTimeRadiation analysis component. I added this as an enhancement.

Ugur, as Abraham mentioned the results makes sense. I wonder if you are concerned about the drastic change in the color? You may want to use legendParameter and set the minimum to 0 so the colors make more sense. Something like image below.

Attachments:

• ANOTHERSURFACES_A_MSR.gh, 90 KB

Permalink Reply by Abraham Yezioro on November 21, 2013 at 1:34am

Hi Mostapha,

What about the question:

For the RealTimeAnalysis component to work well there is need to execute the GenCumulativeSkyMtx for the whole year (analysis period)? In Ugur's case he has only one day, one hour. I'm not sure about the relation between the results, so i'm wondering.

Thanks,

-A.

Permalink Reply by Mostapha Sadeghipour Roudsari on November 21, 2013 at 6:26am

Hi Abraham,

I need your help to understand the question better. genCumulativeSkyMtx only produces the sky for the whole year. You need to select the data for your analysis period using selectSkyMatrix. The result of the radiationAnalysis component shouldn't match the realTime analysis unless you are running the analysis for the same sky. Here is an example:

Did I even answered the right question? :)

Cheers,

Mostapha

Attachments:

• ANOTHERSURFACES_A_MSR_REV01.gh, 91 KB

Permalink Reply by Abraham Yezioro on November 21, 2013 at 6:48am

Hi Mostapha,

This was not my question, but thanks.

If you see the original file from Ugur he selected one specific day for the selectSkyMtx. If you use this same date for the RealTimeAnalysis it is fine. But what about changing this date?

So my general question is: If you want to get truthful results for the RealTimeAnalysis should you set the analysis period for the selectSkyMtx the whole year? providing you want to change the date in the RTA.

Thanks,

-A.

Permalink Reply by uğur özgür on November 22, 2013 at 1:45am

Thanks a lot...

Permalink Reply by Mostapha Sadeghipour Roudsari on November 21, 2013 at 7:45am

Hi Abraham, The answer is no. Regardless of the analysis period the results of the intersection matrix should be the same. It is only a geometrical calculation (True/False based on intersections) so the analysis time doesn't matter. -Mostapha