I still got similar error as the one reported in a previous post when using the split mass components to deal with several particular group of floor surfaces.
The building zone for each floor is created by extruding the floor surface. However, there are always thermal zones missing in the result. The component even reports error in the second case.
May I ask if this is caused by the floor surfaces that are not properly prepared, or due to the limitation of the split mass component itself?
Hi Grasshope, I’m pretty sure this is the component.
I’m also having a lot of issues with it, and have been trying to revise some of the code with Chris: https://github.com/mostaphaRoudsari/Honeybee/issues/497 . Thanks for providing these gh files isolating edge cases causing errors, they do help. I’ll try and dedicate some time to look at this during the weekend.
So I believe I’ve caught what was causing 90% of the problem - an equality condition for certain floor heights getting triggered inaccurately due to the limitations of floating point precision. There are still errors happening when we’re cutting the core/perimeter floors, but this initial fix should take care of all your issues with floor heights at the least.
I’m going to make a pull request with the fix, in the meantime here’s a file with the new SplitBuildingMass components that are being developed, that addresses most of your issues.
Yeap, it should be cutting the floor heights consistently now, but something is occurring in the second function: deriving the perimeter and core geometries, that’s causing certain floors to fail. This appears to be a different problem that I’m hoping to address when I have the time.
