7. Computer-Controlled Machining

For this week I was simply tasked to “make something big”, so I decided to use this week to make progress on my final project.

Link to group project: click here

Outlining Final Project Table

First, I modeled generally what my table for my final project would like in Fusion360, which can be found on my Final Project Page.

Coming Up with Design

Originally, I wanted to create a design that could be “waffled”, a technique shown to us by Ms. Vaneza Caycho Ñuflo. Ms. Ñuflo came to Charlotte from Peru, where she graduated from Fab Academy at Fab Lab Peru and went on to start her own furniture company iFurniture. She very kindly came to our lab and told us all about the 50 digital joints her and her team promote, along with helping each of us to come up with unique ways to fabricate our ideas using only wooden joints. After running through a few iterations of chairs I thought might be cool to build in my head, I eventually realized that I should probably use this week to make progress on my final project. Thus, I decided to create the table that would house the main part of my final project. I wanted this table to have a sort of unique, modern, rounded look to it, with an oval shape on the sides. To get a better idea of the shape I wanted this to be, you can visit my final project page where I have a 3D model I created of this table. As a brief summary, my CNC design would have two side pieces which would have a sort of track shape and would be connected by a long rectangular piece the width of the table running along the outside (save the top flat part as there will later be a separate top piece which will fit here) rounding with the round parts of the sides, along with a long piece running along the inside part.

First Attempt in Fusion 360

I created a Fusion360 file where I began to create the wooden pieces which I would cut out for this table. Because I have documented the processes and tools used in Fusion 360 in-depth prior to this, especially in Week 2, I will not be explaining my use of specific tools here, but merely outlining which tools were crucial to the progression of this project. I started off with the two side pieces, which I created a sketch of, and then extruded. I then started creating the slots which the joints of the inside piece would fit into using a rectangular pattern.

I then began on the outer joints, also using a rectangular pattern, and using a circular pattern on the sides.
I also did this circular pattern on the inside part.
I then added in two long rectangular pieces which would connect in the middle to create the long exterior edge. I spaced out which parts of them would include the kerfing and began to space out the joints correctly.
I created dog bones in the joints on the side pieces by just putting circles in the corners and getting rid of all the excess lines. I know had this design. Although I was a long way from finished, I put this design into Aspire to begin learning this part of the process. At this point, I realized that my design was going to be far too large to fit on the CNC bed, and I would need to downsize it all significantly. Although I had been using parameters for almost everything, there were a few disconnects which made parametrically changing the design not really work. Thus, I decided to restart.

Resizing and Restarting

I initially basically recreated most of what I had before, only sized down, more properly constrained with parameters, and with a sketch in the background which displayed the dimensions of the CNC bed for reference. I added (as you can see in the second image) a control panel as well. Also, I made the difference between the edge and the joint .45 inches for the inside part because I realized that 1/4 inch of plywood would not be able to hold up the amount of things I will have in the main area. As of writing this, I am still not sure if .45 inches will either, but lets hope for the best.

Living Hinge

Initially, I was going to make my hinge using this kerfing pattern method which cuts holes in a pattern fully through the wood and allows it to bend. However, upon further research, I found this method which only cuts most of the way through the wood, allowing for a nicer look on the exterior. I soon started to realize the complexity of the calculation required in the joints on the curved part. Not only did I realize that they would impede the wood’s ability to curve properly and aesthetically because the joints were straight, I also realized that there would be complex and possibly impossible calculations in making the joints fit. You see, with this kind of kerfing, the parts that are not cut out will compress together (you can view my kerfing testing below for a better visual). This makes it so that the length of the top and bottom of the wood are the same when they are laid flat, but if they are curved at 180 degrees, the inside (or top when its cut) will naturally be the material thickness*PI shorter than the outside (or bottom when its cut) of the wood. Additionally, the amount these compress together could not be determined by the tests, nor could it be determined if they will all compress equally if both the cut is uniformly spaced out (which can not be done properly with a manual saw) and the angle of curve is exactly 180 degrees. Theoritically, they should be, however, this would still require an insane amount of calculation, possibly venturing into the realms of trigonometry (of which I already have enough of what with math class), which would all likely prove to be unnecessary anyway because of one thing I found with my testing. Therefore, I decided to neglect to have any joints on the curves. And so, I took them out.

Kerfing Testing

To get a better feel for what I was doing and just to ensure that it worked fine, I played around with some spare plywood on a table saw. My first cut, which had decently spread out cuts, turned out like this:

My second test, which I accidentally broke some of the pieces in between the cuts (which makes it more bendable), bent much more: I was even able to bend it a full 360 degrees after slowly stretching it out for a while: Finally, I did a third test in which I would do multiple simultaneous cuts (ensuring a didn’t leave thin strips), but also leaving larger spaces between these groupings, meaning that it would just be larger gaps and large pieces, making it look a little more uniform and better. However, this led it to be extremely rigid so that it broke when I tried to bend it even a little. Although I already mentioned many of the conclusions I arrived at upon executing these tests in the section above, I also concluded that if the joint’s two endpoints remain the same with relatively equally spaced out cuts, the actual curved part does not move around, but instead will remain rigid and steady. This is what will allow me to remove joints on the curved parts from my design.

Finishing Fusion Model

I now had to finish my design in Fusion 360 based off of this information. I began by making the pattern on the longer piece. I essentially just created a sketch which I extruded and made a rectangular pattern of that extrusion.

I then added these joint slots into the two side pieces (pieces not included in this design will go here and act as walls for the main compartment). After making this one, I just copied it and replaced the control panel hole for the other piece. I added the kerfing onto the smaller piece and all I had left were the joints (many of these steps actually took hours but most of that is comprised of measurement or calculation erros or just forgetting to include something and the subsequent time spent readjusting anything that is broken in the process of fixing these errors).

I added the joints by adjusting the original sketch for the outline of the two rectangular pieces. Of course, more calculations were involved in this process. These calculations were made infinitely times easier by the fact that I was only including joints on flat parts and not curved parts, meaning that almost all dimensions (for rectangular patterns, joint legnths, etc.) involved in making these joints were in whole numbers. After making these joints, positioning it all to fit in the CNC bed area with a precariously concerning amount of space to divide them, and fixing any and all final problems or issues that arose with measurement and dimensions, I had my final design in Fusion 360 ready to send to aspire as a .dxf.

Using Aspire

After spending a very long time in Aspire preparing two separate files (one for the living hinge and one for the actual cut), the estimated time was over 12 hours for the living hinge file alone, so I had to change to a simpler design for the sake of time. After creating a box which included finger joints and spaces for two different ledges (one for the piece to hold the sand and one for the top) as well as the ledges themselves in the file, this is what it looked like:

I put this into Aspire and went through the process again. I first resized the material size to 96x48 inches (the size of the CNC machine). Next, I created dogbone joints on all the corner that a finger joint would be inserted into so that it could fit (otherwise the round bit would have rounded the corner, making it much harder to fit). I had to then use the join tool to make sure all the vectors were properly connected. I then created a new profile toolpath and selected all the lines except for the ones outlining where it would engrave for the ledges. After measuring the thickness of the wood I was going to be using, I set the material thickness in Aspire to .45 inches. Per Mrs. Morrow’s suggestion, I set the cut depth for this toolpath to .46 inches, .01 more than the material thickness. I added tabs in as well. Next, I created a pocket toolpath for where the ledges would go. I selected all of the necessary lines and set the cut depth to .25 inches. For the bit for both, I used the 1/4” bit. Finally, I could export these as an .sbp file.

Milling and Assembly

I used the ShopBot CNC machine at our lab to mill. To do so, I followed this workflow provided by our lab’s instructors. In summary, I got my plywood sheet, secured it to the bed using a brad nail gun, followed the instructions on the workflow to home and perform an aircut, then imported the file again with the z axis reset to perform a real cut. Below is a brief video of my aircut and then my real cut.

Finally, I assembled this with the assistance of a few screws for stability. This is a picture of the box with the first ledge installed and a piece of acrylic (as it will be in my final project) resting on it.

Reflection

I procrastinated finishing this week for a while, but it ended up being a pretty important week to understand as it can be hard to create a large project without the use of the CNC machine. Overall, I hope to use the CNC machine more often and confidentally in the future.

File Download

The files that I used in the week’s processes can be downloaded using this link.

---

Last update: June 26, 2024