Exoskeleton suit replica from the movie Elysium in Autodesk Inventor, 3D printed and investment cast out of Aluminum
Already have an account? Log in.
To make the experience fit your profile, pick a username and tell us what interests you.
I designed the parts with Autodesk Inventor, 3D-printed and re-cast the largest or most complicated parts out of Aluminum 356, and used CNC for the rest. Pistons were machined on a metal lathe and the exosuit was mounted on a tactical plate carrier vest & hips.
The printable STLs are available here and on GitHub. See the assembly instructions if you want to put together one of your own.
My other projects: cyberpunk guitar and drumming robot.
The metal casting process:
Today Ligature Marks released a metal music video featuring this exoskeleton:
Some production shots:
I took the exosuit to Wasteland Weekend, and only encountered minor inconveniences like bolts loosening and falling out. I would say the exoskeleton is now totally wearable and can be considered completed.
Some progress shots from working on exo gloves and reinforcing various parts of the exoskeleton. Most of this work was done during the winter, and relatively minor work is still left to do. The project is on hold until the work space opens.
Repairs after NW Wasteland Decompression party. Replaced plastic rod ends with steel ones from McMaster Carr, electrical wiring in the spine with bungee cords. Replacing finger connectors with elastic straps as well to make the suit appropriate for extended wear.
I got to wear the suit to Wasteland Weekend afterparty (did not make it on time for Weekend itself):
This is the last week of painting, next I have assembly and tuning.
I finished casting all the required parts mid-July and moving onto finishing & painting. So far it's been a month of finishing all the cast parts by using metal filler and modeling primer to sand over imperfections, as well as using a rotary tool to clean up tiny details. I expect much less effort to be required for machined parts. No idea how much time painting will actually take after all the parts are primed & polished.
In other news, I presented this project to high school students at the Pacific Northwest College of Arts:
3D Printing to Metal Casting: Making of Elysium Exoskeleton Prop
This weekend I made a second attempt at casting chest & shin plates. For the next one I will use thin sprues in addition to thick sprues to prevent parts of the model from breaking off with the sprues.
Wax parts before dipping to build shells around them:
First attempt:
Both put together, before cleanup:
I will make another attempt at casting the chest plate to see if I can get it even cleaner, and then redo one of the shin plates before there wasn't 100% fill.
The last mold required to finish this project! There are some optional parts left to make molds for that I might do if I have time left before the Wasteland Weekend.
Assembly Instructions in PDF format: Hackaday (Dark) | Default (Light)
Thanks David, hope it helps you achieve something more awesome
Cool project! I recently started a project that involves designing an electric 3kW forge that is intended to process a large volume of aluminum cans.
Would a desktop sized system for processing soda cans and outputting a continuous stream of similar sized ingots be useful this sort of application?
The metal stock I buy comes from high-purity aluminum from melted down engine blocks, further processed by de-gassing and de-drossing. I doubt soda cans would work - but if they did, you would need a cleaning mechanism installed on the way out that will add a gas tablet, remove the gas, and remove dross particles.
Interesting. Will do some reading on de-gassing. The drossing issue I am familiar with.
If you're interested, I've created a paste dump for links related to information related to recycling, and casting aluminum and copper to form aluminum bronze.
https://old.reddit.com/r/arcticmicrofactory
My reply is based on what I learned on Alloy Avenue forums (seem to be offline currently) and what someone from my hackerspace told me. It seems to match the overall definition for 356 (lower impurities, higher strength). Makes sense that they wouldn't care about quality when making cans, but would care for car parts since they would get sued if the engines exploded all the time.
Very interesting project. I really enjoyed that film. Interesting to see that you are casting your own parts from aluminium. The actual casting part is really difficult, I know because I have tried it. Since you are putting in so much effort into making it from metal, are you not tempted to make some of mechanisms function like a real robotic exoskeleton? With assistive actuation and motors? great project, will follow with interest.
What grade Aluminium did you cast it from? I'm intending to make my own exoskeleton design. However, I'm not sure on the best choice of Aluminium from the extensive variety available.
I am using this right here (closest to 356 alloy):
http://www.ebay.com/itm/ALUMINUM-INGOTS-43-lbs-26-to-30-ingots-Casting-alloy/182166261652
The quality is a lot better than you would expect - I originally bought this as practice stock, but then realized it's actually top-shelf casting stock sold cheaply :) I get as-cast surfaces that are literally polished with no further work required.
Lastly, I did consult with two mechanical engineers here at the hackerspace (one who works at Intel and one who works in a local machine shop), and they both pointed me to 356.
You are casting, right? The alloy is completely different for CNC/machining, 6061.
See https://www.metalsupermarkets.com/7-things-consider-choosing-aluminum-grade/
Ah right. No I haven't decided yet on whether I would cast or CNC/machine it out, while CNC is much more expensive I don't really have the appropriate tools/location to cast.
I've also been looking around for months but can't seem to find those type of pistons/hydraulics used in the joins of your exoskeleton. Is there a specific name or site I've missed? Or did you get those custom manufactured?
I manufactured the pistons, see page 6 of my RPF build thread:
https://www.therpf.com/showthread.php?t=212832&page=6
...through middle of page 7:
https://www.therpf.com/showthread.php?t=212832&page=7
I don't regret learning how to use a metal lathe, but in retrospect this was the wrong decision for the project. Lathe was $500, tube stock I ordered was around $500 as well, and I spent 2.5 months, about 45 hours per week, machining until late night and coming home each day looking like a coal miner. Then I spent another $100 taking them to a painting plant not to mention I had to wake up in the middle of the night to make it back in time for work.
Knowing what I know now, I would say either low-tech them if you are going for appearance and use a convenient size of pipe (2 diff. diameters cut & inserted into each other), or post for help on a CAD forum to learn how to make drawings and submit them to a machine shop (and have them do high-detail matte/gloss indoor electro-coating). Price will likely be the same for machine shop as doing it yourself, BUT you get to spend all that time hanging out with friends instead of machining and crying :)
So with these files being online. I want to ask first. Is this design an open source design. I think it is amazing looking and would like to make one myself. I don't think i can come close to getting this level of detail. Also this suit looks amazing!
Prop replicas are never completely open source, since Sony owns the right to the movie and all props (you are always a voicemail away from Cease and Desist). However, all the files have been created from scratch and a few parts differ from on-screen appearance (thus it's not an exact copy) so as long as nobody's crazy enough to sell it they will likely let it slide.
The build was split between Viper for SLA (http://www.agile-manufacturing.com/product/used-3d-printers/viper-sla.html) and sPro for SLS (http://www.agile-manufacturing.com/product/production-3d-printers/spro-sls-series/spro-140-230-sls-center.html). Parts ordered from ProtoLabs (https://www.protolabs.com/3d-printing/).
Val
Second attempt:
Daniel Nikolajsen
Øystein
rawe
DevLab.nrw
that is DUCKING AWESOME!!!!