Week 03 - Computer-Controlled Cutting

This week we have the following tasks to complete:

• Complete your lab's safety training.
• Characterize your laser cutter's focus, power, speed, frequency, kerf, joint clearance, and joint types.
• Cut a design using the vinyl cutter.
• Design, laser-cut, and document a parametric construction kit while accounting for the laser cutter's kerf. The kit should be capable of multiple assembly configurations.
• Optional: Include non-flat elements.

Group Assignment

How to Use the Laser Cutter?

Our group assignment initially faced some challenges due to suboptimal communication, but ultimately, it progressed smoothly and successfully. Before operating the laser cutter, the following steps must be completed:

• Put on laser safety goggles from the storage cabinet.
• Open the top window and install the ventilation cover by inserting it into the window frame.

• After installing the ventilation system, press the top button to activate the exhaust.
• Press the button on the chiller to activate the cooling system for the laser cutter. If the chiller is not activated, the laser cutter will display an error message.

• Flip the main switch to power on the laser cutter (the laser tube remains deactivated at this stage)..

• Outside the room, above the door, the laser warning sign will light up, indicating that anyone entering must wear laser safety goggles and that the door should remain closed. Unnecessary movement in and out of the room should be avoided.

• Next, place the material inside the laser cutter by opening the large top lid.

• The focus has a significant impact on cutting results, so it is crucial to set the height correctly using the two buttons on the right side of the laser cutter.

• The focus point is correctly set when the focus gauge can barely move freely between the material and the laser carriage when placed on its long side.

• To make positioning and preparation easier, a small light can be activated to indicate the laser's position once the focus is set correctly. To enable this, flip the second lever on the top right of the laser cutter.

• The laser cutter will not operate if the top lid is open. This is ensured by a micro switch that detects the lid’s state (open/closed).

• The laser cutter is controlled via the panel located on the top right side:

I recommend using the left side of the laser cutter because our honeycomb bed is the flattest there.

• Once everything is set up, activate the laser tube and start the laser job from RDWorks.

Characteristics of Our Laser Cutter

Focus

To better understand the focus and beam path, we used a piece of PU foam. This material is easy to cut, even at greater thicknesses. For visibility purposes, the foam needed to be thicker than 20 mm, as this was the thickness of our focus gauge. We positioned the bottom of the laser carriage directly onto the foam to visualize the entire beam path.

The laser cutter then performed a straight cut through the material. We made the cut in the center of the foam to obtain a clearer cross-section without needing to balance the parts later on the table.

After cutting, we placed the pieces on a flat surface and marked the point where the beam appeared to be the narrowest.

The final step was measuring the distance between the surface top and the marked focus point. It is crucial not to flip the parts, as this would result in incorrect measurements.

In our case, we determined that the narrowest point was 15 mm below the top surface, meaning the focus point is 15 mm beneath the bottom of our laser carriage. Our current acrylic focus gauge aligns the focal point 5 mm below the material surface. To optimize future cuts, a 15 mm focus gauge could be used, or, for cutting 3 mm cardboard, a focus depth of 16.5 mm to 18 mm might be preferable to position the focus at the top, middle, or slightly below the material surface. Initially, we forgot to create a new gauge, so both the kerf test and my personal parts were cut with the focus point 2 mm below the bottom of the cardboard and HDF sheets.

Power

The laser cutter has a maximum power output of 80 W at 100% power. However, since our laser tube is relatively old, its actual maximum output is likely lower. For cutting, we typically use power settings between 60% and 90%. Engraving is usually performed at power levels between 10% and 30%.

Speed

We generally avoid speeds above 300 mm/s, as excessive speed can cause significant oscillation. However, higher speeds may be used occasionally, provided the machine operates smoothly. If abnormal sounds occur, the process should be stopped, and the speed should be reduced. Typically, cutting is performed at speeds below 90 mm/s, while engraving speeds are higher.

The following test diagrams for speed and power were created with a 15 mm distance between the laser carriage and the top of the material. As a result, the focal point was positioned directly on the surface of the cardboard. The top diagram shows the results for cutting, while the bottom diagram displays the results for engraving.

In this test, the optimal range of speed and power for engraving was repeated and verified.

Frequency

We discovered an option in RDWorks that allows us to set the laser pulse frequency. Previously, we had never adjusted or activated this setting. The default laser frequency is likely around 20 kHz, but we do not have precise confirmation.

Kerf

To measure the kerf, we designed a test where the laser cutter would cut 20 rectangles (later reduced to 10 for practicality).

w03_kerf_test.dxf

Our instructor, Ferdi, recommended using a material other than cardboard for this test, as cardboard layers can shift slightly above each other, affecting measurement accuracy. We used 3 mm HDF instead. Alternative materials such as acrylic or plywood could also be used. Since we had a scrap piece of HDF available, we proceeded with that.

Before cutting, we needed to determine the appropriate power and speed settings. As mentioned earlier, our typical cutting power ranges from 60% to 90%. I decided to use 90% power and adjust the speed experimentally by assigning different layers to squares in RDWorks.

I tested speeds between 35 mm/s and 5 mm/s in 5 mm/s decrements, using seven squares for an initial classification.

The test results showed that starting from the fifth square, the parts were fully cut.

This corresponded to a speed of 15 mm/s.

To fine-tune the settings, I repeated the test using speeds from 19 mm/s to 16 mm/s in 1 mm/s decrements. The optimal speed was found to be 18 mm/s, which was used for the kerf test.

The actual kerf test was then performed.

Finally, we measured all the cut pieces placed side by side. We calculated the kerf by subtracting the measured total length from the expected length and dividing the difference by the number of parts. In our case:

(10 * 10 mm - 99 mm) / 10 = 0.1 mm

Joint Tolerance

The cardboard material used is slightly compressible, making clearance joints unsuitable for this application, as the parts would fall apart. Instead, an interference fit works best. To confirm this, I designed a small test comb that allows testing in both directions. First, I measured the material thickness, which was 3 mm. Considering that we likely needed an interference fit rather than a clearance fit, I tested slot widths from 3.3 mm down to 2.5 mm.

w03_tolerance_test.dxf

For the final cut, I decided to engrave the numbers rather than cutting them out.

The engraving was slightly faint but still readable.

Using the finished test comb, I tested slot widths ranging from 3.3 mm to 2.5 mm.

The results showed that 2.5 mm provided the best fit, while 2.6 mm was slightly loose, especially upon repeated use.

Laser parameter used:

Installing and Setting Up RDWorks

During the RDWorks installation process, a window similar to the one below will appear. Follow the on-screen instructions. After the general installation is completed, the following window will appear. The settings used in this example are shown. It is important to select the correct laser type. In our lab, the "LaserWork" type works correctly. The machine origin is set to the top-right corner, and we use millimeters (mm) as the unit of measurement.

After completing the installation, RDWorks will launch with the following interface. To establish a connection with the laser cutter, click on "Port Setting".

I chose to connect via the local network, so I selected "Web" and entered the IP address of the laser cutter. This address was obtained from another PC where the machine was already configured.

Finally, select the newly added device.

Initially, the bed preview in RDWorks may not display the correct dimensions. In my case, it appeared approximately 2.5 times larger than the actual size. To correct this, simply load a file into the software and press Start. This action will automatically adjust the bed size to the correct dimensions.

To prevent unintended laser activation, it is recommended to disable the laser tube beforehand. This way, the laser cutter will only move without emitting a laser beam. Once the file is sent to the laser cutter, you can also abort the process if necessary.

Laser Cutting

The goal of this project was to create a modular kit consisting of basic geometric shapes such as triangles, squares and pentagons. These shapes allow the construction of various dice used in Dungeons & Dragons (DnD) and other pen-and-paper role-playing games.

• Triangles can be used to build tetrahedrons, octahedrons, trigonal dodecahedron and icosahedrons. The only requirement is a specific connector to ensure the correct side-to-side angle.
• Squares combined with 90° connectors form a basic cube.
• Pentagons are required to assemble a dodecahedron.

While designing the connectors, I aimed to keep them as simple as possible. All connectors share a similar design, differing only in angle. However, in the process of simplifying the design, I forgot to add chamfers to the parts. As a result, instead of chamfer joints, I only have press-fit joints. These work well but make assembly slightly more delicate.

The design works best with slightly flexible materials, such as cardboard, as this allows for minor adjustments when assembling the final structure. For more rigid materials, wedge joints (with the wedge positioned on the outside) would be a better choice. However, wedge joints are significantly more complex, and my focus was on keeping the design simple.

I designed all the laser-cutting files using Inventor, meaning the final 2d files are only available in DXF format. The parameterized design process was the same as the one I explained in week 02.

Tetrahedron

To assemble it, the following parts are required:

Cube

To assemble it, the following parts are required:

Octahedron

To assemble it, the following parts are required:

Dodecahedron

"Measure twice, cut once." I did not position the laser correctly, which resulted in poor material utilization. As a result, there was not enough material to cut all required parts.

I had to recut the missing parts.

To assemble it, the following parts are required:

Trigonal Dodecahedron

It is not a typical D10, which is usually a pentagonal trapezohedron. To create one, I would need to use kite-shaped faces. Before cutting it, I was uncertain whether the die would function properly, and as expected, it only works with some complications.

When rolling this D10, an edge between two faces always ends up at the top, making it impractical for use as a standard die. Additionally, a typical D10 is numbered either "1 to 0" or "10 to 00", whereas this one is numbered "1 to 10". Due to its shape, it somewhat resembles a UFO.

To assemble it, the following parts are required:

Icosahedron

To assemble it, the following parts are required:

Vinyl Cutting

A tetrahedron always lands on a face, with a corner directly opposite. Typically, the dice number is displayed in this corner. To ensure these numbers do not get confused with the standard numbers in the center, I decided to cut them out of black vinyl and stick them onto the backside of the first four triangular tiles. Additionally, I wanted to create a sticker for a local student club, so I cut that as well.

To design the SVG file, I used Inkscape. The club's logo was only available as a PNG, so I used the "Bitmap Trace" feature to convert it into a vector path. The numbers were typed using the "Text Tool" and then converted into paths using the "Object to Path" function.

After cutting, I carefully removed all unwanted material by peeling off the excess vinyl from the corner, keeping the pull close and parallel to the vinyl sheet.

For transferring the numbers onto the laser-cut triangular tiles, I used a tweezer instead of traditional transfer tape. This is not the standard method, and I generally wouldn't recommend it. However, since I only had small items (the numbers), this method worked efficiently.

For the bc Club sticker, I followed the standard process:

• Apply the transfer tape on top of the vinyl sheet by peeling off the backing of the transfer tape using an inverted version of the known technique.

• Carefully peel off the transfer tape along with the vinyl, using the same technique as when removing excess vinyl.

• Apply the sticker to the desired surface.
• Slowly remove the transfer tape, ensuring the sticker adheres properly.

• Lastly, the transfer tape can be reused a few times.

What Did I Learn This Week?

Key takeaways from this week include:

• A laser cutter should not produce visible fumes above the workpiece. I previously believed that some fumes were acceptable, but this is not the case. By adjusting the settings, I was able to completely eliminate visible fumes, which significantly improved both safety and cut quality.
• When Ferdi first mentioned that we would only be cutting cardboard, I was quite skeptical, as I assumed it would look cheap and flimsy. However, I was pleasantly surprised by how clean and professional the final parts turned out.
• "Measure twice, cut once." I was fortunate that this mistake happened this week. Re-cutting two tiles was quick and easy. In another situation, this could have had a much bigger impact.

What Caused Problems?

When embedding videos in my MkDocs page using the HTML <video> tag, I encountered an issue where the videos did not display correctly on my website. Instead, they appeared like this:

I originally used the following code:

<video width="1080" controls>
  <source src="img/w03/w03_tolerance_test_4.mp4" type="video/mp4">
</video>

However, I had to modify the file path as follows:

<video width="1080" controls>
  <source src="../../img/w03/w03_tolerance_test_4.mp4" type="video/mp4">
</video>

At first, I assumed the same path structure would work because all images and videos for this week are stored in the same folder. When using Markdown, this approach worked as expected. However, when using HTML inside MkDocs, the HTML files are generated in a separate folder within a different directory hierarchy.

To ensure the correct path from the actual HTML file to the media folder, I needed to add "../../" to my usual path.

This solution worked both locally and globally—at least once. However, after the next commit, it stopped working globally, even though the path remained unchanged. Locally, everything still functioned correctly. The issue was likely caused by an incorrect path reference. The solution was to add a leading "/".

The following solution worked again for a single commit:

<video width="1080" controls>
  <source src="/img/w03/w03_tolerance_test_4.mp4" type="video/mp4">
</video>

In the end, I found the following command to work reliably across multiple commits:

<video width="1080" controls="" src="../img/w03/w03_tolerance_test_4.mp4" title="Title"></video>

Design Files

dice_dxf

dice_design_files

To create this page, I used ChatGPT to check my syntax and grammar.

Copyright 2025 < Benedikt Feit > - Creative Commons Attribution Non Commercial

Source code hosted at gitlab.fabcloud.org