Computer Controlled Cutting¶
- Cut something on the vinyl cutter
- Characterize your lasercutter’s focus, power, speed, rate, kerf, and joint clearance
- Design, lasercut, and document a parametric press-fit construction kit, which can be assembled in multiple ways, accounting for the lasercutter kerf
Vinyl cutter¶
Characterization of Silhouette Cameo 3¶
At the Lab, we use a Silhouette Cameo 3, a small vynil cutter. We have to use a proprietary software (Silhouette Studio) and it’s not downloadable on linux, so I have to use a computer from the lab. There is only one blade, that you have to set to the right depth before inserting it in the machine.
You can change the depth manually: you need to insert it into this hole and turn it until the number matches the depth you want.
We have two types of vynil on the lab, with different parameters.
| Classic vynil | Flex (for hot transfer) | |
|---|---|---|
| Depth | 2 | 1 |
| Force | 12 | 8 |
| Speed | 5 | 5 |
The Silhouette Studio software is only reading DXF files, so I have to prepare my designs in Inkscape accordingly. When saving, choose Save as > .DXF
You can now import your design directly into Silouhette Studio and virtually place it on the mat. Don’t forget to change the parameters to match the type of material you want.
Once you are satisfied with the position and the size of your design, you can cut a piece of vynil and place it on the cutting mat. Then, you can load the material into the machine and send the task to the vynil cutter. This is a quick process, the long part comes after!
First test¶
I decided to cut a QR code to paste on my notebook with a direct link to my documentation to test the vynil cutter process.
QR Code generation in Inkscape
I found a really cool and easy way to create a QR code directly in Inkscape. If you have a recent version it’s already integrated in your basic extensions.
Just go to Extensions > Render > Barcode > QRcode, and insert the URL you want to transform. It will generate automatically a vectorial version of your QR code.
The process was tedious, and the transfer paper we own not very good (or am I not patient enough?): I had tested the workflow and it worked, so I decided to make some other designs rather than loosing my mind in transferring something that complicated.
Transfer process¶
I wanted to test the transfer process so I made a simple annotation in the Bauhaus police. It wasn’t easy as the letters were very small. With two tests I managed to transfer the entire annotation on my notebook: it looks great and I learned a new process, I’m happy!
T-shirts
To be continued. I really want to try transfering some cool designs on a t-shirt. Maybe when the 
will return?
Laser cutter¶
I already used a laser cutter in the Carrefour Numérique (a great fablab in Paris) and in the fablab Digiscope. At these occasions I made a characterization test model and cutted Okoumé wood for an easel.
Set up¶
We have a Trotec Speedy 100 in our lab. The first thing to do is to turn it on (button on the back left) and let the plate go down until the bip. After the initialization, you can fix the material of your choice on the plate and make sure it’s flat (we use masking tape). The material must be placed on the origin (top left of the plate).
To make sure the focus is set, we use the focus tool: we place it on the side of the lense and move the plate up until the focus tool falls (see picture above).
Once you have set up the laser cutter, it’s time to transfer your designs on the appropriate software and choose the material parameters.
Parameters on Job Control¶
I keep track of the best parameters on my global fab document.
| CUT | Power | Speed | Frequency |
|---|---|---|---|
| Cardboard 4mm | 80 | 2 | 1000 |
| PMMA 3mm | 100 | 0.80 | 2000 |
| MDF 3mm | 100 | 1 | 1000 |
| CP Plywood 3mm | 100 | 1.85 | 1000 |
| ENGRAVE | |||
|---|---|---|---|
| PMMA 3mm | 100 | 100 | 1000 |
| MDF 3mm | 100 | 60 | 500 |
| CP Plywood 3mm | 50 | 80 | 500 |
In Job Control, the parameters library looks like this.
How to test a new material¶
Most of the materials we have at the lab have been tested already, and we have a good library of parameters. But we don’t have any parameters for 4mm cardboard, so it was a great opportunity to take notes of the workflow.
Tips
- Start with parameters from a similar material
- Prefer low values for a beginning
- Keep close to the laser and be prepared if there is any chance of a fire
To test the cardboard, I started by testing with Power: 50, Speed: 2 and Freq: 1000. It just cut through the first layer of the material. I augmented the power to 75 and it almost worked, I could detached most of the design from the material. I finally settled with 80 for power: it cut through the material without burning it and I can do it in one pass.
Kerf test¶
Now it was time to find out the kerf, which is the laser diameter. Once you figure out what is the kerf, it’s easier to make a press-fit right. I wasn’t expecting to lose so much time though, so for future references, here is the story of the first afternoon with the laser cutter…
Try again… Or the history of one lost afternoon
I thought it was gonna be a simple afternoon of laser cutter characterization. It resulted in three hours of frustration and incomprehension… because of one parameter. In Job Control, none of my designs appeared to be detected: they were showing on the screen but there were no time estimation for the cut. After many tests, we finally realized that an opacity parameter was activated in my Inkscape version and caused all of that trouble. We realized it when studying two files (one created on my computer, the other on the lab computer) with a text editor.
<rect
style="opacity:0.6;
fill:none;stroke:#ff0000;stroke-width:0.26458333;stroke-linecap:round;stroke-miterlimit:4;stroke-dasharray:none;paint-order:markers stroke fill"
id="rect1232"
width="100"
height="20"
x="5"
y="35"
ry="0.15939742" />
After realizing that, it was easily corrected and I could finally characterize the laser cutter.
We made several squares to measure the kerf of the laser cutter, from 20mm to 20.5mm. After measuring all of these with a digital calliper, we concluded that the kerf value was equal to 0.2mm.
| Inkscape measure | Reality |
|---|---|
| 20.0 | 19.62 |
| 20.1 | 19.72 |
| 20.2 | 19.82 |
| 20.3 | 19.94 |
| 20.4 | 20.04 |
| 20.5 | 20.12 |
Comb test¶
I then made different combs to test the best finger joints with our different materials.
| Material | Press-fit |
|---|---|
| CP Plywood 3mm | 2.9 |
| MDF 3mm | 3.0 |
| PMMA 3mm | 2.8 |
Caution
I intially made a parametric comb in FreeCAD, imported it in Inkscape and started to cut a piece of MDF 3mm. But even if in Inkscape everything seamed normal, I realized that the laser passed two times instead of one. It made a big flame and I stopped everyhting immediatly. I need to figure out why FreeCAD to Inkscape is such a mess, because I would like to have a parametric workflow secure!
Joints¶
For a first test and because I wasn’t sure to have the time to modelize every joints type before the week-end, I took the file Neil showed us during class and tried to export it in Inkscape. As I had the double-line problem as explain earlier, I wanted to figure out why. I finally discovered that when I exported my files from the TechDraw workbench in SVG, FreeCAD exported all lines AND a complete sketch of the object. It resulted in a double pass on the laser cutter.
| Double-line invisible | Double-line visible | Underlying sketch |
|---|---|---|
 |
 |
 |
To solve this issue, I had to export my sketches as Flattened SVG and not through the TechDraw workbench. I will circle back on that if I find another solution, because it’s only working for sketches: if I add another sketch on the orginal padded form for example, the flattened SVG option is no longer available and I have to correct the file in Inkscape.
Export a SVG sketch from FreeCAD
Select the sketch in the TreeView, Ctrl + E, Flattened SVG.
I made my own joints because I wanted to train myself to make parametric designs.
Engraving test: QR Code¶
I tested the QR code design I made for the vynil cutter, this time in PMMA 3mm. It works perfectly, even with the transparency of the plexyglass. I just have to avoid too much reflections when scanning it.
Parametric designs¶
I wasn’t expecting to have such trouble finding an idea for the press-fit construction kit. I wanted to find an elegant solution for my final project but couldn’t figure out during the week-end how to do it. I finally made a pretty simple parametric model to have something to show on Wednesday because I started to slightly panick about that.
Press-fit kit¶
I wanted to make a design with as little parameters as possible, so I chose to make a kit with circular and octogonal forms.
| Alias | Value |
|---|---|
| diameter | 20 mm |
| thickness | 4 mm |
| kerf | 0.18 mm |
I designed the simple forms in FreeCAD (see the next section for more details on parametric design) and exported them in Inkscape to send them to the laser cutter.
I then had some fun with the kit 
Final project¶
I wanted to try to design finger joints myself. It was much harder than expected, and I may be using one FreeCAD addon for my final project. Here are the steps I followed to make it (you’ll see I made some mistakes during the initial conception that I corrected afterwards).
Finger joints¶
I first created a Spreadsheet with values needed for the design. I wanted the kit to be adaptable to the final size of my final project in the end, even if my tests were going to be small. I used auto-alias to make my values accessible during the modelling.
| Parameters | Values |
|---|---|
| material_thickness | 4 mm |
| kerf | 0.2 mm |
| height | 30 mm |
| width | 20 mm |
| slot_size | height / 9 |
I then made the first rectangle with basic dimensions, without forgetting to add the kerf value and then padded the rectangle with the material thickness parameter.
I sketched a first slot on the right side, accounting for the kerf value.
With the slot sketch, I made a pocket of the size of the material thickness minus the kerf value.
I then took the pocket feature and applied a linear array to it. I chose arbitrarily to make five pockets (and the slot size is the height of the model divided by ten).
To make the linear array correctly you have to set up the length to Spreadsheet.height - Spreadsheet.slot_size.
I projected the slot sketch and extruded it on the other side to make the opposite operation (this was a mistake, I corrected this step later on by making pockets instead).
I finally made the top and bottom slots with external geometry and symetry to keep the model light.
And I obtained this piece that I cloned to have a visual idea of the kit.
I exported my designs with the TechDraw workbench, corrected the double-line problem in Inscape and send two pieces to the machine. This is when I realized I made some mistakes with my kerf calculations, and I iterated several times to find the perfect values.
I ran out of time to test more of my designs for the entire house, but I learned a lot and it will be easier to prepare the scenery for my diorama after this week. It may not be fully parametric to keep track of proportions. We’ll see 
Engraving a facade¶
We are getting closer to the electronic weeks, and I wanted to have a small prototype of the house to be able to put a light in it and make some simili-real tests. Moreover, I wanted to get ahead of the design for the facades. So I spend some time putting together my design of the houses I made on the Computer Aided Design week with little additions I made in Inkscape.
As I wanted a minimal design, I did’nt think that my windows would be too small for the laser, so I made a little mess (but learned how to clean the honeycomb…)
And here is the result: I can see there are flaws in my design already, so it will be a great base for improvement!
To be continued, I didn’t have time to finish it this week
Bonus: Rimbaud¶
Arthur Rimbaud was a famous poet, born in Charleville-Mézières, my child town. He’s one of the favorite writers of my mother, so I decided to surprise her with a gift made during the Fab Academy. I took one of his picture (CC-BY-SA), vectorized it and cleaned it to be suitable for the laser. I tested the capabilities of the Trace bitmap module in Inkscape and smoothed the different results with Ctrl + L (see screenshot below).
Files¶
Vynil cutter files¶
Laser cutter files¶
Final Project¶
About this week¶
I didn’t think it was going to be that hard, but maybe because I was expecting to make something really good for my final project and was disappointed by my own imagination. Nevertheless, I made it, and it was a great week in the end. I learned a lot about assembling and it will be very helpful for my final project!