Erik's Blog: Discussing IT, personal fabrication and a new and better world

Are you operating or building a 3D printer? (RepRap, Cupcake CNC, RapMan, etc.)

We still need some more responses for the data to be statistically significant. So if you have not done the survey, please do so! (If you have, thanks so much!!)

Why? I can understand that you don't have time for just any survey. But this one is REALLY IMPORTANT:

We're trying to learn how open source software development differs from open hardware. Also, I'm learning how the RepRap community innovates and how this is different from commercial development of physical products. If helping me graduate isn't enough reason: you will help the RepRap project (results are used by me and Adrian Bowyer) and the field of open source research (as I'm directly working with the most cited scholar on this subject, Eric von Hippel from MIT).

Your input is essential! Many others have already filled in the RepRapSurvey.org but we need several more respondents for statistical significance of the findings.

It takes people 15 minutes on average.

Thank you VERY much!

Erik

P.s. Everyone will be given access to what we learn from this study. If you have questions or remarks, don't hesitate to contact me.

From Erik's RepRap parts

For some, the fact that the mini-mendel isn't a finished product is a reason not to try it. For me, it's the opposite. I've now printed all parts for the mini-mendel. I've also found suppliers for (I think) all of the non-printed parts that are needed. I'm waiting on delivery from RS Components (nut's & bolts), Pololu (stepper motor drivers), Smalltec (smooth rods), Adtech (PTFE Bowden cable) and Active-robots (NEMA 14 stepper motors). If a miracle happens I will have all items within two weeks. In the meantime, as parts are coming in, I will see when I enough parts to finish one sub-assembly at a time. I'll keep you updated with the trouble I run into

This is the table with lists of parts. Interestingly, it seems possible to stay within 350 euro target in terms of expenses, for future Mini-mendel's, that is. I will probably be over this target because I won't get everything right (I already ordered some of the wrong rods, there goes 40 euro's):


The Bill of Materials:


This can also be found on the RepRap wiki.

Some answers to questions I got:
Q: Does it use the same timing belt as the regular Mendel?
A: Yes. Only shorter lengths. Hopefully, my BOM shows the right lenghts that you need to order.

Q: How much time did it take to print it?
A: About 15 hours in total, I estimate. I didn't keep track of it. I just turn on a batch and work on other things. I printed it in about 8 sheets because my build platform is warped. I can only use a 20x10 space of it. So it's still pretty labour intensive to print a batch.

Q: How much material is used to print it?
A: All the parts I printed (everything minus 3 pulleys minus a couple of grams) weigh just under 300 g. I'm using 30% infill.

There are so many ways to get to a 3D model. Some are really great if you like programming. It is really powerful to use logic to generate objects, but not for all models. Those who are better with visual modalities will probably not like such a system much, and will like the following. Besides scanning software, modeling software is also rapidly improving in terms of usability.

There are already great pieces of software that can be used for 3D modeling of meshes.


It already shows how you can make 3D modeling accessible and still be very powerful. I would like to think about how we can achieve both with relatively simple means and do it better than in Minority Report.

Blender is another excellent product that's open source. Since version 2.5 the sculpting system has seen a major update. That it's powerful can be understood better by seeing in than by reading about it:


The problem is that the user interface is usually still in 2D and is not intuitive for beginners. A 3D interface would not only drastically reduce the learning curve for working in 3D, it would also be a lot more impressive and work much faster:


It should be possible to implement this with off the shelf parts and open source software (something interactively adjusting blender's perspective).

The wii remotes could provide useful for head and cursor (or finger) tracking. To get a "3D screen" you could use shutterglasses to generate a stereoscopic interface, or you could even more cheaply create a lower fidelity interface by just adjusting the perspective relative to head movements.


Download and print two LED clippy's on your RepRap, or if you don't have a 3D printer at home, make them yourself with some wire and Scotch Tape:


Or buy off-the-shelf LED glasses. Here you should probably still replace the LED's with IR LED's if you want proper contrast with the wii-mote setup.
You could create a custom HID device or USB joystick with an Arduino. Or just put an off-the-shelf 5 euro (including shipping) IR emitter around your wii-mote or webcam.

Years ago I blogged about creative wii interfaces. You don't need expensive equipment to make interesting interfaces 3D like Leonar3Do:


Now the great thing about what you see is that it's not THAT complicated!.

For those of us who have RepRaps or don't hesitate to tinker with hardware, we can make the system as elaborate as you wish, building data-gloves and such. A quick google gives you things like the ZerkinGlove, built in a day and using mostly parts from Sparkfun:


Many of the parts for such an interface could be 3D printed (to exactly fit your body, you should sculpt the second iteration with the system from the first iteration).

But I like the simplicity of a simple '3D mouse' with buttons that has three markers, since most of the solution would be in software and anyone can easily get it working with standard webcams/wii-motes and software they can download. It's easier to replicate such a setup. Another option would be to use a phone (android based) and connect it to you computer directly via bluetooth, or to your arduino prototype rig (via the amarino toolkit).

I finally finished splitting and uploading the video's, so here it is in a Youtube playlist as well:

Source: video.fosdem.org


Some tweets from around the time of the talk:

mvdkleijn At the #reprap .org talk at #fosdem - terrific

ppapadeas Enjoying an excellent so far talk for RepRap at #fosdem

ppapadeas "Make manufacturing more like agriculture!" Start your revolution with RepRap #fosdem

bobylito #reprap : sounds like a wonderful anarchist tool to me

NarcissusBleak I approve of the philosophy involved in the RepRap talk #FOSDEM

BatistLeman Reprap talk at #fosdem! Nice: http://twitgoo.com/db1af

ehn RepRap talk at FOSDEM. Who doesn't want a 3D printer? — at ULB Building H http://gowal.la/s/3D9b

bacharakis http://twitpic.com/11v2xl - At reprap talk! this is a 3d printer!

bartcopp Time for the way too early RepRap talk at #fosdem

Kim_Bruning I'm at the Reprap talk at #fosdem, 1st and 2nd gen machines present in Janson

josephzizys The Reprap is Marxist fantasy made flesh! Talk about controlling the means of production! #Marxism #Maker #Reprap #PhilosophyMadeFlesh

tyll:I like the #reprap talk on !fosdem !fosdem2010 currently running. Nice vision: Use old child shoes to create bigger ones if needed #fosdem from identi.ca.

It was also featured in the Dutch newspapers NRC and NRC.next, Belgian newspapers and blog posts here and there. Also, me and some other Protospace RepRappers were on television last weekend! See it on Youtube here or on the TV station's website here.

From RepRap @ FOSDEM

All-in-all, great publicity for the RepRap project!

RepRap got coverage in a well respected Dutch newspaper, NRC.next. They also made a blog post with an interview, the original post is in Dutch, but here's a (poor) google Translation to English.

The interview is in Dutch, sorry. But you can still promote it and give it a good rating since it's about RepRap


Adrian Bowyer, as well as several active Belgian and Dutch RepRappers can be seen in the background. When the FOSDEM talk by Adrian is available, I will post it here!

You think vanity and Yoda do not go together? The force may be strong in this fella, but he used a very powerful scrubbing lotion to make him look much younger.

Stratasys sent me these sample parts to show of the impressive surface finish you get from a process they call ReadyPart^TM. Over the years I'm involved in RepRap I saw a lot of questions and remarks about polishing parts, in some cases by using acetone. I've tried to see if I could make objects more smooth by putting them in a pot of (diluted) acetone for a short while, but I never got a good result. I had this Yoda sample, but didn't blog about it until I got another question about it from someone at Nophead's post (the one where Chris shows off his smooth surface finishes!)




For more pictures, view my Picasa RepRap development gallery.

ReadyPart^TM is only available for ABS parts, not PC or their PC/ABS blend. This probably tells us that it's a specific solvent. It's probably not acetone since that's very unsafe and environmentally bad. After digging into it they may be using a vapor degreasing method (since that's listed as a consumable of their commercial finishing stations).

This is a good explanation of the process that I found through the wikipedia entry:

"Vapor degreasing is a relatively simple process. A heat source raises the liquid solvent to its boiling point. When the solvent boils, it produces hot, heavy vapors that rise to an established vapor line. At this point, the vapors are condensed on cold circumferential condenser coils, and the vapors rise no higher. Because the solvent vapors are heavier than air, they push the air above the vapor line. Parts at ambient temperature are then introduced into the solvent vapor, and the solvent vapor condenses on the part’s surface. The liquid solvent produced as a result of this condensation dissolves the greases and oils on the part and flushes them away. As the parts are cleaned, more vapors are produced in the boiling sump to replace those that were condensed."

Source: pfonline.com

Vapour degreasing is patented but the patent expired. There are more techniques for which the patent expired. There are probably other patents to consider, so you can probably use it experimentally (thanks to the research clause in patent law) but not for profit. Another option is to buy a Stratasys smoothing station to smooth your RepRapped parts

Even things that are peripheral to the RepRap and makerbot are going through some remarkable evolution! It not that strange, because they can really make life a lot easier. Some accessories, like spindle holders, are worth some trouble. So I decided to scratch the itch (lesson 1 from the Cathedral and the Bazaar). My itch was a jammed supply of filament that ruined a print job. Also, I had the impression that under the wrong circumstances the extruder would be able to pull the entire RepRap from my desk. The plastics from bits from bytes already come on really nice spindles. While more expensive, it's a lot more convenient than in this form:


I had already designed a holder to roll this filament onto. It is a tedious job if you don't want to make a total mess, because the coils will get entangled into each other:


But once you have it rolled onto the spindle you're set for many hours of printing.


After a while, it got picked up by Cidvilas else who made it much better!


Cidvilas mentioned some problems that still needed to be addressed:

• Spindle spins so freely that the filament itself would probably unwind itself slowly. In the future i would like to add a screw that can be tightened which adds friction to the sides. Keeping the spindle from spinning 'too' freely.
• The Spindle is slightly underweight and with enough lateral force could topple. To address this issue simply attach something heavy to the base. Or add an additional acrylic piece perpendicular to the current stand to increase the base size.

At least one of these issues was fixed marvellously in a third iteration. To some it might seem like it's perfect already, but there is often someone who thinks he can even improve on that design and make it better! That's the beauty of open design, it often takes multiple people to make several consecutive improvements, and open design enables and encourages people to permissionlessly improve designs.
.

Now this spindle holder requires a laser cutter to make, it's more convenient if we could print it out! That's what tbfleming came up with:


Very elegant, requires little printing. I wanted to apply this principle to my Bits From Bytes spindle, but still couldn't resist making it more complicated. Still, I like how it came out:



I wondered whether it will be remixed and improved! If you want to keep your eyes on the digital design market for spindle holders, keep your eyes on the 'filament' Thingiverse tag!

Update (26/01/10), another improved version has been published on Thingiverse by tbflemming:

Nick Jones from Gartner made an interesting remark:
"[A]s one of my clients pointed out yesterday, do we really want an affordable domestic fabber? Fabbers will likely “print” objects using some form of plastic. So the inevitable consequence of mass market fabbing will be a huge increase in the amount of non-biodegradable plastic waste clogging up the planet for hundreds of years into the future. Should we maybe ban fabbers before the problem arises? Like most problems there are solutions, like biodegradable plastic. But if we wait until all the problems with a technology are solved before we permit it, then we will waste a decade or two of potential value; and in any case there’s no way we can predict all the social and environmental issues associated with a new technology before it arrives."

I see this as another opportunity to highlight some benefits.

I'm glad that the RepRap already uses bio-degradable PLA as a main material and that we have a proof of concept prototype recycler. I don’t see such a dark future. Plus, we can produce sparse objects more easily than with injection moulding and still have a decent exterior, requiring no assembly or expensive multi-part moulds. I think 3D printing is part of our much more innovative future. It’s not a separate peripheral product that we will have in our homes, it’s a transformative enabling technology, much like the internet, that will catalyze physical product innovation.

Nick is right that we should consider the relevance of thinking about this in time. There are ecological threats, but there are even more opportunities. If you consider banning a flexible production technology, you’re left with artisan production (inefficient, custom made) and mass-production (efficient, limited product ranges because of high up front costs). There would be an important missing element. He doesn't seem to see the repercussions of a ban...

As a RepRap 3D printer owner I (can) speak from personal experience: This week, it printed out another RepRap (all its parts, I need a couple of weekends to put it together). It’s quite efficient compared to having lots of parts shipped from overseas and then realizing you’ve spent 80 dollars on pulleys that do not fit. After doing exactly this, I designed the part and printed it myself. The printed part did fit and I could instantly see if it wasn’t the wrong size before printing out the required batch size. Look in your drawers and closets, they are filled with things that are not entirely or exactly what you needed. What if you could change, upgrade the products that you used? If we had the source code to the product and could generate a version for testing its usefulness on the spot, that would change things!

From Erik's RepRap parts

What about fabbers upgrading themselves?

From Erik's RepRap Development

This saves you from having a new 3D printer every time. Just like we prefer to buy a new inkjet printer once the cartridge runs empty, because it’s hardly any more expensive than just the cartridge. What about people (re)using (junk) electronics combined with printed parts to make new things? We collect so much stuff "In case we need it". Would we still be doing this when we could have it as soon as we really encounter a need?

Think of the size of the replacement market. Not only is the size of this market huge, it is also severely limited by factors that do not apply to 3D printed parts:

• misaligned interests of manufacturers (buying a new product generates more revenue that supporting older ones)
• uncertaintly about demand for the items, because it's unclear in advance what will break


• the stock keeping costs of these huge numbers of parts
• limited revenue possible, yet high transaction costs (shipping and handling)

A category for replacement parts already exists on thingiverse. If a belt buckle breaks, you can now print a new one:

From Erik's 3D printing! - Design by jrombousky & rmadams

These are printable designs. The manufacturing data can be downloaded and executed by a 3D printer or other flexible production technology.

How much would we save on shipping and the related environmental burden:





It allows us to fix things and counter product obsolescence. Also, I regularly print out things that I would otherwise need to go to the store for. This saves gas and CO².

With mass production you end up either under or over producing. To prevent underproduction, you end up making the trade off between production and opportunity costs of not producing. All this overproduction is waste and is inherent in make to stock manufacturing of physical goods. 3D printers are make to order production technology. The result is that products for which there is an actual market will see the light of day, other simply won’t, they’ll stay digital. While this is important for innovation, it clearly doesn’t have the waste problem.

What about open, modular parts where you have ’situated physical products’ that help you with a certain task for a week, and then can be reconfigured to help you with another task. I believe open source aids in this standards formation. Right now all our products have their own engines, batteries/power supplies and processors while we use only a few at the same time. This ’situated hardware solution’ requires a combination of generic and solution specific parts that are designed to be reconfigured. That’s where fabbers (such as 3D printers) come into play. Right now products are so integrated that if one thing breaks, we end up replacing it all.

One more thing: If the question is “Should we ban 3D printing on environmental grounds?” you implicitly assume that you could ban it. While theoretically it would be hard, how would you enforce it? It is hardly possible nor is it beneficial to try.

For a more detailed comparison of the environmental costs of local fabrication (ship homogenous raw materials once) to centralized fabrication, see my post on the open manufacturing list.

We demonstrated the RepRap at NPOX (15 nov. 2009).

It was in the "Beeld en Geluid" building which was a huge building, but beautiful at every scale.

We were positively surprised by the way our project was advertised on pamphlets that were passed around and people seemed to know about it in advance. The visitors were eager to learn what was going on and posed questions ranging from how the technology worked, to how this could impact innovation and how this potentially changes the world. Some also asked how long it would take for this technology to reach the home. I think the RepRap needs some 'domestication' but it shouldn't need to take long. More important drivers and inhibitors are social and dependent on other flexible production technologies and services that might be an alternative.

The RepRap Day at Protospace (14 nov. 2009) was a great success.


(see also, more pictures by Bas)

The plan was to print an entire machine at the event, but we did more talking than actual printing, which was fine. Some whistles were printed by Joost. I printed two Bowden cable clips for a second Bowden extruder. The plan is to use this setup for multi-material printing!

All in all, it was a REALLY nice day with lots of great people who were either very interested in the concept of RepRap, FabLab and/or building a RepRap themselves. We had about 10 machines there (not even all of them were on the table!), including Joost's Makerbot, several RepRaps, an acrylic RapMan and an plyRap (from plywood).

Now I still need to blog the NPOX festival in Hilversum, and Manchester and Nottingham trips... doh!