The Metalbot Project: an open source initiative...

Hi everyone!

I would just like to introduce the Metalbot Project, the first open initiative of it's kind, to design and create a 3D Printer capable of making metal parts (in titanium, tool steel etc)...

http://www.metalbot.org/

I am sure many people here have heard of 3D Printers/printing, perhaps some of you have even had the opportunity to used them.

The two limiting factors with 3D printing have always been a - strength of the printed parts and b - part resolution. We are hoping to solve both of these problems by designing printer that works through a process called 'laser sintering' (although a few other avenues are being pursued).

Ideally we want to make this available to a wide range of people (at the moment there are only a handful of 'commercial' systems and these are way too expensive for the average person).

As an open source project the information is given free and taken free.

We are at the very early stages and need people who are interested in this subject to come forward and add their thoughts on the subject. Already we have had very clever suggestions that have improved the design greatly but there is a mountain of work still. Any knowledge could save a lot of time!

So... we are looking for experts in machining (some precision machining), physics (Laser physics and powder physics), programming and software, where most of you will be comfortable - electrical engineering (specifically an arduino type control system for the powder system timing etc, also if anyone is familiar with laser scanners and scanning software that would be really helpful), mechanical engineering etc...

Here is a link to the software and electronics sub-forum, there is also a laser forum, be sure to check it out!

http://www.metalbot.org/forum/viewforum.php?f=11

I would be more than happy to answer any questions!

Best regards,

Jethro.

I had no idea that 3D metal printing is possible, so I find this amazing.... and it makes me feel that we are very close to having our own replicator just like in Star Treck.

CVMichael, It is indeed an incredible technology. So simple too, just layer upon layer of powder melted on top of each other.

What I find particularly interesting is that commercial companies (currently the only users of metal 3D printers) have reported that printed parts are marginally stronger than their forged equivalents. There are no seams where one layer has been fused to the other, but one contentiously crystalline structure.

Unfortunately these commercial systems (Arcam and EOS) have a base price of $680k and charge $680 per KG of titanium printed.

Clearly our work is nicely cut out...

You need a better website designer!

Your main website comes up totally blank on my web browser, and I tried both javascript on and off. No luck.

To me it seems a bit dumb making a DIY opensource type web page that only works on some (new?) browsers.

Your forum link seesm to work.

Works on Firefox 20.0.1

Ken

THE_RB - Unfortunately we can not afford a website designer at the moment. It is a learning curve for me. May I ask what Browser you are using? I am glad that the forum at least works... the website HTML5. Very frustrating to hear, sorry about that. There is loads of information on the wiki if you are interested: http://www.metalbot.org/metalbot-wiki/index.php?title=Main_Page
The community is very small at the moment and has only been live for 3 days but there is a lot of information on the site.

This video explains the principal of laser sintering:

http://www.youtube.com/watch?feature=player_embedded&v=PHaXX2OoOs4


There are two main categories when dealing with the Metalbot electronics, which is perhaps where most of the knowledge here would be helpful.

1 - The Laser Electronics
2 - The powder Timing Electronics

The laser electronics includes the PSU and also laser scanner electronics.

I am not sure if anyone here is familiar with Laser Scanners. This is the most efficient method of melting a large cross section of powder. It basically is an array of two mirrors connected to two motor shafts. Depending on the angles of the mirrors, you can deflect the laser beam to follow x-y coordinates.

Laser scanners are typically used in engravers and laser show equipment. We are considering hacking a commercial laser engraver because all of equipment is in place. The issue comes with interfacing the scanner software with 3D Printing software. We can also hack show lasers but would need to replace the low power laser with a high wattage one.

As for the powder timing electronics, we need a method of spreading out a thin layer (>100 - 30 microns) of metal powder. This entire action needs to be timed and driven by motors. Perhaps arduino would be a suitable driver?

Best Regards,

Jethro.

In my opinion, using an arduino may be ok for hobbyists. If you want some serious computing power get something like an ARM processor.

Cool stuff:
http://makerblog.anat.org.au/what-is-3d-printing/

I have a friend at a company which uses a Zprinter to produce prototype parts for auto mfrs. Amazing to watch a 3D part appear in front of your eyes. I saw this a few years ago, was a very slow (hours) process.

Has the speed been improved ??

Pulsed said:

THE_RB - Unfortunately we can not afford a website designer at the moment. It is a learning curve for me. May I ask what Browser you are using? I am glad that the forum at least works...
...

Click to expand...

I tried using Opera, it's an older version but works fine for web pages that don't have insistence on flash or java.

If you are doing an open source page for everyone, it's a good idea to use simpler web features that won't stop some people being able to see the site. You really don't need guff like flash, trackbacks, java, psp etc. A web page just needs words and pictures and links.

And sorry for sounding negative re the project, I'm not negative to your project itself. It's just annoying when someone says "Everybody should come down the road and visit this really great bar!" and when I go down the road the bar doorman says "only people wearing the very newest clothes will be allowed in".

@ MrChips, thanks for your suggestion. For powering the timing of a few valves etc arduino is sufficient. It has been done before to drive a wax 3D Printer, and the part driven by arduino is very similar to what we want to do with metal powder... here is the link if you are interested: http://reprap.org/wiki/SLS_wax_printer . Besides, we are hobbyists

@ tubeguy, the trouble is that all of these machines are just that, prototyping machines. That's due to the weakness of the printed 'plastic' parts. Using a metal laser sintering system (3D Metal Printer) the end parts have the same strength specs as their forged equivalents (some case studies have show that there is a marginal increase in the printed parts strength).

As for the speed of production, when considering using the tech for mass production, having them run in parallel is the best option.

Still, when considering that you can print a part with as much complexity as you like (buried coolant tubes, internal cavities etc) The amount of time it would take to machine a part with the same complexity will take much longer if it is even possible.

Basically, complexity does not cost anything (time, money or skill). It will take the same amount of time to make a metal block as it would to make a similarly sized intricate door handle for an airbus 8380.

As for increasing the actual speed of the printing, this is a tough challenge. We are looking to print in high resolution (30-100um cross sections) so you have to calculate the amount of time it will take to spread the powder and then sinter the cross section... multiply that by the number of cross sections your part is comprised of. So in order to increase printing speed you need to spread and melt quicker. That means bigger lasers and more money. As money and cost is a high concern to us as hobbyists we are focusing on thinner layers and therefore smaller lasers (our biggest cost, have you seen how much they cost??).

I hope that answers your question

@ THE_RB, I understand your annoyance! It is frustrating as a website owner to hear that some people can not view the website (I have heard this from a few others also). I will try and think what I can do to fix the situation. One option that has been looked at is completely scrapping the info site in favor of a WordPress site. In any case, the wiki should work well for you (most of the pages need to be created still but most of the crucial information is on the main page.

http://www.metalbot.org/metalbot-wiki/index.php?title=Main_Page

Best regards!

Jethro.

Arduino works fine at the production rate of a 3D printer. RepRap is an opensourse project to make a fused deposition (plastic) 3D prints. The whole thing costs under $500 including upgrades to hard chrome rods and linear bearings. The laser power and heat transfer (part metal vaporization/cooling/sag) will likely be rate limiting for laser sintering additive manufacturing.

There is a National Center for Additive Manufacturing in Youngstown, OH. Has the OP applied for any research funding to support your OpenSourced / GroupSourced project?

Pulsed said:

... One option that has been looked at is completely scrapping the info site in favor of a WordPress site. In any case, the wiki should work well for you ...
http://metalbot.org/wiki/index.php?title=Main_Page
...

Click to expand...

The Wiki page is simple and neat and works great.

If you have not already made it a design goal, it would be good to make the laser output power closed-loop controlled.

Laser power output changes over time due to internal heating etc, it's a problem on cheap laser cutters where cut power changes over the first few mm of each cut.

I'm not sure how the expensive machines monitor laser output power and keep it constant, or how hard it is to do, but I imagine there is some type of light sensor in the control loop.

@ GopherT - True, this is a much more ambitious project.

There is a National Center for Additive Manufacturing in Youngstown, OH. Has the OP applied for any research funding to support your OpenSourced / GroupSourced project?

Click to expand...

I can see issues with that, for one thing I am located in France.

For the moment we are trying to focus on getting data and info, only then can we work towards a working prototype. Funding is not top of the priority list at the moment. Experimentation and tests are what is needed, and this is going ahead.

@ THE_RB - Thanks for the feedback! I love the wiki software, very clean and simple. Unfortunately many of the pages are still under construction, but there is content going up every day.

If you have not already made it a design goal, it would be good to make the laser output power closed-loop controlled.

Click to expand...

Interesting suggestion, we certainly want to maintain consistency throughout the printing process.

One of the design guidelines (if you can even call it that) we are trying to follow is the 'it just works' model. So we want to make good use of sensors, closed loops etc.

I think I mentioned before that we are looking at hacking a commercial YAG engraver which has all the machinery in place already. I am not sure if these machines incorporate a feedback loop.

Thanks for the good suggestion!

Jethro.

FWIW, the website looks fine using Safari on a Mac.

Interesting project. My daughter got to play with 3D printers in her undergraduate engineering work. This would be orders of magnitude more useful. What I wouldn't give at times to make the "right" part.

Just wondering, do you have a target for $/kg? I mean, are there some practical limits that are not likely to be overcome? Price of titanium powder, for instance.

@Pulsed, You may want to make a post like this on another forum that I'm a member at, CNCZone.com. They have a fairly large and active 3D printing and laser communities there.

I happen to live in Youngstown, OH, home to the National Center for Additive Manufacturing.

@ wayneh - Thanks for the feedback. The wiki and forum should have no problem working on any browser.

I wonder which printers they were. No doubt plastic. Most 3D printers by far are used in prototyping, not real world applications simply because of strength (sometimes resolution). So an accessible (I say accessible because metal printers are available but cost way too much - $680 per kg of titanium to be exact) 3d metal printer will change the landscape greatly.

It is an exceptional challenge however, and the laser lies at the heart of this challenge.

As for fixing a price per print, aerospace titanium as fine as we want it (5-3um) is rather expensive (about $100/kg). But 1kg of titanium printed into a complex shape will cost far more to machine.

EOS - The current world leader in this technology quote $680/kg of titanium. I am sure we can improve that!

Right now our biggest hurdle is the cost of powerful lasers.

@ shortbus

Thanks for the suggestion! I have... I will find the link - unfortunately I posted it under the wrong subforum, perhaps I will ask for the topic to be moved.

There is a lot of metal work that needs to be done, and a load of questions to be answered on that front.

Jethro

Heating titanium will cause it to pick up atmospheric oxygen AND nitrogen, thereby reducing elongation of the finished part. Brittle titanium does not meet physical testing requirements of most titanium grades. The workspace will have to be evacuated and back filled with argon (the number of vacuum/back-fill cycles will depend on how well the system is evacuated before each backfill of argon).

You may be much better off trying to focus on stainless steel steel, high nickel alloys (super alloys) and tool steels. Machining costs of super alloys are high and will make your process competitive. Super alloys don't have the headaches of titanium.

Hi GopherT!

The Metalbot 3D Printer frame will hold a gas tight atmosphere of either nitrogen or argon (widely available due to their use in welding).

You may be much better off trying to focus on stainless steel steel, high nickel alloys (super alloys) and tool steels. Machining costs of super alloys are high and will make your process competitive. Super alloys don't have the headaches of titanium.

Click to expand...

We are looking into a whole range of materials... even certain plastics will be compatible with the machine. The basic requirement is that the materials are in a fine powdered form (<5um).

One of our current goals is to document the different sintering thresholds of a number of materials under certain conditions (with a tonne of variables).

Just a quick update for the interested... there are some interesting ideas going around!

We have decided, for now at least, to put a "laser scanner" on the back burner and go with a Gantry. So some form of CNC software in combination with a slicer will be used.

Check out some preliminary designs for the Metalbot printer here... http://www.metalbot.org/forum/viewtopic.php?f=4&t=111

If you have google sketchup you can download the model here... http://sketchup.google.com/3dwarehouse/details?mid=154316e5f0930d093ff7a6479db06f74&prevstart=0

Best!

Jethro.

looks very interesting!