How's progress on Metalicarap?

It looks like things have stalled. Actually it seems like maybe the goal was just way to high here. It's the same mistake that Stallman made when developing GNU; try to do things super well the first time even if it takes forever, and it costs too much and takes too long and net result is nothing gets done I'm afraid.

I think maybe a mini version would be a good intermediate step. People don't join a project until it looks serious, and a working say 1/6 scale prototype that at least demonstrates sucessful building would do that.

It may appear wasteful at first to pursue this sort of thing, but a)it's educational and this is essentially a research project. What is learned will be useful in avoiding costly mistakes and false starts. and b) that's just the way people are. People don't get that a project is significant just by reading about what it would mean etc. very well.

Well, just look at the reprap. It's practically useless, but it has led to a huge amount of effort being directed to the development of something more.

Secondly, a small unit with no subtractive processing capability would still be extremely useful if it could make solid parts, even of poor accuracy.

Do you know what the dimensional accuracy of an arcam EBM (commercial electron beam melting printer) machine is? Roughly 300 microns last I checked. They use >185 micron powders. A third of a millimeter. Useless for nearly anything without extensive post processing. And yet it still has some use.

I think just designing and building a small machine with say a 5 by 5 by 5 build volume as a proof of concept and to gather information about what we are facing would be the best use of funds.

There are still many questions, for instance, that need to be answered:

Does surrounding powder stick to the object being built, either by sintering after the melt is done or in another way that could be a problem?

How hard is it to remove sintered metal from hard to reach areas? Will we be able to do so practically? (maybe e.g. the part could be submersed in an abrasive slurry could be agitated sonically or ultrasonically to remove powder or something. Maybe an abrasive that is hard enough for the powder but not enough to really damage the solid metal?

What sort of build speeds can we get? This has important implications as it may turn out to be a major limiting factor.

How much will thermal distortion and stress issues be a problem? High build temperatures can help but when the part contracts after building it's shape will change in ways that may be hard or impossible to predict, decreasing dimensional accuracy. Especially in the case of protrusions or limbs sticking out, it could cause large changes in the object's dimensions due to the leverage.

... it seems, my first criticism of EBM-complexity was true - much to complex for DIY or semi-proffessional level

Low-power SLS could be a good start for 'rigid-fabbing', even if in the beginning you use plastic powder instead of metal - but here too the entry level (shielding and exhausting) seems to be a killer criterion

---

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

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

GreenAtol Wrote:
> It looks like things have stalled.

Well, its probably because Rapatan is now getting into the nitty gritty of engineering out the power supply. So less time is being used on internet updating, and more time and effort is being put into research and engineering out specific issues.

Rapatan is currently dedicating three days a week to the project.
His primary efforts during the spring has been; looking into the Power Supply design, and trying to engineer out, some of the issues there. No matter the size of a prototype gun, we need a suitable power supply, that can control the electrical/magnetic elements of the gun.

His research so far has resulted in a possible reduction in cost of the High Voltage Power Supply from ~€2K to ~€0,8K. It may also become testable at only i.e. 300V. For a DIY power supply, testing of details like feed-back balances, will be much nicer to do at 300V than at 3-10.000 volts.
As part of this "engineering out" process, we may also have discovered ways to avoid nearly all high voltage components. As you can imagine, insulation issues between components and between circuit board traces in the power supply are quite different if they carry 200V, 2.000V or 20.000 Volts.

I expect Rapatan will get back to these issues when he returns from Holiday.


About the production of an actual prototype of the power supply.
Whenever we get to actually produce the first test power supply, we fortunately have a full PCB production facility available in Labitat (the hackerspace in Copenhagen, where we on Wednesdays often discuss MetalicaRap issues). We have a professional PCB CNC and a professional through-hole bath. I just love hackerspaces!!!



There are also a few others putting in thoughts, suggestions, documentation abilities, time and money to the project. It is however difficult to get experienced engineers and other experts to donate / give away for free, their hard earned knowledge. It is indeed one major issue with all open source projects.


One other major issue with open source projects is; many are happy to reap the benefits of the finished project, but few are willing, or able, to contribute to the development.
I am for instance not able to participate much in the "engineering out" process, as I am simply not able to consume the amount of engineering information that goes into first understanding the details of every component in ie. the power supply, and then researching possible ways to simplify or improve on those details.
So I find other ways to contribute. At one time I printed a handful of RepRap's, sold the parts on ebay, or to people that contacted our hackerspace, and donated 50% of the earnings to the MetalicaRap project. FYI, our paypal account uses this email identity: MetalicaRap@3iii.dk
Another person thought it would be a good idea for us to have access to professional CAD software, as the project might benefit from some serious CAD and simulation ability. So now we are very happy to have been granted a very favorable deal with Autodesk.



GreenAtol Wrote:
> I think maybe a mini version would be a good
> intermediate step.

Once the power supply is up and running, this may indeed be what will happen. I will however leave that decision to Rapatan.



About your other issues with the metal parts.
While we get the prototype gun ready, other projects continue to accumulate information and work on all those same issues. So far, some details from some of those projects have been made available to everyone.
At other times we need to buy a book or report, to get access to the details. This gives those deeply involved with the project a chance to reproduce, reverse-engineer or even improve on the setup. Until we are able to reproduce our own drawings and/or test data, we will however have to respect whatever lincesing rights any particular work carries.


And generally about project/production questions.
Questions are important. Its one of the main resasons some of us even bother to get up in the morning. To find things to explore!
Constructive suggestions for possible solutions are however much preferable in the forum format. They simply have a much more productive core, and with so few people currently seriously involved in the project, we really need solution oriented ideas, links, debats, support and prototyping.


So keep up the good work.

Talk to you all later,
MrAlvin

I'm glad to hear that work is still being done on this. I am an aerospace engineer (recently graduated) going into space vehicle development. A machine like this would be incredibly useful both for prototyping and production, as well as for use in space itself. Once I have a job lined up (in progress...) I may see if there is anything I can offer this project.

---

Help improve the RepRap wiki! Just click "Edit" in the top-right corner of the page and start typing. Anyone can edit the wiki!

How about useing a old CRT tube (tv) cut out the front glass, remove the phospher screen, wire your metal powderbed tol the crt as the anode, place this all in an argon atmosphere high vacume, and just send pictures to the crt to display. you could put a (possibly audio) amplifier in between the signal and electron gun to increase the amps to melt temps.

Interesting? Yet

• all motors/drive shafts/ sensors for bed would have to be isolated to +70KV ,
• cathode filament waiste thernal heat would have no where to go as completely sealled in TV tube
• baffels- an apeture in tungsten plate, waiste thernal heat would have no where to go as completely sealled in TV tube
• vacuum pipe conection to glass could be complicated/ difficult?

thanks
MetalicaRap team

how about something similar to this: [www.optomec.com]

How powerful of a laser will you need (would you need lower wattage laser(s) with smaller powder particle size?) A fiber optic laser seems like a good option.

Hi
Nice machine if you got the power and money!

Your largely self reproducing printer is possible in the commercial setting with either laser or electron beam printing parts due to 3 Phase supply and 100K euro 400W laser part purchase not being a problem(as the commercial machines indicate).

But its a different story in the single phase setting 10K euro home budget world; due to the power transfer inefficiency from wall socket to most common metals via lasers being 50 times worse than in electron beams, available home lasers would print too slowly (power wise an exception to this is laser to steel:80%)( typically a CO2 laser in copper 1.6% energy transfer efficiency, So a 400W energy to metal therefore requires a 25000 W laser, current home build lasers are considered large at 30W), therefore lasers would limit achievable part size to 10's of cm3( e.g. 3cm x3cmx3cm) with a two day print duration, which makes it impracticably slow for self replicating printers. (For further proof try our Laser 808 Build Speed Calculator get from bottom of this linked page here Tool head process discussion SLS vs EBM vs EDM ) .

great printer video though!
kind regards
MetalicaRap team

... if you switch to black plastic powder instead of metal, then you can start with an IR-diode with some Watts of power - fibercoupled IR-diodes with 9Watts of power are around 300 Euros now.

With a CO2-laser you can use all sorts of plastic and colours - here the cheapest cinese laser-tubes starts at 20 Watts with similar prices.

Haven't tried it with wax-powder, but could be even simpler and with lower energies than plastic ... would be perfect for lost-cast moulding ...

---

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 will give you those links again.

Nasa Article

Technical Talk video

Aerospace casting quote

kind regards all
MetalicaRap team

Corrected version
hi

The goal is to have affordable home-manufacturing of solar cells, key electrical parts and milled-quality metal parts.

Unfortuanately your suggestions would mean we would fail to achive a solar cell process with a laser and key electrical parts need to be in fine metal construction which casting is unsuited too.

Given the complex alloy metalurgy behind structural aplications it is not easy to asses a new process in comparison. Any way I suggest this technical lecture from Nasa. Its quite fun.

Technical Lecture Nasa

In the lecture she is compairing this electron beam vacuum metal process with excisting traditional forged and wrought metal processes, and comparing with specific aerospace high strenght structural alloys in terms of strenght, grain structure and residual stress.

Or as a short Nasa article says " Electron beam freeform fabrication (EBF³) is a cross-cutting technology for producing structural metal parts. Developed by researchers at NASA's Langley Research Center as a replacement for forgings.Nasa Article

Great question about why not "lost cast moulding" or other casting processes e.g." investment casting (also called the "lost wax process") for structural metal parts?

Cast parts have some limited uses, as one author put it "if strength and quality of material are the prime factors in a given part, a cast will be unsatisfactory" Aero. Eng. Article , for example the structural elements in a wing of plane are based around forged or wrought metal prepared parts traditionally, with the correct grain structure, alloy mix, temperature history and verifyable strenght.

thanks for the thought provoking suggestion.

kind regards
MetalicaRap team.

P.S.Unfortuanately wax powdered mixed processes that create porous metal parts are not considered structural parts but good for great sculptures!




Understanding structure of each metal involves knowing; its crystal deformation behaviour, its phase diagram for each alloy recipie and its response to metal working processes.
yet it can still be difficult!