Excited with the new AFN component.
The example is easy to follow and understand (have some doubts on the results but this is a different issue). Some of the internal surfaces are defined as outdoors from start, as i understood.
My question here is regarding the workflow on more complex cases, for instance when modeling a building in DF. How to deal the “outdoor” state on internal surfaces?
The question akes sense?
Hey @AbrahamYezioro ,
Sorry for the late response. The Dragonfly schema supports interior windows. You just have to apply the
WindowParamters before you plug the
Rooom2Ds into the “Solve Adjacency” component. So, usually, you should be applying the window parameters in between the “Intersect Room2Ds” and the “Solve Adjacency” component if you want interior windows with Dragonfly. Also, to ensure that the interior windows are operable, I think you’ll want to make sure you are applying the “DF Window Opening” before you solve adjacency.
With that approach, you’ll get a naturally ventilated Honeybee model from Dragonfly that you can pass through the “HB Airflow Network” component before simulation.
But, with all of this said, you will never be able to get nearly as good control over geometry with Dragonfly as you can working directly with Honeybee. So, given how precise of a tool the AFN is, I’m sure there are a lot of cases where it’s better to use the “Honeybee scalpel” over the “Dragonfly sledge hammer.”
I need to think on a case to apply what you recommend. Since one of your original advises was to use DF when you have a multiple floor building (rather than modeling it in HB), i’m trying to follow and see how to implement the AFN. Usually, in DF, i don’t define interior partitions at the apartment/office level. Just i define apartments, offices, core as one space.
Would you say that for such cases AFN is still better (in principle) than the natural ventilation process i used so far?
I didn’t mean to imply that you shouldn’t build large multi-floor buildings with Honeybee. It’s true that Dragonfly lets you work at a higher level of abstraction from Honeybee so that it’s easier to build large models if they abide by certain rules (like all rooms being extruded floor plates). But Honeybee can still do all that Dragonfly does and more so, though it might take a few more components. If its the ability to organize Honeybee models into stories that you are wondering about, there are Honeybee components that separate rooms by floor height and assign story IDs to the rooms, which effectively provide the same functionality as Dragonfly stories:
And, to answer your question about the AFN, the question of the “better” model usually comes down to whichever one you put the most effort into setting up correctly. The AFN can certainly model many more cases of air flow than the simple ventilation objects can model but, if you don’t spend the time to set up the interior openings correctly, then the AFN isn’t likely to be any more accurate than the simple openings. With this said, I have found the AFN to typically have lower overall air flows compared to equivalent models with simple ventilation objects and I think this is to be expected when moving from the simple ventilation’s “all or nothing” assumption about cross ventilation to the AFN’s more sophisticated way of modeling wind-driven flow.