Issues with Imageless Annual Glare in LBT 1.5

Hello all,

since the last update to 1.5, I get this error from the Imageless recipe:

  1. Solution exception:The recipe failed to run with the following summary:

Scheduled 14 tasks of which:

  • 8 ran successfully:
    • 1 CopyGridInfo(…)
    • 1 CopyRedistInfo(…)
    • 1 CreateRadFolder(…)
    • 1 CreateSkyDome(…)
    • 1 CreateTotalSky(…)
  • 1 failed:
    • 1 CreateOctree(…)
  • 5 were left pending, among these:
    • 5 had failed dependencies:
      • 1 AnnualImagelessGlare(…)
      • 1 DaylightGlareAutonomy(…)
      • 1 LetImagelessAnnualGlareFly(…)
      • 1 RestructureDaylightGlareProbabilityResults(…)
      • 1 _Main_5a4cf8eaOrchestrator(…)

This progress looks :frowning: because there were failed tasks

Use the report_out attribute of recipe settings to see a full report.

Has someone encountered this problem and knows of a solution?
Do I need to upload more information about the model and calculation?

Thank you for your help!

Hristo

Hi @Hristo,

I don’t have any issues with the sample file. Have you tried that?
annual_glare.gh (81.8 KB)

Are you able to share the logs file? You can find it in the __logs__ folder.
image

Yes, certainly!
logs.log (49.7 KB)

Thank you Mikkel!

Hi @Hristo,

It seems that the problem is one of the Radiance polygons leading to a fatal error when creating the octree. This tells me that the problem is likely your HB Model* rather than the recipe. If you can share the Grasshopper file (with internalised geometry or with the Rhino file) I will take a look.

*or maybe the HB Model to Radiance converter fails to write the polygon correctly.

Hi Mikkel, thank you for your time. Sending the model to you.

I ran the same calculation on LBT 1.4.0 copying the faces and geometries assigned, and I am able to run it.

@Hristo ,

Just to make sure: does your model pass validation using the HB Validate Model component?

If not, then that component will tell you exactly what is wrong with your model so that you can fix it.

If it does pass, then you have found a bug in Honeybee and I can help investigate it with @mikkel .

1 Like

Hi @chris, it does pass the test with warnings. A quick fix for @Hristo was to pass the problematic geometry through the Grasshopper native Mesh Brep component before adding the geometry to HB Face. I found that the issue is caused at this commit, fix(planarize): Tweak planarization to work at all scales of curvature · ladybug-tools/ladybug-rhino@90fb50a · GitHub, hence why it works fine in 1.4.0. I will share the case with you elsewhere.

1 Like

Hi @Hristo ,

Your starting geometry was not valid as the HB Validate Model component would tell you.

When you have starting geometry that is this messy, you really should be meshing it before you pass it to a Honeybee component. You’ll see that, if you do this, you don’t get any validation issues:

In any event, I loosened some of the restrictions with the Brep planarization process in order to help out this case:

… but I really recommend meshing everything if you don’t plan to go through a cleanup process on your Brep geometry.

1 Like