I am currently working on a project involving the calculation of outdoor comfort using different options for external shading devices. The architect has been providing me with various designs, and I am testing them using UTCI simulations.
I have encountered an issue in my Grasshopper script where inserting meshes as shades significantly slows down the UTCI simulation process. I’m reaching out to inquire if there is a more efficient approach to account for meshes as shades without causing such a slowdown in the simulation.
I have attached a screenshot of my script for reference, which may provide better clarity on the issue.
Thank you in advance for your assistance and guidance.
Using Meshes is generally more efficient than the alternative of starting from Breps. So the answer to this:
Is not that the use of meshes is slowing down your simulation. Rather, it’s that your input geometry is very complex and it’s going to slow down the simulation whether it’s a complex mesh or a complex Brep.
So, if you want a faster simulation, you have to find a way to simplify the geometry. By that I mean using fewer mesh faces (or Brep faces) to represent the geometry.
I don’t know what this geometry looks like but, as an example, if your shade is like a pergola or has a bunch of small louvers that are all much smaller than the grid size of your UTCI comfort map (or a human subject), you probably don’t need to model the invidual shade faces. Just model the whole pergola as a single flat shade geometry and assign an ep_trans_sch_ to it that represents the percent openness of the shade.
Yea, that’s going to be a very long energy simulation since you’ll have 4,000 different shade geometries in the IDF that’s generated under the hood of the recipe.
An alternative suggestion is to just remove the Rooms from your Model and replace that geometry with Shade geometry. With no Rooms in the model, the surface temperatures of all of the geometry are going to be at the same temperature as the EPW air. So it’s not as accurate and it might not represent well the cases of heavy shading vs. sun. But that will cause the recipe to not run the energy simulation, which will cut the overall runtime by a lot. In that situation, all of the temperature variation that you see in the UTCI map is the result of shortwave solar and longwave sky temperature, which are likely the largest factors contributing to UTCI comfort.
For this:
It depends entirely on what the geometry is. If you can share a screenshot or a model, I can make some suggestions.