Tips on optimizing 3D tiles to reduce computational time on ShapeDiver

Grasshopper ShapeDiver
1

Hi!

I have been working on a ShapeDiver version of a script I did for our company. Basically, it’s a script that produces a patterned tile layout using 2 types of 3D tiles. It works but it quickly reaches its computational limit when the wall length goes over 6 meters.

image
image426×122 5.12 KB

We’re hoping we can set this to allow for at most a 10-meter-long wall.

Before this I have already optimized the script to not include any boolean operations and avoid trimming altogether. I have split the 3D mesh model of the tiles to minimize triangular faces as much as possible (see the screenshot below for the current topography of the 3D mesh tiles), and used texture files no larger than 100kb in size.

image
image1920×1080 112 KB

I’m just hoping if anyone here have more tips I can use to further optimize this model?

The (private) link for the model view is here: https://www.shapediver.com/app/m/canvas-r8-2025-04-23-web-3

Thanks in advance!

2

You definitions outputs thousands of copies of the same mesh (which contains about 300 vertices), which quickly hits the limits of your account, as well as reasonable limits for browser interactions.

Instead of sending copies of the same mesh, you can send it once along with an array of transformations, which will make your model faster while also staying largely under size limits. Check out this video to learn how to use transformations: https://www.youtube.com/watch?v=d2Nkk7gKn4o

3

Hi Mathieu! Thank you for your feedback and for sharing the tutorial. I was able to pick something from that and now the script and model view is more optimized.