Handling Complex Geometry in Energy Modeling with Ladybug Tools – Need Guidance

old forum
1

Hi everyone,

I’m working on an architectural design that includes some pretty complex geometry—non-orthogonal surfaces, curved façades, and multiple split levels. I’m using Ladybug Tools (mainly Honeybee) for energy modeling, and while it’s a powerful suite, I’ve hit a bit of a wall in terms of accurately translating the geometry into the simulation model.

I’ve read some posts about simplifying geometry for better performance, but I’m unsure how much simplification is too much, especially when accuracy is key. Are there workflows or best practices for managing intricate shapes in energy simulations without losing too much detail?

On a related note, I’ve been exploring what is generative AI and wondering if anyone’s used AI tools to help optimize or even generate building forms compatible with Ladybug workflows. Could be an interesting blend of tech and sustainability.

Any guidance, tips, or resources would be amazing. Really appreciate the community here—such a great space to learn and share!

Thanks

2

Hi @oliverlewi ,

It is an interesting question and personally I think it is hard to establish too many ‘rules’ regarding how much simplification is ‘too much’. Certainly some standards like ASHRAE 90.1 Appendix-G try and provide guidelines in places (merging residential zones by orientation, etc…) and so you might try taking a look at those?

Personally, the way I tend to think about it is to first consider the ‘type’ of building I am working on: ‘envelope dominated’, or ‘internal gain dominated’. Depending on which case I am working on, I might choose to spend my time focused on different aspects:

1) Envelope Dominated:

The smaller the building and the lower the internal loads (people, equipment, lighting) - the more the ‘envelope’ dominates the energy profile. In these cases, it is relatively more important to get the envelope ‘just right’ in order to get an accurate result.

Though do also consider that ‘geometry’ is only one aspect of getting the envelope ‘right’. In these cases it is just as important to get the assembly u-values calculated properly, consider thermal bridges, calculate window u-values correctly, etc… so even in these types of buildings, the geometry is only one piece of the puzzle.

2) Internal Gain Dominated:

By contrast, the larger (well…maybe more specifically: the more ‘compact’ the building - ie: more internal floor area, less exterior surface) the building, and the higher the internal gains - the less important the envelope becomes. In these cases, it is relatively more important to spend your time getting the loads, schedules, internal zones, and equipment ‘just right’. (not that you can totally ignore the envelope :wink: I’m just saying about where to put the ‘focus’ of your effort)

It also depends a bit on the ‘phase’ you are in. During early project phases, you may reduce the complexity significantly in order to get faster results, knowing that many things will change, but as the project develops and items are finalized - typically the ‘resolution’ and definition of the model will increase as well.

You might also consider the Pollination Rhino Plugin which, I am told, is a big help when modeling complex geometry in Rhino. I have not used it myself, but something to consider for sure.

So yeah: I think it is hard to say there is any ‘rule’ (as far as I know), but I would start by considering the use-type, and the size/compactness. The more ‘envelope dominated’ the building is, the more ‘exact’ the geometry should be in order to get useful results.

hope that is helpful food for thought,
@edpmay

2 Likes