This set of white RepRap parts were created in molds, instead of being printed by another RepRap. [Mark A. Ganter] of the University of Washington admits that this breaks the idea of a 3D printer that is self-replicating. But the molds – which were created by tweaking Prusa Mendel parts to be mold friendly – have the ability to produce every plastic part necessary to build your own RepRap and they can do it much faster. Once the molds were completed [Mark] and his students were able to produce a full set of parts in just 30 minutes, cutting as much as 14 hours off of the time it would have taken to print the parts. Still not convinced? How about this: the molds can be created by a 3D printer or by using a high-resolution power printing method like they have here.
The process starts by printing master parts, then creating a silicone RTV mold from them. Once the molds are ready, [Mark’s] team pours polyurethane into them and waits for it to harden. They plan to share the STL files in less than a week so that you can make your own molds to use to build your RepRap army.
So you don’t have a 3d printer that prints itself, you have a printer that you use to make molds then you have to (very carefully) mix up RTV silicone (which is notoriously difficult to work with and quite expensive) and then let that set and then hope you didn’t get any bubbles in it and that you got a solid mix and the cure didn’t inhibit, etc and then do the exact same thing with urethane. Did I mention silicone can flex, throwing off tolerances. I like the concept but just don’t expect this to be like OMGZHHZHZZ I COPIED A REPRAP ITS TURNING INTO GREYGOO AHHHH because that isn’t the reality of how this works.
> high-resolution power printing method
thats POWDER not power
There are very specific benefits for 3d printing as a manufacturing method. there has been homebrew molding for a long while, and it is totally awesome. but you can’t make the same kind of parts.
one example is the square parts on the right. they mount a motor, but where they mount to the system is through two long bores from left to right on the top and bottom of the part. (you can see small nubs in the mold below) now yes, you made a part that looks like the 3d printed version, but now you have to go in and drill a rather long 8mm hole, twice, through each part. kind of defeats the purpose.
the beauty of 3d printing is that you only build what is necessary, and your finish work is limited.
i’d also like to point out that when the reprap community started, much of the first prototypes were cast, since there wasn’t a reprap to print it on at that point.
Also – silicone shrinks a bit (maybe 1%) or so when it cures, depending on the type. Urethane shrinks a bit more but each type shrinks different amounts. This may not matter for RepRaps but it is something to bear in mind. Also, please if you do this, take precautions with the urethanes when you are working with them. The iso side is pretty nasty.
that being said, the part on the left is totally fine and usable (no holes in other planes), so yes, molding is a valid manufacturing technique. at least for that part.
If you are going to work with silicones or urethanes, this link is highly relevant and extremely useful:
http://lcamtuf.coredump.cx/guerrilla_cnc1.shtml
In theory, one could also try to use Oogoo instead of a pricey silicone – http://www.instructables.com/id/How-To-Make-Your-Own-Sugru-Substitute/step3/Mixing-Oogoo/
The oogoo isnt great on detail and the solvents to get it to flow add lots of shrinkage. Better on detail and workability is silicone (off the shelf) with some glycerin and acrylic paint. Works better for home molding than the oogoo.
about the process, its less time, but more manual work, not exactly a golden grail
btw homemade (and professional) CNC machines are made from Polymer Concrete (Epoxy Granite), this mold idea is going this way.
I wonder why don’t we have faster 3D printers with “successive approximation” addition, starting with coarse addition, using wide nozzles, then adding progressively finer details using twice narrower nozzles in each pass. Perhaps even picking and placing industrial polymer granules or even some sort of ready made primitives (like Lego, only without disassembly) and pasting them to each other by molten polymer thread (“printing” between them).
Oh, and if it isn’t asking too much, can we have “un-printers”: CNC machines which melt, suck away quanta of plastic from their working object in controlled fashion, and form the plastic thread from it to be reused in 3D printers? That way we could reuse, repair and fine-finish our creations, even do recycle-at-home, putting plastic garbage to a good use instead of letting it float to oceans’ great gyres.
It is hard to print huge verticl walls with a nozzle.
Unprinters are called CNC machines. They cut and mill rather than suck.
Epoxy is rigid. Casting rigid into rigid is very hard to demold.
@salec.
yes, yes. today, the thinking on 3d printing is focused on the one-touch machine. it is a miracle that still needs to get in touch with reality.
but what you say should be possible, no?
Hardest crowd to please ever……
This isn’t a replicator like on star trek, but its still a valid foundation. Last I read, there are still issues with mendel’s that haven’t been readily fixed…like 14 hour print times, or variances between prints because most of the non printed parts are scavenged and may not always be 100% spec….
but lets keep splitting hairs since we can all do way better than any one else…sheeesh…try to give some credit when its due..for the record this is innovative use of the mendel, maybe not be fully inline with their mission statement, but innovative none the less. and no it isn’t the holy grail of star trek replication nor is it the answer to life, the universe and everything (that would be 42!)
@salec
Well.. we were waiting for you to invent them. So what’s keeping you..
Seriously though.. This stuff is still taking it’s first steps. And the affordable end of the wedge is only just getting here.
I think the RepRap project is amazing. In concept. The execution and reality…(make ANYTHING! It will be GREAT!) doesn’t quite live up to the hype (actually, you can’t).
I deeply encourage it but I remain convinced that the RepRap is inherently limited by its design. It is a big glue gun. It will never be able to reach the level of precision of some other 3d printing methods. Like SLA or Polyjet or projector based UV curing. I may be biased because I own these types of machines but when I see outputs that my machines put off and compare them to the RepRaps…. well there is a reason you get what you pay for. The precision just isn’t there with the RepRaps. I wish it was. But even if it was, you can’t fill in square “pixels” with round “glue guns”. There are practical limits to these machines – which is completely fine. I want people to improve on them. I want people to make awesome things with them. I just don’t want people to think that they will be able to do it out of the box with no effort or hassle and with perfect parts every time. This community is obviously able to roll up its sleeves and figure out problems which is good because the RepRap needs a lot of that until you really dial it in and even then it isn’t like a printer you can buy at a store.
Yay. share your STL files so we can all pay you a buttload of money to print this crap and then pay some silicone foundry a buttload of money to turn them into silicone molds.
just sell the molds for cheap.
Well, I just finished printing my own Prusa-Mendel on a hot-rodded cupcake (took about a day unsupervised). I have to say this is a neat idea, but I’ll stick with the tried and true method. Working with molds is great if you’re not modifying anything, but the flexibility of being able to say “hey, let’s add an X to the Y so it can also do Z” and then tacking that onto the file is just much better for my purposes.
We have cupcakes and mendels at Metrix Create:Space, and we’ve printed mendels, and it’s great, but it doesn’t make for ‘quick replication’ or all that many more 3d printers. The way to get more printers is to make more printers. Yes, it’s a moving target, so mass production is a miss, but distributed small scale production (which turns out, is mass production) is totally within reach.
We have a powder printer at the shop now, and we are about to start selling molds (we have already started the printing process) as soon as these files are released. We will also be providing monthly mold diffs for when the target moves. Just because it is a mold, doesn’t mean you get left behind when things change. 3D printing a mold is a *lot* like 3D printing a part, but it takes the same amount of time for far greater output.
For those who feel this is crazy and anti-reprap, the STLs are open source, and what we’re really doing is speeding up this process. If you have a powder printer, you should be doing this.
Printing *parts* directly is -slow-
If you have molds, you can pour 13 clonedels in the time it takes to print 1 prusa.
Additionally, complaining about what the quality of molded parts is a trap if you think about it. If you have the molds, you can pour whatever you want in them. Want a floppy mendel? Go ahead. To harden things up, or make the material have greater tensile strength, you just pour something else or use an additive. There are already PLA mendels, ABS mendels, wood mendels and aluminum. Molds expand the possibilities beyond what you can extrude.
Self replicating aside, the exciting thing about the reprap project is the accessibility of cheap and capable XYZ cartesian robots. Right now, electronics are still the most expensive and out of stock part of this problem. Addressing that is the next step.
Lets stop the religious argument of self replicating, make more robots and get back to designing cool things to make on them. Or, get them into the hands of those that can.
What we need to do is make a printer that:
A) Prints a mold set
B) Extrudes material into the molds
C) Regulates humidity and temperature during curing
D) Safely demolds the pieces
A) Go To B
Or, more likely, two machines: one to print the molds, one to use them.
Sorry, but although it’s awesome, using casting with a reprap isn’t new.
Here was used liquid silicone ( mold ), and polyurethane ( final object) to copy a reprap printed part to produce a duplicate …. IN 2007!
http://www.reprap.org/wiki/SoftToolingForCasting
Here was used a higher-temperature method to create metal copies of repraped objects with plaster-of-paris ( mold ) and pewter ( final object).
I Hope this helps some of you!
Buzz
From the http://www.reprap.org/wiki/Admin team.
Anyone thinking that parts for a mendel HAS to be printed is missing the point of a mendel.
It is a great thing that the reprap project exists and its goals are honorable, but the point of a mendel is not to be fruitful and multiply. The medel is a product of the reprap project and has many uses outside of reprap.
It’s called prototype development and was never meant to produce production quantity parts.
Molded polyurethane IS meant to produce parts in quantity and any mold maker worth his/her salt can and do make excellent quality molds from which excellent quality part emerge which require little or no finishing. So someone has to drill a hole because a part may be too complex for a mold. So what? If it take 10 minutes to make the part, 2 minutes to drill the hole you’re way ahead if the part takes 2 hours to print.
These guys are not strictly pushing the reprap agenda, they’re pushing the mendel agenda.
Get over it.
PS: I agree, maybe you should sell the molds but just having the STL files is enough. they are easier to ship than all that silicon.