Update: If you're here because you bought a 3D printed jig on Gunbroker, that jig was printed and sold without the seller asking permission to use my design.
So that...actually..worked? I'm always a little surprised when my harebrained schemes actually work the way they're supposed to, and my 3D printed CETME folding jig worked nearly perfectly. There are going to be a lot of pictures in this post...
Here are the main pieces of the jig. As I said in my last post, the mandrel portion is made from more pieces than it needs to be because I changed my mind halfway through printing and didn't want to reprint everything, so I just printed the extra bits and glued them on. The lower jig has clearance cutouts that should work for CETME and G3 flats, although I goofed on the ejection port location and had to do some grinding on the jig. I'll update the files for that before I post them. Each part has two 1/2" hardwood dowels connecting and aligning the pieces, and depending on your printer's tolerances after printing you will probably have to chase the holes with a 1/2" drill bit. Based on a user review, I recommend not gluing things together so that you can adjust the length of the jig in case the alignment holes in your flat are off. The holes on the ends of the jig are sized for 1/4" bolts, through the mandrel and threaded into the lower jig.
It's printed in eSun PLA+, standing on end, .16 layer height, 8 walls, 15% Gyroid infill, .45mm line width(with .4 nozzle), 107% wall flow and 120% infill flow to "overstuff" the layers for better layer adhesion, a higher than normal 220°C and only 70% cooling fan, again for better layer adhesion. These are the basic setting for all my "strong" parts(the only difference with my printed receivers is that I use 99% infill there). Using these setting I get significantly stronger prints than the "standard" print settings, at the cost of a not quite as smooth surface finish. With these settings, if my prints fail, they crack through the part, or just kind of mush over, I do NOT get any layer separation. People tend to think of PLA as a weak and brittle filament, but the reality is that it has nearly twice the tensile strength of ABS and PETG, and more compression strength than either of the others. It's biggest drawback is that because it is so rigid, it tends to crack in situations where ABS or PETG would just flex.
There are four more pieces to the jig not pictured above. The top half of the mandrel portion of the jig is undersized so that the sheetmetal has room to bend. But, the whole point of this jig is to keep the flat straight as it bends, so there are some thin filler pieces to fill the gap between the mandrel and the sheetmetal in the later stages of the bending process. Here's where they fit on the mandrel:
Here's how the mandrel fits the sheetmetal without and with the fillers:
Ok, now the real fun starts. I used a short bar of aluminum to help spread the load across the top mandrel. A longer bar would have worked better, but this is what I had.
I pressed a little, then moved the bar and pressed some more. A little bit on this side...
A little bit on that side...
I went back and forth about 10 times, trying to keep things even. I actually did very little pressing right in the middle, most of it was at the ends. Once I got most of the way there, about what you see in the pic above, I put the filler pieces into the jig. It was a tight fit and I had to (very)carefully tap them in with a hammer.
With the spacers in, it was back to the press for more back and forth. Once the magwell starts to close up, you've got to start getting creative with your press tools. "Whatever is closest and will fit" is the method of tool selection that generally I use. You want to try to keep as much surface area as you can, the more the load is spread out, the better.
With pressure on the jig, I gently tapped around the whole thing with a rubber mallet, just helping the sheetmetal form around the mandrel a little more. At this point, I decided that it was as pressed as it was going to be.
The flat and mandrel were pressed tight against the bottom jig and there
was no where left to go.
Since the whole point of all of this was to try to keep the bend straight and even, and keep the receiver shaped properly, how'd we do? I'd say pretty good...
After taking the bolts out, I had to use a rubber mallet to get everything apart, and I had to hammer the mandrel out of the receiver. I got a slight wave in the sheetmetal above the ejection port, but that's mostly due to my putting the cutout in the wrong place on the jig. It's also much more noticeable here than it is in real life. Look Ma, I made a gun! (from a legal standpoint anyway)
It turns out that the lower jig is also great for holding things while you're welding too. I put my recovered G3 stock attachment weldment in the back of the receiver, used a piece of copper tubing as a backer for the front portion, clamped on a whole lot of Vice Grips, and welded it all up.
Even though it should be Tigged to be correct, I Mig welded everything because that's what I have at the moment. I've spent a whole lot of time doing sheetmetal work on cars, so welding up a receiver this thick was no problem. The welds weren't factory Tig pretty, so I ground them all down and this is what I ended up with:
The true test of a CETME/HK folded flat is the bolt drop test. Basically, you drop the bolt carrier through the receiver and it should freely drop through with no interference or drag. I checked it after every step along the way. So how'd I do? Fresh off the jig: Perfect. Tack welded: Perfect. Fully welded: Perfect. As best as I can tell, this receiver is as close to perfectly shaped as a home build can get.
And what of our jig, how did it hold up? Pretty good. One of the add on end pieces I superglued on the mandrel came off, but it was a glue failure, not a print failure(the files I'll post will have that as one piece). The lower just has a few scrapes in it, and a mushed spot where the ejection port flares out and that's about it, no other damage(and I'll fix the ejection port cutout in the file before I post it). This thing worked so well that it could probably be done in a bench vice or with C-clamps, if you have some big enough, or possibly even by using 1/4" All Thread and some nuts to pull the two parts together with C-clamps for the middle. Total cost for this jig was around $20 in filament, and about 3 days of print time.
If you want to make one for yourself, you can find my STLs here:
https://www.thingiverse.com/thing:4614385