Daylight result changes with grid size

Hi guys ,I meet some strange problem when I use honeybee to do daylight simulation.
At first, everything seen ok ,the result also seen right.But when I changed the grid size directly ,something goes wrong.Though most of areas works well,but the simulation illuminance result of some areas are too high.And I also compared the new result with the old one,the domains are similar.So I guess the new result is correct roughly, just some parts are wrong.The same thing also appeared when I changed the date directly.
Here are the figures of my result and my gh file.


grid size:3

grid size:6

test.gh (509.2 KB)

It is only a matter of keeping a consistent scale.

6 meter grid

1 meter grid

Hi devang, thanks for you reply.
Are you means keeping the legend in the same scale? It works but I am still confuse about why the result of some areas are so high ?I am not sure whats you simulate values are. If you connect your illuminance values with panel and see whats the result values.Whether there are some values much larger than others ,I can sure these values are wrong.

Yes. Keeping the legend scale same. I don’t see any problem with values. Here’s how you can see values on the grid points. This should help you understand where what value of illuminance is received.

I keep the legend scale same this time.And also show the values on the grid points.

You can see when the grid size is 6m ,all values seen ok. But when I change the grid size into 3m ,you can see the values I outline are too large, almost 30,000 lux.

Also, are you use the same weather file and setting with me ? It seen like our result values are different.

6m grid:

3m grid:

Thanks you reply again.

I used Abu Dhabi weather file as the weather file was not internalized in your file.

The radiation values are calculated at the center point of each grid. Having a high resolution grid is giving you more accurate results, of course.
Imagine that you have only one grid for your whole area. The value you’ll get is the one at the center of the room. You are going to miss what is happening close to the window.
This is exactly the reason you get those high values. On 6 m grid size you are probably missing the real values close to the facade/windows.

Hope this makes sense for you.
-A.

1 Like

Thank you very much. It makes sense a lot for me now.