I've always wanted to create my own custom t-shirt designs, since I like drawing and graphic design. T-shirts designs are often made using screen printing, in which a screen is covered in emulsion which is sensitive to UV light. Then the design is placed on top of the wet emulsion, and the emulsion not covered by the design hardens. The emulsion around the design is then removed by washing the screen out with water. When the screen is placed on a shirt, ink is wiped across it, and the sink seeps through the porous areas of the screen onto the shirt below, forming the image.

This process can be finicky as there is variation in the cure time, emulsion application, and light sensitivity. It's also expensive, since a new screen needs to be purchased for every design. I decided to attempt a new way of manufacturing screens using laser etching, which turned out to be much cheaper but had inconsistent results. The shirts I created are featured at the bottom of this project page.

Stencils have long been used to create designs applied with paint or ink. Stencils are also very easy to cut on a laser. However, stencils have some major limitations in that you cannot have any internal islands that are not connected to the stencil frame. This makes designing images for stencil based art much more difficult.

Instead, I decided to use laser etching to create a screen. While typically screens start empty and then gaps are filled with emulsion, I would instead start with a solid material and laser etch holes to create a screen effect.

When a laser etches something, it does so by moving back and forth across the part and burning tiny dots onto the surface. This is exactly what I needed to create my screen.

I purchased a pack of thirty 10mil Mylar sheets off of amazon. These were ideal because they were very thin, but still somewhat rigid and most importantly safe to laser cut.

Originally I had wanted to be able to print multiple shades of the same color ink by varying the dot density in a process called dithering. The laser software I was using (RD Works) could perform this shade to density conversion automatically. However, I soon found that this concept of multiple shades was not as easy as I had imagined.

The main limitation of this was the heating of the plastic. When the laser cuts a dot, it needs to generate a certain amount of heat to melt through the plastic sheet. When it does this, some heat is distributed to the area around the dot on the plastic sheet. This makes cutting the next dot easier for the laser, since some of the heat it needs is already in the plastic. This means it takes less laser power to cut high density dots, and more laser power to cut low density dots.

When running in one operation, the laser used a single power setting. This meant that in order to cut the low density holes, I had to raise the power. However, this meant I was severely overpowered on the high density holes, which caused a few problems. First, the high density holes became larger, and thus the amount of plastic material left to hold together the screen was significantly less, and so some areas of the screen became to disintegrate. The second problem was warping, which occurred because too much thermal energy was being pumped into the material. My experience with warping is detailed in the next section.

Ultimately I found that I could not generate consistent results with dithering, and instead settled for solid colors determined by the ink color used.

RD Works method of dithering shades of grey. The result is a nonuniform distribution of pixels, which results in uneven uniform and warping

Closeup of the pixelated pattern. Extra black pixels have been added along vertical edges, but not on the top and bottom edges

Each Mylar screen was taped to a thin, laser-cut, wooden frame to help give it structure. The screens were then placed onto the surface of a shirt and clamped in place. A healthy amount of screen printing ink was squeegeed on to the surface of the shirt. The ink was pressed through the holes in the Mylar and into the fabric of the shirt. The screen was then removed, and the ink allowed to dry.

Sadly, the screens did not typically yield good results after 2 or more prints. However, my final method of laser etched screen printing was perfect for one-off, fast shirts. It only took me about 30-45 minutes to produce a screen and print a shirt once I had created my design. This made my method perfect for creating my own personalized shirts, but not for high volume production.

All of the shirts below are ones that I created using the laser etched Mylar method. The shirts were made at various stages of the my iteration process, so they have varying quality. Most of the designs feature original elements, renditions, or combinations of logos and designs found online.