Alternative approaches to the age-old standard perf pattern?

[rjohnson1 (Rjohnson)]

I’ve been using Grasshopper for over 13 years, with my first built application being the standard circular perforation pattern (In fact, it’s the first thing I taught when I taught a university grasshopper class - so it may have been lots of our first application of computational design)

(Pic of my first every grasshopper script that was fabricated - click here for more inf about the project)

Since then, I’ve frequently been tasked with designing custom laser-cut metal patterns - I bet its a dozen projects at this point. I am desperate to move away from “plain old circles”, but I feel I’ve hit a creative wall… I’m trying to find something innovative (and frankly interesting) that still respects the realities of industrial fabrication.

The primary constraint is the laser-cutting process. As many of you know:

1. Structural Integrity: We need a specific amount of support material to remain. It also depends on the thickness of the material as well as proximity to edges and if its bent, etc…

2. No Islands: We cannot have “islands” of material that would simply fall out once cut.

3. Kerf Limitations: I’ve explored “kerf-only” cutting or single-line “etching”, but the few industrial fabricators I’ve spoken to on other projects have noted that the kerf is so much smaller than consumer laser cutter and is much too thin to be visually effective at scale. (would love to hear your thoughts if you have had success or alternative applications of this)

I’ve tried a few things on real projects to date, but most have been more graphical in nature and usually have all these islands that either have constraints making it not possible. (I’ve linked to my work for examples where possible)

• KD-Trees: Currently using these for random but evenly spaced nodes (off-grid).

  

• Typography: Used letters by intentionally removing the islands, but this requires a large scale to be legible and isn’t always the right aesthetic.

  

• Dithering/Halftones: I love the look of true halftones and dithering (which I’ve used for vinyl graphics), but these typically don’t translate to laser cutting because they require dots to blend or vary in ways that compromise the metal’s structure.

  

• Graphical Blocks: Interesting for film/print, but usually full of floating islands - so not sure how to do with laser cutting.

  
  

Whenever I find something cool that seems like it could potentially be applicable to my work as a computational designer I add it to my random list on my website. (Link to my list here). The problem is that they are like an unformed thought… I haven’t been able to translate anything from these ideas and concepts I find interesting to real work. For reference, here are some of the things I’ve come across:

• Grids & Tiling: Organic Stålberg grids, Wang Tiles, Aperiodic sets, and the 3D modularity of Erwin Hauer.

• Mathematical Logic: Bernoulli Percolation, the 4-Color Theorem, and Poisson Disc Sampling.

• Graphic Algorithms: Dithering (Sierra/Floyd-Steinberg), Voronoi-based stippling (StippleGen), and Genetic Algorithms for image reconstruction.

• Distribution: Blue Noise and barcode logic.

I’m looking for new terms, theorems, project examples, that can translate somehow be fabricated from laser cut material to get me out of my rut:

• How are you moving beyond standard Voronoi or circular perforations?

• Are there specific mathematical theorems (e.g., related to maze-generation, pathfinding, or continuous-line patterns) that you’ve successfully applied to metal fabrication?

• Are there ways to make Voronoi or other cellular patterns “novel” again through specific constraints?

Thanks!

[old_taz]

Have you seen some of the projects that Zahner does? They do a lot with perf metal patterning.

Products & Systems - Zahner

Louvered ZIRA™ - Zahner

[Tom_Newsom]

Slots that respect the “direction” of the source pattern. Like woodcut or engraving but with dashed lines to maintain integrity.

Some fabricators might have a 5 axis machine, so you can experiment with slanted holes if the material is thick enough. Could do some fun things with changing patterns based on viewpoint.

Or make cool moiré patterns with multiple layers.

[Volker_Rakow]

If it ain’t broke, don’t fix it. I mean you’re lasering holes into sheet metal. A hole is a hole, and as you’ve noticed, islanding is a problem.

To give a sense of direction, maybe start using lens shapes / ellipses with some vector fields? I like @Tom_Newsom idea of layers – obviously more expensive than a single sheet – but you can get around some islanding problems with simply two sheets and also begin to work with alternative material reveal.

Have you tried bending out cut sections like fins?

There is only so much you can do within the constraints of the machining and material.

[Cole_Howell1 (Cole Howell)]

Something that I’ve been thinking about is to design a tiling system, then using wave function collapse or Wasp to aggregate tiles to laser cut. I’ve used a similar method for lattice structures, but your likelihood of creating islands is pretty high.

Naturally something like reaction-diffusion joining some perforations and creating paths along vector fields could give the work a sense of flow that might be interesting, not sure if that would compromise the metal though.

[rjohnson1 (Rjohnson)]

@old_taz and @Tom_Newsom - looks like Zahner has a decent example of what you were suggesting - it’s a great idea!

Curious if you all have ideas on how to generate the linework? obviously you can do what zahner is proposing , and draw some organic like shapes. You also could even combine it with @Cole_Howell1 's suggestion of wave function collapse, or something similar.

I’ll need to look into reaction diffusion to see how that could translate to something architectural.