Taking a Different Route
You may have been wondering where all the updates are on the electric trike are. Unfortunately you’ll have a little while longer to wait as my attentions lately have been focused on a second project, my CNC Router.
Like many an engineer I’ve been fasinated with CNC Routers and Mills since the day I learnt they existed. The things some of the big commercial 5 axis CNC mills can do defy belief, see [this one]. Now I’m not kiding myself, this level of performance requires huge precision and mass to obtain the required rigidity. Precision and mass both equate to big $$. So I’ve settled on something a little more sedate, enter my CNC Router:
The frame will be constructed from welded 65x35x3mm RHS steel. The linear guides are THK SR 25 rail with SW 25W-Y bearings which I lucked upon moments before they were thrown out. In order to actually move things three 1.9Nm stepper motors will drive C7 rolled ball screws, available from my [favourite direct from China store]. The rails and screws are mounted to 10mm thick aluminium plate with the rails on 12mm standoffs. The intention of this is to allow tweaking of the alignment by shimming of the back plates with respect to the frame to make up for my less than stellar welding abilities.
The end result may look simple but it is the result of hundreds of hours of research, thinking and playing around in Solidworks. At first I had intended to make a vertical mill, but came to the conclusion that I wouldn’t be able to make something solid enough within my budget and limited machining resource. I played around with three or four fundamentally different designs before finally settling on a fixed gantry design. This sacrifices work area and ease of construction compared to a moving gantry design, but hopefully with the reward of greater rigidity.
A welded steel frame was chosen over bolted Aluminium (which would have been easier to construct) in order to make the project more affordable and hopefully more rigid. The reason rigidity sat so high on my list is that reading through [cnczone] and [mycncuk] it quickly became apparent that the majority of people who built routers wished they were more rigid, and as my primary goal is to mill aluminium I knew it was important to attack this from the get go.
And then this happened: No sooner had I written the above sentence when I had an idea which may change some of the details in the attached image. I had previously considered using a piece of 20mm thick aluminium for parts of the build but, the lack of access to a mill lead me to abandon those ideas. I just realised (probably because I can be a bit slow at times) that I have a router, I have some linear tracks, surely I can make something to allow me to “mill” straight sections in aluminium, after all the majority of the CNC routers I’ve seen on the net are just using normal wood routers as their spindles. 20 mins out in the garage and I’d proved my point by squaring up a piece of 10mm scrap using nothing but a straight edge (THK track), a wood router, and a straight 2 flute 1/2″ wood router bit.
Two of the edges were already parallel so I just squared up the remaining two edges. The first edge I made ~4mm passes with a 2mm depth of cut (DOC), this worked reasonably well and only required 5 passes to get through the whole 10mm. The second side I tried something I knew would be pushing it a bit and made a single pass removing ~1mm but at the full 10mm DOC. The first edge gave a reasonable though not great finish, certainly good enough to build the router though. The second edge resulted in a lot of chatter and a fairly rough edge. That being said it was nothing I couldn’t fix with a file and 2 minutes.
Other things I learnt from this little experiment:
- 1kW should be sufficient for milling aluminium at a reasonable DOC and feed rate.
- A 2 flute cutter in a 25k rpm router with a 1m/min feed rate seems like a good starting point, though I should actually do the proper calculations to confirm this.
- Swarf gets EVERYWHERE! And by god is there a lot of it. I hadn’t really considered it too much, though it’s become apparent that I’ll need to develop some sort of swarf management system in parallel with the router. And I thought saw dust was bad.
So, watch this space, I’ll be back with an updated, easier to build (though probably more expensive) version of my CNC router soon.