Post-008
Jan. 19th, 2022

Machining adventures part 2: facing and fixturing a disc.

Using the HAAS as the world's most expensive paint scraper. 

IMG_2069.HEIC

So today we're going to try some new and fun HAAS things — we want to take a disc that Andrew + Evan in the lab will use for a Futura Pendulum project they're working on and just do some facing to one side of the pendulum. We're going to keep everything simple and just use the 10mm flat end-mill we were using yesterday.

The goals are simple, there's blue paint on the top surface of the part, we want to fixture the part to the HAAS, remove the paint from the top to create a smooth surface finish, and then remove the part from the HAAS. We're only going to face off a few millimeters. We're also going to try to go ahead and face around the outside of the part, just make the surface a little nicer.

We're going to face "down" around 1.5mm and then face inwards around 1mm.

Screen Shot 2022-01-19 at 12.02.02 PM.png

We start with our setup of course, we model the cylinder's final dimensions that we want and then we add a stock cylinder around it in setup. In this case we offset the stock from the bottom of the cylinder because we want to face only the top and sides. Since it's a cylindrical piece of stock, we want the origin to be in the center and we'll have to calculate the x/y zeros when probing by hand. 

Now it's time to CAM the operations, we first want to do a facing operation. Aluminum's SFM with carbide is roughly 500-1000 SFM so let's go smack in the middle with 800SFM. We're also going to use flood coolant for this. Most of the other default settings are OK, the important thing to check is the bottom height and that the step over + step down is not too large. In this case, we aren't stepping down very far so it should be OK.

Screen Shot 2022-01-19 at 12.10.26 PM.png
Screen Shot 2022-01-19 at 12.13.33 PM.png
Screen Shot 2022-01-19 at 12.12.13 PM.png

For the second tool path we're going to simply choose an adaptive clear with "rest machining." Rest machining is basically telling the tool-path to do what's left but not things that were done before. The only thing we adjusted on this second tool path besides the SFM was the bottom height which we wanted to be around 0.05in below the bottom so we knew the whole side would be faced. Now it's time to export this G-Code as a program and setup the part on the HAAS and run it!

But first, we're going to try a new fixturing technique which involves green PET tape and CA Glue. We tape the bottom of the part and the top of the fixture, then we CA glue the tape together to fix the fixture and part together, and we clamp the fixture, below are images of this process. 

IMG_2060.HEIC
IMG_2061.HEIC
IMG_2062.HEIC

Some tips for this, you want to leave some tape off on the sides so that you can peel the part off when you're done. You also want to be careful not to have any gaps between rows of tape but at the same time you want no overlap. Overlap causes a non-flat fixture and gaps could cause the metal to glue to the metal. Then add very small amounts of CA, you DON'T NEED A LOT, to hold the parts together. 

The main idea is here is to take a block (or something we know we can fixture to the HAAS) and glue it to something we can't really fixture in order to machine the thing we can't really fixture. Circles are hard to fixture therefor we're employing this technique. When we fixture this to the vise in the HAAS, be careful not to hammer the disk, hammer the block itself so we don't "glue it more than it needs to be glued."

IMG_2063.HEIC
IMG_2064.HEIC

Ok so some tips for zeroing a circle on the HAAS — it's not actually as intuitive as we might thing. The technique is to probe two horizontal points (on the X-axis) and two vertical points (on the Y-axis) and the point in the middle is actually the zero. You can do this many ways but here's how we did it.

  • First, probe one of the x, points and hit "part zero set" in the offset table of the HAAS.

  • Go and probe the other x-point being very careful not to change the y-coordinate of the machine and subtract the x-coordinates.

  • divide by two to get the distance from the sides you probed to the center of the circle.

  • ADD this value to the offset that you set in the G54 table when you hit "part zero set," this translates the math into machine coordinates automatically. 

  • do the same for the y-coord. 

Other options include working only in machine coordinates (which I prefer now). You probe the two sides of the part and look at the machine coordinates, then average the machine coordinates (find the coordinate half way) and make that the offset in the G54 work offset table.

We're going to set the tool-offset the same way we did yesterday using the paper technique again because precision isn't the biggest priority here (yet). Remember again to use the G-Code test where we use G-Code to send the HAAS to Z0.1 to ensure the tool offset is set correctly. Now we're ready to run the program! There's two coolant modes on our HAAS, flood and mist, we'll be using flood today. This program should not take long to run. We'll also go slow initially then speed it up.

Here are the compulsory running videos, the whole tool path took like 10 minutes max, not even I think because the facing operation was super quick. The thing that took the longest was zeroing which was a little new to figure out in the X/Y direction. 

I think at some point I will write a guide on G54 work offsets from start to finish on the HAAS and machine coordinates and so on, for my education but also because it's super useful to go from start to end on that. That will also help us determine a better zeroing technique for the tools for the future lab machine shop. 

IMG_2067.HEIC
IMG_2068.HEIC
IMG_2071.HEIC

Now, of course, some evaluations that will help us determine the next steps and lessons learned from all of this. We'll go step-by-step.

Feeds/Speeds for Top Facing Al — these went well. We were running around 800SFM and the top facing came out very smooth to the touch. The visual finish wasn't as uniform as we hoped but we think that has to do with the step-over and the pattern rather than the actual SFM itself.

Feeds/Speeds for Side Facing Al — these went less well. We can see in the picture above for facing the sides the disk with the bit penetrating into the fixture, the feed might have been a little aggressive. We want to fix those in the future.

Fxituring of the Part w/ Tape + Glue — this was super creative (so thank you to the FSAE people who taught me this trick). Inside the HAAS the glue and tape were rock solid and didn't let anything move around while milling. But when we took a small screwdriver, wedged it between the pieces, and tapped it with a hammer, the top fell right off with just one stroke no damage to the part at all. 

This was all super exciting because we got to run our first aluminum parts on the HAAS! The next steps are probably to re-CNC the press fit holder we CNC'ed out of Delrin in aluminum, that will teach us more about feeds/speeds but mostly tool-pathing. We'll try that later this week. We also want to do some machine shop cleaning and maybe order a FAT32 flash drive.

#machining #disk #tape-and-glue #fixturing