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

BoingBoing blogged about reproducing keys from distant photographs. This made me think back off what we were doing at Hacking At Random '09 with our RepRaps. Ray had created an STL model (3D design file) of a police handcuff key. He had also made it based on a photo. He printed it back in Germany, where it worked in a set of handcuffs that he owned. On my RepRap we printed another couple of keys during the event. It's a tiny fiddly key, so its strength varied from print to print and needed some more tweaking, but it actually worked. During HAR they asked policemen to give it a try. Most refused, but then one of them tried it and it worked. And Ray told me that it was one of the locks that you won't open with a paper clip (too bad, because I also printed those).

Courtesy: Blackbag.nl

Still, I will try not to give the police a reason to get caught. I don't think that I am heroic/stupid enough to use it to escape from the police, though.

From 3D printing!

Similarly, some people had printed out some HAR 10 euro coins.

inflation through technology from Leiwandville on Vimeo.

It wasn't used for real, but it did work. If someone would use it, the people at HAR would find out afterwards. They would know it came from someone with a 3D printer. Each home-built 3D printer has its own 'signature', so they could end up with me. Then again, the real reason I won't print coins for profit is that I don't want to ruin an event that I care about.

The fact that you CAN do bad things with a technology is nothing new. It's the same with the internet in general, where people can exchange ideas about bad things just as easily as good things. Now with 3D printing this capability is extended to the physical realm, you can also exchange 'bad' or 'good' things yourself. As far as the use that is beneficial to society far outstrips the detrimental use, we can be glad about this. I'm confident that this is the case, and that good ideas will propagate faster, simply because more people have the incentive to spread and use them. Having the knowledge or capabilities to do bad things should never be illegal per se. Doing bad things (like escaping from custody or creating counterfeit money) should be, however.

So with great power comes great responsibility...

I've designed a little part that allows you to connect the extrusion drive mechanism, which pushes the plastics, to the extruder which it is pushed through.



Here you can see a printed version, attached to the extrusion head:


More pictures: here.

It is a holder for a 4.67 mm PTFE cable that serves as a guide for the filament. The plastic feedstock will travel from a driving mechanism to the extrusion head, through this flexible tube. The PTFE, a.k.a. Teflon, a very low friction. This solution will allow you to have multiple extruders on your machine without the moving mass becoming too heavy. The lighter it is, the quicker you can move the heads, e.g. when you're not extruding. The machine will need less current to run the motors, it reduces stringing when not extruding. Even more so, it reduces the weight of the X-axis so much that a much simpler and lighter Cartesian bot is possible. The steel rods wouldn't necessarily be needed, allowing a much larger ratio of self-replication. The Bowden cable itself could be printed in several modules that snap together. This guide could be lined with telfon-tape (available in every hardware store) that can be replaced when worn.

It has one major drawback. The need for a control mechanism that accounts for pressure on the filament is needed. The tube is a bit flexible, and so is the plastic filament. This makes the assembly a bit springy. But when you use a flow-control system that accounts for pressure, you will end up with much better quality. The extrusion starts and stops very sharply. You can do this by pulling back after extruding a segment. After that, you push it in as much as you pulled it back and simply start extruding again. You do need stepper based extrusion to adequately be able to control how much the filament is pushed into the extruder and pulled back out.

I've been using the cable for a while now, but I need to write a batter algorithm that manages the tension that is put on the filament. Actually I did write one, but it's not perfect yet since I ran into another bump. Right now, this algorithm is causing a bug in my RepRap's firmware when I try to rotate the stepper motor in the opposite direction (using negative extrusion values).

Here you can see it being printed:


You can find the STL and source files of the object on Thingiverse, the online digital repository for physical objects.