Intro/background

This project started out with a desire to print, assemble and get working an 'Ada Hand' By Open Bionics in Bristol, an open source designer and manufacturer of prostheses. I don't think I would have embarked on this project unless I had watched this talk and visited them online so have a look here for some inspiration:

https://www.openbionics.com/

These guys have files on their website for a full robotic hand that can be downloaded and printed; this was the initial idea. I like the design because it removes a lot of fiddly mechanical parts/fixings and uses the material properties to achieve the movement.

Initial issues and project redefinition

Having read through the data sheets etc I came across a number of potential issues.

Firstly, the recommended material is called 'Ninjaflex', very cool flexible filament that you should be able to run on most quality printers. From what I gather it has a plastic like quality but is incredibly flexible and hardwearing. I couldn't get hold of this through university channels (which I had to go through if I wanted funding for the project). I could get hold of some flexible rubber filament called 45flex but although they were both flexible, that is where the similarity ends, on paper at least. The Ninjaflex is marketed as being low friction, the 45flex is marketed as a printable rubber (implying higher friction). This could be a problem with actuating cables/ribbons snagging.

Secondly, the 45flex is softer and more elastic so potentially the structural properties of the Ada Palm could be compromised.

With this in mind, and that the palm would need around 26 hours to print, I was not keen to just press on and hope for the best. My experience is that the larger the part, the higher chance of a misprint.

My thoughts at this stage were to convert the stl file for the palm to an stp file, load into solidworks, chop a finger off and do a test print. This didn't happen because of the anatomically correct nature of the palm; it was so geometrically complex that every single computer I tried it on just crashed.

The next step was to install Blender 3D (Open source modelling software used by Open Bionics) and try and amend the model in that. A quick go with blender and I realised this software was totally different to what I was used to, and maybe I would be better improving my Solidworks skills than trying to learn a whole new program.

At this point I abandoned the Ada Hand and decided to design my own hand, it will be heavily inspired by it though. One thing I would like to do is to break it down into smaller parts to mitigate the risk of misprinting a big part. This would also mean that more parts could be made solely from PLA/ABS, reducing print time and complexity.

New goals

Design goals: Design and make a working finger, then a palm, then a thumb, then the rest of the fingers. Primary focus is on mechanical performance, rather than being a coding/electronics exercise.

I had a chat with one of the technicians at uni and found that he had acquired a Makerbot 2X with 2 print heads. This was for printing dissolvable supports but having done some research it may be possible to print with PLA or ABS on one head, and 45flex on the other. I couldn't find anyone who has done this at uni, or online so it's going to be a learning exercise.

The first step will be to do some test prints to see if this can be done. If it can then I intend to design the hand to be printed as one piece, but of two materials. The benefit of this is firstly that I will get the structural properties of the PLA/ABS where I want it, and the flexible, grippy properties of 45flex where I want them (joints, palm and fingertips).

Before I get into the minutae of the design, I think it would be a good idea to work out the manufacturing process.

Combining .stl files

In order to perform a dual extrusion on a 3d printer, there needs to be 2 .stl files merged together on the build plate, the extrusion heads are then each...