8. Computer controlled machining

Group Assignment

  • Test runout, alignment, speeds, feeds, and toolpaths for your machine

Individual Assignment

  • Make (design+mill+assemble) something big

For this week, I wanted to make a functionable piece of furniture and found this Makezine Article that included a bunch of potential furniture ideas. I thought that the italic shelf was a really cool idea and I decieded to base my idea on a modular shelf.

CAD Design

In Fusion 360, I sketched out a vertical support and a board. The width of the slots in the vertical support and the width of the board are controlled by the “board_width” parameter. The thickness of the slots is controlled by the “thickness” parameter. The distance between the slots is controlled by the “shelf_clearance” parameter.

Next, I extruded the sketches by the “thickness” parameter. I converted the bodies into components, duplicated the components and assembled the components together with rigid joints.

However, I realized that the shelf was structurally weak in the middle because only two vertical supports existed there. With no triangular supports, any give on between the slot and the board would cause the shelf to sway back and forth. I decided to model some bigger vertical supports that would run along the two sides of the shelf. By holding the board through four points of contact, I believed that I could mitigate the amount of swaying for the shelf.

Then, I parametrized the board length and margins of the slots. However, when I tested these parameters, the assembly incorrectly updated. I realized that the problem was that the joint offsets were literal values instead of parameters. So, I changed the joint offsets to parameterized values.

Now, I exported two versions of the sketches. One version had the parameters cut in half for a cardboard model. Another version had the “thickness” parameter set to the caliper readings of the wood thickness.


To prepare the sketch, I imported the DXF files into CorelDraw and duplicated the pieces enough to make the full assembly. I exported each file as an SVG.


For the cardboard model, I first cut out one board and one vertical support in order to test the fit.

Since the fit worked, I cut out the other pieces and assembled the model. I also noticed that without the supports the shelf significantly swayed more.


For the actual wood pieces, I imported the SVG into Aspire to generate tool paths.

I added T-Bone fillets to the slots because it would be impossible for a bit to create a true profile inside the slot. I set tool radius to .125 in. because I was going to use a .25 in. bit.

Next, I selected all the outlines of the parts, excluding the slots, and used the 2D Profile Toolpath tool. I set the tool to the .25 in. end mill.

I made sure to added some tabs to keep the peices in place.

Then, I selected the slots and used the Pocket Toolpath tool. I set the tool to the .25 in. end mill.

Here are previews of the toolpath and the cut:


I exported the toolpaths as the Shopbot format and started the cut on the Shopbot.

While the Shopbot was working on the big vertical supports, the tabs broke and I had to stop the cut. I realized that the tabs were too small and I modified the tab size and length in Aspire.

However, while the next cut was sucessful, I found out that the thickness of the slots was too small as the board couldn’t fit into the slot. I had to vigorously sand down a slot for a board to barely fit through. I was considering cutting another batch of pieces, but the coronavirus forced the lab to shut down.

Lab Opened

With the lab opened after 2 months, I added .01 in. tolerances to the slots to fix the issue. I recut the pieces and, fortunately, the tolerances allowed the peices to fit together. Finally, I could assemble the shelf.

Group Site