4. Laser cutting
We all remember the first time we rode a bike, perhaps the day that I learned to use a laser cutter might live on in my memory like that too. Just in case it doesn't, below is a guide for how to use Waag's laser cutter.
Henkette: Our laser cutter
Our laser cutter is a BRM, model number: BRM90130. Apparently it wasn't named yet, so we dubbed it Henkette.
So, where does Henkette's power come from? The large, water-cooled glass tube at the back of the machine.
-
Electricity is sent through the tube to excite a gas mixture (usually C02 w/ nitrogen + helium + etc.)
-
Mirrors at either end bounce the light back and forth (making the light that's created even stronger)
-
One of the mirrors lets some of the light escape, bouncing it off of other mirrors which are directed down the focal point
Parts
When we fire up a laser cutter, the laser, a chiller, and an air suction and air filtering system are all turned on. Before we get into the process of using the machine, here are some of the parts:
The laser tube.
The air and carbon filters.
The nozzle / focal point, the lense and fastener.
The air tube, which blows the particulate away from the lense (and also helps with cooling).
Preparing to cut
The steps for using the laser cutter at Waag are as follows:
You turn the red knob to the on position.
Next, flip the auxillary switch on. This will activate the laser, the chiller, and the air filter.
Then, we reset the machine.
Make sure the computer is on. Insert your memory stick here, when ready.
There is an on/off switch for the chiller below the computer (NB. Flip that before you start cutting).
After you've placed the material on the honeycomb table, make sure to adjust the focal point by placing the acrylic depth measurer underneath it and releasing the tightening screws. The focal point will drop to the acrylic underneath. At which point, re-tighten the screws and remove the acrylic piece.
If the focal point is set higher than the acrylic piece the laser cut lines will be blurred. Neil mentioned that sometimes blur is useful, but, as of right now I don't have plans to use it.
Lightburn
The app that we use to line up and initiate laser cuts is called Lightburn. When running our laser cutting tests, we imported our SVG files into Lightburn and edited them to our specifications for the task at hand. NB. Import .svg files, Open won't work.
Below is an example of the main screen in Lighburn.
In Lightburn, editing the shapes is easy, however, it's important to always consider the dimensions of your sketch against the material you're planning to laser cut.
On the right hand side of the screenshot above (at the bottom) is the control panel. That's where you access functions like Frame, Start, and Stop.
Under the Move tab is a GUI to move the laser cutter's focal point.
Importantly, notice the rowed section (on the right + above). Lightburn allows for layer-by-layer setting customisation. During testing, we relied on this feature to examine how the laser cuts different sections.
Double clicking on a row will open an expanded view of these settings. There are also switches in each row that can be toggled on/off. The Output switch tells the laser cutter which objects to cut, or not.
To add an object to a new layer, click on the object and then select one of the colored boxes at the bottom of the screen.
Each cut can be updated based on a few key settings. We mostly looked at the following:
- Speed
- Max Power
- Min Power
- Mode (we mostly used Line)
- Pass count
Unless it's essential, leave the Job origin setting unchanged (set to bottom-left). There are lots of checks to make before working with a laser cutter, so, by leaving the Job origin setting unchanged, you're decreasing the likelihood of mistakes.
NB. Once the laser cutter has started, don't leave the aread until after it has finished.
Laser cutter testing
Before I jump into this section, it's important to mention that the work of former students played a big role in guding us through testing the laser cutter. Notably, Sam and I often referred back to Nadieh's blog (Waag Fab Labe 2021).
In order to learn more about how the laser cutter works with different materials, we conducted a few tests. Varying results were achieved by adjusting the Speed, Max Power, Min Power, and Pass count.
Cardboard
Power + Speed test
The 2021 Waag class used a matrix to test the varying effects that happened across different speed - power combinations. The SVG file for our test is listed in my Digital files at the bottom of the page.
The first image is a screenshot of our Lighburn set up for the carboard power-speed test. The colors of the lines indicate which layer they're assigned to. At this stage, we had only programmed the first row. Each box was programmed to match the corresponding speed and power settings according to their column + row.
After the cut, flipping over the cardboard reveals how the material reacted to those settings. It's also worth noting that the Power: 10 column didn't do anything because there wasn't enough power to properly activate the laser.
After this test, it became apparent that there was an issue in our methods. For instance, why hadn't the Power: 25, Speed: 60 cut out the rectangle, but the 80 speed had? The same can be asked of the Power: 100, Speed: 60 -- but lower power attempts had.
The answer was that our Min Power settings weren't matched to our Max Power settings. Min Power is the power at the corners of the boxes. We hadn't paid attention to that and so the whole test was a fluke.
We fixed our mistakes and ran half of the test again to assess the power-speed combinations we were most interested in.
Ultimately, the tests we ran weren't perfect. In future tests, for projects, I feel strongly that I'd get more accurate results that would help aid with important decisions.
Kerf test
The kerf test is an interesting test, in my opinion. The “kerf width” is the width of material that the laser cutter removes as it makes a cut. The amount of material removed will alter the measurements of the objects you're working on. Therefore it's necessary to consider how much material is lost to the kerf of your machine. The amount should be factored into your calculations when designing models.
In order to figure out the kerf width, we cut a 100mm long rectangle into 10 smaller rectangles. Because the kerf size is quite small, the accuracy of the kerf width measurement increases as more more rectangles that are cut. A compounding effect.
Nadieh's notes were really helpful here. The cardboard needed to be cut in a certain direction to avoid them collapsing in on themselves, which would make measuring them difficult:
"The cardboard pieces were collapsing into each other with our first test. We should’ve cut the strips in the opposite direction of the cardboard, because of the wave structure within the cardboard."
The strips, when lined up and measured were not equal to 100mm. We use the difference to help calculate the kerf.
10 fingers = 11 cuts
Final measurement: 97.45mm
Difference after cutting: 100 - 97.45 = 2.55mm
Each cut removes material equal to the laser's kerf (the diameter of the laser beam).
/10 = .2318mm kerf per cut (aka diameter of the laser beam)
In CAD softwares you offset by the radius of the kerf
/2 = .1159mm
And after doing a bit more research, it turns out that's not totally correct. Waag's 2023 student, Michelle Vossen had this to say in her documentation:
By using 11 lines, the formula becomes easy as I have 9 inner lines that count for a full kerf, and then two outer lines that I only have to count half of, because the other half of the kerf is made on the leftover material.
Curve test
The curve test is designed to show the laser cutter's abilities in cutting curves and sharp edges at different speeds and power settings. Below is the image we decided to use for our test. The .svg file is listed below in my Digital files.
Noticed the difference in how the laser cut the material.
Focus test
The focus test is aimed at measuring blur. We place something underneath on end of the material, this is in order to the height different across the entire range of one axis.
Make sure to set the height of laser from the middle of that path. This makes sure to test the laser cut from near and far.
Acrylic
During the cardboard power + speed test we noticed that the Power: 10 setting didn't activate the laser, so we opted not to include it in our further tests. This was to save both time and materials.
Acrylic kerf test issue
During the kerf test, the laser was cutting the outline first and then the ten pieces. The problem there was that the first cut fell inwards before the smaller rectangles were cut. That was definitely going to affect the results.
To fix this, we split the lines in Lightburn's editor -- setting up the inside lines on one layer, and making sure the border was cut second.
It's worth pointing out that when setting up a kerf test, make sure not to create an svg file where the boxes are copy pasted next to one another, as that will unecessarily double the width of the lines -- and thus alter the results.
We also learned about the frequency causing waviness in acrylic cutting.
Focus test considerations
For our focus test, we tried to be too conscious of material wastage and ended up ruining the results. We tried to squeeze the cuts in a small area of leftover material. One of the cuts was too close to the edge and the thin edge warped into the empty space, meaning we couldn't really judge the if there was blur.
Wood
When corner testing the wood we noticed that there were differences in how burnt the material got based on the size of the object that was being cut. The same settings at smaller distances affected how easily the wood was cut / if it was burnt.
We cut a big circle at the same settings that we used for this little wing, and the wing ended up pretty charred.
Below is a photo of us cutting 9mm wood. The issue with this depth is that it requires multiple passes, which burns the material.
Baguette
A few friends of mine recently started a bakery and I told them I wanted to make a baguette lamp for them. I let a few baguettes go stale and covered them in acrylic spray to keep out any moisture.
On Friday, Henk gave me the green light to lazer cut the bakery's logo into one of them. The results were pretty decent. I will probably still hollow out the baguette and fit a light in there, but that's for another documentation page.
The first step was to open Inkscape, so that I could create an SVG of the bakery's logo.
Sam showed me a great trick that he had learned earlier that day. Pick an image with text > Path > Trace bitmap > adjust the Threshold toggle to the desired results.
We ran two cut attempts over the bread and the results were ok. Eventually, we decided to step up the power a bit and lower the speed a bit (Power: 80, Speed: 150).
For the third cut, Henk had noticed that the line interval setting could be improved. We changed those, as you can see below, and the results were better!
Laser cutter testing (digital files):
Parametric construction project
Documentation of the previous section took a lot of time to complete. I think that's mostly because I did a lot of work and documented after. For this next step, I'm going to try document along with the process of doing.
My idea
The ideation phase was a lot of fun. The project that stood out the most to me was to make a laptop stand. Ergonomics holds a special place in my heart (and back). Currently, at Waag, I'm using the cutout from a previous year as my laptop stand.
Making a laptop stand that is a cube is the simplest solution. It is also the most realistic, but we are here to dream, no?
I decided that I would try and make a laptop stand that resembles the facade of the Waag academy's building in Nieuwmarkt.
Parametric design
What are the advantages of using parametric design?
Every time you change the value of a parameter, its changes everything related to it according to the new value.
It saves you a lot of time, and is more likely to be accurate. It creates efficiency, and a work that you can easily use as many times as you need in other design projects.
Using Fusion
Found the width and depth dimensions of my laptop. Made a rectangle. Made parameters for that rectangle called macdepth and macwidth. Offset the sketch by a bit so that it's a bit bigger than the laptop (I can adjust this later). Extruded the new shape by macdepth (which is basically the height of my laptop when open). This will be the Waag cafe section of my design (aka the base).
I realised that this way wouldn't allow me to link the parts of the building parametrically, so I'll have to create objects that aren't offset.
I needed to make the upstairs section that would form the backboard for my laptop to rest on. I called this section upstairs. The upstairswidth was set to equal cafewidth, which is 10% bigger than the macwidth.
upstairsdepth is only 30% of of cafedepth. The other 70% will be where the opened laptop sits. I can alter the amounts because the design is parametric.
It was around the stage that I made it to the end of my "spiral one" plan that I realised I had hit a big hitch. My 3D modeling was working in blocks rather than thin boards (like cardboard). I'm not sure if there's a way to integrate what I've done with cardboard, but I decided to push on with my parametric design.
Each new sketch and extrusion measurement was linked to previous measurements. The design is divded into three levels: the cafe, the upstairs, and the roof. I found this really useful for adjusting the size of certain elements in the model.
Sunday, modeling with cardboard
Since I will be dealing a lot with parameter values, I set up a shortcut to open the parameters panel: shift+p.
Since I don't know the exact height of the material I'll be using, I made sure to give the cardboard a height parameter. That way I will be able to update all of the relevant objects simultaneously: cardboard.
I also don't know where will be the best place for the joints to be, so I set up spacing parameters for those.
Pressing d is the shortcut for opening the dimensions tool
After making the top layer of the cafe, I made the joint holes cutting a negative extrusion by the value of cardboard.
I'd like to test snap joints as the connections for the base of the cafe. Most aspects of the snap joint design are linked to other aspects in the model's design, but, as is true with some other elements, like the spacing of the joints from the borders, not everything is linked. By not over-linking, I hope that I will have the freedom to adjust certain parts as I test characteristics like strength.
To calculate the back and front sides of the base / cafe structure, I initially used the length of the cafe, less the carboard on either side.
I realized that was incorrect and that I had to add in the horizontal spacing that the depth joints were going to be from the edge of the cardboard. The updated calculation is below:
When I realized my error, and made the corrections, it messed up the position of the joint I had created. I realized that this was a case for better parameter setting, so I obliged:
Monday morning
I printed some examples of my designs and I realized that they were too complicated to edit and poorly named. I also realized that my design wasn't actually parametric... sigh.
Note: Export designs in DXF format, to use in Lightburn.
I started again and decided to set up more UserParameters, rather than ModelParameters.
The red box, above, is highlighted as it was causing my model to not be parametric. I tested this by altering the macbookWidth and macbookDepth parameters. Below you can see what the problem looked like:
After solving that problem, there was an issue when I adjusted the height of the model. I hadn't constrained the joint boxes to the sketch properly, so they didn't adjust when the parameters were changed. Results below:
Eventually I got everything working! In the screenshot I updated the macbookWidth and macbookDepth parameters to 11cm and 8cm and the whole sketch updates parametrically. I then used the scaled down size to laser cut a small version of the project.
Coming in on Monday and realizing that my model wasn't actually parametric was quite a big turning point in my understanding of this week's project. It was on that morning that I truly understand what parametric means. The big difference in how I did my sketching was that as I went along I made sure to test the model by shrinking certain parameters to see if the shapes function together properly.
Here is the digital file (version 9) of the final model.
NB. The constraining of my designs was not finished yet (refer to the following screenshot and link).
Irja explains about how to constrain an object.
PressFit kit
The design and assembly of my PressFit kit actually took place in week 5, after I had finished all of my work for that week. I had misunderstood the laser cutting assignment, and so I went back to complete the task.
Due to time constraints, I decided to make a simple (and very random) structure made of circles and long rectangles. Most of my time went into two processes:
- Making sure the shapes were parametric
- Comb testing the material (operating the laser cutter)
Comb test
I'll start with point #2. I made a few mistakes when printing the comb. Firstly, I tried to use an SVG from Nadieh's documentation that was way too small. Then, I designed my own one in Fusion, but the increasing incremental joints heights were too big and the combs wobbled around. I then made the joint heights decreasing from 6mm to 5mm, and it worked.
The width of the cardboard I was using was 6.18, and when I measured 4 pieces stacked on top of each other, the width/4 = 6.08. There's a chance that my initial measurement was wrong, but I forged onward with 6.08 as the material thickness, in any case.
At 5.3mm the two combs connected nicely with one another. I put that into my design under a parameter called cardboard.
Parametric design (again)
Since I had already done a lot of parametric design, making the PressFit kit was good for practicing what I learned. Along the way I also picked up a few new tricks.
Firstly, the pattern tool is useful for duplicating shapes around an object. For my PressFit kit I needed to make evenly spaced joint notches in a circle. The pattern tool made that a breeze.
My initial plan was to add the notches to my circle sketch, but I didn't feel confident that the object was constrained properly, so I decided to make a second sketch for the three notches. While in sketch mode, I:
- Pressed
p - Selected the circle's edge
- Created one notch
- Constrained it
- Patterned it around the circle
- Finished the sketch
- Extruded the notches as a cut through the circle.
With Neil's words from the previous week's review still ringing in my ears, I decided to chamfer one of the joint entrances. The notable aspect of this process was that I opted for a Two Distance chamfer, as it seemed the most efficient / effective thing to do.
NB. I created a Meme directory to store some that I've made.
Drawing view
Michelle told Sam, and Sam told me, about this lil trick that makes exporting objects from Fusion, for LightBurn, a bit nicer.
On the Fusion toolbar, where it says Design (usually), select Drawing. Then, use the settings in the screenshots below, and export as a DXF file.
PressFit final outcome
I am a bit frustrated that I didn't get to make something cooler, but I guess that's a fun kind of frustration. But after staring at the finished product for a while, I realized that the object could be made useful, and so I was happy again. The kit became a holder for my MCU after I laser cut some rectangles in the middle of the circles so that I could fit the USB C cable through them.
Digital files:
Helpful resources:
Vinyl cutting
The Roland GX 24 was giving issues, so we ended up not using it in the end. Here's Henk showing us the blade. The Cricut and the Roland vinyl cutters use the same kind of blade.
The open source tool we'll use to control the Roland GX 24 is called Mods. The Mods site allows us to control the vinyl cutter through our web browser -- although it only works through Chrome.
Cricut
Since the Roland machine wasn't working, we moved on to the Cricut Maker 3. Henk doesn't like it because it requires that you set up an account, but other than that it was pretty easy to use.
The Cricut comes with a mat that has a stickiness on it's surface. That is really useful for feeding smaller pieces through the machine that would otherwise not be picked up by the roller mechanisms. The roller mechanisms are essentially what move the vinyl pieces up and down it's y-axis, while the blade runs from left to right.
Overall, sticker printing was fun, and a lot easier than parametric design. The two prints that I'll highlight are the Neil sticker and the Farine sticker. Oh, and we also made a sticker for the naming of our laser cutter:
Farine is the bakery, whose logo I laser cut into a baguette. I figured it would be cool to make a vinyl cut sticker of their logo. I was able to test how well the Cricut machine could handle small, detailed cuts. As you might guess, the small letters didn't turn out well. The rest was great though!
I covered the larger logos in transfer paper for safer transport and so that when they get used, it'll be easy to manage all of the tiny bits.
The second sticker project was making the Neil stickers. Irja and I both followed different paths designing the stickers of Neil's silhouette.
CuteCutter
I removed the background in Adobe Express and then used a site called CuteCutter to edit the photo. I thought it worked remarkably well. You can control the Blur, Edge contrast, Threshold, and Speckles. My goal when adjusting those settings was to reduce the amount of noise in the silhouette while maintaining a likeness to Neil.
There is also a great Touchup option that acts as an eraser, and also rounds off erased edges in a way that keeps the image looking natural.
NB. It helps to start with an image that has strong contrasts.
Here's another CuteCutter attempt of mine.
