When I first discovered the RepRap project about a year ago, I was excited. The idea of a self replicating manufacturing device is a very powerful one indeed. The potential is endless. It could eventually lead to a world where people could manipulate the physical world as easily as they can manipulate information today.

One year later, my enthusiasm has wained. What RepRap is right now is an affordable open source 3D printer. One with some serious limitations, but also one with potential for great things. I do not however, see this as ever becoming the transformational device that I was hoping for. Why? Several things really.

First there is the obvious point. For anyone following the project, it can't have escaped your notice that for all the effort put into this device, the amount of a RepRap that can be self-replicated is not increasing. It is still heavily reliant on traditional manufacturing technology, and there doesn't appear to be any concerted effort to improve the situation. I think the community is by and large ignoring this issue, and simply focusing on making the most of what RepRap is currently capable of, rather than expanding it's ability. It's pretty clear that in order for RepRap to achieve it's goal of self-replication, a lot more effort is needed. Now I'm aware of the inherent difficulties in self-replicating many of RepRap's parts, but that brings me to an interesting question.

Is what is being attempted here even possible? The operating principles of a 3D printer are significantly different than any known self-replication process. As I see it, there are a few brick walls, which may keep RepRap from ever advancing much beyond it's current state.

One is the extruder mechanism. Because it relies on heat, and the bonding principles of heated molecules, it can never really print with any material that has a greater or equal melting temperature than itself. That means the the device would always have the limitation of not being able to fully self replicate, without relying on some external mechanism to produce the extruder. That in and of itself is not a deal-breaker. If the only thing left in the world that needed to be manufactured by a traditional process was the extruder, it'd be pretty hard not to call that a win.

But then there is the issue of material strength. If the RepRap is never able to print with anything stronger than plastic, that would be a deal-breaker. A lot of things are made out of plastic, but giving people the ability to print self-designed plastic trinkets is hardly going to be a world changing accomplishment. In order to make this device really useful, it must be able to print with much stronger materials. As metal is just about the only thing strong enough that could easily be melted down and extruded, it would seem like solving the issues involved with extruding metal should be a priority. However as was mentioned in this thread, metal tends not to conveniently bond with itself when extruded like plastic does. There are multiple ways to get around this, such as implementing some sort of casting process into the device, or redesigning it so it melts metal down into blocks and etches away at them in a two stage process. I don't see this issue as being fundamentally unsolvable. It will just require a hell of a lot of effort. Perhaps even a complete ground-up redesign. RepRap might even need to become a multiple device project in order to overcome it. Either way it's a pretty big issue that's not being addressed, but that's not even the half of it.

Lastly there is the most important issue of all, the use of complex electronics. The RepRap, such as it is, is not even remotely capable of reproducing these things. Even the larger electronic components are well beyond it's abilities, given that it can't print with any conductive materials. And even if you could get it printing the larger components, what about the integrated circuits? If you could get it to the point where everything else was self replicable, perhaps you could scale the print mechanism down far enough to be able to print them. There would also be the possibility of multilevel printing, which may make it possible to fully reproduce the electronics at a significantly lower resolution than the current off the shelf parts are manufactured with. Of course, this is all entirely theoretical, and so far beyond the devices current capabilities that it's probably not with considering yet. But that's the project's end game, in order to achieve complete self replication.

Self-replication is certainly an interesting goal, although I can't say for certain if it's possible. But if this is what RepRap is being designed to do, so many things still need to happen that aren't even being worked on. At least not as far as I can tell. Maybe people are working on it behind the scenes, but I haven't seen anything being publicly released. So I'm just curious, is anyone working on these things? Is any tangible progress being made? Has the self-replication objective already been for all intents and purposes been dropped?

Despite what I've said, I still think that in it's current state, RepRap is a great project with a large amount of usefulness. It just seems to me that it ultimately needs to decide what it wants to do with itself. If all RepRap wants to be is an open source 3D printer, then that's great, and the current direction is fine. Just free the RepRap of the design limitations that have been imposed to facilitate self-replication, and the sky is the limit. If the RepRap still wants to be a self-replicating device though, the project may be a complete tear down. I think that's ok too. Making mistakes is how you learn. The problem RepRap is attempting to solve is worth solving, no matter how many attempts it takes.

... my thoughts were similar, when i joinded reprap - but i'm planning on a wider time-scale

When i first dived into 3D-RP around 1987 there was only CNC-milling, Stereolitho and first starts of LOM-fabbing around ...

The last years i made some progress with pastes mixed from plastics, ceramics and metalls, what's really promising in terms of rigidity and conductivity ... and possible accuracies are slightly below 0.1mm ...

Now i'm tinkering with lasers and have the possibility to cut/engrave or build complex parts with laminating ... or even selective laser sintering with accuracies down to 0.01mm in nearly all possible materials theat will absorb the beam (i have to colour bright or clear materials, so they will melt and/or evaporate in the beam).

This isn't a really shortcut to 'self-replication', as you will encounter troubles building lasers, lenses or silicone chips ... but i can make precise moulds and etching masks, so some DIY-microtech is possible too.

And for some other essential 'vitamines' (e.g. chips) you'll have to look into new technologies as printing organic electronics and 3D-chips made from laminated OLED-structures ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

The things you said you were working on do sound promising, but is there any realistic game plan or time frame for integrating that stuff into RepRap's existing design?

... i'm developing this primary for my own needs (sometimes co-funded by individuals or companies interested in special solutions), but most of the results could be integrated, if needed.

I think most reprappers are busy with their plastic-extruders and the 'official' path is concentrated on extruders too, but with time there sould be more crossover

There are some single developers like me around powder-printing, paste-extruding and laser-sintering, but the 'critical mass' wasn't hit until now. Actually i've solded two 'naked' diodes and two complete 5Watt-IR-modules ... and it should be only a matter of time, when some other folk will start with this too, so the word will be spread

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

the key to getting reprap development expanded is getting more repraps in service and developing. There are a lot of developers working on various issues you mentioned, MrKim's oled circuits, unfold's ceramic extrusion, etc, I personaly am working on both the ceramic extrusion and organic printing of circuitry. Im also working on expanding on jeri ellisworths home microchip manufacturing such that it could be applied to reprap. all of these are long term goals. Another of my long term projects is developing a ebf3 print head for a reprap, though this would probably require a major redisign, but since my goal is a reprap then by definition it will become a reprap.

The work is being done but most of the current work goes into making current technology better/faster/stronger (breaks into daft punk)

I was excited when I read the reprap home page and design goals. But before diving in and buying plastic parts for a plastic extruder, I thought about it. That's why I bought a mill instead.

A mill is a device conforming to the reprap ideal, because it can be used to create pretty much all of its own parts (except the motors and electronics). The mill can also make more of the parts of Mendel than the Mendel itself can.

Now, there are often more steps in manufacturing something using the mill. There are some shape limitations on what you can directly manufacture. But the diversity of workable materials and the precision with which they can be worked, gives it a flexibility that I feared not having if I spent my money on the Mendel. You can even mount a plastic extruder in place of the standard spindle.

One thing I can do with the mill that I haven't seen the reprap community doing much, is making additional tools to take advantage of different manufacturing techniques. If I wanted to make a thousand sets of identical plastic parts, for example, I'd mill some molds and a plastic injection rig, then use injection molding. I have the impression that the "reprap community" would instead just undertake to extrude each and every part, one after the other, on their Mendel machines. Which would take forever, and produce inferior parts.

Every finished product is a realization of a design goal, achievable with the right tools. The right tools, likewise are themselves finished products, and the realization of design goals which were achievable with other tools. But usually you don't have to go more than a few steps down that chain to find that you could achieve a given finished product starting with a mill.

So anyway, while plastic extruders are cool, if I'd been given the reprap manifesto I'd have immediately concluded that the reprap machine ought to have been some kind of mill.

... it's a bit more complex than that - a mill needs sturdier mechanics and better/expensiver bearings, so the cost to build a reliable CNC-mill from scratch is much more compared to a 'plastic-n-sticks-frame' what's the RepRap-basic structure.

Following RepRap over 3 years now and don't own a printer - mostly because of the 'poor' quality of 0.1mm ... while i need resolutions and accuracies of 0.01mm and even finer ... i'm achieving this with aditional toolheads ty my CNC-mill or developing of special tools and methodes ...

But the basic idea of RepRap is great! - and with time there will be some improvement, so this limitation will be adressed too

Actually i'm 'repstrapping' my CNC-mill by replacing the controller with a Gen3-motherboard and fixing an extruder to the toolhead - mostly to develop kits for repstrapping common CNC-mills ... and for developing/offering aditional toolheads like paste-dispensers, laserheads, fabbing methods and more ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

I have to say the 0.1mm tolerances in finished pieces were the main sticking point for me too.

Almost every Thing I want to make is a mechanical part I can't buy. And mechanical parts need to be pretty darn precise to work well, work reliably, and work efficiently. If I'd been able to get 0.002mm precision, which I can get easily from my mill, I'd probably have gone for it, including making gears etc out of plastic that I'd prefer to have in metal.

I had to look at what I wanted to make (gearboxes, transmissions, pistons, cylinders, robot parts, etc) and I just found that I couldn't use the output of the plastic-extruder reprap. It's a cool idea, but so far, it's just not good enough for what I want to do.

Plastic injection molds off my mill, on the other hand, make plastic parts I can actually use.

Bear

... my focus is aimed on DIY-able technologies for high precision fabbing/manufacturing.

Here i have tools and support for:
- milling ...0.0125mm with my old mill, up to 0.001mm with linear stages i have to assemble for another 3-axis mechanic
- paste-dispensing ... down to 0.06mm nozzles, but with common pastes best usable 0.1mm
- LOM-fabbing ... depends on sheet-heights and cutting accuracy
- diodelaser-heads (up to 20Watts) ... around 0.1mm focus diameter with IR (808nm, 975nm), maybe 0.05mm diameter with 445nm and 405nm for:
- - cutting thin sheets (e.g. for LOM-fabbing)
- - engraving plastics and wood
- - SLS - laser-sintering of powders
- - UV-curing epoxies ... maybe 0.03mm focus with 405nm diodes
- fiberlaser (50Watts) ... down to 0.01mm focus diameter, but not really in the DIY-leaque ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Lets take this to the next step - wouldn't it be cool if the rep rap could work with self replicating materials...

;-)

CS

Personally I see the RepRap as the first step on a very long top down road to the ultimate* desktop fabricator along Drexlarian Nanotechnology lines, such as the desktop Nanofactory:

[www.youtube.com]

In this context the RepRap is one part of a larger replication machine or factory at a macro level. It might take several desktops and years more development to get close to a true replicator even at the macro level, but the information gained, especially in the programming of such systems, will be invaluable for the Nanofactory. See the primitive Nanofactory by Chris Phoenix of the CRN (free PDF download of the paper under the JET paid for version):

[crnano.org]

Julian


*Yes, I'm aware it's always dangerous to say anything is ultimate....

Does RepRap assemble itself? If not, than it is simply not self-replicating.

In which case, you have not achieved your goal at all.

... so what?

There is the nontrivial point of 'vitamines' needed for completing all common system (be it a reprap, a CNC-mill or a building) - and the assembling is one 'vitamin' more, that's adressed to be replaced with time

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

I tend to not care about "vitamins" and self-reproducibility, as long as the vitamins are cheap and abundant.

There's always the three bins "what I can make", "what I have to buy", and "what I have to buy and is expensive and/or hard to find".

In the "what I can make", best time in my whole making life (the exhilarating sensation the first time you get power generation out of the system you've built: priceless!), some years ago, out in the rainy woods: a full pico-hydro system. 120 m of water pressure, 1 to 2 liters per second, 400m of 100mm reinforced plastic tubing, right in the midst of nowhere. 120 mm pelton wheel, big car alternator, deep cycle batteries, sinusoidal inverter. 13.5 KWH per day, more than enough for two houses, being careful: lighting, telecoms, even the washing machine. With more electronics knowledge, could have made a more efficient direct AC generator from a big motor, but there was no time for that, "out in the woods" means many, many concurrent making tasks.

So how does that influence the design of projects? you tend to overbuild, just to make d... sure you won't be bothered by early failure. Wood building pieces are cut heavier and thicker so they'll take longer to rot and you won't have to deal again with 'em too soon, hopefully they should last more than a generation.etc. It is a wholly different mindset that the one you get with the commercials at hand.
But with a family, it wasn't sustainable.

So now out of the woods, I tend to invest my time in the parts that are either interesting for learning (the essential added value), or expensive/customized/hard to get.

Plywood or MDF are way cheaper than plastic, I also like the precision of routing/milling. Not the noise, but I'll have a more apt place soon.
Have yet to find the exact place of extruded plastic in my ecosystem, but it is fun (and more musical than the router).

Tooling up is an iterative process.
I am glad I started the process upstram with a CNC router instead of any sort of reprap kit.
Maybe next year fabrication will include a real mill (I'll have to buy it) or a laser cutter (that one I can build). The most fascinating is that Metallicarap thing, but for now getting into that seems totally out of budget.

Conclusion: I couldn't care less about the differences between a reprap and a repstrap.

Reprap to me means completely self fabricated.

Looks like I took the same route as Lantham above. I build a CNC rounter in metal. In order to build that router I had to learn to MIG weld. Both of these fabrication techniques were then appled to my repstrap. Corner pieces, servo plateforms and such were built from 3/4" 7 ply plywood (I didn't use the top grade plywood) Other connection points like the 8mm rods that intersect the threaded rod are inserted in tubes that are welded at right angles. Most of the extruder is welded, except for the gears that are printed. Wood is a great construction media as long as you don't go overboard and try to make everything out of it.

Reprap to me means econimical 3D printer for the home.

I find that the CNC router is more practical for just about anything made from wood but fabracation is more a process then the 3D printer. Since the router is 2 1/2 D is sometime tricky to get depth. The printer is facinating in that anything you find in STL (considering when printing unsupported surfaces) can be printed. See it on the page ---> print it. Think up a configuration that you can design in a CAD system ----> print it.

Self replicating is not important to me simply because I don't believe printed parts perform the best for every component of the printer. And as someone pointed out above, a printer will never fully self replace (and also be the best at printing) so why fuss about the concept.

... and for a 'really self replicating device' you should take a look into Eric Drexlers 'Universal Assembler'

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]