Procedural support attachment points

Is there a way to define support points procedurally? I have a series of tiles, about 4cm x 4cm x .5cm. While the set of tile shapes is fairly consistent, there are some 500 tile surface variations so effectively I have 500 unique tiles. I’d like to be able to define where the support tips are attached myself, i.e. take one tile, have PreForm do the best it can with it, modify the supports slightly if need be, test print, then feed that information into PreForm directly for the next 499. Obviously I don’t want to make the same hand adjustments 500 times over. Is there a way to supply a file that specifies support attachment points?

I don’t think there’s such a function in Preform, on the other hand, since you designed your own tiles, why not design the supports right in, and forgo, Preform’s supports altogether.

Yeah, I was thinking of that too, though there’s a fair amount of know-how in the geometry of the supports that would be easy to miss if I was recreating them from scratch.

It did get me thinking, though, could I make a support system like they have on plastic model airplane kits, i.e. have not just a single layer of tiles, but 3 or 4 tall? One of the challenges with my setup is there are a lot of unique pieces. If I only print a single layer I can get 18 of them at once, but if I could do 4 layers, that’s almost 80 tiles in a single print run.

That would be efficient, but I would worry about betting such a huge batch on one run. With more models and complex supports there are more points of failure which might ruin the whole or most of the batch. But I don’t want to be a scare crow, it would be great to see something like this work.
Meshmixer might be an option to try for building custom supports.

Stacking the tiles is not a good idea, since the supports for the upper layers will go through the lower ones. You will almost certainly be able to print more in a run by printing the tiles at a slight angle and on one edge. By choosing an optimal orientation for each tile, you will need to move fewer points to get the results you are looking for. And with 500 tiles, you will learn the best placements quicker than any AI that could be added on to Preform.

I would model your supports in your 3D application, then you could just copy and paste it for each tile and export that as-is to PreForm and essentially print to the platform without Preform adding its own supports. You could also use Preform to figure out support shapes and try to replicate that in your 3D software.

This seems to be the consensus, namely creating the supports in my 3D app. Does PreForm have any hidden options to save out the geometry with the supports in .obj or .stl format?

While PreForm doesn’t have functionality for procedural support generation, stacking objects atop each other is totally doable. We used this for the Durable Resin joints in the Fuse Pavilion. The easiest way to go about this is to stack the objects atop each other in your CAD software and then export as a single STL.

We’ve seen the suggestion for procedural support generation a few times in regards to regenerating similar supports when importing an updated version of a model with slight changes. This is something we’re looking into though as you mentioned, the support generation algorithm is nuanced and there are a few quirks that make this challenging.

As I was exploring creating supports myself, what I wanted to be able to do was limit where PreFrom could attach supports. A simple solution (from the user’s point of view) would be an attachment group specified via the .obj file: if group exists, run supports only to those polys. I also thought of having a group to tell PreForm “don’t attach here, these faces are precious”.

What is unfortunate when making ones own supports is losing all the great smarts in the truss building of the supports. As previously suggested, I used PreForm’s supports as guides to modeling my own attachment points: short tapered tubes with the protruding end rounded, a sort of “come attach here!”. But right now it’s an all or nothing approach as I can’t designate the butts of those tubes to be the only attachment points without building my own entire truss structure.

Continuing to expand the capabilities of support structures and improve ease of removal and finishing is a top priority that we’re continuing to explore. There could be a few different implementations of a sort of masking algorithm that selectively adds supports only to highlighted areas. For now, manually removing supports may be easiest and you can bulk remove supports by dragging your cursor which might help to speed things up.