The monsters keep dancing in my head is a study and exploration of a number of things - a study of the phenomenon of persistence of vision - the process that allows static images, played at a specific speed to take on animated qualities (for those more scientific purists - phi phenomenon and beta movement.) Itís a first attempt for me at working with gears and the transfer of motion, and itís a look into 3d printing technology and prototyping/manufacturing - allowing us to take the ideas in our head and massage them into existence - including the design, design improvement and prototyping processes.
In thinking on this project, I was struck by the thought that most 3d printing projects fall into (very generically speaking) one of two categories - either utilitarian (gears, replacement parts, fixes to an issue) or tchotchkes (reproducing pop items like your favorite star wars figurine, or 3d printing a vase or lampshade) part of this is undoubtedly, that like all new technologies, until we figure out what weíre doing, we do what we know. I find in my situation, the one thing I want to do and cant, is prototype things - I'll have a great idea, but no feasible way to produce it or check its validity as a functional design. This project - part tchotchke, part utilitarian - was designed to stretch not only my artist and design chops, but to stretch my prototyping/fabrication knowledge. The laws of physics, the available print area on desktop 3d printers, and over-all visual esthetics were central design considerations used when creating this project.
In its completed state "the Monsters" is a 3D persistence of motion sculpture, featuring a hand crank driven wheel with 12 three-dimensional sculptures mounted to it, each posed in a position that makes up 1/12th of a dance animation loop. The wheel is buried inside a 3D sculpture of my head, the model of which was created with 123D catch, then edited to support the internal wheel and drive mechanisms. To operate, a user turns the external crank - when the proper speed is reached (the equivalent of between 12 and 24 fps, depending on how smooth you want the animation to appear.) POV takes over and it stops looking like a spinning wheel, and becomes a dozen monsters dancing in place around on the surface of my head.
Step 1: Are you sure you know what you're doing?
Step one, before getting too deep into creating materials, was to prove that this thing would even work. After some investigation into POV, early cinematic "toys" and current film projection practices, I discovered 12 frames per second (FPS) is about the minimum rate one can move static images and pull off something that appears animated. From a manufacturing point of view, I knew I wouldnít want to have to print a million statues anyway, and, as there's limited real-estate on the rim of the wheel, the more statues, the smaller each would have to be, eventually killing the over-all readability of the effect.
To test feasibility of the 12 statue model, I created a simple test sequence in a 3D program. Start in a drawing program like Illustrator and lay out a wheel with 12 equidistant spokes - these will mark the location for each of your statues. To simplify the test, I made a series of 12 circles, each slightly different size than the previous to stand in as my statues - the idea being that if the approach worked, I should see what looks like one circle growing (and shrinking) in size when the hub circle revolves at the correct speed. After placing the model circles in the appropriate spaces around the wheel, group the drawing and export as paths for your 3d software.
The import and trial of the wheel worked so well, I had to stamp an arrow shaped hole in the disk to make sure i was actually seeing the rotation. While at it, I added a mat shaped like the outline of the human head to see a better simulation of what would happen if we focused the viewers attention by blocking out a number of the statues.
Step 2: Time to start laying out the prototype parts!
Once the demo showed promise, it was time to start prototyping parts. When laying out parts, remember the first consideration is the size of the machine you'll be using to print the parts - donít lay out files too big for your printer! One of my fondest memories from childhood was snapping together brightly colored plastic pieces to assemble toys, so designing individual parts seemed like something I could figure out. Returning to the original test wheel in Illustrator, I then broke the wheel into sections, keyed the spokes so they would interlock, added a square opening for the central hub that would turn the wheel, and began exporting parts paths so I could render them into 3d objects.
I lucked out on the gear creation portion of this project in two ways - first, I found an online gear creator page (http://woodgears.ca/gear_cutting/template.html) so I didnít have to geometrically tax myself in illustrator to create paths. Second, the apparatus was hand cranked - this eliminated me having to do the math on what gear ratios, motor speed, etc., would be required to reach the rotational equivalent of 24 FPS. Simply let each user crank the handle till they see what looks like a static statue dancing in place. My only real contribution in this case was "when a big gear turns a little gear = speed" We donít want anyoneís arm falling off before they reach the proper rotational speed! Once all wheel parts, gears, cranks and support items were laid out, a quick import of the paths into my 3D software and voilaÖparts!
Once the internal mechanics looked promising, I did a little additional manipulation to the parts, hoping to help lock this collection of rotating parts together - thereby avoid a centrifugal force accident, and someone losing an eye. I dovetailed the pins that the statues would connect to - on "bridged" sections (two pieces that split a spoke,) sliding on the matched dovetailed statue base would help lock the pieces together.
I also added a flange on the smaller end of each wheel pie piece. These would then be locked together with a square retaining ring.
Step 3: Modeling the statues
While trying to settle on the final "monster" character I would use, I decided it wouldnít hurt to run an actual trial of dancing statues - after all, circles growing and shrinking are one thing, but dancing humanoids is another ball game.
Again, using the tools available, I opened DAZ 3d, another modeling package that I use on occasion. This one is perfect because it has a character animation section. I placed a character on the stage, applied the dance move I wanted, making sure my character's last move lead back to its first move (for looping purposes) then divided the short animation roughly into 12 sections, as each pose should be 1/12 further into the sequence. One at a time I set the timeline on each of these key frames and exported the posed model as a 3D obj file.
I then imported these statues into my other program and placed them around the circumference of the wheel, each on the appropriate spoke pin. Some quick math told me if I have 12 items going around 360 degrees of a wheel, each one would need to be rotated 30 more degrees than the previous (12 statues x 30 degrees = 360 degrees) to appear to be standing flat footed on the wheel rim. One quick note - make sure you have everything else set in position before starting to import statues - mine were pretty intense from a geometry (parts) point of view so moving things around once they are all in the scene is pretty processor intense!
Step 4: Modeling my HEAD!!!!!
In the process of creating two previous Instructables - http://www.instructables.com/id/Cool-AutomataFaux-Steampunk-inspired-CD-Case/ and http://www.instructables.com/id/123D-Make-inspired-Acrylic-Lamp-Mashup/ I discovered Autodeskís suite of 123d products - all of which are free. If you havenít looked at them yet - you should they are crazy good for the price. For this Instructable, I decided to take a shot at using 123D catch - a software package that stitches together multiple photos of an object to create an actual 3d model
It took three passes to get a good model, mostly because we made some incorrect assumptions about what part of the picture the software keys off of to build its dimensions. Attempt number three did the trick, mostly (it did some weird things to my nose, hair and chin) but we ended up with a useable model :) I exported the model from 123D catch as an obj file and imported into my 3D modeling program for editing
I also took the key frontal face photo used to develop the model and imported it into Illustrator to layout the internal structure to connect the wheel apparatus, crank, etc., to the model of the head.
Step 5: The Final (almost) unveiling
With all the parts successfully modeled, it was time for a (visual) dry run. As I donít have access to a printer, all of this is theoretical, but I believe it to be sound. I you'd like to print one - contact me and we'll give it a go - though you should probably reshoot the head model to look like you.
the insertion of the red and green internal support panels into the model of the head provide a framework for inserting the pin and crank of the internal wheel mechanism. Extending the base of the head model shell, or building up the support legs of the wheel mechanism would help stability, but at this point we know the unit can be mounted properly into the head cavity.
There are still a number of Items I'd like to clean up and refine:
First I still havenít decided what actual monster model I want to use. I've also been kicking around the idea of changing the name to "the robots keep running through my head" or "the aliens are stealing my brain" and using an alien or robot model instead.
Second, the back of the sculpture needs additional shrouding to hide some of the mechanics, and a base plate and leg extensions to attach the gear support frame to the bottom of the head model to tie the unit together, and provide stability.
The original concept of having animated eyes showing through the open eye sockets (attaching photos of my eyes in different "poses" then adhering them to the circular shaped lock keys half way up the spokes so as the wheel is cranked, it looks like my eyes are following the monsters) was scrapped - the photos of the eye would set too far back inside the head model to be visible from any angle other than straight on.
Early on, the choice was made to not have the hand crank coming out of my mouth - now that it lines up with my nostrils, the mouth isnít looking like such a bad option...
Lastly, the shell of the head model would need to be sectioned - or printed on a rather large printer. A center seam down the face, then fusing the red and green backer plates - one to either half - would probably be the best approach for holding the full item together.