Week 07: Computer-Controlled Machining

Learning outcomes:

Have you:

Source: Fab Academy Assessment - Computer-Controlled Machining


My Process

Background

I have a little bit of experience machining with a CNC mill - I took a Shopbot class at Artisans' Asylum a few years back. But this will largely be new for me.

Getting Started

I started by redesigning my CAD file from Week 02 to prep it for CNC milling. I reworked the file so that it would be truly parametric, putting in values for various parameters, including dimensions for the cradle mattress I was thinking of using, the angles of the sides, and, crucially for machining, the thickness of the material.

redesigned CAD part Redesigning my part in Fusion 360

parameters for the design Design parameters included material thickness, side angles, and mattress dimensions

After scratching my head a bit to figure out the dimensions for my tabs and slots, I realized I needed to bust out my old trigonometry skills.

graphin calculator and trigonometry sketch I used some trigonometry to come up with a formula for my tab depth

sketch showing trig formula in design In Fusion 360, I entered the formula for my tab depth at a given angle (sa)

Laser Cutting a Test Model

I felt a little uncertain about these calculations so I wanted to do a scaled-down test cut out of cardboard on the laser cutter.

2d export of my design file I exported all the part profiles to DXF format and prepped them for the laser cutter

Fusion 360 can export DXF files. I had some significant challenges getting the DXF files to read out correctly on the laser cutter - it kept giving me an "empty file" error. Finally I found an app called Kabeja which allows for the import/export of DXF files to various formats. I used kabeja to export my DXF into an Inkscape SVG and from there was able to use my normal laser-cutter workflow and get it cutting.

lasercut parts My scaled-down lasercut parts

I had designed my part for the 1/2" plywood we ordered at Dassault (I measured it to be 12.15mm ±0.1). The cardboard we had on hand measured 4.05mm, or exactly 1/3rd the thickness of the plywood, so I did a 33.333% scale transformation on the file in Inkscape to get my scaled down model.

lasercut parts assembled My scaled-down cradle assembled

another view of the assembled cardboard cradle Another view

another view of the assembled cardboard cradle

another view of the assembled cardboard cradle

It matched my model quite nicely, and I was very happy with the aesthetic.

Testing the CNC machine

Our lab has a Shopbot CNC mill. In order to test it out, we grabbed a piece of scrap MDF and created a test part to cut out.

vcarve software showing test part Our test part in the VCarve CAM software

test part milled out of MDF Our test part milled out of MDF

This test part showed four different styles of "dog bone" fillets as well as some nice press-fit chamfers. It also varied the slot size for testing press-fit snugness, and allowed us to practice using several types of millbits (v-bit, 1/4" 2-flute upcut bit, 1/16" 2-flute upcut bit).

Machining the Cradle

The lasercut scale model seemed good, so I moved on to making the full sized version on the shopbot. I started with the same DXF export of the part profiles, and imported that into the VCarve CAM software on the Shopbot machine.

laying out the parts in Vcarve Laying my parts out in VCarve software

I used the Fillet > "Dog Bone" tool to make dogbone cutouts in the appropriate spots.

inserting dogbone fillets Inserting "dog bone" fillets in my test parts

And I prepared toolpaths for the machine, selecting the endmill that we were going to use.

prepping toolpaths Selecting our endmill (Note: we later determined that these feed and speed rates were off and much slower than necessary)

making dogbones Completing dogbone fillets in the rest of my parts

I also made a screwhole pattern - essentially a bunch of small, shallow circles, to mark where on my material it would be safe to set hold-down screws throughout my material. I then laid down my material and used the machine to mark the screwhole pattern.

I then drilled screws into the plywood at each marked hole to hold the material firmly onto the sacrificial material bed.

I started with some small test parts to test out my workflow and see if/how the material fit together. It all seemed to come out reasonably well, so I decided to go for it with the large pieces. I prepped all my toolpaths in VCarve, saved the toolpaths for Shopbot, and got to cutting.

I thought everything was going swimmingly (despite a couple small headaches here and there), but sadly, I made a grievous error in preparing my toolpaths. I erroneously set my final pass, which used an upcut bit to try to make a nice cut on the far side of the stock material, to be an inside toolpath. As a result, much of my piece was destroyed.

piece destroyed by offset error An error in setting my toolpath wrecked much of my work

offset error 2

offset error 2 You can see that the tabs were more-or-less completely ruined by the offset.

Thankfully, my classmate Sal noticed the error before it destroyed all my parts. I was able to salvage my end pieces and recut my side piece. My final assembled cradle came together OK.

cradle assembled 01

cradle assembled 02

cradle assembled 03

cradle assembled 04

cradle assembled 05

Unfortunately, the ultimate result of this week is the determination that this cradle body is much too big for our living space, and so I have some major redesigning to do. Sigh.

Design Files

Cradle Body v02 (.F3D file)

Cradle Fullsize Layout for cutting on 48"x96" plywood (.DXF file)

Cradle DXF exported to SVG with Kabeja for lasercutting (.SVG file)