Mylar Inflatables

Published February 04, 2018

Our task this week was to familiarize ourselves with the materials and techniques for prototyping inflatable structures. We are using aluminized Mylar, sealed with a various heating implements, to create these prototypes. This thin (5 mil) Mylar film has been coated on one side with aluminum, and is most commonly seen in the form of drug store balloons – specifically the round, star, and otherwise non-balloon-shaped balloons. The aluminum coating allows us to easily heat-seal different sections of Mylar together by without gumming up our tools with melted plastic.

Working with Mylar in this way is relatively new to me. I had previously used an even thinner (3 mil) film in the production of electrostatic headphone diaphragms (which you can read more about here) but had always used adhesives to seal the Mylar to a dissimilar material. My very non-technical opinion of the material (as compared to cling-wrap or other food films) is that it is both incredibly strong and surprisingly inelastic (for a more technical understanding, you can look to Dupont’s datasheets. When thinking of applications in inflatables, these two properties make it especially desirable to work with, because the inflated shape will not differ drastically from the cut shape.

Attempting to recreate this flat, pleated actuator mentioned in this article from Advanced Robotics, I first made the pleated form, but did not seal each pleated section to the underlying layer:

Because inflatables behave unlike any material I have dealt with before, I don’t have an innate sense of how or why mechanisms perform the way they do. I can reason that this form will bend because one side has a greater potential volume than the other, but the specific design criteria which ensure that type of action are still unclear to me. So, unsatisfied with the somewhat arbitrary-seeming inflation, and to more closely match the inspiration, I tried sealing the edges of the pleated sides together:

I can’t say the improvement was drastic, or that this mechanism now seemed to very clearly bend in one specific direction, but I did begin to understand how constraints determine type and direction of movement in inflatable structures. I would like to further this exploration by attaching the inflatable to an ‘elbow’ armature, and see if it will exert any force to bend it.

Second, inspired by some of the inflatable buildings I had seen online, I attempted to recreate the simplest of inflatable structures: a hanger. I created a sealed rectangular structure with internal seals to separate ‘tubes’ running along the shorter axis. This would, I imagines, avoid the whole structure ‘ballooning-out’ in the center, instead more closely mimicking the separate tubes of a bouncy-castle type structure. Then I added constraints between the two lower edges of the structure, such that it would naturally form an upside-down “U” shape, and hold itself up. Unfortunately, several of the internal baffles separated, creating large, inflexible sections, but the basic idea is seen below: