The class objective was simple: CNC machine two small matching objects out of wood and finish appropriately. On one hand, I like the simple motivation: shorten the distance between vector file and finished part. On the other hand, I didn’t think I would learn much. I’d appreciate experience with the Shapeoko, but I doubt I’d use a limited program like Carbide Create in the future.

 

personal

We just got a CNC mill where I work. Exciting! It’s the size of a bookcase and comes with plenty of presents, including an automatic tool changer. I’m dying to use it, but I need to get comfortable with CAM (Computer Aided Machining) as I already am with CAD (Computer Aided Design). Fusion, with its CAD and CAM features, is the industry standard program for CNC design, and the OEDK has a machine, The Carvera desktop CNC, with its own automatic tool changer and Fusion integration. Perfect.

 

 

SOLUTION

The Carvera will teach me CAM, but I need a part—something practical and straightforward, requiring multiple tools, curved shapes, and maybe some text. I decided to make a holder for the 2” vise used with our new trunnion (a 4th and 5th axis attachment). The vice is a reality expensive piece of equipment, and we’ll need to set it aside somewhere when not in use.

 

I desire:

• simple flat shape for machining on a flat bed
• easily machined material with even thickness
• a large circular depression to hold the trunnion vise
• a hole to hold the vise handle
• a feature for the mounting screws

 

 

CAD

In Fusion, we start with a sketch to position everything. Material-wise, MDF is an easily machinable wood product with an even thickness. Sounds perfect.

Here is the basic model. I like a healthy chamfer since things will get banged around. For the screw feature, I choose a cylindrical depression to toss the screws in like a loose change dish. Also, this let me practice machining a curved surface.

I also added some text, even though it’s not necessary.

 

CAM

Since I am using MDF, I don’t need to face the top surface. The first step is to hog out the bulk of the material with adaptive cuts. Adaptive cuts are modern CAM toolpaths that drive bits to a certain depth and move around, altering speed and rpm to keep cutting efficiently. Even with increased efficiency, hollowing out the large depression takes a while, essentially 30 minutes or more for every 5 mm of extra depth. So I removed only 5mm of depth to keep the machine time reasonable.

We need to change bits and cut chamfers after the adaptive rough passes and contour finishing passes. I decided to cut the text using the chamfer bit as well. Notice, I didn’t cut the edge of the board, but instead defined the edge so I could later cut things to shape in the band saw. This saves time, and I can just sand the edges things flat.

For most of the board, roughing, finishing, and chamfers cover everything, but the cylindrical screw dish is still rough. Here, I switch to the final bit, a taper ball nose end mill. This 1 mm diameter bit with a round end makes several small passes to smooth the floor.

Overall, the machining time is still too long. Carvera’s default settings are far too conservative, and if I had the confidence, I could have cut this time in half. However, my main goal was to practice CAM with multiple bits.

 

MAKE

Originally, I wanted to cut through masking tape. In theory, this would leave me with a natural stencil for painting, but in actuality, the tape fouled the bit. I had to clear it. This was for the best, as my tests showed MDF so easily absorbs paint; I could never have painted with clean lines, even if I sanded back the front face.

Enjoy these lovely videos, and notice the bit changes in the automatic tool changer. For fixturing, I just used double sided tape, and it worked like a dream.

 

FINISH

After machining both pieces, I band-sawed the edges and then sanded them flat. After finding I couldn’t paint a clean line on MDF, I resolved to just apply a clear coat. The text will just have to stay mildly readable.

 

 

 

MDF drinks paint like a sponge, but after several clear coats, it forms a hard layer.

 

RESULT

Ehhhhh. I think it looks okay, but the text is still illegible. You’ll notice some lines in the chamfer of the screw dish. This is because I cut the chamfer in the wrong order and with too small a bit. However, the floor of the screw dish is delightfully smooth. A good learning experience.

 

COST

For the cost calculation this time I cheated a litte, but forgive me. In my work, I often need to price made-to-order parts we design in CAD, and hopefully, in the future, we can CNC these parts ourselves. Regardless, made-to-order services like Fictiv, Hubs, or SendCutSend, take a drawing or CAD model file from you and provide an instant quote in return.
Since I designed the part myself, I sent my CAD model to Hubs. Now, they don’t cut wood, but services like JMP do. For example, I chose aluminum, which is not a material I would choose for this part but is even cheaper than HDPE, which I might choose. I think this price is in the ballpark for a wooden one. It was both cheaper and more expensive than I expected.

Cost Type	Cost	$/#	Source	#	Tot($)
Materials, Labor, and Machine Time	Make to Order	113.2	Protolabs Network	2	226.44
Software	Fusion2	680	Autodesk.com	0.003 yr	2.04
Design	CAD/CAM	45	Indeed.com	2 hrs	90
				Total Cost	318.48

 

CLEAN

I left the Carvera better than I found it.