New to the reprap crowd, have a few questions.

Recently I have come across the reprap concept and it has peaked my interest.

Has anyone experimented with this technology by combining it with a wire welder unit that will print out structural steel parts?

I have previous CNC experience from college but that was over 12 years ago, how easy is it to learn the new technology to operate these machines?

Has there actually been an industrial demand for these devices or are they mainly a benchtop toy for the mechanically minded?

And lastly, I've read in a few different topics that there can be an issue with the printed out parts sagging due to excess heat remaining in the material, what methods have been tested or proven to eliminate this issue?

Has anyone experimented with this technology by combining it with a wire welder unit that will print out structural steel parts?

It's been proposed. You'd probably want to also use a milling head as per
[objects.reprap.org]
but with steel rather than plastic.

Academic-types have done a project or two with this, but there's been no comms back and forth.

I have previous CNC experience from college but that was over 12 years ago, how easy is it to learn the new technology to operate these machines?
CAD it up, send it to the printer. Go make coffee.

Has there actually been an industrial demand for these devices or are they mainly a benchtop toy for the mechanically minded?
No industrial demand yet. We're a research platform and a desktop toy.

This is still a pipe dream:
[objects.reprap.org]

I'm waiting on the Spoolhead group doing their release talking it up.
[objects.reprap.org]

Basically, we're where computers were 30 years ago. Big iron, and these funny hobbyist toys that don't do anything 'useful' ...

What is your proposal for the RepRap killer ap? And, how do you want to help. Ceramics? Laser cutting? Milling head?

[objects.reprap.org]

And lastly, I've read in a few different topics that there can be an issue with the printed out parts sagging due to excess heat remaining in the material, what methods have been tested or proven to eliminate this issue?

[objects.reprap.org]
also, We're still researching support material.

Edited 1 time(s). Last edit at 03/08/2010 11:36PM by SebastienBailard.

Economies of scale preclude reprap from producing commercial quantities of things, but those same economies of scale mean reprap is perfect for one-offs or test runs. The commercial rapid prototyping machines are far more expensive, but have similar limitations, so we're the underdog in an existing market.

However, don't be fooled- this is still very much a research project. Attempts to restructure it into a commercially viable offering have been shouted down many times in these forums.

I believe there's been talk in the forum of fitting a mig welding head to a reprap, but I also don't think anyone has done it. Feel free to be the first if you want I think the most difficult problem here is removing the final object from the bed...

Skeinforge has numerous modules to fix numerous problems. I believe one of them causes the head to orbit the build platform between small layers to give the layer time to solidify.

If you want to see what's possible with DIY rapid prototyping, nophead's blog has some lovely examples. casainho's blog has some neat examples too. There are numerous others who have posted excellent work, check out the 'reprap blogs' link on the front page of reprap.org for a comprehensive selection.

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

oh, it's also worthy to note that there are about as many different software/firmware/hardware combinations as blog contributors at the moment- we're still deeply in the experimental stage where nothing has been shown to work best for almost everyone.

Personally, I'm making mendel's RP parts from wood as faithful to the originals as I can get away with (except for using M6 threaded rod instead of M8), but I have a totally custom electronics and firmware setup. Haven't yet built extruder or finished frame, so can't tell you how well it works yet.

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

I've already thought out an idea that will elliminate permanently welding the printed object to the bed when using a wire welder setup.

The bed will need to be made out of high temp ceramic and there will need to be a grounded starter plate of about 1/2 inch (13mm) square that's 1/8th inch (3.5mm) thick that is some how firmly attached to the ceramic on the outside edge of the table. The welding wire will be programmed to start on this location then slowly weld it's way to the center of the table where it will begin the actual 3D buildup of the part.

I'm sure there are several logistical problems to over come and someone would have to develope a way to convert a wire feed welder for use with ultrathin welding wire maybe in the .006 to .015 inch range other wise the parts will not be able to be made with any sort of precision since the average weld is 1/4 inch (6mm) or larger.

However, don't be fooled- this is still very much a research project. Attempts to restructure it into a commercially viable offering have been shouted down many times in these forums.

Is that what's happening?

I have no problem with people selling 2048 kits a day. And I think a commercially-viable kit is even better. I mean, if someone wants to define and start maintaining a 'stable branch', that's great.

But I haven't seen a viable attempt to 'restructure it into a commercially viable offering'. One or two people did say 'you should run this like a commercial company'. But it was more along the lines of trying to police generic user-devs' behavior. Or a 'design and build this as I tell you, along these lines'. So ... doomed to failure. I mean, if you start a good project on RepRap.org with good design, start building it, and talk it up, it will take off.

It's a lot like herding cats, except, dev costs are higher for reprap than for generic software, and people who sell kits may be too busy to make super easy for folk to replicate their work, so it's more like herding dragons.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

Too much heat only becomes a problem on small parts where the heat is coming in from the hot plastic faster than it can leave by convection, etc.

If you have a stepper based extruder you can simply slow down when the layer is small. I drop to 8mm/s when the layer takes less than 10s.

Cooling with a fan also works, but is not compatible with a heated bed. It would have to blow warm air.

Having a pause between layers works as long as your extruder is not incontinent.

---

[www.hydraraptor.blogspot.com]

SebastienBailard Wrote:
-------------------------------------------------------
> I have no problem with people selling 2048 kits a
> day. And I think a commercially-viable kit is
> even better. I mean, if someone wants to define
> and start maintaining a 'stable branch', that's
> great.

I also have no problems with commercialisation of our research as long as the licenses are obeyed- it's the attempts to govern user/developers which you mention that I was referring to

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

As for the commercial viability of the reprap I just didn't know if there was a commercial application already in use. So that question was just there to satisfy my own curiosity.

I have already seen that up to this point in the development of the reprap devices that they are not an affordable option for high volume production scenarios. CNC machine tools and high volume mold tooling beat them out hands down.

Having spent time in the production of composite aircraft parts I could see a well engineered reprap device being capable of making master molds that could then be used to recreate copies of the molds out of tougher composite materials.

While I was working for the aircraft company all of the molds were hand made, which meant that there were difficulties getting the lefts and rights to match correctly. A reprap could have made the RH/LH molds mirror images of each other.

One of our lower wing molds cost over $250,000 to produce, it took six months to complete then an engineer was fired from the company when it was discovered that the mold was constructed with one leading edge angling just a fraction of a degree off which was enough to create an out of tolerance overbite when the upper wing section was bonded into place. There was no way to fix the mold so we ended up with a $250,000 yard ornament.

I have already seen that up to this point in the development of the reprap devices that they are not an affordable option for high volume production scenarios. CNC machine tools and high volume mold tooling beat them out hands down.

True. Folk are working on:
[objects.reprap.org]
[objects.reprap.org]
or maybe (vaporware)
[objects.reprap.org]
for CNC/subtractive. They are research topics rather than mature releases.

For low volume mold tooling, people will print the molds out of thermoplastic or make them out of rubber.

Here are mendel injection molding people:
[dev.forums.reprap.org]

I find it very interesting that injection molding people are entering this market. I don't know if it is possible to recoup costs on injection molds before would-be RepRap user-developers shift their focus to mendel-apollo or a post-mendel.

Plus, they're participating in a market where eventually people will run their repraps for for 4 days in order to sell a set of parts at (materials cost + $40) because 'hey, $40 and all I had to do was box it and ship it'.

I have been posting all over the place, but I have some newcomer questions, too

What 3D CAD drawing program are we using? Is there any prefered one, or does everyone scrape up what they can as long as it output DXF files?

Once we have a 3D model, how does it get converted from DXF to something that a reprap can use? I hear a lot about g-code, and I am guessing that g-code runs on the microcontroller board that controls the stepper motors, but that is only a guess. I have also heard mention of skeinfold, Blender, and a few other things. What software should I be loading on my computer to start designing parts that I can eventually make? What will I have to customize or write on my own to get my modified McWire to start printing Mendel parts?

What is the process from first thought to software model to processed file to reprap commands to stepper motor commands?


I was also looking at starting a wiki page for the wire-cable powered McWire I am building, and also a wiki page for the Team Open Air design I want. I see that Sebastien has created a very nice default example page, but I do not know how to copy it and rename it and get it into the wiki index.

So many questions, so little time!


Mike

---

Team Open Air
Blog Team Open Air
rocket scientists think LIGHTYEARS outside the box!

@SebastienBailard

How easy would it be to add a small forum specifically for the newby's like me, so that we can learn about the basic specifications and functions of the reprap devices.

Maybe treat it like the military does with their instruction manuals, describe what the machine is, what it does, how it does it and so forth in not so large paragraphs using language that a 5 year old could follow.

Call the forum "First time at Reprap.org?" or something similar.

I'm sure you know that there are already wire welding robot arms. I remember a research paper that made a pot out of metal using that technique in the 80's I believe.

The problem was cost of fabrication. Welding wire and the energy to melt it is usually very inefficient in terms of thermal loss, so its prohibitively expensive.

Personally, I have some ideas about doing something like this, but it would require inventing a metal deposition system that was more energy efficient than whats on the market today.

So we need a sticky FAQ for new people it sounds like.

Yeah. I was thinking the new forum for the newbys would be a great place for them to ask the dumb questions and the regulars could answer said questions if they are feeling generous with thier experiences with the Reprap machines.

rocket_scientist Wrote:
-------------------------------------------------------
> I have been posting all over the place, but I have
> some newcomer questions, too
>
> What 3D CAD drawing program are we using? Is there
> any prefered one, or does everyone scrape up what
> they can as long as it output DXF files?

STL actually, not DXF. I use blender. Others use AOI or solidworks or openscad. I want to try openscad but I need to do a bunch of software wrangling first.

STL is a list of triangles in 3d space, so it's a very primitive format but it does what we need. GTS has been suggested as a replacement, since it supports colours (different materials) and other features.

> Once we have a 3D model, how does it get converted
> from DXF to something that a reprap can use? I
> hear a lot about g-code, and I am guessing that
> g-code runs on the microcontroller board that
> controls the stepper motors, but that is only a
> guess. I have also heard mention of skeinfold,
> Blender, and a few other things. What software
> should I be loading on my computer to start
> designing parts that I can eventually make? What
> will I have to customize or write on my own to get
> my modified McWire to start printing Mendel
> parts?

STL gets converted to gcode, which tells the hardware platform what to do- literally move to this location, set heater to such and such temperature, extrude this much extrudate while moving to next location etc.

The STL describes a shape, whereas gcode tells a particular machine how to construct that shape. SO, we can all share the STL but the gcode must be regenerated for each machine.

There are various programs to convert STL to gcode. skeinforge is one that seems to be getting lots of attention lately.

> What is the process from first thought to software
> model to processed file to reprap commands to
> stepper motor commands?

concept -> 3d model -> stl file -> slicer program plus machine settings -> gcode file -> machine -> stepper pulses/heater/etc -> physical object

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

Thanks Triffid. That clears up some confusion of mine. I have a student copy of Solidworks, but sometimes use Google Skectup. It looks like both have at least some ability to export in STL.

Mike

how well & fine can you make the finish on the r/r creations?can it be made from different colors filaments or painted over?sanded?thnx!

Sebastian,
I have actually been toying with the idea of making a flat-pack kit available especially for newcommers who wish to get started immediately are not quite developers, at least not yet. However, the very self-generating aspect of the Mendel makes it hard to make the plastic parts any other way. And making a flat-pack, software and video included of the SarterBot I am designing and planning on working on next would be quite feasible, but is it worth it to get people started in reprap with a non self-replicating kit?

At least I am comforted to hear that there is no bias against someone going commercial as long as the hardware and software are still open.

Mike


SebastienBailard Wrote:
-------------------------------------------------------
> I have no problem with people selling 2048 kits a
> day. And I think a commercially-viable kit is
> even better. I mean, if someone wants to define
> and start maintaining a 'stable branch', that's
> great.

---

Team Open Air
Blog Team Open Air
rocket scientists think LIGHTYEARS outside the box!

but is it worth it to get people started in reprap with a non self-replicating kit?
...hmmm... are you trying to make money, or trying to get more machines out there?

Random musings:

Tight now the RepStrap market is hot, but all the RepStrap companies are betting against an exponential, so ... good money but a self-cannibalizing market.

Right now a decent set of imperial-hardware plans for a 'wolfstrap' would be good.
[reprap.org]

Doing kits is tedious, unless you're in it for the money.

Don't try to reinvent every part of the toolchain. Get your hardware out first, using 'vanilla electronics' and 'vanilla software'. Then go crazy hacking software, but only change one variable at a time.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

I am mostly doing it to get more machines out there, and get people with lesser technical, geeky backgrounds to get their toes in the door. But if course, I would rather make money than lose it.

The very thing that makes a Mendel a self replicating 3D printer, the very high parts count of specialized plastic parts, make it hard to mass produce by standard techniques. I do not know much about manufacturing, but injection molding would seem to be the best choice for mass manufacturing a true reprap that is not yet fast or widespread enough to make all its own parts. But shooting from the hip, the traffic on these forums would make me think that the first 100 Mendel kits would fly off the shelves, but the second hundred would require word-of-mouth news spreading to lots of friends and relatives, and finishing the first thousand might require outside advertising, such as being featured in a national (or international) news broadcast.

A second point is that a thousand Mendels in the hands of developers, dreamers, and engineers would be the best thing that could happen to the group effort tight now. But 10,000 Mendels in the hands of Joe-Q-sixpack could hurt the project by creating a negative public impression of hardware that is not quite ready for prime time.

I agree with you that getting a flat-pack of hardware, and using a plain vanilla tool chain for either a Windows machine or Linux (possibly boot if I am feeling generous). The extruder needs to be either prefabricated, or drilled and tapped and ready to assemble, probably with a geared down drive to provide lots of torque. Plus all electronics already soldered and connectors mounted, cables, ATX swithiching power supply, nuts and bolts. Ad I would take one of the first kits to my sister and hammer out the instructions until she can build one unassisted. But unless I can find a Chinese factory willing to tool up for a small run, I don't see a way to even break even on all this work.

Mike
P.S. Sorry for topic stealing. Sebastian, perhaps you can pull my last two posts, plus all the others that you feel have jumped topic to selling kits, and start a whole new thread "Is anyone willing to pay for flat-packing SarterBots and/or Mendels?" I am not sure how to split threads on this forum, and preferably don't have the rights.

---

Team Open Air
Blog Team Open Air
rocket scientists think LIGHTYEARS outside the box!

A second point is that a thousand Mendels in the hands of developers, dreamers, and engineers would be the best thing that could happen to the group effort tight now.
Mendel went viral some time ago. Those developers, dreamers, and engineers have their machines, both RepRap and RepStrap. And they're all using thingiverse to contribute to RepRap.

But 10,000 Mendels in the hands of Joe-Q-sixpack could hurt the project by creating a negative public impression of hardware that is not quite ready for prime time. If you're hucking them on tv or at CES, selling them as turn key machines, that is one thing. But if it is clear they are kits, it's all good. And if someone is a beer drinker, and a reprap-user-developer, then it's a little silly to disparage their reprap-fu.

But shooting from the hip, the traffic on these forums would make me think that the first 100 Mendel kits would fly off the shelves, but the second hundred would require word-of-mouth news spreading to lots of friends and relatives, and finishing the first thousand might require outside advertising, such as being featured in a national (or international) news broadcast.

Makerbot ~ 10K machines? All of them making whistles? And the occasional Mendel?

And all of those Mendels making > 1 Mendel each? And all of those Mendels making > 1 Mendel each? And all of those Mendels making > 1 Mendel each? And all of those Mendels making > 1 Mendel each? And all of those Mendels making > 1 Mendel each? And all of those Mendels making > 1 Mendel each? And all of those Mendels making > 1 Mendel each? And all of those Mendels making > 1 Mendel each?

Your math teachers did warn you to never bet against an exponential, right?

But unless I can find a Chinese factory willing to tool up for a small run, I don't see a way to even break even on all this work.
We don't need a factory. We just need a guy in Shenzen with a Darwin, lots of steppers and filament, and a couple years of free time.

P.S. Sorry for topic stealing. Sebastian, perhaps you can pull my last two posts, plus all the others that you feel have jumped topic to selling kits, and start a whole new thread "Is anyone willing to pay for flat-packing SarterBots and/or Mendels?" I am not sure how to split threads on this forum, and preferably don't have the rights.

Too much work. Do up a napkin sketch of your flatpack-repstrap-which-can-be-made-with-a-4x8-cnc-router, and start a new thread.

---

-Sebastien, RepRap.org library gnome.

---

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything!

Quote
Sebastian
Makerbot ~ 10K machines? All of them making whistles? And the occasional Mendel?

I guess my research is way out of date. I thought they were selling dozens per month, not hundreds. With those numbers, it might be worth while trying to put a kit package together.

Mike