Revisiting a senior thesis: smart structures
Imagine having a lazy Sunday and laying out in the sun, but never having to get up and move your shade umbrella to a more optimal place throughout the day. This kind of technology is possible when structures and technology combine to make “smart” structures. Your umbrella could be a structure that senses the location of the sun through the solar panels on its covering, and depending on the amount of sunlight available, create optimal umbrella structure shapes for you.
How can this be done?
“Smart structures” are in fact highly possible. For my senior thesis at Princeton, I studied these adaptive sun shading structures, and my built model was composed of a pre-stretched dielectric elastomer adhered to an inextensible compliant frame. At a small scale, these built flexible models could furl and unfurl predictably. However, these built models were very small and labor intensive. What if there was a way to numerically compute possible “smart” structure shapes to more quickly iterate through different designs? In addition to verifying the built flexible models, I strove to understand if a computational method called dynamic relaxation could be employed for the analysis of dielectric elastomer minimum energy structures (DEMES).
So can dynamic relaxation be employed for the analysis of DEMES?
The short answer is yes, it seems like it can.
Dynamic relaxation, a common structural form finding method, was chosen as a numerical simulation technique to simulate the curling action the DEMES. Dynamic Relaxation introduces a fictitious inertia and damping terms into the equations of motion, formulating a static system as the equilibrium state for a group of damped vibrations.
Comparing computed shapes to the physically modeled stretched elastomer structures, there was a noted correlation between equilibrium angle and applied voltage for a biaxial stretch when using a modified Dynamic Relaxation algorithm with bending and clustered elements. Overall, while I found that a numerically modified structure is influenced by material uncertainties approximated input values, the Dynamic Relaxation technique was found capable of predicting the shapes and elastic energy of DEMES.
Continue reading “Revisiting a senior thesis: smart structures”