Just came across this in my email:
I saw that too, but Anthony was faster 
It looks like the author is using an aluminium block for the lower part.
1 Like
Yeah, dies are easier because fewer constraints on anvil width, I think. A vice wasnāt designed specifically to be a brake. 
I saw the headline and had it on my list to watch after work.
His are probably a better design.
1 Like
Not much to add, just that this was really interesting to read.
2 Likes
I will say, on reflection, that I think he makes two mistakes with the magnets:
- I donāt think that the magnets against the vice faces are actually necessary; so they just add complexity and cost and reduce strength without benefit
- The magnets on top look like they have a substantial amount of plastic between the magnet and the vice. Those inverse exponential terms really bite! This is why I experimented with the minimum amount of plastic I could leave under the magnet and still hold the parts to the vice. That substantially reduces the size of magnet needed. (Whatās the equation for magnetic force between a permanent magnet and an induced magnet? I vaguely recall it having more terms in it than I naĆÆvely expected, which isnāt saying muchā¦)
Ultimately, it makes me think about modeling a gooseneck that isnāt a finger die. Iāll look at MRs in gitlab if anyone wants to experiment with me on this one.
1 Like
One thing I have found useful on these sort of prints is to model a tiny pinholes or slits near areas of stress concentration so it forces the slicer extra perimeters only in those areas for strength.
Iām up for playing around if you got some ideas of what you want to model.
1 Like
I printed mostly solid; I donāt remember how many perimeters, but I think I used 90ā° rectilinear infill at 90-100%. I just pushed an update that adds a groove to wrap spectra around the anvil after printing. I havenāt tested this yet. Itās .75mm deep and 1.5mm wide. I am having trouble finding a reference for the strength of spectra line by cross-sectional area, so I donāt know what 1.125mmĀ² of spectra will do; it was just a wild guess based on using probably less than that to sew hammock straps that have held up three people sitting in a single hammock. 
Saw that. 5 tons is impressive.
Interesting, wouldnāt have thought the lateral compression would have helped that much.
Though, I guess it depends on how it was printed.
I found this topic in Internet, found quite interesting based on fact that we are tool supplier. The one thing for all printed tools (and Iāve seen a lot of videos and tryouts) the limitations with tonnages and worklife (as I understood near 100 bends in general). But for thin parts and fast solution it is good, no doubt. Yes, want to renew this conversation because it came more and more actual.
I saw a post in the past few weeks about punch forming with printed PLA forms. Thereās clearly ongoing interest.
I think if I were going to pick this back up, Iād make much simpler integral designs, and I would try to āannealā the PLA (itās actually nearly the opposite of real annealing). My design above is not at all robust enough for use. But the general idea seems sound.
1 Like
Have you tried any other filaments from a safety POV? E.g. do PETG or ASA fail more gradually under this type of load? I know theyāre softer, so thereās clearly a tradeoff here. I expect PET would shatter dangerously, but am not sure about PC. Anyway, Iāll probably have to try at some pointā¦
I havenāt tried any other filament types. Iād expect PETG not to fail as explosively but also to deform more easily. Similarly PC, since it is known to be conducive to cold-forming; just about the opposite of what you want here, I think.
If I do anything with this later, Iāll consider using PET-CF, because of its combination of dimensional stability and stiffness.
Agree WRT PETG, but then the whole thing is not supposed to work with plastic, right?
I would expect PET-CF to fail as explosively as PLA, if not moreso. At least the test pieces Iāve broken cracked wildly.
To be clear, I would expect to want plastic that is hard, not tough, for this purpose. I donāt want the plastic to deform. Iāll accept the risk of catastrophic failure (and prepare for it!) in order to get more precise forming. When I tested the PLA to failure, I was intentionally causing failure; it was in a hydraulic press. The question wasnāt whether it was going to break, it was what it would behave like when it did. The fact that it let loose suddenly was a good sign; it meant that it would hold its form better before it broke. 
2 Likes