Preface: This particular costume much
harder than it would appear (Despite appearing hard already) largely due to accuracy (This was aimed to be as accurate as possible above all else), space constraints (The wings are proportionally MASSIVE compared to the back section, EVEN for wings 1/4 their size), money, weight, and tool constraints (I don't have a workshop, "beginner tools" is what I'd use to describe what I have). Simplified design, less accurate ones, less mobility, more tools (this REALLY needs a workshop to do), more money, more knowledge/experience (First time doing 95% of this), etc would have made it notably easier. The mechanical aspect made it over 10x harder to make, if it were fixed, I'd rate it "quite difficult", trying what I'm aiming for I'd rate "extremely difficult". This is also my first, so I've not had a chance to evaluate all methods fully.
I designed parts to be a lot stronger than needed since I could only estimate forces involved. Almost all parts worked as designed, but excess weight and some parts needed some testing. Having done it I can give some recommendations though:
Eva foam over aluminium frame construction is great for large scale, but requires proper tools. Aluminium alloys are stronger than you'd think, I used 2-3mm thick in places where 1 or 1.5mm would have done. Tracing paper as diffuser behind acrylic worked better than expected. Contact adhesive to glue foam it absolutely brilliant, but takes a more practice than you'd think to get it perfect. Use a ceramic sharpening rod to keep disposable box-cutter blades sharper for longer (I would have gone through over 60 blades if not for honing). Lipo batteries this powerful can output over 200amps, instantly turning small wires to flames, so be careful.
Do you have a making-of website for it? I'd like to know more about the materials you used to create it and the process you followed to design it.
I don't have one at the moment, although I've been asked quite a bit, but still deciding if I should start doing it. This is my first, so I've not got any place to post them.
Materials-wise (Partly summarized since there's so many):
- Square tube aluminium for most of the frame (1.5 and 2mm thick walls, 20,25,30mm square); Sheet aluminium (1.5, 2, and 2.5 mm thick), 2 large aluminium bars for main wing cross, 10mm aluminium round tube to support wings.
- Over 100 M4 bolts and nuts (largely 30mm and 40mm), M4 grub screws for locking sections to motors, various M5 and M6 bolts, 4x shoulder bolts for main pivot points, some computer screws, nylock nuts for heavy rotation parts
- Wiring contains about 4 meters of ethernet cable (8wire) for controlling servos and detactable control pad, About 7m of 2 wire Audio cable reused as separate 12v and 6v lines for lighting/motors and servo power, various small wires
- Arduino Mega microcontroller, Various resistors/caps/diodes (mostly pulled from old hardware), n channel mosfet (pulled from old psu) for flashing lights, power switches, buttons, limit microswitches, 10-turn potentiometer for measuring main feather lifting motor rotation, voltage regulator for 6v, L298N H-bridge motor controller, arduino protoboard
- 11-12m of led strips (60/m) so around 660-720 leds, several separate leds
- Around 120 square feet of EVA foam (Sold as Anti-fatigue floor matting), which is 32 60x60cm sheets (about 4 times of what an average cosplay uses)
- 2 sheets of 100cm x 50cm and 2x A4 sheets clear blue acrylic, tracing paper as a diffuser
- 2x 500ml and 1x 250ml cans of evo-stik impact (Contact adhesive), 8 tubes of superglue, 2.6litres of Mod Podge (Glue, but also used to finish foam), 1 tube of flexible decorators caulking as filler
- Many sheets of 120-400 grit sand papers
- 5x 400ml spray cans of primer (proved to not be enough), 5x 400ml spray cans of elegant copper (Note: Actually looks closer to gold, which was intended. I didn't want a yellow gold look), 5x satin clear coat spray, tubes of black and silver acrylic paints.
- 140 x about 4cm piping for sword
- 2x 5000mah and 1x 2500mah 3s lithium polymer batteries, lots of power connectors
- Mild steel box section for the main hinge (Took an angle grinder to cut it)
- Teflon sheets for bearing material
- 6 servos (will have more used when finished), 2x 50kg/cm torque (very high) motors (will have 2 more when complete)
- 6 linear bearings for the end of wing extension
- Dyneema as wing lifting power transfer lines (Extreme high strength string)
- 4x4 touch pad to control different aspects (Lighting mode, opening/closing, pulsing speed, raising/lowering, enabling/disabling servos,etc.
- w/e else... there's quite a lot of things not listed, but you get the idea
Essentially the process was:
- Extract model from game, scale to my size, minor adjustments to make it physically possible to build
- Research and evaluate different existing construction techniques for price and ease of manufacture
- Research effects and finishes that could be used and how they'd look
- After choosing design method, adapt model to fit basic concept, design frame to fit within it
- Design mechanics of joints and lifting actions while evaluating cost and ease of manufacture of parts
- Cut frame parts to size, drill hundreds of holes, bolt together
- wire all electronics and powerlines, soldering hundreds of things,
- Program microcontroller (Just over 1000 lines of code)
- Measure parts in 3d and transfer to foam while taking into account heat forming, offsets, etc. Note: The majority of full costumes work from premade templates, mine had never been done like this before, and contains large number of organic curves. Every edge was measured and adjusted to compensate for 3d molding largely through estimation, which means most parts are within 5%. This requires a very good ability to visualize working from a 3d to 2d then back to 3d via molding/bending/etc. A lot of costumes are done with flat surfaces which are much easier.
- cut out, heatform, glue, trim, fill gaps
- Heat form acrylic parts with heat gun by gloved hands (i.e. how to burn yourself 101)
- Heat seal foam, lightly sand, Mod podge coat, sand, 3 more coats, sand, prime, sand, paint, clear coat.
I would give more, but trying to list 1000+ steps. I can't really list all the sources, as it's over 100 hours of researching, but can give some specifics if needed.
Is this something you do professionally as well? Something that elaborate looks like it'd be very difficult for someone to do unless they had some sort of professional background.
I have no professional background in any aspect of it, except programming and even then the language is actually a bit different to what I'm used to.
The amount of knowledge required to do this is in the order of 2 years of a university degree. I just happen to already know quite a bit about physics, electronics, computer stuff, material engineering, parts manufacturing, etc through curiosity, but no where near enough, so it required a lot of research.
For a non-professional, non-experienced average person I'd rate this project about 9-10 out of 10 for difficulty. I happen to know enough and have the mindset to do this kind of thing, and exceptionally good at DIY-type tasks, it would be impossible otherwise.
I make elaborate Halloween costumes every year (spend a month or two on each), but I've never done anything motorized. FWIW, I can't link to it yet because I'm a newbie, but if you google "g42 jonsnow" you'll find my website where you can see the sort of costumes that I make.
Ah yeah I've found them, pretty hillarious ones.
When you add motorized movements to your outfits, do you have to develop them from scratch and do soldering and the like or are there programmable systems that you could get off the shelf?
It was designed to be motorized. I chose the actions I wanted it to do and designed the frame accordingly. The frame and mechanics designs were done from scratch, none of the designs on the internet would work for mine since it's very unique limitations. In fact, most wings even 1/4 of the size had larger connecting areas.
Honestly, adding basic motorization on very low load aspects is fairly simple, the hard part about that is the hinges and linkages. Heavy load, multiple joint motions with strict space requirements make it MANY times harder.
Example: If you want little flaps to move. I'd do it with a battery (Lipo if lots of power required or weight limited), voltage regulator, arduino (Anyone would do, although the smallest ones do need separate programmers since they don't have usb), servos, a hinge on the joint, a little connecting arm, wires, buttons or sensors to control it. That could be added to an existing design fairly quickly with some knowledge.