Disclaimer and Social

This page may contain affiliate links. As an Amazon Associate I earn from qualifying purchases.

Sunday, January 21, 2018

Good Enough to Start With - Cut and Run

I got something right on the first try, then spent a full day trying and failing to make it better - only to realize it didn't really need to be improved. But like most dead ends, I learned a lot of useful stuff along the way.

Last week I posted about a 3d printed template for cutting, slashing and modifying fabric. The design was a success, both because it was easy to customize and print new designs with only a few minutes work for each pattern, and because it actually helped making clean straight cuts in tee-shirts.

Not satisfied, I continued to think of new ways to use this approach until I reached the limits of the first design. I wanted bigger better, smaller faster, simpler yet more complex designs. Are you beginning to suspect I might have gotten carried away?

Sure, the template worked. But I noticed that any design with the slots closer than about 10mm started to create problems. Common tee-shirt fabric, being all stretchy and flimsy, was particularly problematic. It would stretch, gather and distort when the cutting slots got too close together. This really sucked because slashing old tee-shirts was the original impetus behind the entire concept.

I also noticed that very thin or slippery fabrics like sateen, or very thick fabrics like denim didn't always work very well. If I was careful, I could do both. But I had to take care to hold the template just right and position and lay out the fabric correctly.

A simple, first try success wasn't good enough. I had to make it better. I had a serious case of designer fever.


What I needed were some cleats or ridges and matching grooves to really clamp down on the cloth and hold it in place.


So I started a redesign of the entire template system. I spent half of a day thinking about it and drawing simple sketches. What if I put cleats or ridges on the surfaces, maybe add matching indentations to really clinch the cloth and keep it from moving.

I did several sketches like that. I got a feel for a way to draw the "features" easily. And then started thinking about how to automate the process so that I could type in a few numbers and the pattern would magically draw itself.


The first (successful) version simply extruded shapes up from the ground plane, the cut or joined the shapes to create two slotted plates. But this new cleated, wavy system wouldn't work like that. Now, I had to extrude on two or even three planes and somehow coordinate the origins and geometries across multiple drawing steps.






That was kind of annoying. I had to redo all my old work/ And, because I wasn't sure which, if any. of the new shapes would work, I  needed to find the simplest method of automating multiple styles of cleats and clips and waves.

I couldn't find any one style that worked for all shapes, so I wound up generating several different files with completely different automation strategies.

Getting a flat face to match with another flat face is easy. getting curved and notched faces to mate with other curved faces with varying distances between them, well that isn't nearly as easy. Especially if the surfaces curve in all directions at once.

Automating that process is even harder. Boy did I learn a lot.


Or what if I made the entire surface wavy - yeah, that would really lock them together.


I spent about a full day concepting then drawing and automating these new designs. I was getting closer, but also getting frustrated. Then I went back to look at the results from the first tests and realized something.

I didn't need to do all this.

Seriously, I'm never going to be cutting two hundred strips of cloth that are only 2mm wide. That's just ridiculous. Look at any example of slashed tee-shirt mods. They have a few (10-20) cuts and the strips are at least an inch wide, not a hundred tiny cuts.

This project was supposed to be about slapping down a template on a worn out tee and making 10 or 20 relatively straight cuts in a few seconds.

I had gotten carried away and was trying to create a process that would rival the precision of a laser cutter. Why do that?

So I stopped the design and tried to feel better about the day I had just lost to stupidity. It was probably the easiest part of the day. I did actually learn a lot, and any time you find new ways of doing things it's a win. If not today, then sometime soon.

There are parts that I can still use in this design. I can use some of the new ideas in other projects. I'm working on a metal and paper embosser where these techniques are useful. A lot of these ideas are also applicable for sliding joints for lids and cases. And this exploration helped me get a better grasp on basing multi-plane geometries off a single origin point.

So like most failures, I got a lot of useful information to use. So maybe that day wasn't a complete loss after all. Now I'm off to cut up more old shirts.





Also find me on: FACEBOOK