Right, so now we're into CNC machining. Not for everyone, in fact I don't suppose there are many here who have ever contemplated doing it, fewer still who actually do it. Feel free to move on. But if you're even mildly curious, stick with it. It's a steep curve enlivened with some bumps along the way.
First of all, and to answer at least one question, this is what the mill looks like. To give you an idea of size, the tooling plate that the work is mounted on is about 250 x 100 mm. A lot smaller than any industrial mill. But you can't see those because they are in closed cabinets and sprayed with coolant while they are working.
But let's back up. The first thing is to define the cutting path that the tool is to take. Regular readers will know I'm a user of F360 for creating solid models of parts. Whaddaya know, F360 does tool cutting paths as well. You create a solid model, give it lots of information about the tool size and the feed rate and the stock to remove, and it figures out the tool path and downloads a file that the milling machine controller can use.
Wow! So much complexity and sophistication in a free package of software. I can only think they left all these features in the free package because they knew very few hobbyists would ever use them.
Since this thread is about a loco not a milling machine, let me say something about the latest component. One of the strangest features on this tender is a plank of wood mounted on top of the rear buffer beam. Not any old plank, it is specially shaped and has filler pieces on the bottom that wrap around the buffer stocks.
I thought initially it was for someone to stand on, but there aren't any steps up to it. I think it was there to locate the tank and stop it moving around on the frame, but any other insight from the LNWR experts is welcome. Anyway, it's a 3D shape and too thick to be easily cut by hand.
My first attempt was to define it as a 3D model and cut it out in one go. I made a couple of mistakes. The first was to use too large a cutter on the basis that the larger it is, the more rigid and less likely to break. Which overlooked that larger means more metal to cut away and the cutting force goes up. Nope, smaller is sometimes better. The other mistake was to do it all in one. That required a lot of metal to be cleared, which increased the machining time a lot. So I upped the feed rate and the depth of cut, and that increased the cutting force to the point where it started making loud protests (sometimes you have to be courageous) and eventually pushed the workpiece aside.
On a larger and heavier machine you could do it, but that wasn't the way to go. I scrapped a piece of brass but didn't break the tool. The cost of brass means it was probably more expensive than the tool.
So I redrew the model in three parts - the base plank which is a constant thickness and therefore a essentially 2D, and the two extensions that fit over the buffer stocks.
Now I talked about toolpaths a few minutes ago. F360 can show you the toolpath and even animate the tool movement. This is the toolpath for the base plank. It's what F360 calls a 2D contouring strategy.
Very simple in machining terms, just a 2D curve repeated at several depths to cut out the profile. If you define the profile, there are lots of apps that will create the toolpath from it. F360 adds the approach and withdrawl but that's about all. The filler piece is more complicated.
This is a 3D Adaptive Clearing strategy. There are several other strategies to choose from. You can experiment with them on the computer.
It's using a lot of computing to devise that path. There are movements to optimise the rate of metal removal, and the curved face is approximated using a series of fine steps. That's what a modern CAD package can do. I'm just grateful, and not a little amazed, that you get it all for free. Anyone here from Autodesk - you aren't reading this.
So download the toolpath files and squirt them at the machine controller. Here is a picture of the base plank and one of the filler pieces cut out.
Separated from the stock, the filler piece looks like this.
Yes, that curved surface looks a bit of a mess but in fact it cleans up easily using a half round file, so eventually I had the three parts ready to assemble.
Quick work with the soldering iron and here it is in place on the buffer beam.
And there we are, another component done and a lot of experience gathered. Well, it wouldn't be half the fun if you didn't learn from it, would it?
Nick
