When we think of using a press to form metal we think of large stamps with custom made metal dies under unimaginable hydraulic pressure. It’s unlikely we’d e think of anything 3D printed. And in a commercial environment we’d be right. But your average garage hacker is far more likely to have access to a bench vise and a 3D printer. It’s in this context that [The Shipping Container Garage] has spent considerable time, effort, and money perfecting a process for pressing copper parts with 3D printed dies, which you can watch below the break.
In the quest to make a custom intake manifold for his project car, [The Shipping Container Garage] first made 3D printed jigs for cutting out a manifold flange that bolts to the cylinder head. It’s a process he calls Analog CNC, as all the cutting is done by hand.
Buoyed by his success, he proceeded with the next step: making manifold runners. His metal of choice was copper. While softer than many metals such as steel, he found it too hard. In the video, he describes his method for annealing the copper. Once cooled, two 3D printed dies are pressed into the copper tubes to progressively shape them. Watch the video to find out one of the neatest details of the die itself: how he gets it out!
Of course no matter how clever this all is, it’s useless if it produces poor results. And that’s where the most astounding part of the build is: The parts are all the same within 0.006 inches (0.15mm) of each other, and the parts fit the manifold flange they were made for. Additionally, the die can be used for the duration of the project at hand. For low volume production, this appears to be a viable method. It’ll be interesting to see what others do to iterate these processes to even more advanced stages.
You may also like to see 3D printing used in leather working and in jigs for beautiful circuit sculptures. A big thanks to [JapanFan] for the tip! If you have your own pressing hacks to share, let us know via the Tip Line!
I’ve been playing with this and have found that if you use a set of progressive dies, with each one only stretching the metal like 2mm, you can have quite long-lifetime dies and do some pretty amazing forming. Not just bending, either: I’ve been putting 25mm diameter hemispheric (as in 12mm deep) depressions in 16 ga galvanized _steel_ sheet. It’ll do annealed 1mm thick aluminum all day long, so now I can form hollow aluminum balls with two press passes and some welding. I’m really impressed by what dies can do if you’re careful with design and don’t try to do too much.
Impressive! Would you mind sharing a few details of what you have found works best for making your dies?
Specifically for hemispheres: 6 male dies, one female. The first male die is 2mm deep and matches almost the full width (like 24mm) so it just slightly indents the metal with a large radius on the edge. So, OD is 24mm, the flat middle section is like 18mm, with a 3mm radiused area. All the stretch to the metal happens in that area. Next one is 4mm deep, and similar, so it barely restretches the previous stretched area, and instead stretches another section inboard of it, and so forth down to the last one, which is full radius and stretches the whole surface just a little. When I bisect one and measure the metal thickness it’s a fairly consistent 30% thinner than it started. With aluminum it needs annealing halfway through. With steel, it was just fine without annealing. I was really surprised it stood up to steel.
Dies PLA with 60% infill, 0.2mm layer height, on a prusa 3S.
After 20 punches the active (large radius) area of the die was visibly scuffed but was not measurably (using calipers) deformed compared to after the first punch. From new to first punch there was some die change, but then it seems to stay pretty solid afterwards.
I lack the time to pursue this but I suspect that it would be possible to sink a form to slightly oversize using simple male/female dies, and then press (say) a recognizeable human face into thin metal with a final pair of dies. I was gonna try Han in carbonite (sans hand sticking out) but haven’t gotten back to that project.
I did a presentation on this for the east coast reprap festival in 2020.
BTW these were all two-piece dies: a fixed female and a fixed male die clamped in a vise. With this setup there is some crumpling of the surrounding flat metal, although the form itself had no ripples or folds in it.
The right way to do this is have a fixed female die, and a two-part male die, with a flat part that clamps to the female die and holds the metal in tension, and a second interior part that you then press in to progressively draw the metal.
I aligned my two dies with a 3d printed fullering spring that holds them in the right place x,y but hinges open in z, because I was really interested in the simplicity of just using a vise as the clamping system.
The three-part die needs something more sophisticated for alignment/clamping.
Silversmiths do this by bolting the dies together with holes through the silver sheet for the bolts.
Aluminum can manufacturers do this using hydraulics to hold the dies together but not quite clamp-tight, just tight enough to prevent metal crumpling, as they rely on drawing metal down into the punch from the surrounding area to get deep-drawn forms that are 5-10x as deep as they are wide. That needs pulling in lots of metal, and I haven’t tried that though I’d like to. (That’s highly abrasive/erosive to the female die where the metal is being compressed and restretched, so would almost certainly need a steel insert… that could easily be formed by the same die pressing process.)
In the past, 3d printing has been used to make molds to make the cast to pour a metal die. Printing the die itself and having it survive is an impressive feat.
Of course they can. It’s just a question of your tolerance specs.
With decent infill even PLA parts can withstand a surprising amount of compression stress. I made a set of upper mounting brackets to mate with the bolt-on aluminum lower bracket for attaching a pair of hoverboard wheel motors to an extruded 20-4040 beam and they happily and reliably carried the weight of an adult holding the stems of the motors tight when compressed between the M8 nuts on one side and the aluminum bracket on the other.
What was the guy that was doing this in a sheet metal brake to make complex bends? Those were surprisingly robust – really heavy gauge steel and he was pushing waaaay beyond where I thought it would have failed
Personally I’d make these in taulman 910 or a similar nylon. Nylons are somewhat self lubricating and should accel at this type of use.
What a fantastic waste of time.
He isn’t making 1000 of them. So why go through the trouble of making the worst possible, yet still technically functional dies?
Does EVERYONE with a 3D printer say to themselves…
“How can I make the worst version of X, while also spending 10x the time, and wasting a bunch of plastic and electricity too?”
Because I sure don’t.
Prototyping and Art.
If you are 3D printing anything else, you are wrong and should be publicly discouraged, not given attention…
(Unless you are a real engineer, with a real technical problem that can ONLY be solved with additive manufacturing.)
Cliffnotes for the above: “No fun allowed”
I really wish people didn’t think “CNC” meant cutting something with a spinning tool or that “a CNC” is a specific type of machine tool instead of a method of control. I twitch a little when people say they’re “going to CNC” something or “throw a part on the CNC.” Some day I’ll have to show off all the parts I made on my manual.
Yes, I know I’m a crank.