Skip to content

7. Computer Controlled Machining


Weekly Assignment:


Group assignment:

  • Do your lab's safety training
  • Test runout, alignment, fixturing, speeds, feeds, materials, and toolpaths for your machine

Individual assignment:

  • Make (design+mill+assemble) something big (~meter-scale)
  • Extra credit: don't use fasteners or glue
  • Extra credit: include curved surfaces

Introduction


Since most of did not frequent the ShopBot, Dr. Taylor gave an introductory lecture on CNC (Computer Numeric Control) and the machine.

Term Definition
Profiles outer or inner contours of a part that are cut to define its final shape
Pockets recessed areas that are cut into the material without going all the way through
Flutes the edge of the bit that cuts
Upcut End Mill spirals upward, pulling chips away from the material
Downcut End Mill spirals downward, pressing chips into the material
Collet something that holds onto the shank that has the cutting surface
Chip material that has been cut and kicked out
Straight bit doesn't have a helix and makes it chip rough
Chip load the amount of material that is removed by each cutting edge of a tool during a single revolution of the cutter
Feed rate how fast the cutting tool (not the spindle) moves across the material as it’s cutting
Ramping a more graduated approach to cutting something
Runout the rotation inaccuracy of a cutting tool or spindle. It occurs when the tool or spindle does not rotate perfectly true, causing it to wobble or oscillate slightly during rotation
Dogbone fillet a specific type of fillet used to round off or smooth out corners where two perpendicular edges meet

We set the zero position as the machine bed and the red dot is the origin. Dr. Taylor warned us about getting x and y confused. Also take into consideration material thickness and kerf.

I followed along on Aspire, generating the toolpaths for a profile and pocket cut. I drew a square and added dogbone fillets.

When I opened the Toolpaths bar, I found options for operations. Pocket shapes out the shaded area inside something while profile only cuts out the line. Both of them need vectors to be fully closed.

Note: The object has to be in a pink, dotted line for the toolpath settings to be applied to it.


Profile


I went with a profile cut first. The first setting is the cutting depth, which I would set to the thickness of the material.

Onto the tool, you can go into edit and choose the database, material, and then bit. Then, it will automatically tell you how many passes is needed given the cutting depth and tool data.

To get to the tool database, I needed to sign in to the Fab Lab's account. Once I was synced with what we had available for each machine, I could choose the bit I wanted. We were all cautioned not to click apply or change the settings, as it would complicate things for all the computers.

Then there are the different machine vectors: outside/right, inside/left, on. Each requires different calculations

Next, we need tabs. If it is a clean cut that goes through, the cutout piece is free to fling out, which is a safety hazard. Tabs prevent this from happening by ensuring that it does not detach from the larger piece that you are cutting out of.

Note: Never put tabs in corners. More tabs is always better than too few.

Finally, ramping is like the path assuming a spiral path, where it goes lower in gradations. This is to put less pressure and strain onto the bit and also the wood.

When you are satisfied with the distance and angle settings, do a double check of what objects you have selected and have a pink dotted line. Then, click calculate. You should see some multicolored lines, the red being how the bit travels above the material and the blue being where it interacts with the wood. When you play the simulation, you can see a projection of what it might actually look like.


Pocket


Pocket is similar, but also has directional settings: conventional and climb.

It also does not require tabs, as nothing large should fling out.

When you go back to the general toolpath's bar, you can access the Toolpaths summary, which gives you a synopsis of the time and all the active toolpaths.

To save your toolpaths, you go to Save Toolpaths, which tracks. This is the point where you make sure everything is in order, literally.

Note: Order is important, always raster before vector. This is to make sure that your surface is secure and does not move when you raster. The machine runs the toolpaths by the hierarchy they are in when you export them all. It goes top down.

The toolpaths are saved as .sbp files.

Making Something Big


Design


I knew I wanted to go with a bookshelf, because its something that I feel like I've needed for a long time and have just been putting off purchasing. To the left was was my original idea, but I knew it was an ambitious one. Since I'm on a time crunch, I pivoted to some other designs and I ended up going with the one on the right. This is the site I found it on.

I did some more research on the way my shelf would be supported and what joints to use. This was a helpful sight for me to visually see how each could work and which stye would be most suitable for my bookshelf's design. I wanted mines to be press fit, and the original wasn't, so I had to make some changes in suit.

I figured that most of my design was the same by reflection. This was the first set of pieces I came up with:

I generally followed the same four different 'types' of pieces as I kept going. From left to right, they are the vertical support near the corner (10), the vertical supports on the side that alternate (10 for this set), the center plates (4), and the plates at the top and bottom (2).

I saved all the .stl files and imported them into a new project. When I went to assemble it, though, I realized that it was not as simple as I made it out to be and that with my current set, there were some pieces that just couldn't fit.

Up until now, I relied entirely on mental visualization, because I didn't think it would require all too much planning and sketching. But after so many failed designs, I decided to stop rawdogging it and actually made blueprints.

This was honestly the most frustrating part of the process, I found myself just gazing into the distance at times, visualizing, failing to do so, then forcing myself to think harder about it.

And even so, it wasnt perfect. I assembled another simulation of it and found one edge jutting out. So I made some minor adjustments.

I added the appearance of wood and then saved all the files I needed, labelling them with how many I needed to make steps easier in the future.

I also saved them all as .dwg files for CorelDraw.

And .dxf files for Aspire.


Laser-Cut Model


I pulled all the files I needed into CorelDraw, and was immediately grateful to my past self for labelling them with the count of each I needed.

Importantly, I needed to scale them all down to the size of the cardboard, considering kerf. I measured its thickness as 0.17. I referred to the past week (so useful!) where we measured kerf here and got the kerf of the machine to be 0.015 inches.

I designed my tabs to be 0.5 inches. So, after some simple math, I scaled everything down around 2.7 times so that the length of the tabs was 0.185 inches.

It didn't fit on one piece of 30 by 20 inch cardboard, so I had to break in a new sheet. After it all came out, I assembled it and it turned out just as I imagined.


ShopBot CNC Milling


I moved into Aspire, importing all the files just like I did before. I was a bit worried about how I would make everything fit, since we were running low on wood by the time I got to this step. I only had a sheet reserved for my project, and I didn't want to encroach onto another piece. Luckily, it worked out.

For the dogbone’s radius, i was told to use half the diameter of the drill bit. I plan on using the ⅜ compression bit, so its 0.1875. But I had to join vectors because at some vertices, the check mark that indicates I can add a dogbone didn't show up.

Last time, it fully worked the first time around. however, I was still told that there were two that were open. I knew that it had to be on a piece that came in pairs, which was only one type of piece: the top and bottom type pieces. I was really lost as to what could have happened, though, and I didn't see any point of inconsistency even after I zoomed in and tracked all along the profile.

Angel brought up the astute idea that it might have been due to a double up, which was likely because there was a lot of copying and pasting I had to do. Sure enough, there it was.

I didn't know how to remove it though, because no matter how hard I tried to seperate them, they stuck together, treated as the same object. Wilson faced something similar and introduced me to the Interactive Vector Trim. I used it to cut the bad section off, and then used node editing mode to resnap the vertice.

After that, it started working. All the tabs that go inward would have wood fitted in, so I put dogbones on them all.

However, I had an issue when I got to the 0.5 inch wide smaller tabs, where there wasnt enough width for there to be two 0.1875 dogbones. So, I decreased the ones for those to 0.176, which means it'll probably require manual sanding later.

I prepared the profile toolpath, and arranged everything so that it fit. But this was the point where I realized my middle boards design was a bit off. One side was wider than the other and would jut out.

I fixed this error and was sure I got it right this time around, so I didn't do a second cardboard model. Of course, it also meant I needed to do some rearranging. It still worked out though. I would actually be surprised if it didn't because my new piece was smaller than before.

When I went to mill, I did a little practice one first in an area of my larger piece that was empty, just to make sure all of my settings would work out and it would be a good press-fit.

I did an air cut first. After, in the process of cutting it out, I was told to make sure the drill was activated by turning a key.

Before letting it go, I also needed to click the green button and a sound should be heard. Then, I could click OK to start the cut.

It came out fine, although a little messy. Once I took it out, though, I found that there was still a little bit that wasn't completely cut out, and was still attached to a tab. I took a mental note of that to increase the cut depth next time.

It fit fine parallel, but when I tried to fit it perpendicular, it was a hard, if not impossible, fit.

It was at this point that Mr. Nelson told Angel and me about the allowance offset setting. It could be found right under Toolpaths in the Toolpaths bar. Apparently, setting it to a negative value could make our joints a better fit, given that it was currently too tight. I did two more practices, one with -0.05 in, which didn't seem to be enough, and then -0.01 in, which seemed to work.

I used the -0.01 value on the rest of my toolpaths and began my final cut. When it was done, I used a vaccum to clean it up before using an oscillating multi-tool to cut the tabs from my pieces.


Assembly


It still required some sanding, but it was of consolation that if I hadn't done the practice tests, it would be much, much worse. I plan on painting the pieces, so I haven't fully assembled them all yet.

I learned that I should paint the fully assembled shelf because painting each individual piece and then assembling it would change the thickness of the wood slightyly and might affect my press-fitness. I finished assembling it all and took it home. When I was loading it into my mom's car, I realized that it wouldn't fit because her trunk was occupied with some other things. This was when I was really grateful it was all press fit, because I just gently pried it in half and we made it work.

I reassembled it in my room, and this was the finished bookshelf without paint.

My room has been a mess for almost a year with my precarious stacks of books, as seen to the left. I have another table that's even messier, so I'm so glad to finally have a bookshelf!


Group


The group assignment this week was to doo our lab's safety training and test runout, alignment, fixturing, speeds, feeds, materials, and toolpaths for your machine. Our documentation can be found here


Overview


I thought this week was really fun. Using the ShopBot was intimidating at first because I only used it once last year and it was with extensive help from teachers. It was a much more intense experience doing it without as much support, but I think it was good for me to be able to use it independently. I really liked the idea of designing something and having it become a reality. The bookshelf has been something I've needed for a long time because currently my books are in a pile and I've been too lazy to look for a bookshelf online. So this was definitely one of the weeks where I'll be able to use Fab products in my actual life and I love that concept.


Files




Last update: March 26, 2025