Grasshopper

algorithmic modeling for Rhino

Hello All,

I am facing a problem in running an energy simulation using a model with curved surfaces.

I am attaching the script and the model that should give you the same error I get.

The error appears as soon as I introduce the glazing surfaces in the geometry and it is the following:

1. Solution exception:'hb_EPZoneSurface' object has no attribute 'coordinates'

I have been trying to solve it for a while and I have noted the following:

- I am using the approach of building the zones using the create HBSfrs component. However, nothing changes if I use the Honeybee masses2zone component (error remains);

- I am introducing the glass surfaces using geometries built in Rhino. However nothing changes if I use the HB component to automatically build the windows based on glass to wall ratio (error remains);

- This problem in the simulation has been introduced only once I have updated all the components to the latest versions of LB and HB. The geometry was working fine with the versions from Feb 2017. Also, I don't expect that curved geometries should cause problems as they have been used in simulations for quite a while now.

Finally, there is another problem in the simulation which relates to the HB context surfaces. Even if the simulation runs with the curved geometry, as soon as I add the roof as a shading element I am again not able to go further. The roof is composed also of curved surfaces and I could eventually simplify them but I am surprised that they are not 'digested' as they are quite simple and 'light' in the end.

I hope that I am missing something here and that the problem can be solved easily. It would be great to have your input!

Thanks a lot in advance and let me know if you have any questions on this!

Carmelo

Views: 459

Attachments:

Replies are closed for this discussion.

Replies to This Discussion

Hi Carmelo,

Using Honeybee masses2zone instead of using hb_EPZoneSurface did the trick and I could run your definition.

Definition attached.

-Devang

Attachments:

Hello Devang,

Thank you for looking into this. Unless I am missing something here, you did't include the glazed areas in the script which is when the problem occurs. I could make it work with only the opaque envelope. If you have any other tips that would be great!

Best,

Carmelo

Carmelo,

Thanks for reporting and thank you, Devang, for helping narrow down the probelm to the glazing.  It seems there was an old unnecessary check in the code pathway taken by curved surfaces with glazing tat was causing this to fail.  I just fixed it here:

https://github.com/mostaphaRoudsari/honeybee/commit/5c936aedfb41b92...

... and you can see that the attached version of your file works.

Also, I would still really recommend planarizing the geometry.  I know it's a little extra manual work but it will make the calculation so much faster.  This said, if you hadn't tired this method, then we never would have found the bug.  So I am glad that you tried this first.

-Chris

Attachments:

Thanks, Chris. I was at it but couldn't find a remedy. Excellent work as always!

Hello Chris,

Thanks a lot for your help, the definition works fine now!

There was another issue I mentioned in my first post; if you try to include the geometry of the roof in the context, then the simulation does not conclude properly (please see the image attached).

I feel this is again due to the geometry of the roof which I could simplify. However, I was wondering if you were able to understand what is causing the problem?

Best,

Carmelo

Attachments:

Carmelo,

It looks like the issue with the roof is the result of the doubly-curved elements that you have in between the top and the bottom pieces, which really don't affect the simulation and are just going to increase your calculation time.

It looks like they are causing the simulation to fail because you are hitting the limits of the Rhino meshing algorithm to the point that it makes strange mesh faces (like meshes with two sides).

So I would just model the top and bottom surfaces and then everything will run fine.

-Chris

Thanks a lot for looking into this Chris and being very helpful!

Best,

Carmelo

Hello Chris,

Thanks for your help so far, if possible I would like to get your thoughts and those of the community on few more errors I get when running the simulation.

The simulation concludes correctly and results seem to make sense; however I get the following error that I have never seen before:

The simulation has not run correctly because of this severe error:
** Severe ** Temperature (high) out of bounds (272.45] for zone="FRONT_ZONE", for surface="E75D3CEDA5DC4B80A14F_GLZP_29"

This happens when I change the weather data to use the one from Bangkok (so, I much warmer climate compared to London).

There are also the following two errors:

The simulation has not run correctly because of this severe error:

** Severe ** Problem in interior solar distribution calculation (CHKBKS)

The simulation has not run correctly because of this severe error:
** Severe ** DetermineShadowingCombinations: There are 1 surfaces which are casting surfaces and are non-convex.

I know from previous discussions on the forum that the two errors above are actually not severe and I remember you explained what they refer to.

You were mentioning that the temperature of the surfaces would not be accurate. Does this mean that I may get not accurate results if I want to assess the thermal confort in one of the zones, for example the one close to the glass facade?

Do you think this errors will be fixed at some point in future versions of HB?

Is the only way to fix them to make my geometry composed of planar surfaces? I don't have L-shaped zones in the model.

Thanks,

Carmelo

Carmelo,
In most cases, a CHKBKS error is not severe but it seems you have found a case where this error actually is severe. A 272.45 degree Celcius surface surface temperature is definitely not realistic and it's likely that this is caused by the solar distribution error. I suggest changing the solar distribution type to be FullExteriorWithReflections using the Energy Sim Par component as I state in the other discussions that you reference.
And, while you don't have an L-shaped zone, it is a concave zone and this is what causes the error (an L-shaped zone is just one type of concave zone). My suggestion above will get your simulation running correctly but this error is a limitation of E+ and not Honeybee. So it's not something g that we can fix on our end and we have to hope that the E+ developer team eventually addresses this better.
-Chris

I will also try to divide the surfaces into rectangular (or non-convex) ones. It works for me sometimes. For instance a L shape, instead of being one surface it can be 2 rectangular ones.

-A.

RSS

About

Translate

Search

Photos

  • Add Photos
  • View All

© 2019   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service